From dashing.meng at gmail.com Sun May 1 07:03:56 2011 From: dashing.meng at gmail.com (littlebat) Date: Sun, 1 May 2011 13:03:56 +0800 Subject: [FFmpeg-user] Installation Problems In-Reply-To: References: Message-ID: <20110501130356.3714c1d2.dashing.meng@gmail.com> On Wed, 27 Apr 2011 22:17:16 -0400 Grant Smith wrote: > I'm trying to install ffmpeg on Ubuntu 10.04 using the tutorial found > @ http://ubuntuforums.org/showpost.php?p=9868359&postcount=1289 I tested tutorial for installing a static linked ffmpeg with the style of Ubuntu's deb package. I found I must execute "sudo mkdir -p /usr/local/share/doc/lame" before "sudo checkinstall ..." when install LAME, or it will fail. From andreluizmbm at bol.com.br Sun May 1 08:35:36 2011 From: andreluizmbm at bol.com.br (Andre) Date: Sun, 1 May 2011 06:35:36 +0000 (UTC) Subject: [FFmpeg-user] mpeg2 file Message-ID: I have 2 files. One is the video, file.m2v, the other is the audio, audio.wav. How do I join these files to make a mpeg2 file. What extension should I use: .mpg, .vob? From krueger at signal7.de Sun May 1 08:57:13 2011 From: krueger at signal7.de (=?iso-8859-1?Q?Robert_Kr=FCger?=) Date: Sun, 1 May 2011 08:57:13 +0200 Subject: [FFmpeg-user] ffmpeg and x264 In-Reply-To: References: <61386.97286.qm@web86408.mail.ird.yahoo.com><237821.66460.qm@web86407.mail.ird.yahoo.com> Message-ID: On Apr 30, 2011, at 17:39 , Dave Pope wrote: > Thanks! The change makes perfect sense, I just had to dig around for while before figuring it out. Do you know if there's a patch in the works to make 'ffmpeg --help' report something about these? It seems like it might require building some "handshaking" between them that might not already exist in general form. > > ________________________________ > > From: ffmpeg-user-bounces at ffmpeg.org on behalf of Robert Kr?ger > Sent: Sat 4/30/2011 4:04 AM > To: FFmpeg user questions and RTFMs > Subject: Re: [FFmpeg-user] ffmpeg and x264 > > > > > On Apr 29, 2011, at 17:37 , Dave Pope wrote: > >> Yeah, I hit this one too. Use "-preset" and "-profile" switches instead >> of -vpre. "-preset fast -profile baseline" works OK for me. Annoying >> that it's not reflected in the --help output; I don't see a way to list >> the available values yet either. >> > > Check x264 docs. As far as I understood the purpose of the change was not to duplicate efforts from the x264 project and that is documented quite well. > > try x264 --fullhelp > No, I'm not aware of anything like that. And BTW, please don't top-post (http://en.wikipedia.org/wiki/Posting_style). It is a rule the list has given itself to make following discussion threads easier. From belcampo at zonnet.nl Sun May 1 11:13:35 2011 From: belcampo at zonnet.nl (belcampo) Date: Sun, 01 May 2011 11:13:35 +0200 Subject: [FFmpeg-user] mpeg2 file In-Reply-To: References: Message-ID: <4DBD243F.3010001@zonnet.nl> On 05/01/11 08:35, Andre wrote: > > I have 2 files. One is the video, file.m2v, the other is the audio, audio.wav. > How do I join these files to make a mpeg2 file. What extension should I > use: .mpg, .vob? Both are acceptable mpg is nroamlly used, .vob when it would become a part of a DVD structure. > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From mjs973 at optonline.net Sun May 1 13:52:33 2011 From: mjs973 at optonline.net (Mike Scheutzow) Date: Sun, 01 May 2011 07:52:33 -0400 Subject: [FFmpeg-user] mpeg2 file In-Reply-To: References: Message-ID: <4DBD4981.8070301@optonline.net> Andre wrote: > I have 2 files. One is the video, file.m2v, the other is the audio, audio.wav. > How do I join these files to make a mpeg2 file. What extension should I > use: .mpg, .vob For future reference, there is documentation at http://www.ffmpeg.org/ffmpeg.html I would try a command like: ffmpeg -i file.m2v -i audio.wav -vcodec copy -acodec copy out.mpg -map 0.0 -map 1.0 Mike Scheutzow From jfrotscher at gmail.com Sun May 1 14:16:53 2011 From: jfrotscher at gmail.com (Johannes Frotscher) Date: Sun, 1 May 2011 13:16:53 +0100 Subject: [FFmpeg-user] Copyright issues iFFmpeg Message-ID: <42F9BCDC-E783-4371-8653-34010804B71F@gmail.com> Hi there, I was just checking whether the team is aware that someone is corrupting the GPL and selling this software for profit of $10, $15 $23.32: http://www.macupdate.com/app/mac/35846/iffmpeg http://www.macupdate.com/app/mac/8988/ffmpegx http://www.macupdate.com/app/mac/21888/visualhub According to copyright law, if a software uses another backend software, and it makes up the main part of the software in order to function, that software cannot be incorporated without prior consent to the originators of the backend license agreement. They have been doing that for quite sometime, some as far as almost 10 years. And I don't understand why nothing happened over these years. Plus what is the Hall of Shame for on the ffmpeg website? Yours Jack From h.reindl at thelounge.net Sun May 1 19:40:17 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Sun, 01 May 2011 19:40:17 +0200 Subject: [FFmpeg-user] Copyright issues iFFmpeg In-Reply-To: <42F9BCDC-E783-4371-8653-34010804B71F@gmail.com> References: <42F9BCDC-E783-4371-8653-34010804B71F@gmail.com> Message-ID: <4DBD9B01.2050503@thelounge.net> Am 01.05.2011 14:16, schrieb Johannes Frotscher: > Hi there, > > I was just checking whether the team is aware that someone is corrupting the GPL and selling this software for > profit of $10, $15 $23.32: > http://www.macupdate.com/app/mac/35846/iffmpeg > http://www.macupdate.com/app/mac/8988/ffmpegx > http://www.macupdate.com/app/mac/21888/visualhub > > According to copyright law, if a software uses another backend software, and it makes up the main part of the > software in order to function, that software cannot be incorporated without prior consent to the originators of the > backend license agreement. > > They have been doing that for quite sometime, some as far as almost 10 years. And I don't understand why nothing > happened over these years. Plus what is the Hall of Shame for on the ffmpeg website? please would you explain where is your problem? this is simply a garphical user interface for ffmpeg not more, not less -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From jfrotscher at gmail.com Sun May 1 21:57:25 2011 From: jfrotscher at gmail.com (Johannes Frotscher) Date: Sun, 1 May 2011 20:57:25 +0100 Subject: [FFmpeg-user] Copyright issues iFFmpeg Message-ID: <77159931-AFC5-46D7-8F77-351F4E770414@gmail.com> I do not have a problem, it is not my development work of decades getting sold out for nothing, I do not care much just as to inform you what illegal practices are going on. A mere GUI for a solid program it is, precisely the point, you cannot sell software which uses the main functionality of another software and effectively braking the license agreement and copyright originator. It is therefore an infringement and making money of it, the GNU and EEF foundations would not be very pleased with it. I thought FFmpeg was OpenSource, but what do I care, do what you want, I have no problem. Its yours to do what ever you want with it. Bye From jfrotscher at gmail.com Sun May 1 22:00:54 2011 From: jfrotscher at gmail.com (Johannes Frotscher) Date: Sun, 1 May 2011 21:00:54 +0100 Subject: [FFmpeg-user] Copyright issues iFFmpeg Message-ID: And I just found on the ffmpeg website itself lol on the big heading of "Legal": http://www.ffmpeg.org/legal.html "Q: Is it perfectly alright to incorporate the whole FFmpeg core into my own commercial product? A: You might have a problem here. There have been cases where companies have used FFmpeg in their products. These companies found out that once you start trying to make money from patented technologies, the owners of the patents will come after their licensing fees. Notably, MPEG LA is vigilant and diligent about collecting for MPEG-related technologies." From h.reindl at thelounge.net Sun May 1 22:05:41 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Sun, 01 May 2011 22:05:41 +0200 Subject: [FFmpeg-user] Copyright issues iFFmpeg In-Reply-To: <77159931-AFC5-46D7-8F77-351F4E770414@gmail.com> References: <77159931-AFC5-46D7-8F77-351F4E770414@gmail.com> Message-ID: <4DBDBD15.8090009@thelounge.net> Am 01.05.2011 21:57, schrieb Johannes Frotscher: > I do not have a problem of course, your problem is not understand what you speaking about > I do not care much just as to inform you what illegal practices are going on this is bullshit > A mere GUI for a solid program it is, precisely the point, you cannot sell software which uses the main > functionality of another software and effectively braking the license agreement and copyright originator. where does it break anything? it does not selling ffmpeg, it sells a gui > You might have a problem here. There have been cases where companies have used FFmpeg in > their products. These companies found out that once you start trying to make money from > patented technologies, the owners of the patents will come after their licensing fees. > Notably, MPEG LA is vigilant and diligent about collecting for MPEG-related technologies." you should try to understand what you read no line here says that any ffmpeg-copyright is the problem the problem ist that some inside ffmpeg useses algoritms which where somebody like MPEG LA holds patents and as long ffmpeg is downloaded as source there is no problem - if you include FFMPEG AS BINARY even with totally untouched source you can have a problem with MPEG LA no ffmpeg-license again: you are speaking bullshit as long we are speak about a GUI -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From jieyunfu at gmail.com Sun May 1 22:14:13 2011 From: jieyunfu at gmail.com (Jieyun Fu) Date: Sun, 1 May 2011 16:14:13 -0400 Subject: [FFmpeg-user] Copyright issues iFFmpeg In-Reply-To: <4DBDBD15.8090009@thelounge.net> References: <77159931-AFC5-46D7-8F77-351F4E770414@gmail.com> <4DBDBD15.8090009@thelounge.net> Message-ID: Let's turn down the volume here and stop yelling at each other. I believe Johannes started the discussion for good will. Johannes: In my humble opinion, the difference is that iffmpeg does not contain any ffmpeg binaries. It does not try to "steal" ffmpeg code or algorithms. It is just selling the ffmpeg GUI. Correct me if I were wrong. Thanks. On Sun, May 1, 2011 at 4:05 PM, Reindl Harald wrote: > Am 01.05.2011 21:57, schrieb Johannes Frotscher: > > I do not have a problem > > of course, your problem is not understand what you speaking about > > > I do not care much just as to inform you what illegal practices are going > on > > this is bullshit > > > A mere GUI for a solid program it is, precisely the point, you cannot > sell software which uses the main > > functionality of another software and effectively braking the license > agreement and copyright originator. > > where does it break anything? > it does not selling ffmpeg, it sells a gui > > > You might have a problem here. There have been cases where companies have > used FFmpeg in > > their products. These companies found out that once you start trying to > make money from > > patented technologies, the owners of the patents will come after their > licensing fees. > > Notably, MPEG LA is vigilant and diligent about collecting for > MPEG-related technologies." > > you should try to understand what you read > > no line here says that any ffmpeg-copyright is the problem the problem ist > that some inside > ffmpeg useses algoritms which where somebody like MPEG LA holds patents and > as long ffmpeg > is downloaded as source there is no problem - if you include FFMPEG AS > BINARY even > with totally untouched source you can have a problem with MPEG LA no > ffmpeg-license > > again: you are speaking bullshit as long we are speak about a GUI > > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > From h.reindl at thelounge.net Sun May 1 22:22:18 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Sun, 01 May 2011 22:22:18 +0200 Subject: [FFmpeg-user] Copyright issues iFFmpeg In-Reply-To: References: <77159931-AFC5-46D7-8F77-351F4E770414@gmail.com> <4DBDBD15.8090009@thelounge.net> Message-ID: <4DBDC0FA.5030308@thelounge.net> Am 01.05.2011 22:14, schrieb Jieyun Fu: > Let's turn down the volume here and stop yelling at each other. I believe > Johannes started the discussion for good will. yes but totally wrong beginning with the first line > I was just checking whether the team is aware that someone is corrupting the GPL GPL != LGPL = FFMPEG > Johannes: In my humble opinion, the difference is that iffmpeg does not > contain any ffmpeg binaries. It does not try to "steal" ffmpeg code or > algorithms. It is just selling the ffmpeg GUI. http://en.wikipedia.org/wiki/GPL http://en.wikipedia.org/wiki/LGPL so you can not only use the ffmpeg-cli for your GUI what would even possible if ffmpeg would be GPL you can even use and link the ffmpeg-libs in your code becuase they are LGPL -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From hardik.sharma22 at yahoo.com Sun May 1 22:33:25 2011 From: hardik.sharma22 at yahoo.com (Hardik Sharma) Date: Sun, 1 May 2011 13:33:25 -0700 (PDT) Subject: [FFmpeg-user] Fw: Unable to get correct decoded file Message-ID: <906776.36347.qm@web46209.mail.sp1.yahoo.com> Thanks Leo but still I am not getting desired output. My decoded video output is with very slow frame rate (even after mentioning frame rate as 30fps) and there is some problem in video frames as per appearance concern. It's different from the input yuv video even without much loss. Please tell me if I should include more format option while encoding or decoding video sequence. I really appreciate your help. Thanks. Regards, Hardik Sharma ? ? --- On Sat, 30/4/11, Hardik Sharma wrote: >From: Hardik Sharma >Subject: Unable to get correct decoded file >To: "ffmpeg" >Date: Saturday, 30 April, 2011, 12:16 AM > > >Hi, > >I am a new bee with Ubuntu and ffmpeg both so please help me with this issue. I encoded the yuv video to h264 by following command-? >?ffmpeg -f rawvideo -r 30 -b 256k -s 352x288 -i silent_cif.yuv -vcodec libx264 -b 256k -s 352x288 -preset slow -f h264 -threads 0 silent264.h264 > >But I don't think I am getting correct encoded file as after decoding back to yuv video, I am only getting 2 frames and that too wrong frames. > >Command line for decoding- >?ffmpeg -r 30 -b 256k -s 352x288 -i output.h264 -vcodec libx264 -b 256k -s 352x288 -preset slow -f rawvideo -threads 0 slient.yuv > >Please tell me my mistake in command line. I really appreciate your help. Thanks. > >Regards, >Hardik Sharma? > > From gavr.mail at gmail.com Sun May 1 22:35:48 2011 From: gavr.mail at gmail.com (Kirill Gavrilov) Date: Mon, 2 May 2011 00:35:48 +0400 Subject: [FFmpeg-user] decoding subtitles Message-ID: Hi, this moment I'm trying to understand how should I parse subtitles using FFmpeg libraries. I call avcodec_decode_subtitle() and investigate results in AVSubtitleRectstructure. Each time I got SUBTITLE_ASS type (even for STR subtitles), .text field is empty and .ass field filled as "ASS/SSA compatible event line" like this: Dialogue: 0,0:37:01.26,0:37:03.16,Default,,0000,0000,0000,,He couldn't have gone far. I found some basic description for ASS format but via AVSubtitleRect I got only Dialogue events and no any Format description. So I can not retrieve even text because Text field position is unknown for me (it should be last but delimiter , can be used in text). Could someone explain me how should I retrieve Format for ASS subtitles given by FFmpeg? Bitmap subtitles currently out-of-interest for me. Thanks! ----------------------------------------------- Kirill Gavrilov, Software designer. From leo.izen at gmail.com Mon May 2 01:12:49 2011 From: leo.izen at gmail.com (Leo Izen) Date: Sun, 1 May 2011 19:12:49 -0400 Subject: [FFmpeg-user] decoding subtitles In-Reply-To: References: Message-ID: On Sun, May 1, 2011 at 4:35 PM, Kirill Gavrilov wrote: > Hi, > > this moment I'm trying to understand how should I parse subtitles using > FFmpeg libraries. > I'm not sure, but questions like this usually go to the libav-user at ffmpeg.org mailing list. Use that: http://ffmpeg.org/mailman/listinfo/libav-user Though if someone knows they can still answer here. From cs_palkar at yahoo.com Mon May 2 01:27:50 2011 From: cs_palkar at yahoo.com (Charu Palkar) Date: Sun, 1 May 2011 16:27:50 -0700 (PDT) Subject: [FFmpeg-user] Request for advise on performance - transcoding on the fly In-Reply-To: <906776.36347.qm@web46209.mail.sp1.yahoo.com> References: <906776.36347.qm@web46209.mail.sp1.yahoo.com> Message-ID: <717280.70640.qm@web65612.mail.ac4.yahoo.com> Hi, I am new to the forum and would like guidance. System : Server : Rackable CPU : AMD Opteron (2GHZ) Dual CPU RAM : 2 Gb Disk : SATA ( 250GB x 4 disks ) - 3 disks configured as RAID0 (3Ware) Network : 1 Gig OS : Slackware - 13.1 Command : ffmpeg -benchmark -i /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -threads 4 -target ntsc-dvd -aspect 16:9 test.mpg NOTES : Only one CPU has 90% (sys + user), others 4% (sys + user), no wait on I/O There is no swapping, initial page faults is 11 and remains constant for the entire time Both files input and output are located on the RAID partition The time it takes to transcode is 20% higher than the run-time of the media file Nothing else is running on the server ffmpeg - Built with ARCH=x86_64 CPU=opteron QUESTIONS : Q : What else can I do to speed up the transcoding to take less time than run-time of media file ? + To use for on-the-fly transcode with a media server. Background : Installed media server 'serviio' and transcoding on-the-fly-enabled. The following command is invoked. ffmpeg -i /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -y -threads 4 -vcodec mpeg2video -sameq -r 23.976 -g 15 -copyts -acodec ac3 -ab 192k -ac 6 -map 0:0 -map 0:1 -sn -f mpegts /mnt/samba/trans_tmp/Serviio/transcoding-temp-2-MPEG2TS.stf CPU Usage is 60% per core and no wait on I/O, swapping. Viewing using WDTV Live streaming stops after 31 secs consistently Viewing using XMBC on WinXP stream starts buffering after 7 mins or 3 mins of play, takes about 8 to 10 seconds to catch up Cabled network 100 MB connection on WinXP the network usage varies between 7% to 50% Q : What can be done to make transcoding on-the-fly continuous ? Any and all help is greatly appreciated. Thanx Charu Palkar -- From jfrotscher at gmail.com Mon May 2 01:54:08 2011 From: jfrotscher at gmail.com (Johannes Frotscher) Date: Mon, 2 May 2011 00:54:08 +0100 Subject: [FFmpeg-user] Copyright issues iFFmpeg Message-ID: Another case of stupid abuse of fundamentals, "Free Software GPL", yeah right as in free to plunder, especially the intelligence of users and developers. Steve Jobs: "Good artists copy, great artists steal.", classic paradigm. I know I should not start a discussion of legality here, but this is one of the most prominent cases going well beyond anything. GPL is shooting itself in its left foot if you ask me, what outrageous stupidity. GPL is really a virus, as stated in the Wikipedia article of yours: http://en.wikipedia.org/wiki/GPL#Criticism How else can there be more and even more commercialism, and misinformation, miseducation as on the web right now. And don't tell me its because of the increasing number of users, that is the true bullhog here. The clear distinction of free and unfree software from its origin has been contaminated, by its own people. FreeBSD is the only true freedom act that ever happened to all Unices. Its also the reason why everything is getting patched and cracked beyond degree out there, because its an anarchic state, people wake up. And to correct again: http://www.gnu.org/licenses/gpl-faq.html#MereAggregation The very source of the definition my friends, and if those programs use numerous ever-never-stop-expanding-since-10-years-APPLE-APIs-of- which-iFFmpegGUI-was-using-ffmpeg-selling-it, which have their own copyrights, and license agreements, then I am in fact extending the functionality of the program, thank you no applause please. BTW nice cases of its former Hall-of-Shame here: http://news.ycombinator.com/item?id=790316 Intelligence is conceit, wisdom is bliss. The funny thing is, its actually like the even cranker counter-part of the capitalist markets out there, almost pinpointing its weaknesses. Its not just a matter of intellectual property, its a matter of the legal-economic framework, which define the finale of politics and ideology, hence the dynamics of our future. From h.reindl at thelounge.net Mon May 2 02:22:51 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Mon, 02 May 2011 02:22:51 +0200 Subject: [FFmpeg-user] Copyright issues iFFmpeg In-Reply-To: References: Message-ID: <4DBDF95B.8040708@thelounge.net> Am 02.05.2011 01:54, schrieb Johannes Frotscher: > GPL is shooting itself in its left foot if you ask me, what outrageous stupidity. > > GPL is really a virus, as stated in the Wikipedia article of yours: > How else can there be more and even more commercialism, and misinformation, miseducation > FreeBSD is the only true freedom act that ever happened to all Unices. > because its an anarchic state, people wake up. > The very source of the definition my friends, and if those programs use numerous > ever-never-stop-expanding-since-10-years-APPLE-APIs-of-which-iFFmpegGUI-was-using-ffmpeg-selling-it > which have their own copyrights, and license agreements, then I am in fact extending the functionality > of the program, > thank you no applause please. > Intelligence is conceit, wisdom is bliss jesus christ what are you trying to tell us after some hours ago you did not know about the difference GPL/LGPL and quoted a paragraph of the ffmpeg-legal-page without understanding a word you read there? -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From developer at noknok.net Mon May 2 03:06:35 2011 From: developer at noknok.net (NokNok Developer) Date: Sun, 01 May 2011 21:06:35 -0400 Subject: [FFmpeg-user] Request for advise on performance - transcoding on the fly In-Reply-To: <717280.70640.qm@web65612.mail.ac4.yahoo.com> References: <906776.36347.qm@web46209.mail.sp1.yahoo.com> <717280.70640.qm@web65612.mail.ac4.yahoo.com> Message-ID: <4DBE039B.3070306@noknok.net> Neither of those streams you are doing are "live transport" streams. For example the WD LIVE stops because when it issues its 1st GET, it also gets the CURRENT size of the file, at that point in time, and then thats all it asks for. You need to use a live streaming container/player. eg: HLS (Apple M3U8), or FLV with Flash Player, or ASF Streaming, etc. There are other approaches, as to streaming in chunked http, etc. Shawn On 5/1/2011 7:27 PM, Charu Palkar wrote: > Hi, > > I am new to the forum and would like guidance. > > System : > Server : Rackable > CPU : AMD Opteron (2GHZ) Dual CPU > RAM : 2 Gb > Disk : SATA ( 250GB x 4 disks ) - 3 disks configured as RAID0 (3Ware) > Network : 1 Gig > OS : Slackware - 13.1 > > Command : > ffmpeg -benchmark -i > /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -threads 4 -target > ntsc-dvd -aspect 16:9 test.mpg > > NOTES : > Only one CPU has 90% (sys + user), others 4% (sys + user), no wait on > I/O > There is no swapping, initial page faults is 11 and remains constant for > the entire time > Both files input and output are located on the RAID partition > The time it takes to transcode is 20% higher than the run-time of the > media file > Nothing else is running on the server > ffmpeg - Built with ARCH=x86_64 CPU=opteron > > QUESTIONS : > Q : What else can I do to speed up the transcoding to take less time than > run-time of media file ? > + To use for on-the-fly transcode with a media server. > > Background : Installed media server 'serviio' and transcoding > on-the-fly-enabled. The following command is invoked. > > ffmpeg -i > /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -y -threads 4 > -vcodec mpeg2video > -sameq -r 23.976 -g 15 -copyts -acodec ac3 > -ab 192k -ac 6 -map 0:0 -map 0:1 -sn -f mpegts > > /mnt/samba/trans_tmp/Serviio/transcoding-temp-2-MPEG2TS.stf > > CPU Usage is 60% per core and no wait on I/O, > swapping. > > Viewing using WDTV Live streaming stops after 31 > secs consistently > Viewing using XMBC on WinXP stream starts buffering > after 7 mins or 3 mins of play, takes about 8 to 10 seconds to catch up > Cabled network 100 MB connection on WinXP the network > usage varies between 7% to 50% > > Q : What can be done to make transcoding on-the-fly continuous ? > > Any and all help is greatly appreciated. > > Thanx > > Charu Palkar > -- > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From ludovic.artru at etu.univ-lyon1.fr Mon May 2 09:14:33 2011 From: ludovic.artru at etu.univ-lyon1.fr (ARTRU LUDOVIC p0802080) Date: Mon, 2 May 2011 09:14:33 +0200 Subject: [FFmpeg-user] Question about the LGPL In-Reply-To: <3F0605D5DEBC204F91753FF87A147DC9022ED5F551@BV-LAME-13.univ-lyon1.fr> References: <3F0605D5DEBC204F91753FF87A147DC9022ED5F551@BV-LAME-13.univ-lyon1.fr> Message-ID: <3F0605D5DEBC204F91753FF87A147DC9022ED5F553@BV-LAME-13.univ-lyon1.fr> Hello I'm a French student doing a work experience placement and I have to program a software using videos. But the videos are in .avi and I have to convert it to .flv That's why I found your software. I download it with Windows XP and it works (I haven't already writte a line of code) But the finality of my work experience placement is commercial software. I have some question about the right of using your software. Can I use the ffmpeg.exe if I only use it to convert videos to .flv? (The software will be write in VB.net). Thanks. ARTRU Ludovic IUT Informatique 2G2 IEM From arnstrb at yahoo.com Mon May 2 14:25:43 2011 From: arnstrb at yahoo.com (Arnie Bearak) Date: Mon, 2 May 2011 05:25:43 -0700 (PDT) Subject: [FFmpeg-user] How to convert a set of pgm files to an rtsp streamable mp4 file In-Reply-To: <4DBD4981.8070301@optonline.net> References: <4DBD4981.8070301@optonline.net> Message-ID: <604548.15849.qm@web121705.mail.ne1.yahoo.com> I am hoping someone can help me with this issue as I have not been able to create an rtsp streamable mp4 file (streaming from darwin server on my linux box to quicktime or vlc on my windows box). I run the ffmpeg command ffmpeg -r 1 -i frame_%05d.pgm -b 18000 -vcodec libx264 -y -threads 0 -vpre slow Arn.mp4 and I can create an mp4 file that when copied to the my windows box can be played with VLC or quicktime. However, if I try to stream the file, I get an indication of unsupported media type. Can anyone tell me the appropriate parameter to add to my command line to get the valid tags inserted into the file so that it can be rtsp streamed? Thank You, Arnie Bearak From grant.smith at envent-tech.com Mon May 2 14:44:49 2011 From: grant.smith at envent-tech.com (Grant Smith) Date: Mon, 2 May 2011 08:44:49 -0400 Subject: [FFmpeg-user] Installation Problems In-Reply-To: <20110501130356.3714c1d2.dashing.meng@gmail.com> References: <20110501130356.3714c1d2.dashing.meng@gmail.com> Message-ID: Interesting.... I don't think my LAME build is failing but I will give this a try and let you know. Thanks, Grant Smith A+, Network+, MCP x 2, BSIT/VC, MIS Phone: +1.317.560.4457 On Sun, May 1, 2011 at 1:03 AM, littlebat wrote: > > On Wed, 27 Apr 2011 22:17:16 -0400 > Grant Smith wrote: > > > I'm trying to install ffmpeg on Ubuntu 10.04 using the tutorial found > > @ http://ubuntuforums.org/showpost.php?p=9868359&postcount=1289 > > I tested tutorial for installing a static linked ffmpeg with the > style of Ubuntu's deb package. > > I found I must execute "sudo mkdir -p /usr/local/share/doc/lame" before > "sudo checkinstall ..." when install LAME, or it will fail. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From cs_palkar at yahoo.com Mon May 2 15:18:20 2011 From: cs_palkar at yahoo.com (Charu Palkar) Date: Mon, 2 May 2011 06:18:20 -0700 (PDT) Subject: [FFmpeg-user] Request for advise on performance - transcoding on the fly In-Reply-To: <4DBE039B.3070306@noknok.net> References: <906776.36347.qm@web46209.mail.sp1.yahoo.com> <717280.70640.qm@web65612.mail.ac4.yahoo.com> <4DBE039B.3070306@noknok.net> Message-ID: <875257.26231.qm@web65609.mail.ac4.yahoo.com> Hi Shawn, But what about the buffering issue I see with XMBC with the same same transcoding. Any explaination for that ? Thanx Charu -- ----- Original Message ---- From: NokNok Developer To: ffmpeg-user at ffmpeg.org Sent: Sun, May 1, 2011 9:06:35 PM Subject: Re: [FFmpeg-user] Request for advise on performance - transcoding on the fly Neither of those streams you are doing are "live transport" streams. For example the WD LIVE stops because when it issues its 1st GET, it also gets the CURRENT size of the file, at that point in time, and then thats all it asks for.? You need to use a live streaming container/player. eg: HLS (Apple M3U8), or FLV with Flash Player, or ASF Streaming, etc.? There are other approaches, as to streaming in chunked http, etc. Shawn On 5/1/2011 7:27 PM, Charu Palkar wrote: > Hi, > > I am new to the forum and would like guidance. > > System : >? ? ? Server : Rackable >? ? ? CPU :? AMD Opteron? (2GHZ) Dual CPU >? ? ? RAM : 2 Gb >? ? ? Disk? : SATA? ( 250GB x 4 disks )? -? 3 disks configured as RAID0? (3Ware) >? ? ? Network :? 1 Gig >? ? ? OS : Slackware - 13.1 > > Command : >? ? ? ? ffmpeg -benchmark -i > /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -threads 4 -target > ntsc-dvd -aspect 16:9 test.mpg > > NOTES : >? ? ? ? Only one CPU has 90% (sys + user),? others 4% (sys + user),? no wait on > I/O >? ? ? ? There is no swapping, initial page faults is 11 and remains constant for > the entire time >? ? ? ? Both files input and output are located on the RAID partition >? ? ? ? The time it takes to transcode is 20% higher than the run-time of the > media file >? ? ? ? Nothing else is running on the server >? ? ? ? ffmpeg - Built with ARCH=x86_64? CPU=opteron > > QUESTIONS : >? ? ? ? Q : What else can I do to speed up the transcoding? to take less time >than > run-time of media file ? >? ? ? ? ? ? ? ? ? + To use for on-the-fly transcode with a media server. > >? ? ? ? Background : Installed media server? 'serviio' and transcoding > on-the-fly-enabled. The following command is invoked. > >? ? ? ? ? ? ? ? ? ? ? ? ? ? ffmpeg -i > /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -y -threads 4 > -vcodec mpeg2video >? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? -sameq -r 23.976 -g 15 -copyts -acodec ac3 > -ab 192k -ac 6 -map 0:0 -map 0:1 -sn -f mpegts > >? /mnt/samba/trans_tmp/Serviio/transcoding-temp-2-MPEG2TS.stf > >? ? ? ? ? ? ? ? ? ? ? ? ? ? CPU Usage is 60% per core and no wait on I/O, > swapping. > >? ? ? ? ? ? ? ? ? ? ? ? ? ? Viewing using WDTV Live? streaming stops after 31 > secs consistently >? ? ? ? ? ? ? ? ? ? ? ? ? ? Viewing using? XMBC on WinXP? stream starts >buffering > after? 7 mins or 3 mins of play, takes about 8 to 10 seconds to catch up >? ? ? ? ? ? ? ? ? ? ? ? ? ? Cabled network 100 MB connection on WinXP the >network > usage varies between? 7% to 50% > >? ? ? ? Q : What can be done to make transcoding on-the-fly continuous ? > > Any and all help is greatly appreciated. > > Thanx > > Charu Palkar > -- > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user _______________________________________________ ffmpeg-user mailing list ffmpeg-user at ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user From ranjiniraguu at gmail.com Mon May 2 16:15:18 2011 From: ranjiniraguu at gmail.com (ranjini raguu) Date: Mon, 2 May 2011 19:45:18 +0530 Subject: [FFmpeg-user] to .mkv conversion fails. Message-ID: Hello im using ffmpeg for conversions... my prob is whn ever i try to convert any video files to .mkv it fails.... this is the parameter im passing for the conversion. -vcodec libx264 -b 1250kb -acodec libmp3lame -ar 44100 -ab 160kb -ac 2 is there any prob with this conversion in default???? thanks for any replies.... From ranjiniraguu at gmail.com Mon May 2 16:15:18 2011 From: ranjiniraguu at gmail.com (ranjini raguu) Date: Mon, 2 May 2011 19:45:18 +0530 Subject: [FFmpeg-user] to .mkv conversion fails. Message-ID: Hello im using ffmpeg for conversions... my prob is whn ever i try to convert any video files to .mkv it fails.... this is the parameter im passing for the conversion. -vcodec libx264 -b 1250kb -acodec libmp3lame -ar 44100 -ab 160kb -ac 2 is there any prob with this conversion in default???? thanks for any replies.... From flora.giannone at gmail.com Mon May 2 16:16:31 2011 From: flora.giannone at gmail.com (flora giannone) Date: Mon, 2 May 2011 16:16:31 +0200 Subject: [FFmpeg-user] h264 encoding/decoding GPU/cuda acceleration Message-ID: Hello I would like to know if exists somehow to run encoding / decoding h264 with GPU/cuda acceleration. Thank you very much flora From rodney.baker at iinet.net.au Mon May 2 17:24:03 2011 From: rodney.baker at iinet.net.au (Rodney Baker) Date: Tue, 3 May 2011 00:54:03 +0930 Subject: [FFmpeg-user] to .mkv conversion fails. In-Reply-To: References: Message-ID: <201105030054.04007.rodney.baker@iinet.net.au> On Mon, 2 May 2011 23:45:18 ranjini raguu wrote: > Hello im using ffmpeg for conversions... my prob is whn ever i try to > convert any video files to .mkv it fails.... this is the parameter im > passing for the conversion. > > -vcodec libx264 -b 1250kb -acodec libmp3lame -ar 44100 -ab 160kb -ac 2 is > there any prob with this conversion in default???? thanks for any > replies.... > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user I use ffmpeg with libx264 for converting m2t to mkc regularly and it works fine. You need to post your full command line and full output from ffmpeg (with whatever error messages you're getting). "It fails" does not help anyone diagnose the problem. My guess, though, is that you're getting an "invalid size or bitrate" message because you're not passing any presets to libx264 (and the ffmpeg default parameters for libx264 are badly broken, apparently). After -vcodec libx264 add -preset [-profile ] [-tune ] The parameters in <> you can find by doing x264 --help (or x264 --fullhelp for more detail). The parameters in [] are optional. You need at least the preset, and you will probably want the profile too. If you need more help than that, like I said, post your full command line along with the full output. -- =================================================== Rodney Baker VK5ZTV rodney.baker at iinet.net.au =================================================== From belcampo at zonnet.nl Mon May 2 18:36:29 2011 From: belcampo at zonnet.nl (belcampo) Date: Mon, 02 May 2011 18:36:29 +0200 Subject: [FFmpeg-user] h264 encoding/decoding GPU/cuda acceleration In-Reply-To: References: Message-ID: <4DBEDD8D.7010905@zonnet.nl> On 05/02/2011 04:16 PM, flora giannone wrote: > Hello > I would like to know if exists somehow to run encoding / decoding h264 with > GPU/cuda acceleration. > Thank you very much If you search the mailing list you'll find answers, it's requested more than once. > flora > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From arnstrb at yahoo.com Mon May 2 18:38:39 2011 From: arnstrb at yahoo.com (Arnie Bearak) Date: Mon, 2 May 2011 09:38:39 -0700 (PDT) Subject: [FFmpeg-user] Can anyone explain the error I am seeing when trying to do rtsp streaming Message-ID: <804313.46930.qm@web121712.mail.ne1.yahoo.com> When I try to run the following command to do rtsp streaming ffmpeg -r 1 -i frame_%05d.pgm -b 18000 -vcodec libx264 -y -vpre slow http://condor1:8090/feed1.ffm I get the following results Input #0, image2, from 'frame_%05d.pgm': Duration: 00:00:24.00, start: 0.000000, bitrate: N/A Stream #0.0: Video: pgm, gray, 3332x332, 1 fps, 1 tbr, 1 tbn, 1 tbc [buffer @ 0x12b11050] w:3332 h:332 pixfmt:gray [scale @ 0x12b11430] w:3332 h:332 fmt:gray -> w:640 h:160 fmt:yuv420p flags:0xa0000004 [buffer @ 0x12b14b20] w:3332 h:332 pixfmt:gray [ffsink @ 0x12b16400] auto-inserting filter 'auto-inserted scaler 0' between the filter '(null)' and the filter 'out' Impossible to convert between the formats supported by the filter 'auto-inserted scaler 0' and the filter 'out' Error opening filters! [flv @ 0x12b16d60] automatic thread number detection not supported by codec, patch welcome Output #0, ffm, to 'http://condor1:8090/feed1.ffm': Stream #0.0: Video: flv, yuv420p, 640x160, q=1-5, 200 kb/s, 1000k tbn, 15 tbc Stream #0.1: Video: [0][0][0][0] / 0x0000, (null), q=0-0, 102400 kb/s, 1000k tbn Stream mapping: Stream #0.0 -> #0.0 Stream #0.0 -> #0.1 Error while opening encoder for output stream #0.0 - maybe incorrect parameters such as bit_rate, rate, width or height *** glibc detected *** /home/champ/deps/ffmpeg-HEAD-60c68c0/ffmpeg: free(): invalid pointer: 0x0000000012b13220 *** From flora.giannone at gmail.com Mon May 2 19:53:20 2011 From: flora.giannone at gmail.com (flora giannone) Date: Mon, 2 May 2011 19:53:20 +0200 Subject: [FFmpeg-user] h264 encoding/decoding GPU/cuda acceleration In-Reply-To: <4DBEDD8D.7010905@zonnet.nl> References: <4DBEDD8D.7010905@zonnet.nl> Message-ID: Hi I tried to search in the archives of the mailing list (user & ffmpeg-devel) but I have not found answers to my question. Can you send me some links? Thanks Flora 2011/5/2 belcampo > On 05/02/2011 04:16 PM, flora giannone wrote: > >> Hello >> I would like to know if exists somehow to run encoding / decoding h264 >> with >> GPU/cuda acceleration. >> Thank you very much >> > If you search the mailing list you'll find answers, it's requested more > than once. > >> flora >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user >> > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > -- FG From news at achim-hofmann.com Mon May 2 20:05:42 2011 From: news at achim-hofmann.com (Achim Hofmann) Date: Mon, 02 May 2011 20:05:42 +0200 Subject: [FFmpeg-user] Copyright issues iFFmpeg In-Reply-To: <4DBDF95B.8040708@thelounge.net> References: <4DBDF95B.8040708@thelounge.net> Message-ID: <4DBEF276.1090504@achim-hofmann.com> On Mon, 02 May 2011 02:22:51 +0200, Reindl Harald wrote: > > > Am 02.05.2011 01:54, schrieb Johannes Frotscher: > >> GPL is shooting itself in its left foot if you ask me, what outrageous stupidity. >> >> GPL is really a virus, as stated in the Wikipedia article of yours: > >> How else can there be more and even more commercialism, and misinformation, miseducation > >> FreeBSD is the only true freedom act that ever happened to all Unices. > >> because its an anarchic state, people wake up. > >> The very source of the definition my friends, and if those programs use numerous >> ever-never-stop-expanding-since-10-years-APPLE-APIs-of-which-iFFmpegGUI-was-using-ffmpeg-selling-it >> which have their own copyrights, and license agreements, then I am in fact extending the functionality >> of the program, > >> thank you no applause please. > >> Intelligence is conceit, wisdom is bliss > > jesus christ what are you trying to tell us after some hours ago > you did not know about the difference GPL/LGPL and quoted a paragraph > of the ffmpeg-legal-page without understanding a word you read there? Harald: Please, don't feed the troll. Johannes: *plonk* From developer at noknok.net Mon May 2 21:14:38 2011 From: developer at noknok.net (NokNok Developer) Date: Mon, 02 May 2011 15:14:38 -0400 Subject: [FFmpeg-user] Request for advise on performance - transcoding on the fly In-Reply-To: <875257.26231.qm@web65609.mail.ac4.yahoo.com> References: <906776.36347.qm@web46209.mail.sp1.yahoo.com> <717280.70640.qm@web65612.mail.ac4.yahoo.com> <4DBE039B.3070306@noknok.net> <875257.26231.qm@web65609.mail.ac4.yahoo.com> Message-ID: <4DBF029E.1080900@noknok.net> Im not all that familiar with XMBC, but its probably not buffering as you would think it to be, but more of each time XMBC goes back to request data, its getting a new filesize (as it is growing), so it needs to reseek to the end of the file to gather timestamp information, then reseek back to where it was, and it may be that its sucking down the whole thing from beginning again. Best would to do an wireshark trap of the exchange and im sure you will see whats going on. again, the issue is that the containers you are using are NOT live streamable. They can be made to be, but in the form you are using it, they are not "natively". XMBC is a normal media player, and it will see the size in the HTTP response, and say OK, so its 500,000 bytes, let me seek to the end, get the timestamp info so I know how long in time it is, then will attempt to seek back, and so on. On 5/2/2011 9:18 AM, Charu Palkar wrote: > Hi Shawn, > > But what about the buffering issue I see with XMBC with the same same > transcoding. > > Any explaination for that ? > > Thanx > > Charu > -- > > > > ----- Original Message ---- > From: NokNok Developer > To: ffmpeg-user at ffmpeg.org > Sent: Sun, May 1, 2011 9:06:35 PM > Subject: Re: [FFmpeg-user] Request for advise on performance - transcoding on > the fly > > Neither of those streams you are doing are "live transport" streams. > > For example the WD LIVE stops because when it issues its 1st GET, it > also gets the CURRENT size of the file, at that point in time, and then > thats all it asks for. You need to use a live streaming container/player. > > eg: HLS (Apple M3U8), or FLV with Flash Player, or ASF Streaming, etc. > There are other approaches, as to streaming in chunked http, etc. > > Shawn > > On 5/1/2011 7:27 PM, Charu Palkar wrote: >> Hi, >> >> I am new to the forum and would like guidance. >> >> System : >> Server : Rackable >> CPU : AMD Opteron (2GHZ) Dual CPU >> RAM : 2 Gb >> Disk : SATA ( 250GB x 4 disks ) - 3 disks configured as RAID0 > (3Ware) >> Network : 1 Gig >> OS : Slackware - 13.1 >> >> Command : >> ffmpeg -benchmark -i >> /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -threads 4 > -target >> ntsc-dvd -aspect 16:9 test.mpg >> >> NOTES : >> Only one CPU has 90% (sys + user), others 4% (sys + user), no wait on >> I/O >> There is no swapping, initial page faults is 11 and remains constant > for >> the entire time >> Both files input and output are located on the RAID partition >> The time it takes to transcode is 20% higher than the run-time of the >> media file >> Nothing else is running on the server >> ffmpeg - Built with ARCH=x86_64 CPU=opteron >> >> QUESTIONS : >> Q : What else can I do to speed up the transcoding to take less time >> than >> run-time of media file ? >> + To use for on-the-fly transcode with a media server. >> >> Background : Installed media server 'serviio' and transcoding >> on-the-fly-enabled. The following command is invoked. >> >> ffmpeg -i >> /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -y -threads 4 >> -vcodec mpeg2video >> -sameq -r 23.976 -g 15 -copyts -acodec > ac3 >> -ab 192k -ac 6 -map 0:0 -map 0:1 -sn -f mpegts >> >> /mnt/samba/trans_tmp/Serviio/transcoding-temp-2-MPEG2TS.stf >> >> CPU Usage is 60% per core and no wait on I/O, >> swapping. >> >> Viewing using WDTV Live streaming stops after 31 >> secs consistently >> Viewing using XMBC on WinXP stream starts >> buffering >> after 7 mins or 3 mins of play, takes about 8 to 10 seconds to catch up >> Cabled network 100 MB connection on WinXP the >> network >> usage varies between 7% to 50% >> >> Q : What can be done to make transcoding on-the-fly continuous ? >> >> Any and all help is greatly appreciated. >> >> Thanx >> >> Charu Palkar >> -- >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From cs_palkar at yahoo.com Mon May 2 22:46:50 2011 From: cs_palkar at yahoo.com (Charu Palkar) Date: Mon, 2 May 2011 13:46:50 -0700 (PDT) Subject: [FFmpeg-user] Request for advise on performance - transcoding on the fly In-Reply-To: <4DBF029E.1080900@noknok.net> References: <906776.36347.qm@web46209.mail.sp1.yahoo.com> <717280.70640.qm@web65612.mail.ac4.yahoo.com> <4DBE039B.3070306@noknok.net> <875257.26231.qm@web65609.mail.ac4.yahoo.com> <4DBF029E.1080900@noknok.net> Message-ID: <804529.28914.qm@web65612.mail.ac4.yahoo.com> Hi Shawn, I am new to multi-media streaming and need to gain more knowledge. Any suggested reading material ? How does one convert to 'live streaming' ? I am considering using MediaTomb as media server does it do live streaming. Thanx in advance Charu -- ----- Original Message ---- From: NokNok Developer To: ffmpeg-user at ffmpeg.org Sent: Mon, May 2, 2011 3:14:38 PM Subject: Re: [FFmpeg-user] Request for advise on performance - transcoding on the fly Im not all that familiar with XMBC, but its probably not buffering as you would think it to be, but more of each time XMBC goes back to request data, its getting a new filesize (as it is growing), so it needs to reseek to the end of the file to gather timestamp information, then reseek back to where it was, and it may be that its sucking down the whole thing from beginning again. Best would to do an wireshark trap of the exchange and im sure you will see whats going on. again, the issue is that the containers you are using are NOT live streamable. They can be made to be, but in the form you are using it, they are not "natively". XMBC is a normal media player, and it will see the size in the HTTP response, and say OK, so its 500,000 bytes, let me seek to the end, get the timestamp info so I know how long in time it is, then will attempt to seek back, and so on. On 5/2/2011 9:18 AM, Charu Palkar wrote: > Hi Shawn, > > But what about the buffering issue I see with XMBC with the same same > transcoding. > > Any explaination for that ? > > Thanx > > Charu > -- > > > > ----- Original Message ---- > From: NokNok Developer > To: ffmpeg-user at ffmpeg.org > Sent: Sun, May 1, 2011 9:06:35 PM > Subject: Re: [FFmpeg-user] Request for advise on performance - transcoding on > the fly > > Neither of those streams you are doing are "live transport" streams. > > For example the WD LIVE stops because when it issues its 1st GET, it > also gets the CURRENT size of the file, at that point in time, and then > thats all it asks for. You need to use a live streaming container/player. > > eg: HLS (Apple M3U8), or FLV with Flash Player, or ASF Streaming, etc. > There are other approaches, as to streaming in chunked http, etc. > > Shawn > > On 5/1/2011 7:27 PM, Charu Palkar wrote: >> Hi, >> >> I am new to the forum and would like guidance. >> >> System : >> Server : Rackable >> CPU : AMD Opteron (2GHZ) Dual CPU >> RAM : 2 Gb >> Disk : SATA ( 250GB x 4 disks ) - 3 disks configured as RAID0 > (3Ware) >> Network : 1 Gig >> OS : Slackware - 13.1 >> >> Command : >> ffmpeg -benchmark -i >> /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -threads 4 > -target >> ntsc-dvd -aspect 16:9 test.mpg >> >> NOTES : >> Only one CPU has 90% (sys + user), others 4% (sys + user), no wait >>on >> I/O >> There is no swapping, initial page faults is 11 and remains constant > for >> the entire time >> Both files input and output are located on the RAID partition >> The time it takes to transcode is 20% higher than the run-time of the >> media file >> Nothing else is running on the server >> ffmpeg - Built with ARCH=x86_64 CPU=opteron >> >> QUESTIONS : >> Q : What else can I do to speed up the transcoding to take less time >> than >> run-time of media file ? >> + To use for on-the-fly transcode with a media server. >> >> Background : Installed media server 'serviio' and transcoding >> on-the-fly-enabled. The following command is invoked. >> >> ffmpeg -i >> /mnt/samba/share/test/Baraka.1992.BRrip.H264.AAC.ITS-ALI.mp4 -y -threads 4 >> -vcodec mpeg2video >> -sameq -r 23.976 -g 15 -copyts -acodec > ac3 >> -ab 192k -ac 6 -map 0:0 -map 0:1 -sn -f mpegts >> >> /mnt/samba/trans_tmp/Serviio/transcoding-temp-2-MPEG2TS.stf >> >> CPU Usage is 60% per core and no wait on I/O, >> swapping. >> >> Viewing using WDTV Live streaming stops after 31 >> secs consistently >> Viewing using XMBC on WinXP stream starts >> buffering >> after 7 mins or 3 mins of play, takes about 8 to 10 seconds to catch up >> Cabled network 100 MB connection on WinXP the >> network >> usage varies between 7% to 50% >> >> Q : What can be done to make transcoding on-the-fly continuous ? >> >> Any and all help is greatly appreciated. >> >> Thanx >> >> Charu Palkar >> -- >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user _______________________________________________ ffmpeg-user mailing list ffmpeg-user at ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user From belcampo at zonnet.nl Mon May 2 23:01:04 2011 From: belcampo at zonnet.nl (belcampo) Date: Mon, 02 May 2011 23:01:04 +0200 Subject: [FFmpeg-user] h264 encoding/decoding GPU/cuda acceleration In-Reply-To: References: <4DBEDD8D.7010905@zonnet.nl> Message-ID: <4DBF1B90.1040404@zonnet.nl> On 05/02/11 19:53, flora giannone wrote: > Hi > I tried to search in the archives of the mailing list (user& ffmpeg-devel) > but > I have not found answers to my question. Can you send me some links? Search the list for CrystalHD and/or Using video card chipset to encode Januari 15th and February 12th-14th 2010 > Thanks > Flora > > 2011/5/2 belcampo > >> On 05/02/2011 04:16 PM, flora giannone wrote: >> >>> Hello >>> I would like to know if exists somehow to run encoding / decoding h264 >>> with >>> GPU/cuda acceleration. >>> Thank you very much >>> >> If you search the mailing list you'll find answers, it's requested more >> than once. >> >>> flora >>> _______________________________________________ >>> ffmpeg-user mailing list >>> ffmpeg-user at ffmpeg.org >>> http://ffmpeg.org/mailman/listinfo/ffmpeg-user >>> >> >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user >> > > > From mahakcay at gmail.com Tue May 3 02:11:28 2011 From: mahakcay at gmail.com (Mahmut Akcay) Date: Tue, 3 May 2011 01:11:28 +0100 Subject: [FFmpeg-user] Using RTP format for H264 streaming Message-ID: I'm having problem to re-stream H264 with RTP format. Below is the command line params and output. ./ffmpeg -i rtmp://flashmedia.nic.in:80/live/dd1 -vcodec copy -an -f rtp rtp:localhost:9000 ffmpeg version git-N-29571-g7d70d19, Copyright (c) 2000-2011 the FFmpeg developers built on May 3 2011 00:51:01 with gcc 4.2.1 (Apple Inc. build 5664) configuration: --enable-libfaac --enable-libmp3lame --enable-libx264 --enable-nonfree --enable-gpl libavutil 51. 2. 0 / 51. 2. 0 libavcodec 53. 3. 0 / 53. 3. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 4. 0 / 2. 4. 0 libswscale 0. 14. 0 / 0. 14. 0 [flv @ 0x101012400] Estimating duration from bitrate, this may be inaccurate Input #0, flv, from 'rtmp://flashmedia.nic.in:80/live/dd1': Duration: N/A, start: 0.000000, bitrate: 64 kb/s Stream #0.0: Video: h264 (Baseline), yuv420p, 400x300 [PAR 1:1 DAR 4:3], 24 tbr, 1k tbn, 48 tbc Stream #0.1: Audio: mp3, 22050 Hz, stereo, s16, 64 kb/s Output #0, rtp, to 'rtp:localhost:9000': Metadata: encoder : Lavf53.0.3 Stream #0.0: Video: libx264, yuv420p, 400x300 [PAR 1:1 DAR 4:3], q=2-31, 90k tbn, 24 tbc Stream mapping: Stream #0.0 -> #0.0 SDP: v=0 o=- 0 0 IN IP6 ::1 s=No Name c=IN IP6 ::1 t=0 0 a=tool:libavformat 53.0.3 m=video 9000 RTP/AVP 96 a=rtpmap:96 H264/90000 a=fmtp:96 packetization-mode=1; sprop-parameter-sets=Z0KAH5ZSAyE/3gKhAAADAAEAAAMAMOBgAVXAABGMP8Y4wMACq4AAIxh/jHDtChUk,aMuNSA== Press [q] to stop encoding frame= 1540 fps= 27 q=-1.0 size= 6kB time=59.14 bitrate= 0.8kbits/s When I play the sdp file printed above I've got following: ./ffplay ./test.sdp -debug 5 ffplay version git-N-29571-g7d70d19, Copyright (c) 2003-2011 the FFmpeg developers built on May 3 2011 00:51:01 with gcc 4.2.1 (Apple Inc. build 5664) configuration: --enable-libfaac --enable-libmp3lame --enable-libx264 --enable-nonfree --enable-gpl libavutil 51. 2. 0 / 51. 2. 0 libavcodec 53. 3. 0 / 53. 3. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 4. 0 / 2. 4. 0 libswscale 0. 14. 0 / 0. 14. 0 [NULL @ 0x10203b200] Format sdp probed with size=2048 and score=50 [sdp @ 0x10203b200] video codec set to: h264 [NULL @ 0x101863e00] RTP Packetization Mode: 1 [NULL @ 0x101863e00] Extradata set to 0x10102af80 (size: 58)!Unsupported bit depth: 0 [sdp @ 0x10203b200] Could not find codec parameters (Video: h264) [sdp @ 0x10203b200] Estimating duration from bitrate, this may be inaccurate ./test.sdp: could not find codec parameters I've tried to transcode the stream to mpeg4 and it's working, but with h264 it doesn't work. Am I missing anything? Thanks in advance mahmut From pawel1987 at gmail.com Mon May 2 13:53:44 2011 From: pawel1987 at gmail.com (pawel) Date: Mon, 2 May 2011 04:53:44 -0700 (PDT) Subject: [FFmpeg-user] DV100 encoded MXF file? In-Reply-To: <1304337224049-942321.post@n4.nabble.com> References: <6c7493650812291832m11899d25t8d874d9997cd1fca@mail.gmail.com> <1304337224049-942321.post@n4.nabble.com> Message-ID: <1304337223955-3489835.post@n4.nabble.com> Was this issue fixed? I have an essence in DVCPRO HD and need to wrap it to MXF. Is this possible with ffmpeg? -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/DV100-encoded-MXF-file-tp942319p3489835.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From dujun at qvod.com Tue May 3 11:21:47 2011 From: dujun at qvod.com (John Du) Date: Tue, 3 May 2011 17:21:47 +0800 Subject: [FFmpeg-user] No ffplay generated on ffmpeg 0.6.90-ro Message-ID: Dear FFmpeg team I am compling ffmpeg-0.6.90-rc0, and meeting the configure error on fedora os. But if i use ffmpeg 0.6.1, no such problem. #configure yasm not found, use --disable-yasm for a crippled build If you think configure made a mistake, make sure you are using the latest version from Git. If the latest version fails, report the problem to the ffmpeg-user at ffmpeg.org mailing list or IRC #ffmpeg on irc.freenode.net. Include the log file "config.log" produced by configure as this will help solving the problem. Pls refer to the config.log Even I disable-yasm to compling . but the ffplay is not generated finally. How can I build and generate the ffplay? Regards, John Du Shenzhen QVOD technology Co. Ltd. 22A Building No.3, China Academy of Science & Technology Development, High-tech. South Street No.1, South District of High-tech Industrial Area, Shenzhen City. Tel: 86 3363 2699 Ext. 8110 Fax: 86 3363 2696 Cell: 86 150 1381 9009 E-mail: dujun at qvod.com Website: http://www.qvod.com -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: config.log URL: From stefano.sabatini-lala at poste.it Tue May 3 13:31:41 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Tue, 3 May 2011 13:31:41 +0200 Subject: [FFmpeg-user] No ffplay generated on ffmpeg 0.6.90-ro In-Reply-To: References: Message-ID: <20110503113141.GA22860@geppetto> On date Tuesday 2011-05-03 17:21:47 +0800, John Du encoded: > > Dear FFmpeg team > > I am compling ffmpeg-0.6.90-rc0, and meeting the configure error on fedora os. But if i use ffmpeg 0.6.1, no such problem. > > #configure > yasm not found, use --disable-yasm for a crippled build > > If you think configure made a mistake, make sure you are using the latest > version from Git. If the latest version fails, report the problem to the > ffmpeg-user at ffmpeg.org mailing list or IRC #ffmpeg on irc.freenode.net. > Include the log file "config.log" produced by configure as this will help > solving the problem. > > Pls refer to the config.log > > Even I disable-yasm to compling . but the ffplay is not generated finally. > > > > How can I build and generate the ffplay? Make sure that you have libsdl headers and libraries installed on your system before to run configure. From poliva at wtelecom.es Tue May 3 14:50:25 2011 From: poliva at wtelecom.es (Pedro Oliva) Date: Tue, 3 May 2011 14:50:25 +0200 Subject: [FFmpeg-user] RTSP keep-alive issue Message-ID: Hi, I'm recording an IP camera RTSP stream. My app should record the stream indefinitely. Sometimes, ffmpeg stops recording after five or six minutes. I've found out that the IP camera rtsp server doesn't support GET_PARAMETER method (server returns 405 response), which is used to test client liveness. Perhaps this could be the reason of this behavior, but I don't understand why ffmpeg stops recording only in a few cases. Could anyone give me an explanation? From arnstrb at yahoo.com Tue May 3 17:11:30 2011 From: arnstrb at yahoo.com (Arnie Bearak) Date: Tue, 3 May 2011 08:11:30 -0700 (PDT) Subject: [FFmpeg-user] ffmpeg/ffserver run causes ffmpeg crash In-Reply-To: References: Message-ID: <534785.79355.qm@web121719.mail.ne1.yahoo.com> I am trying to encode a set of pgm files and stream them out to ffserver using the following command listed below. I start ffserver, and when I run the ffmpeg command, it starts to do the encoding but eventually takes a segmentation fault. It is producing an ffmpeg warning that I don't understand [libx264 @ 0xad2ffc0] VBV buffer (655360) > level limit (300000) ffmpeg -r 1 -i frame_%05d.pgm -vcodec libx264 -y -threads 0 -b 230k -vpre slow http://condor1:8090/feed1.ffm The output of the ffmpeg run is Input #0, image2, from 'frame_%05d.pgm': Duration: 00:00:24.00, start: 0.000000, bitrate: N/A Stream #0.0: Video: pgm, gray, 3332x332, 1 fps, 1 tbr, 1 tbn, 1 tbc [buffer @ 0xad2c0f0] w:3332 h:332 pixfmt:gray [scale @ 0xad2c4d0] w:3332 h:332 fmt:gray -> w:3344 h:336 fmt:yuv420p flags:0xa0000004 [libx264 @ 0xad2ffc0] VBV buffer (655360) > level limit (300000) [libx264 @ 0xad2ffc0] using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.1 Cache64 [libx264 @ 0xad2ffc0] profile High, level 5.1 Output #0, ffm, to 'http://condor1:8090/feed1.ffm': Metadata: encoder : Lavf52.103.0 Stream #0.0: Video: libx264, yuv420p, 3344x336, q=3-69, 18 kb/s, 1000k tbn, 1 tbc Stream mapping: Stream #0.0 -> #0.0 And the bt for the core dump is 0x0000003aac27275e in free () from /lib64/libc.so.6 #1 0x0000003fcf6a08b5 in avformat_free_context (s=0xad25cd0) at libavformat/utils.c:2599 #2 0x000000000040511c in av_bitstream_filter_filter () #3 0x000000000040aac8 in av_bitstream_filter_filter () #4 0x0000003aac21d994 in __libc_start_main () from /lib64/libc.so.6 #5 0x0000000000404a09 in av_bitstream_filter_filter () #6 0x00007fff392c3f18 in ?? () #7 0x0000000000000000 in ?? () My ffserver.config file looks as follows: (there are lots of things commented out here) # Port on which the server is listening. You must select a different # port from your standard HTTP web server if it is running on the same # computer. Port 8090 RTSPPort 5454 # Address on which the server is bound. Only useful if you have # several network interfaces. BindAddress 0.0.0.0 # Number of simultaneous HTTP connections that can be handled. It has # to be defined *before* the MaxClients parameter, since it defines the # MaxClients maximum limit. MaxHTTPConnections 2000 # Number of simultaneous requests that can be handled. Since FFServer # is very fast, it is more likely that you will want to leave this high # and use MaxBandwidth, below. MaxClients 1000 # This the maximum amount of kbit/sec that you are prepared to # consume when streaming to clients. MaxBandwidth 1000000 # Access log file (uses standard Apache log file format) # '-' is the standard output. CustomLog - # Suppress that if you want to launch ffserver as a daemon. NoDaemon File /tmp/feed1.ffm FileMaxSize 5M # the source feed Feed feed1.ffm # the output stream format - ASF #Format asf Format mp4 #VideoCodec msmpeg4 VideoCodec libx264 AVPresetVideo slow # this must match the ffmpeg -r argument VideoFrameRate 1 # generally leave this is a large number VideoBufferSize 80000 # another quality tweak #VideoBitRate 200 VideoBitRate 18 # quality ranges - 1-31 (1 = best, 31 = worst) #VideoQMin 1 #VideoQMax 5 #VideoQMin 0 #VideoQMax 69 #VideoSize 352x288 #VideoSize 3332x332 VideoSize 3344x336 # This sets how many seconds in past to start PreRoll 0 #we have no audio Noaudio Format status # Only allow local people to get the status #FaviconURL http://pond1.gladstonefamily.net:8080/favicon.ico If anyone has ideas on why this is happening, I would appreciate some advice. Thanks, Arnie From jshupert at pps-inc.com Tue May 3 17:11:55 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Tue, 03 May 2011 11:11:55 -0400 Subject: [FFmpeg-user] jpeg2000 mxf Message-ID: <4DC01B3B.4040408@pps-inc.com> Q: Can ffmpeg play and or transcode to and or from an mxf with jpeg2000 video & pcm audio. by means of my googling i think i know the answer to my Q. I think it cannot - that support for jpeg2000 is anUnfinished 2007 summer of code item. I thought i would just ask here , so that i can feel like i have 'tried everything' thanks From dave at avpreserve.com Tue May 3 17:23:02 2011 From: dave at avpreserve.com (Dave Rice) Date: Tue, 3 May 2011 11:23:02 -0400 Subject: [FFmpeg-user] jpeg2000 mxf In-Reply-To: <4DC01B3B.4040408@pps-inc.com> References: <4DC01B3B.4040408@pps-inc.com> Message-ID: <9FB72DD6-4B13-4912-937C-C49379CEB1EB@avpreserve.com> Hi Jim, Misty De Meo at Canadian Museum for Human Rights did some initial research including ffmpeg as a component here: http://groups.google.com/group/archivematica/browse_thread/thread/267a008ac423b820?pli=1. Also work on jpeg2000 in ffmpeg is a current SOC task as well. Dave Rice avpreserve.com On May 3, 2011, at 11:11 AM, Jim Shupert wrote: > > Q: Can ffmpeg play and or transcode to and or from an mxf with jpeg2000 video & pcm audio. > > by means of my googling i think i know the answer to my Q. > I think it cannot - that support for jpeg2000 is anUnfinished 2007 summer of code item. > > I thought i would just ask here , so that i can feel like i have 'tried everything' > > thanks > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From mike.scheutzow at alcatel-lucent.com Tue May 3 20:59:49 2011 From: mike.scheutzow at alcatel-lucent.com (Mike Scheutzow) Date: Tue, 03 May 2011 14:59:49 -0400 Subject: [FFmpeg-user] How to convert a set of pgm files to an rtsp streamable mp4 file In-Reply-To: <604548.15849.qm@web121705.mail.ne1.yahoo.com> References: <4DBD4981.8070301@optonline.net> <604548.15849.qm@web121705.mail.ne1.yahoo.com> Message-ID: <4DC050A5.2000806@alcatel-lucent.com> Arnie Bearak wrote: > I am hoping someone can help me with this issue as I have not been able to > create an rtsp streamable mp4 file (streaming from darwin server on my linux box > to quicktime or vlc on my windows box). > > I run the ffmpeg command > > ffmpeg -r 1 -i frame_%05d.pgm -b 18000 -vcodec libx264 -y -threads 0 -vpre slow > Arn.mp4 > and I can create an mp4 file that when copied to the my windows box can be > played with VLC or quicktime. > However, if I try to stream the file, I get an indication of unsupported media > type. > > Can anyone tell me the appropriate parameter to add to my command line to get > the valid tags inserted into the file so that it can be rtsp streamed? Darwin requires that additional metadata tracks be present in the .mp4 file; they call this procedure "hinting". You can not do this with ffmpeg, you have to obtain a separate utility. I use mp4box to add hints for Darwin. Example: $ mp4box -hint Arn.mp4 Mike Scheutzow From leo.izen at gmail.com Tue May 3 21:28:55 2011 From: leo.izen at gmail.com (Leo Izen) Date: Tue, 3 May 2011 15:28:55 -0400 Subject: [FFmpeg-user] jpeg2000 mxf In-Reply-To: <9FB72DD6-4B13-4912-937C-C49379CEB1EB@avpreserve.com> References: <4DC01B3B.4040408@pps-inc.com> <9FB72DD6-4B13-4912-937C-C49379CEB1EB@avpreserve.com> Message-ID: On May 3, 2011, at 11:11 AM, Jim Shupert wrote: > > Q: Can ffmpeg play and or transcode to and or from an mxf with jpeg2000 > video & pcm audio? > There's FFmpeg does have an mxf demuxer and a pcm decoder. But if you want a jpeg2000 decoder, you need to compile with --enable-libopenjpeg, which can be obtained from this command: svn checkout http://openjpeg.googlecode.com/svn/trunk/ openjpeg/ and built like any other standard open source library (i.e. configure and make). From arnstrb at yahoo.com Tue May 3 22:06:18 2011 From: arnstrb at yahoo.com (Arnie Bearak) Date: Tue, 3 May 2011 13:06:18 -0700 (PDT) Subject: [FFmpeg-user] How to convert a set of pgm files to an rtsp streamable mp4 file In-Reply-To: <4DC050A5.2000806@alcatel-lucent.com> References: <4DBD4981.8070301@optonline.net> <604548.15849.qm@web121705.mail.ne1.yahoo.com> <4DC050A5.2000806@alcatel-lucent.com> Message-ID: <517137.63378.qm@web121708.mail.ne1.yahoo.com> Thanks Mike. Did you run mp4box on a linux box or a windows box? Arnie ________________________________ From: Mike Scheutzow To: FFmpeg user questions and RTFMs Sent: Tue, May 3, 2011 2:59:49 PM Subject: Re: [FFmpeg-user] How to convert a set of pgm files to an rtsp streamable mp4 file Arnie Bearak wrote: > I am hoping someone can help me with this issue as I have not been able to >create an rtsp streamable mp4 file (streaming from darwin server on my linux box >to quicktime or vlc on my windows box). > > I run the ffmpeg command > ffmpeg -r 1 -i frame_%05d.pgm -b 18000 -vcodec libx264 -y -threads 0 -vpre >slow Arn.mp4 > and I can create an mp4 file that when copied to the my windows box can be >played with VLC or quicktime. > However, if I try to stream the file, I get an indication of unsupported >media type. > > Can anyone tell me the appropriate parameter to add to my command line to get >the valid tags inserted into the file so that it can be rtsp streamed? Darwin requires that additional metadata tracks be present in the .mp4 file; they call this procedure "hinting". You can not do this with ffmpeg, you have to obtain a separate utility. I use mp4box to add hints for Darwin. Example: $ mp4box -hint Arn.mp4 Mike Scheutzow _______________________________________________ ffmpeg-user mailing list ffmpeg-user at ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user From dave at avpreserve.com Tue May 3 22:25:01 2011 From: dave at avpreserve.com (Dave Rice) Date: Tue, 3 May 2011 16:25:01 -0400 Subject: [FFmpeg-user] jpeg2000 mxf In-Reply-To: References: <4DC01B3B.4040408@pps-inc.com> <9FB72DD6-4B13-4912-937C-C49379CEB1EB@avpreserve.com> Message-ID: On May 3, 2011, at 3:28 PM, Leo Izen wrote: > On May 3, 2011, at 11:11 AM, Jim Shupert wrote: >> >> Q: Can ffmpeg play and or transcode to and or from an mxf with jpeg2000 >> video & pcm audio? >> > > There's FFmpeg does have an mxf demuxer and a pcm decoder. But if you want a > jpeg2000 decoder, you need to compile with --enable-libopenjpeg, which can > be obtained from this command: > > svn checkout http://openjpeg.googlecode.com/svn/trunk/ openjpeg/ > > and built like any other standard open source library (i.e. configure and > make). There is a ticket here related to using libopenjpeg with jpeg2000 in MXF: https://roundup.libav.org/issue2309. Decode of jpeg2000 with MXF using libopenjpeg gives an error that only the first component is used. SMPTE 422M defines how the jpeg2000 color data is stored in MXF and I don't think the MXF demuxer incorporates this. If it helps gstreamer can demux the jpeg2000 via mxfjpeg2000.c. Dave Rice avpreserve.com From baptiste.coudurier at gmail.com Wed May 4 01:43:28 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Tue, 03 May 2011 16:43:28 -0700 Subject: [FFmpeg-user] DV100 encoded MXF file? In-Reply-To: <1304337223955-3489835.post@n4.nabble.com> References: <6c7493650812291832m11899d25t8d874d9997cd1fca@mail.gmail.com> <1304337224049-942321.post@n4.nabble.com> <1304337223955-3489835.post@n4.nabble.com> Message-ID: <4DC09320.5000301@gmail.com> Hi, On 05/02/2011 04:53 AM, pawel wrote: > Was this issue fixed? I have an essence in DVCPRO HD and need to wrap it to > MXF. Is this possible with ffmpeg? It is not currently possible. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From mroper at kinect.co.nz Wed May 4 12:39:25 2011 From: mroper at kinect.co.nz (mroper) Date: Wed, 04 May 2011 22:39:25 +1200 Subject: [FFmpeg-user] trying to work out ffmpeg and avi conversion Message-ID: <4DC12CDD.2030806@kinect.co.nz> hi, hoping someone can help, i'm fairly new to ffmpeg, trying to work out how to convert an HD 1280x720 AVI file to mpeg2 so my tv via linux media tomb can play it. When i try doing.... ffmpeg -i "$filename" -target pal-dvd -b 25000k /destpath/$filename i get .....Error while opening encoder for output stream #0.0- maybe incorrect parameters such as bit_rate, rate, width or height i've tried searching on the error and drawn a blank, lots of information but nothing I can relate to it. likewise the man pages don't give me a clue. reason why I have the -b 25000k is I'm trying to maintain the quality of the orginal avi, if I don't set this it works fine. likewise if I set a lower rate it also works fine (but at lower quality). The orginal avi when doing a ffmpeg -i filename shows up Stream #0.0: Video: mjpeg, yuvj422p, 1280x720, 30 tbr, 30 tbn, 30 tbc Metadata: strn : FUJIFILM AVI STREAM 0100 Stream #0.1: Audio: pcm_s16le, 44100 Hz, 2 channels, s16, 1411 kb/s Where as the best I can get is via ffmpeg -i "$filename" -target pal-dvd -s hd720 /destpath/$filename and produces Duration: 00:00:02.00, start: 0.500000, bitrate: 7176 kb/s Stream #0.0[0x1e0]: Video: mpeg2video, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 9000 kb/s, 25 fps, 25 tbr, 90k tbn, 50 tbc Stream #0.1[0x80]: Audio: ac3, 48000 Hz, stereo, s16, 448 kb/s Which has a much lower quality/bitrate (you can see the difference on screen). Any ideas? Thanks Miles From koxaniy at mail.ru Wed May 4 20:08:24 2011 From: koxaniy at mail.ru (Tuuls) Date: Wed, 4 May 2011 11:08:24 -0700 (PDT) Subject: [FFmpeg-user] Problem with FFmpeg git-a304071 In-Reply-To: References: <4DB88A0F.8030509@ffmpeg.org> <4DB8AF18.8000100@ffmpeg.org> Message-ID: <1304532504488-3496362.post@n4.nabble.com> Hi Baptiste ! problem with encoding video just me, or are still in the source code? -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/Problem-with-FFmpeg-git-a304071-tp3479411p3496362.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From hardik.sharma22 at yahoo.com Wed May 4 21:33:16 2011 From: hardik.sharma22 at yahoo.com (Hardik Sharma) Date: Wed, 4 May 2011 12:33:16 -0700 (PDT) Subject: [FFmpeg-user] Fw: Unable to get correct decoded file Message-ID: <478332.47926.qm@web46206.mail.sp1.yahoo.com> Hi guys, Even after using following suggestions my decoded output file is still not correct. Decoded yuv video is with very slow frame rate as compared to original and with gap in between every frame. Please let me know if I am missing anything or any parameter in my command line. For encoding my command line is- ffmpeg -f rawvideo -r 30 -b 256k -s 352x288 -i silent_cif.yuv -vcodec libx264 -b 256k -s 352x288 -preset slow -f h264 -threads 0 silent264.264? For decoding- ffmpeg -i in.h264 -vcodec rawvideo -f yuv4mpegpipe out.y4m Thanks, Hardik Sharma >Command line for decoding- >>ffmpeg -r 30 -b 256k -s 352x288 -i output.h264 -vcodec libx264 -b 256k >-s 352x288 -preset slow -f rawvideo -threads 0 slient.yuv > You should use -vcodec rawvideo for decoding. I also suggest this: Your full command line for decoding should be this: ffmpeg -i in.h264 -vcodec rawvideo -f yuv4mpegpipe out.y4m I have removed most unnecessary options, but left some to show a point. Generally you should store rawvideo that is in some yuv pixel format in the yuv4mpegpipe (y4m) container, as it is simply a wrapper for the video, but has data like size and frame rate, so the user of the program reading it doesn't have to specify this data. Note: this only stores these pixel formats: yuv444p, yuv422p, yuv420p, yuv411p and gray. y4m files also can't contain audio, or any non-raw video. But it is very useful for video. ----- Forwarded Message ----- From: Hardik Sharma To: ffmpeg Cc: "leo.izen at gmail.com" Sent: Sunday, 1 May 2011 1:33 PM Subject: Fw: Unable to get correct decoded file Thanks Leo but still I am not getting desired output. My decoded video output is with very slow frame rate (even after mentioning frame rate as 30fps) and there is some problem in video frames as per appearance concern. It's different from the input yuv video even without much loss. Please tell me if I should include more format option while encoding or decoding video sequence. I really appreciate your help. Thanks. Regards, Hardik Sharma ? ? --- On Sat, 30/4/11, Hardik Sharma wrote: >From: Hardik Sharma >Subject: Unable to get correct decoded file >To: "ffmpeg" >Date: Saturday, 30 April, 2011, 12:16 AM > > >Hi, > >I am a new bee with Ubuntu and ffmpeg both so please help me with this issue. I encoded the yuv video to h264 by following command-? >?ffmpeg -f rawvideo -r 30 -b 256k -s 352x288 -i silent_cif.yuv -vcodec libx264 -b 256k -s 352x288 -preset slow -f h264 -threads 0 silent264.h264 > >But I don't think I am getting correct encoded file as after decoding back to yuv video, I am only getting 2 frames and that too wrong frames. > >Command line for decoding- >?ffmpeg -r 30 -b 256k -s 352x288 -i output.h264 -vcodec libx264 -b 256k -s 352x288 -preset slow -f rawvideo -threads 0 slient.yuv > >Please tell me my mistake in command line. I really appreciate your help. Thanks. > >Regards, >Hardik Sharma? > > From sergey.forum at gmail.com Thu May 5 00:50:50 2011 From: sergey.forum at gmail.com (Sergey Kurdakov) Date: Thu, 5 May 2011 02:50:50 +0400 Subject: [FFmpeg-user] Force first frame to be full, non blurred I frame Message-ID: Hi, I develop silverlight application which can play h264 files which I produce with ffmpeg from recorded avi s, still, due to used underlying MediaElement ( which I would prefer not to force to move to some better pos for still image ) the first (still) frame shows blurry, I found that it is possible to set force_key_frames and I do following ffmpeg -i input.avi -an -vcodec libx264 -vpre veryslow -r 30 -b 400k -force_key_frames 0 -threads 0 OUTPUT.mp4 (actually two passes ffmpeg -i input.avi -an -pass 1 -vcodec libx264 -vpre veryslow_firstpass -r 30 -b 400k -force_key_frames 0 -threads 0 OUTPUT.mp4 ffmpeg -i input.avi -an -pass 2 -vcodec libx264 -vpre veryslow -r 30 -b 400k -force_key_frames 0 -threads 0 OUTPUT.mp4 ) but still I get blurry first frame (-force_key_frames -1 also does not give needed result ) , so the question is - which number (time) to set to -force_key_frames ( or set some other param ) so that first frame ( or first frames ) was non blurry I-frame ( otherwise video quality is excellent - but exactly the first frame is really blurry. Some hints - opening resulting video in avidemux shows that the first frame in sequence is B frame then followed by P frame and then 3rd frame is my first blurry I frame - though avidemux might be wrong, still it could indicate a problem that B frame affects what I get as resulting I frame). Regards Sergey From bdogs4ever at gmail.com Thu May 5 17:15:50 2011 From: bdogs4ever at gmail.com (Jay Lucke) Date: Thu, 5 May 2011 08:15:50 -0700 Subject: [FFmpeg-user] ffmpeg and x264 In-Reply-To: References: <61386.97286.qm@web86408.mail.ird.yahoo.com> <237821.66460.qm@web86407.mail.ird.yahoo.com> Message-ID: I have another question related to ffmpeg/x264 This is from my programmer this morning: "The main issue is that I can get really great quality movie which is larger and it seems not it's not working in all platforms (Mac, iPad, ...) or the quality is really terrible." Sound familiar? Jay ----------------- Jay Lucke On Apr 30, 2011, at 11:57 PM, Robert Kr?ger wrote: > > On Apr 30, 2011, at 17:39 , Dave Pope wrote: > >> Thanks! The change makes perfect sense, I just had to dig around for while before figuring it out. Do you know if there's a patch in the works to make 'ffmpeg --help' report something about these? It seems like it might require building some "handshaking" between them that might not already exist in general form. >> >> ________________________________ >> >> From: ffmpeg-user-bounces at ffmpeg.org on behalf of Robert Kr?ger >> Sent: Sat 4/30/2011 4:04 AM >> To: FFmpeg user questions and RTFMs >> Subject: Re: [FFmpeg-user] ffmpeg and x264 >> >> >> >> >> On Apr 29, 2011, at 17:37 , Dave Pope wrote: >> >>> Yeah, I hit this one too. Use "-preset" and "-profile" switches instead >>> of -vpre. "-preset fast -profile baseline" works OK for me. Annoying >>> that it's not reflected in the --help output; I don't see a way to list >>> the available values yet either. >>> >> >> Check x264 docs. As far as I understood the purpose of the change was not to duplicate efforts from the x264 project and that is documented quite well. >> >> try x264 --fullhelp >> > > No, I'm not aware of anything like that. > > And BTW, please don't top-post (http://en.wikipedia.org/wiki/Posting_style). It is a rule the list has given itself to make following discussion threads easier. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From yangtzj at hotmail.com Thu May 5 17:27:23 2011 From: yangtzj at hotmail.com (JiangJie) Date: Thu, 5 May 2011 23:27:23 +0800 Subject: [FFmpeg-user] Has the ffserver's "av loose sync" problem been fixed? Message-ID: Hi all, I'm considering building a stream media server based on ffmpeg + ffserver. The video will come from a v4l2 device and the audio comes from an alsa device, i.e, /dev/dsp. I'd like to stream both the audio and video simultaneously. However, from http://www.ffmpeg.org/ffserver.html, it says that "The audio and video loose sync after a while". Has this problem been fixed in the latest release (such as ffmpeg-0.6.3, or 0.7-rc1)? Is it possible to build a product-level stream media server based on ffmpeg + ffserver? Regards, Robbie From av at bsbc.nb.ca Thu May 5 20:49:40 2011 From: av at bsbc.nb.ca (Anthony Brown) Date: Thu, 05 May 2011 15:49:40 -0300 Subject: [FFmpeg-user] Has the ffserver's "av loose sync" problem been fixed? In-Reply-To: References: Message-ID: <4DC2F144.7080505@bsbc.nb.ca> On 11-05-05 12:27 PM, JiangJie wrote: > > Hi all, > > I'm considering building a stream media server based on ffmpeg + ffserver. > The video will come from a v4l2 device and the audio comes from an alsa device, i.e, /dev/dsp. > I'd like to stream both the audio and video simultaneously. > > However, from http://www.ffmpeg.org/ffserver.html, > it says that "The audio and video loose sync after a while". I wondered about this, too. I can tell you that I use ffmpeg/ffserver every week to stream HDV that is sourced by firewire to VLC on an XP computer, sent via http to ffserver and then streamed out over http again after a 10-ish minute delay. Instant time shift. I have not yet seen an AV sync problem. I use the version in ubuntu lucid-bleed, whatever that it. -- Anthony Brown Audiovisual coordinator Brunswick Street Baptist Church Telephone: (506)-458-8348 (leave message) Email: av at bsbc.nb.ca -------------- next part -------------- A non-text attachment was scrubbed... Name: av.vcf Type: text/x-vcard Size: 163 bytes Desc: not available URL: From sergey.forum at gmail.com Thu May 5 21:38:34 2011 From: sergey.forum at gmail.com (Sergey Kurdakov) Date: Thu, 5 May 2011 23:38:34 +0400 Subject: [FFmpeg-user] Force first frame to be full, non blurred I frame In-Reply-To: References: Message-ID: Hi, >I get blurry first frame found a solution - a patch from Ying Bian which adds qpfile option from x264 which is specially used in such cases to ffmpeg (patch is still is missing from ffmpeg code - but it proved to be useful for me). Regards Sergey From houndeyex at gmail.com Thu May 5 21:39:34 2011 From: houndeyex at gmail.com (James O.) Date: Thu, 5 May 2011 15:39:34 -0400 Subject: [FFmpeg-user] WMV to x264 => non monotone timestamps error Message-ID: Hello all, This is my first time posting to this list, so please forgive me if I break etiquette. I have a few "problem" WMVs that come through our system from time to time that I am trying to get resolved. I've tried everything I can think of at this point, and been all over the internet looking for solutions. We run two transcodes on the files that come in: one preparing it for the web (flash streaming media server), another preparing it for mobile (iPhone, Android, etc). The odd thing is that the mobile works fine, while the web has problems. I'm going to list a few things down here now. Hopefully, some of it will help make sense of what's happening. 1. MediaInfo dump from the offending WMV file 2. Complete web transcode log 3. Complete mobile transcode log Thank you in advance for any help you may be able to provide. If I can do anything else to help, please let me know. James --- MEDIA INFO DUMP --- General Complete name : C:\Users\jocull\Desktop\dt_test\dsp.wmv Format : Windows Media File size : 46.3 MiB Duration : 1mn 33s Overall bit rate mode : Variable Overall bit rate : 4 130 Kbps Maximum Overall bit rate : 11.6 Mbps Encoded date : UTC 2011-04-10 03:58:33.598 Video ID : 2 Format : WMV2 Codec ID : WMV2 Codec ID/Info : Windows Media Video 8 Description of the codec : Windows Media Video V8 Duration : 1mn 34s Bit rate mode : Variable Bit rate : 12.3 Mbps Width : 640 pixels Height : 480 pixels Display aspect ratio : 4:3 Frame rate : 30.000 fps Bit depth : 8 bits Bits/(Pixel*Frame) : 1.334 Stream size : 138 MiB Language : English (US) Audio ID : 1 Format : WMA Format version : Version 2 Codec ID : 161 Codec ID/Info : Windows Media Audio Description of the codec : Windows Media Audio 9.2 - 64 kbps, 44 kHz, stereo (A/V) 1-pass CBR Duration : 1mn 33s Bit rate mode : Constant Bit rate : 64.0 Kbps Channel(s) : 2 channels Sampling rate : 44.1 KHz Bit depth : 16 bits Stream size : 735 KiB (2%) Language : English (US) --- WEB TRANSCODE LOG --- ./ffmpeg -y -i ./8a1e1e8fd0304832a79889c0ae0cf9ae-input.wmv -crf 25.0 -vcodec libx264 -acodec libfaac -ar 48000 -ab 128k -coder 1 -flags +loop -cmp +chroma -partitions +parti4x4+partp8x8+partb8x8 -me_method hex -subq 6 -me_range 16 -g 250 -keyint_min 25 -sc_threshold 40 -i_qfactor 0.71 -b_strategy 1 -threads 0 -s 650x360 -vpre slow ./8a1e1e8fd0304832a79889c0ae0cf9ae_x264-650x360.mp4 FFmpeg version 0.6, Copyright (c) 2000-2010 the FFmpeg developers built on Oct 21 2010 11:30:55 with gcc 4.1.2 20080704 (Red Hat 4.1.2-48) configuration: --enable-gpl --enable-version3 --enable-nonfree --enable-postproc --enable-libfaac --enable-libmp3lame --enable-libx264 --enable-x11grab --disable-shared --enable-static --enable-runtime-cpudetect --disable-ffplay --disable-ffserver --disable-debug --extra-libs=-static --extra-cflags=--stat libavutil 50.15. 1 / 50.15. 1 libavcodec 52.72. 2 / 52.72. 2 libavformat 52.64. 2 / 52.64. 2 libavdevice 52. 2. 0 / 52. 2. 0 libswscale 0.11. 0 / 0.11. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, asf, from './8a1e1e8fd0304832a79889c0ae0cf9ae-input.wmv': Metadata: WMFSDKVersion : 12.0.7600.16385 WMFSDKNeeded : 0.0.0.0000 IsVBR : 1 VBR Peak : 112 Buffer Average : 564 Duration: 00:01:36.20, start: 3.000000, bitrate: 4035 kb/s Stream #0.0: Audio: wmav2, 44100 Hz, 2 channels, s16, 64 kb/s Stream #0.1: Video: wmv2, yuv420p, 640x480, 1k tbr, 1k tbn, 1k tbc Warning: not compiled with thread support, using thread emulation [libx264 @ 0x12a67520]using cpu capabilities: MMX2 SSE2Fast FastShuffle SSEMisalign LZCNT [libx264 @ 0x12a67520]profile High, level 5.1 [libx264 @ 0x12a67520]264 - core 106 - H.264/MPEG-4 AVC codec - Copyleft 2003-2010 - http://www.videolan.org/x264.html - options: cabac=1 ref=5 deblock=1:0:0 analyse=0x3:0x113 me=umh subme=8 psy=1 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=1 sliced_threads=0 nr=0 decimate=1 interlaced=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=2 b_bias=0 direct=3 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=50 rc=crf mbtree=1 crf=25.0 qcomp=0.60 qpmin=10 qpmax=51 qpstep=4 ip_ratio=1.41 aq=1:1.00 Output #0, mp4, to './8a1e1e8fd0304832a79889c0ae0cf9ae_x264-650x360.mp4': Metadata: encoder : Lavf52.64.2 Stream #0.0: Video: libx264, yuv420p, 650x360, q=10-51, 200 kb/s, 1k tbn, 1k tbc Stream #0.1: Audio: libfaac, 48000 Hz, 2 channels, s16, 128 kb/s Stream mapping: Stream #0.1 -> #0.0 Stream #0.0 -> #0.1 Press [q] to stop encoding frame= 20 fps= 0 q=649.0 size= 0kB time=3.26 bitrate= 0.1kbits/s frame= 42 fps= 41 q=649.0 size= 0kB time=5.50 bitrate= 0.1kbits/s frame= 51 fps= 30 q=30.0 size= 44kB time=2.21 bitrate= 162.2kbits/s [mp4 @ 0x12a664e0]st:0 error, non monotone timestamps 2208 >= 2208 av_interleaved_write_frame(): Operation not permitted --- MOBILE TRANSCODE LOG --- ./ffmpeg -y -i ./8a1e1e8fd0304832a79889c0ae0cf9ae-input.wmv -f mp4 -vcodec mpeg4 -b 600kb -r 24 -s 480x320 -aspect 16:9 -acodec libfaac -ab 128kb -ar 44100 ./8a1e1e8fd0304832a79889c0ae0cf9ae_mpeg4-480x320.mp4 FFmpeg version 0.6, Copyright (c) 2000-2010 the FFmpeg developers built on Oct 21 2010 11:30:55 with gcc 4.1.2 20080704 (Red Hat 4.1.2-48) configuration: --enable-gpl --enable-version3 --enable-nonfree --enable-postproc --enable-libfaac --enable-libmp3lame --enable-libx264 --enable-x11grab --disable-shared --enable-static --enable-runtime-cpudetect --disable-ffplay --disable-ffserver --disable-debug --extra-libs=-static --extra-cflags=--stat libavutil 50.15. 1 / 50.15. 1 libavcodec 52.72. 2 / 52.72. 2 libavformat 52.64. 2 / 52.64. 2 libavdevice 52. 2. 0 / 52. 2. 0 libswscale 0.11. 0 / 0.11. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, asf, from './8a1e1e8fd0304832a79889c0ae0cf9ae-input.wmv': Metadata: WMFSDKVersion : 12.0.7600.16385 WMFSDKNeeded : 0.0.0.0000 IsVBR : 1 VBR Peak : 112 Buffer Average : 564 Duration: 00:01:36.20, start: 3.000000, bitrate: 4035 kb/s Stream #0.0: Audio: wmav2, 44100 Hz, 2 channels, s16, 64 kb/s Stream #0.1: Video: wmv2, yuv420p, 640x480, 1k tbr, 1k tbn, 1k tbc Output #0, mp4, to './8a1e1e8fd0304832a79889c0ae0cf9ae_mpeg4-480x320.mp4': Metadata: encoder : Lavf52.64.2 Stream #0.0: Video: mpeg4, yuv420p, 480x320 [PAR 32:27 DAR 16:9], q=2-31, 600 kb/s, 24 tbn, 24 tbc Stream #0.1: Audio: libfaac, 44100 Hz, 2 channels, s16, 128 kb/s Stream mapping: Stream #0.1 -> #0.0 Stream #0.0 -> #0.1 Press [q] to stop encoding frame= 21 fps= 0 q=4.6 size= 155kB time=3.27 bitrate= 387.0kbits/s frame= 44 fps= 43 q=5.3 size= 322kB time=5.69 bitrate= 464.4kbits/s frame= 67 fps= 44 q=4.6 size= 496kB time=8.10 bitrate= 501.5kbits/s frame= 90 fps= 44 q=5.1 size= 680kB time=10.70 bitrate= 520.6kbits/s frame= 113 fps= 44 q=4.0 size= 851kB time=12.93 bitrate= 538.8kbits/s frame= 137 fps= 45 q=3.7 size= 1036kB time=15.35 bitrate= 552.8kbits/s frame= 160 fps= 45 q=4.2 size= 1216kB time=17.76 bitrate= 560.7kbits/s frame= 181 fps= 44 q=3.3 size= 1402kB time=20.18 bitrate= 569.1kbits/s frame= 204 fps= 45 q=4.7 size= 1597kB time=22.41 bitrate= 583.7kbits/s frame= 227 fps= 45 q=3.6 size= 1792kB time=24.64 bitrate= 595.9kbits/s frame= 250 fps= 45 q=3.3 size= 1993kB time=27.05 bitrate= 603.7kbits/s frame= 273 fps= 45 q=4.4 size= 2178kB time=29.28 bitrate= 609.4kbits/s frame= 296 fps= 45 q=3.7 size= 2401kB time=31.70 bitrate= 620.6kbits/s frame= 319 fps= 45 q=4.1 size= 2597kB time=34.11 bitrate= 623.8kbits/s frame= 341 fps= 45 q=3.8 size= 2828kB time=36.71 bitrate= 631.1kbits/s frame= 364 fps= 45 q=4.3 size= 3012kB time=39.13 bitrate= 630.7kbits/s frame= 387 fps= 45 q=4.4 size= 3214kB time=41.35 bitrate= 636.7kbits/s frame= 410 fps= 45 q=3.7 size= 3428kB time=43.77 bitrate= 641.6kbits/s frame= 433 fps= 45 q=3.5 size= 3630kB time=46.18 bitrate= 643.9kbits/s frame= 456 fps= 45 q=3.6 size= 3832kB time=48.41 bitrate= 648.4kbits/s frame= 479 fps= 45 q=3.7 size= 4028kB time=50.64 bitrate= 651.6kbits/s frame= 503 fps= 45 q=3.6 size= 4221kB time=52.69 bitrate= 656.3kbits/s frame= 525 fps= 45 q=4.8 size= 4442kB time=55.10 bitrate= 660.4kbits/s frame= 548 fps= 45 q=3.6 size= 4658kB time=57.52 bitrate= 663.4kbits/s frame= 571 fps= 45 q=3.5 size= 4858kB time=59.74 bitrate= 666.2kbits/s frame= 594 fps= 45 q=4.9 size= 5057kB time=61.97 bitrate= 668.5kbits/s frame= 617 fps= 45 q=4.6 size= 5237kB time=64.02 bitrate= 670.2kbits/s frame= 640 fps= 45 q=4.4 size= 5454kB time=66.06 bitrate= 676.3kbits/s frame= 663 fps= 45 q=3.7 size= 5681kB time=68.48 bitrate= 679.7kbits/s frame= 686 fps= 45 q=3.9 size= 5883kB time=70.89 bitrate= 679.8kbits/s frame= 709 fps= 45 q=3.7 size= 6060kB time=73.12 bitrate= 678.9kbits/s frame= 732 fps= 45 q=3.8 size= 6257kB time=75.35 bitrate= 680.3kbits/s frame= 755 fps= 45 q=4.0 size= 6464kB time=77.58 bitrate= 682.6kbits/s frame= 779 fps= 45 q=3.7 size= 6650kB time=79.81 bitrate= 682.6kbits/s frame= 802 fps= 45 q=4.4 size= 6853kB time=82.04 bitrate= 684.3kbits/s frame= 825 fps= 45 q=3.3 size= 7065kB time=84.45 bitrate= 685.3kbits/s frame= 848 fps= 45 q=4.4 size= 7265kB time=86.87 bitrate= 685.1kbits/s frame= 872 fps= 45 q=3.8 size= 7435kB time=88.72 bitrate= 686.4kbits/s frame= 896 fps= 45 q=3.3 size= 7653kB time=90.95 bitrate= 689.3kbits/s frame= 912 fps= 45 q=3.2 Lsize= 7930kB time=93.46 bitrate= 695.1kbits/s video:6426kB audio:1460kB global headers:0kB muxing overhead 0.562451% ./qt-faststart ./8a1e1e8fd0304832a79889c0ae0cf9ae_mpeg4-480x320.mp4 qtfs_8a1e1e8fd0304832a79889c0ae0cf9ae_mpeg4-480x320.mp4 ftyp 0 28 free 28 8 mdat 36 8074976 moov 8075012 45423 patching stco atom... patching stco atom... writing ftyp atom... writing moov atom... copying rest of file... From Cecil at decebal.nl Thu May 5 21:54:16 2011 From: Cecil at decebal.nl (Cecil Westerhof) Date: Thu, 05 May 2011 21:54:16 +0200 Subject: [FFmpeg-user] Why does this half the filesize Message-ID: <8762poc49j.fsf@Compaq.site> When executing: ffmpeg -y -sameq -i original/resolutie.avi resolutie.avi The resulting file is half as big as the original one. Why is this? The reason I do this, is that I capture something from XWindows and I have to cut the last few seconds. If it is useful: ffmpeg -i original/resolutie.avi FFmpeg version SVN-r201104161305, Copyright (c) 2000-2011 the FFmpeg developers built on Apr 16 2011 11:36:21 with gcc 4.5.1 20101208 [gcc-4_5-branch revision 167585] configuration: --shlibdir=/usr/lib --prefix=/usr --mandir=/usr/share/man --libdir=/usr/lib --enable-shared --enable-libmp3lame --enable-libvorbis --enable-libtheora --enable-libspeex --enable-libxvid --enable-postproc --enable-gpl --enable-x11grab --extra-cflags='-fomit-frame-pointer -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gsm' --enable-debug --disable-stripping --enable-libgsm --enable-libschroedinger --enable-libdirac --enable-avfilter --enable-libvpx --enable-version3 --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libx264 --enable-libdc1394 --enable-pthreads --enable-librtmp libavutil 50. 40. 1 / 50. 40. 1 libavcodec 52.119. 1 / 52.119. 1 libavformat 52.108. 0 / 52.108. 0 libavdevice 52. 4. 0 / 52. 4. 0 libavfilter 1. 79. 0 / 1. 79. 0 libswscale 0. 13. 0 / 0. 13. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, avi, from 'original/resolutie.avi': Metadata: encoder : Lavf52.108.0 Duration: 00:00:53.93, start: 0.000000, bitrate: 2809 kb/s Stream #0.0: Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 15 tbr, 15 tbn, 15 tbc Stream #0.1: Audio: pcm_s16le, 44100 Hz, 2 channels, s16, 1411 kb/s At least one output file must be specified ffmpeg -i resolutie.avi FFmpeg version SVN-r201104161305, Copyright (c) 2000-2011 the FFmpeg developers built on Apr 16 2011 11:36:21 with gcc 4.5.1 20101208 [gcc-4_5-branch revision 167585] configuration: --shlibdir=/usr/lib --prefix=/usr --mandir=/usr/share/man --libdir=/usr/lib --enable-shared --enable-libmp3lame --enable-libvorbis --enable-libtheora --enable-libspeex --enable-libxvid --enable-postproc --enable-gpl --enable-x11grab --extra-cflags='-fomit-frame-pointer -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gsm' --enable-debug --disable-stripping --enable-libgsm --enable-libschroedinger --enable-libdirac --enable-avfilter --enable-libvpx --enable-version3 --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libx264 --enable-libdc1394 --enable-pthreads --enable-librtmp libavutil 50. 40. 1 / 50. 40. 1 libavcodec 52.119. 1 / 52.119. 1 libavformat 52.108. 0 / 52.108. 0 libavdevice 52. 4. 0 / 52. 4. 0 libavfilter 1. 79. 0 / 1. 79. 0 libswscale 0. 13. 0 / 0. 13. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, avi, from 'resolutie.avi': Metadata: encoder : Lavf52.108.0 Duration: 00:00:53.93, start: 0.000000, bitrate: 1458 kb/s Stream #0.0: Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 15 tbr, 15 tbn, 15 tbc Stream #0.1: Audio: mp2, 44100 Hz, stereo, s16, 64 kb/s At least one output file must be specified -- Cecil Westerhof Senior Software Engineer LinkedIn: http://www.linkedin.com/in/cecilwesterhof From h.reindl at thelounge.net Thu May 5 22:16:57 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Thu, 05 May 2011 22:16:57 +0200 Subject: [FFmpeg-user] Why does this half the filesize In-Reply-To: <8762poc49j.fsf@Compaq.site> References: <8762poc49j.fsf@Compaq.site> Message-ID: <4DC305B9.3040206@thelounge.net> because the bitrates are different and sameq is useless before input-file Am 05.05.2011 21:54, schrieb Cecil Westerhof: > When executing: > ffmpeg -y -sameq -i original/resolutie.avi resolutie.avi > > The resulting file is half as big as the original one. Why is this? > > The reason I do this, is that I capture something from XWindows and I > have to cut the last few seconds. > > If it is useful: > ffmpeg -i original/resolutie.avi > FFmpeg version SVN-r201104161305, Copyright (c) 2000-2011 the FFmpeg developers > built on Apr 16 2011 11:36:21 with gcc 4.5.1 20101208 [gcc-4_5-branch revision 167585] > configuration: --shlibdir=/usr/lib --prefix=/usr --mandir=/usr/share/man --libdir=/usr/lib --enable-shared --enable-libmp3lame --enable-libvorbis --enable-libtheora --enable-libspeex --enable-libxvid --enable-postproc --enable-gpl --enable-x11grab --extra-cflags='-fomit-frame-pointer -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gsm' --enable-debug --disable-stripping --enable-libgsm --enable-libschroedinger --enable-libdirac --enable-avfilter --enable-libvpx --enable-version3 --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libx264 --enable-libdc1394 --enable-pthreads --enable-librtmp > libavutil 50. 40. 1 / 50. 40. 1 > libavcodec 52.119. 1 / 52.119. 1 > libavformat 52.108. 0 / 52.108. 0 > libavdevice 52. 4. 0 / 52. 4. 0 > libavfilter 1. 79. 0 / 1. 79. 0 > libswscale 0. 13. 0 / 0. 13. 0 > libpostproc 51. 2. 0 / 51. 2. 0 > Input #0, avi, from 'original/resolutie.avi': > Metadata: > encoder : Lavf52.108.0 > Duration: 00:00:53.93, start: 0.000000, bitrate: 2809 kb/s > Stream #0.0: Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 15 tbr, 15 tbn, 15 tbc > Stream #0.1: Audio: pcm_s16le, 44100 Hz, 2 channels, s16, 1411 kb/s > At least one output file must be specified > > > > ffmpeg -i resolutie.avi > FFmpeg version SVN-r201104161305, Copyright (c) 2000-2011 the FFmpeg developers > built on Apr 16 2011 11:36:21 with gcc 4.5.1 20101208 [gcc-4_5-branch revision 167585] > configuration: --shlibdir=/usr/lib --prefix=/usr --mandir=/usr/share/man --libdir=/usr/lib --enable-shared --enable-libmp3lame --enable-libvorbis --enable-libtheora --enable-libspeex --enable-libxvid --enable-postproc --enable-gpl --enable-x11grab --extra-cflags='-fomit-frame-pointer -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gsm' --enable-debug --disable-stripping --enable-libgsm --enable-libschroedinger --enable-libdirac --enable-avfilter --enable-libvpx --enable-version3 --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libx264 --enable-libdc1394 --enable-pthreads --enable-librtmp > libavutil 50. 40. 1 / 50. 40. 1 > libavcodec 52.119. 1 / 52.119. 1 > libavformat 52.108. 0 / 52.108. 0 > libavdevice 52. 4. 0 / 52. 4. 0 > libavfilter 1. 79. 0 / 1. 79. 0 > libswscale 0. 13. 0 / 0. 13. 0 > libpostproc 51. 2. 0 / 51. 2. 0 > Input #0, avi, from 'resolutie.avi': > Metadata: > encoder : Lavf52.108.0 > Duration: 00:00:53.93, start: 0.000000, bitrate: 1458 kb/s > Stream #0.0: Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 15 tbr, 15 tbn, 15 tbc > Stream #0.1: Audio: mp2, 44100 Hz, stereo, s16, 64 kb/s > At least one output file must be specified > -- Mit besten Gr??en, Reindl Harald the lounge interactive design GmbH A-1060 Vienna, Hofm?hlgasse 17 CTO / software-development / cms-solutions p: +43 (1) 595 3999 33, m: +43 (676) 40 221 40 icq: 154546673, http://www.thelounge.net/ http://www.thelounge.net/signature.asc.what.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From seandarcy2 at gmail.com Thu May 5 22:19:57 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Thu, 05 May 2011 16:19:57 -0400 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? Message-ID: I have an interlaced dv file. I'm transcoding it with x264 to mp4. 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 If I just leave it like that, is out.mp4 interlaced or progressive? 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 Here I assume out.mp4 is progressive. The ffmpeg documentation says: "The alternative is to deinterlace the input stream with `-deinterlace', but deinterlacing introduces losses." Is this still true? Given this note about losses, am I right we should never deinterlace? Almost never? When is deinterlacing required/better? 3. ffmpeg -i file.dv -an -vcodec libx264 -b -flags +ilme+ildct out.mp4 Here I assume out.mp4 is interlaced. How is this different from 1. above? Is it different? The FAQ mentions +alt as a flag. When should this be added? Does it hurt to use it all the time? sean From belcampo at zonnet.nl Thu May 5 22:36:04 2011 From: belcampo at zonnet.nl (belcampo) Date: Thu, 05 May 2011 22:36:04 +0200 Subject: [FFmpeg-user] Why does this half the filesize In-Reply-To: <8762poc49j.fsf@Compaq.site> References: <8762poc49j.fsf@Compaq.site> Message-ID: <4DC30A34.5060608@zonnet.nl> On 05/05/11 21:54, Cecil Westerhof wrote: > When executing: > ffmpeg -y -sameq -i original/resolutie.avi resolutie.avi > > The resulting file is half as big as the original one. Why is this? > > The reason I do this, is that I capture something from XWindows and I > have to cut the last few seconds. > > If it is useful: > ffmpeg -i original/resolutie.avi > FFmpeg version SVN-r201104161305, Copyright (c) 2000-2011 the FFmpeg developers > built on Apr 16 2011 11:36:21 with gcc 4.5.1 20101208 [gcc-4_5-branch revision 167585] > configuration: --shlibdir=/usr/lib --prefix=/usr --mandir=/usr/share/man --libdir=/usr/lib --enable-shared --enable-libmp3lame --enable-libvorbis --enable-libtheora --enable-libspeex --enable-libxvid --enable-postproc --enable-gpl --enable-x11grab --extra-cflags='-fomit-frame-pointer -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gsm' --enable-debug --disable-stripping --enable-libgsm --enable-libschroedinger --enable-libdirac --enable-avfilter --enable-libvpx --enable-version3 --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libx264 --enable-libdc1394 --enable-pthreads --enable-librtmp > libavutil 50. 40. 1 / 50. 40. 1 > libavcodec 52.119. 1 / 52.119. 1 > libavformat 52.108. 0 / 52.108. 0 > libavdevice 52. 4. 0 / 52. 4. 0 > libavfilter 1. 79. 0 / 1. 79. 0 > libswscale 0. 13. 0 / 0. 13. 0 > libpostproc 51. 2. 0 / 51. 2. 0 > Input #0, avi, from 'original/resolutie.avi': > Metadata: > encoder : Lavf52.108.0 > Duration: 00:00:53.93, start: 0.000000, bitrate: 2809 kb/s > Stream #0.0: Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 15 tbr, 15 tbn, 15 tbc > Stream #0.1: Audio: pcm_s16le, 44100 Hz, 2 channels, s16, 1411 kb/s > At least one output file must be specified > > > > ffmpeg -i resolutie.avi > FFmpeg version SVN-r201104161305, Copyright (c) 2000-2011 the FFmpeg developers > built on Apr 16 2011 11:36:21 with gcc 4.5.1 20101208 [gcc-4_5-branch revision 167585] > configuration: --shlibdir=/usr/lib --prefix=/usr --mandir=/usr/share/man --libdir=/usr/lib --enable-shared --enable-libmp3lame --enable-libvorbis --enable-libtheora --enable-libspeex --enable-libxvid --enable-postproc --enable-gpl --enable-x11grab --extra-cflags='-fomit-frame-pointer -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gsm' --enable-debug --disable-stripping --enable-libgsm --enable-libschroedinger --enable-libdirac --enable-avfilter --enable-libvpx --enable-version3 --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libx264 --enable-libdc1394 --enable-pthreads --enable-librtmp > libavutil 50. 40. 1 / 50. 40. 1 > libavcodec 52.119. 1 / 52.119. 1 > libavformat 52.108. 0 / 52.108. 0 > libavdevice 52. 4. 0 / 52. 4. 0 > libavfilter 1. 79. 0 / 1. 79. 0 > libswscale 0. 13. 0 / 0. 13. 0 > libpostproc 51. 2. 0 / 51. 2. 0 > Input #0, avi, from 'resolutie.avi': > Metadata: > encoder : Lavf52.108.0 > Duration: 00:00:53.93, start: 0.000000, bitrate: 1458 kb/s > Stream #0.0: Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 15 tbr, 15 tbn, 15 tbc > Stream #0.1: Audio: mp2, 44100 Hz, stereo, s16, 64 kb/s > At least one output file must be specified > "Normally" one specifies what ffmpeg has to do: ffmpeg -i source.ext -acodec somecodec somecodec-params -vcodec somecodec somecodec-params resulting.ext. Where 'somecodec somecodec-params" in above case should/could be copy. Resulting in: ffmpeg -i source.ext -acodec copy -vcodec copy result.ext Maybe reading the manuals and/or looking for samples the next time ?? From baptiste.coudurier at gmail.com Thu May 5 22:36:54 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Thu, 05 May 2011 13:36:54 -0700 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: Message-ID: <4DC30A66.3040006@gmail.com> Hi, On 05/05/2011 01:19 PM, sean darcy wrote: > I have an interlaced dv file. I'm transcoding it with x264 to mp4. > > 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 > > If I just leave it like that, is out.mp4 interlaced or progressive? progressive. By default encoding is progressive. > 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 > > Here I assume out.mp4 is progressive. The ffmpeg documentation says: > > "The alternative is to deinterlace the input stream with `-deinterlace', > but deinterlacing introduces losses." Correct, it is progressive. Use -vf yadif instead of -deinterlace Deinterlacing may be a bit destructive, especially if the input is _not_ interlaced. > [...] > > Given this note about losses, am I right we should never deinterlace? > Almost never? When is deinterlacing required/better? You have options: if the receiving end playback interlaced (CRT tv): encode interlaced else if the receiver is going to deinterlace if the file is marked as interlaced and you trust this deinterlacer, then you may encode interlaced (deinterlacing will take cpu time) otherwise you should deinterlace yourself using a good deinterlacer. I suggest always deinterlace using -vf yadif if the source content is interlaced > 3. ffmpeg -i file.dv -an -vcodec libx264 -b -flags +ilme+ildct out.mp4 > > Here I assume out.mp4 is interlaced. How is this different from 1. > above? Is it different? Correct. > The FAQ mentions +alt as a flag. When should this be added? Does it hurt > to use it all the time? +alt only applies to mpeg4 and mpeg2. It's supposed to improve compression for interlaced content. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From belcampo at zonnet.nl Thu May 5 22:47:33 2011 From: belcampo at zonnet.nl (belcampo) Date: Thu, 05 May 2011 22:47:33 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: Message-ID: <4DC30CE5.5040308@zonnet.nl> On 05/05/11 22:19, sean darcy wrote: > I have an interlaced dv file. I'm transcoding it with x264 to mp4. > > 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 > > If I just leave it like that, is out.mp4 interlaced or progressive? > > 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 > > Here I assume out.mp4 is progressive. The ffmpeg documentation says: > > "The alternative is to deinterlace the input stream with `-deinterlace', > but deinterlacing introduces losses." > > Is this still true? > > Given this note about losses, am I right we should never deinterlace? > Almost never? When is deinterlacing required/better? It depends, as always. If the final display is a TV, which is capable to deinterlacing, Most except 720p, which is rare, television broadcast is interlaced. If the final display is a computer-screen or an iPhone/AndroidPhone or something like that, these devices don't deinterlace by themselves so the source has to be deinterlaced, or has the deinterlace flag set at the moment of display. Most people don't know that their computer mplayer/vlc/WindowsMediaPlayer has an deinterlace option and will see a very bad interlaced picture. > > 3. ffmpeg -i file.dv -an -vcodec libx264 -b -flags +ilme+ildct out.mp4 > > Here I assume out.mp4 is interlaced. How is this different from 1. > above? Is it different? > > The FAQ mentions +alt as a flag. When should this be added? Does it hurt > to use it all the time? > > sean > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From bouke at editb.nl Thu May 5 23:18:01 2011 From: bouke at editb.nl (Bouke) Date: Thu, 5 May 2011 23:18:01 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? References: <4DC30CE5.5040308@zonnet.nl> Message-ID: <002c01cc0b69$edcc1bb0$4301a8c0@hpkantoor> ----- Original Message ----- From: "belcampo" > If the final display is a TV, which is capable to deinterlacing Oi, big misunderstanding.... a good catholic (injoke) TV is NOT capable of de-interlacing. It is capable of displaying interlaced images one field after another, thus making pretty moving pictures. (don't get me started about 100 Hz sets...) Bouke From jsd at cluttered.com Thu May 5 23:18:29 2011 From: jsd at cluttered.com (Jon Drukman) Date: Thu, 5 May 2011 21:18:29 +0000 (UTC) Subject: [FFmpeg-user] avi file "operation not permitted" Message-ID: /usr/local/bin/ffmpeg -loglevel quiet -v 0 -i list.avi -s 480x352 -an -pass 1 -vcodec libx264 -vpre fast_firstpass -b 300k -bt 300k -threads 0 list.avi.mp4 ffmpeg version UNKNOWN, Copyright (c) 2000-2011 the FFmpeg developers built on May 5 2011 11:01:27 with gcc 4.1.2 20080704 (Red Hat 4.1.2-48) configuration: --enable-gpl --enable-libfaac --enable-libmp3lame --enable-libx264 --enable-pthreads --enable-static --disable-shared --disable-network --enable-nonfree libavutil 51. 2. 0 / 51. 2. 0 libavcodec 53. 3. 0 / 53. 3. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 4. 0 / 2. 4. 0 libswscale 0. 14. 0 / 0. 14. 0 list.avi: Operation not permitted What does it mean? I built ffmpeg from source this morning using last night's tarball/snapshot. From mark at mdsh.com Fri May 6 00:32:22 2011 From: mark at mdsh.com (Mark Himsley) Date: Thu, 05 May 2011 23:32:22 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC30A66.3040006@gmail.com> References: <4DC30A66.3040006@gmail.com> Message-ID: <4DC32576.2010203@mdsh.com> On 05/05/2011 21:36, Baptiste Coudurier wrote: > Hi, > > On 05/05/2011 01:19 PM, sean darcy wrote: >> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >> >> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >> >> If I just leave it like that, is out.mp4 interlaced or progressive? > > progressive. By default encoding is progressive. I'd like to clarify that answer. I agree 100% that the mp4 will be encoded progressive. The problem is that it could be badly encoding interlaced material. What I mean is, it will be using progressive frame encoding techniques to encode a frame that might be carrying interlaced material, which would therefore display with comb edges on movement etc. [...] >> 3. ffmpeg -i file.dv -an -vcodec libx264 -b -flags +ilme+ildct out.mp4 >> >> Here I assume out.mp4 is interlaced. How is this different from 1. >> above? Is it different? > > Correct. It is different from 1 because it encodes the two fields as temporally different half-frames (if you don't mind my over-simplification). In your option 1, the encoder can reduce bit-rate by throwing-away/hiding stuff on every line of the picture, in one go. Because interlacing works by every other line being from a different temporal snapshot, that throwing-away/hiding could move pixels from one temporal snapshot to another. Than can look really horrible. In your opting 3 the encoder knows that the image is made up of the two temporal snapshots and will not do that, but can use its knowledge that the second snapshot may contain the same image as the first snapshot, only with some/all parts moved, to reduce the bit-rate - by encoding the second field as deltas from the first field. -- Mark From mark at mdsh.com Fri May 6 00:41:18 2011 From: mark at mdsh.com (Mark Himsley) Date: Thu, 05 May 2011 23:41:18 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <002c01cc0b69$edcc1bb0$4301a8c0@hpkantoor> References: <4DC30CE5.5040308@zonnet.nl> <002c01cc0b69$edcc1bb0$4301a8c0@hpkantoor> Message-ID: <4DC3278E.30506@mdsh.com> On 05/05/2011 22:18, Bouke wrote: > > ----- Original Message ----- > From: "belcampo" > >> If the final display is a TV, which is capable to deinterlacing > > Oi, big misunderstanding.... > a good catholic (injoke) TV is NOT capable of de-interlacing. Although, there are many non-cathode-ray-tube displays that do de-interlace (by line doubling fields), the ?3500 JVC TFT monitor next to me and the ?350 Sony telly at home being just two examples... > It is capable of displaying interlaced images one field after another, thus > making pretty moving pictures. but, yes that is definitely a better, if somewhat wordy, description that sits better with my pedantic way of thinking. > (don't get me started about 100 Hz sets...) vile, horrid, turn it off NOW. Do manufacturers not understand the time domain? -- Mark From bouke at editb.nl Fri May 6 00:46:17 2011 From: bouke at editb.nl (Bouke) Date: Fri, 6 May 2011 00:46:17 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? References: <4DC30CE5.5040308@zonnet.nl><002c01cc0b69$edcc1bb0$4301a8c0@hpkantoor> <4DC3278E.30506@mdsh.com> Message-ID: <003001cc0b76$41ea1c90$4301a8c0@hpkantoor> ----- Original Message ----- From: "Mark Himsley" > Although, there are many non-cathode-ray LOL, You just took a good joke to the next level! Bouke From seandarcy2 at gmail.com Fri May 6 01:34:12 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Thu, 05 May 2011 19:34:12 -0400 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC30A66.3040006@gmail.com> References: <4DC30A66.3040006@gmail.com> Message-ID: On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: > Hi, > > On 05/05/2011 01:19 PM, sean darcy wrote: >> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >> >> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >> >> If I just leave it like that, is out.mp4 interlaced or progressive? > > progressive. By default encoding is progressive. > >> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 >> >> Here I assume out.mp4 is progressive. The ffmpeg documentation says: >> >> "The alternative is to deinterlace the input stream with `-deinterlace', >> but deinterlacing introduces losses." > > Correct, it is progressive. Use -vf yadif instead of -deinterlace > Deinterlacing may be a bit destructive, especially if the input is _not_ > interlaced. > >> [...] >> >> Given this note about losses, am I right we should never deinterlace? >> Almost never? When is deinterlacing required/better? > > You have options: > if the receiving end playback interlaced (CRT tv): > encode interlaced > else if the receiver is going to deinterlace if the file is marked as > interlaced and you trust this deinterlacer, then you may encode > interlaced (deinterlacing will take cpu time) > otherwise you should deinterlace yourself using a good deinterlacer. Right. I knew that! Just passed right out of my mind, though. > > I suggest always deinterlace using -vf yadif if the source content is > interlaced > Well I found http://guru.multimedia.cx/deinterlacing-filters/ so I thought I'd try: -vf "yadif=3:0,mp=mcdeint=2:0:10" (dv is bottom-field first, right?) That generates a lot of perplexing output: [snow @ 0x230df40] pass:4mv changed:1384 [snow @ 0x230df40] pass:0 changed:1083 [snow @ 0x230df40] pass:1 changed:407 [snow @ 0x230df40] pass:2 changed:147 [snow @ 0x230df40] pass:3 changed:50 [snow @ 0x230df40] pass:4 changed:17 [snow @ 0x230df40] pass:5 changed:11 [snow @ 0x230df40] pass:6 changed:2 [snow @ 0x230df40] pass:7 changed:1 [snow @ 0x230df40] pass:8 changed:1 [snow @ 0x230df40] pass:9 changed:1 [snow @ 0x230df40] pass:10 changed:1 [snow @ 0x230df40] pass:11 changed:2 [snow @ 0x230df40] pass:12 changed:2 [snow @ 0x230df40] pass:13 changed:2 [snow @ 0x230df40] pass:14 changed:0 [snow @ 0x230df40] pass:4mv changed:1864 I realize the filter comparison is from five years ago, and yadif may have changed significantly since then. Does mcdeint still add anything to yadif? sean From baptiste.coudurier at gmail.com Fri May 6 01:42:16 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Thu, 05 May 2011 16:42:16 -0700 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: <4DC30A66.3040006@gmail.com> Message-ID: <4DC335D8.5080504@gmail.com> On 05/05/2011 04:34 PM, sean darcy wrote: > On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: >> Hi, >> >> On 05/05/2011 01:19 PM, sean darcy wrote: >>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>> >>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>> >>> If I just leave it like that, is out.mp4 interlaced or progressive? >> >> progressive. By default encoding is progressive. >> >>> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 >>> >>> Here I assume out.mp4 is progressive. The ffmpeg documentation says: >>> >>> "The alternative is to deinterlace the input stream with `-deinterlace', >>> but deinterlacing introduces losses." >> >> Correct, it is progressive. Use -vf yadif instead of -deinterlace >> Deinterlacing may be a bit destructive, especially if the input is _not_ >> interlaced. >> >>> [...] >>> >>> Given this note about losses, am I right we should never deinterlace? >>> Almost never? When is deinterlacing required/better? >> >> You have options: >> if the receiving end playback interlaced (CRT tv): >> encode interlaced >> else if the receiver is going to deinterlace if the file is marked as >> interlaced and you trust this deinterlacer, then you may encode >> interlaced (deinterlacing will take cpu time) >> otherwise you should deinterlace yourself using a good deinterlacer. > > Right. I knew that! Just passed right out of my mind, though. >> >> I suggest always deinterlace using -vf yadif if the source content is >> interlaced >> > > Well I found http://guru.multimedia.cx/deinterlacing-filters/ > > so I thought I'd try: > > -vf "yadif=3:0,mp=mcdeint=2:0:10" > > (dv is bottom-field first, right?) > > That generates a lot of perplexing output: > > [snow @ 0x230df40] pass:4mv changed:1384 > [snow @ 0x230df40] pass:0 changed:1083 > [snow @ 0x230df40] pass:1 changed:407 > [snow @ 0x230df40] pass:2 changed:147 > [snow @ 0x230df40] pass:3 changed:50 > [snow @ 0x230df40] pass:4 changed:17 > [snow @ 0x230df40] pass:5 changed:11 > [snow @ 0x230df40] pass:6 changed:2 > [snow @ 0x230df40] pass:7 changed:1 > [snow @ 0x230df40] pass:8 changed:1 > [snow @ 0x230df40] pass:9 changed:1 > [snow @ 0x230df40] pass:10 changed:1 > [snow @ 0x230df40] pass:11 changed:2 > [snow @ 0x230df40] pass:12 changed:2 > [snow @ 0x230df40] pass:13 changed:2 > [snow @ 0x230df40] pass:14 changed:0 > [snow @ 0x230df40] pass:4mv changed:1864 This is some debug messages, ignore them. > I realize the filter comparison is from five years ago, and yadif may > have changed significantly since then. Does mcdeint still add anything > to yadif? I think nothing has changed much since then :) -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From seandarcy2 at gmail.com Fri May 6 01:42:14 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Thu, 05 May 2011 19:42:14 -0400 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC32576.2010203@mdsh.com> References: <4DC30A66.3040006@gmail.com> <4DC32576.2010203@mdsh.com> Message-ID: On 05/05/2011 06:32 PM, Mark Himsley wrote: > On 05/05/2011 21:36, Baptiste Coudurier wrote: >> Hi, >> >> On 05/05/2011 01:19 PM, sean darcy wrote: >>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>> >>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>> >>> If I just leave it like that, is out.mp4 interlaced or progressive? >> >> progressive. By default encoding is progressive. > > I'd like to clarify that answer. > > I agree 100% that the mp4 will be encoded progressive. The problem is > that it could be badly encoding interlaced material. What I mean is, it > will be using progressive frame encoding techniques to encode a frame > that might be carrying interlaced material, which would therefore > display with comb edges on movement etc. > > [...] >>> 3. ffmpeg -i file.dv -an -vcodec libx264 -b -flags +ilme+ildct out.mp4 >>> >>> Here I assume out.mp4 is interlaced. How is this different from 1. >>> above? Is it different? >> >> Correct. > > It is different from 1 because it encodes the two fields as temporally > different half-frames (if you don't mind my over-simplification). > > In your option 1, the encoder can reduce bit-rate by > throwing-away/hiding stuff on every line of the picture, in one go. > Because interlacing works by every other line being from a different > temporal snapshot, that throwing-away/hiding could move pixels from one > temporal snapshot to another. Than can look really horrible. > > In your opting 3 the encoder knows that the image is made up of the two > temporal snapshots and will not do that, but can use its knowledge that > the second snapshot may contain the same image as the first snapshot, > only with some/all parts moved, to reduce the bit-rate - by encoding the > second field as deltas from the first field. > Thanks. Very helpful. I have realized I need to deinterlace, but now I know what to do if I'm keeping the stream interlaced. From rt.mxa.csc at gmail.com Fri May 6 03:49:19 2011 From: rt.mxa.csc at gmail.com (RT mxa) Date: Thu, 5 May 2011 18:49:19 -0700 Subject: [FFmpeg-user] Error in configure (libspeex enabled) on RHEL Message-ID: Hi, I get the following error when i try to build a static binary on RHEL with libspeex enabled (I do not get this error for libmp3lame) Here is the configure command I am running ./configure --disable-mmx --disable-mmx2 --enable-libmp3lame --enable-libspeex --extra-cflags='--static' --extra-libs=-static ERROR: libspeex not found This is the specific error in the config.err file BEGIN /tmp/ffconf.npz27351.c 1 extern int speex_decoder_init(); 2 int main(void){ speex_decoder_init(); } END /tmp/ffconf.npz27351.c gcc -D_ISOC99_SOURCE -D_POSIX_C_SOURCE=200112 -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE --static -std=c99 -fomit-frame-pointer -c -o /tmp/ffconf.jYa27360.o /tmp/ffconf.npz27351.c gcc -o /tmp/ffconf.Liz27354 /tmp/ffconf.jYa27360.o -static -lz -lbz2 -lm -lmp3lame -lm -lspeex /usr/lib/gcc/x86_64-redhat-linux/3.4.6/../../../libspeex.a(speex.o)(.text+0x1c0): In function `speex_decode_int': /auto/pulse/FFmpegBuild/speex-1.2rc1/libspeex/speex.c:172: undefined reference to `floor' collect2: ld returned 1 exit status ERROR: libspeex not found Thanks in advance for the help From hardik.sharma22 at yahoo.com Fri May 6 03:58:06 2011 From: hardik.sharma22 at yahoo.com (Hardik Sharma) Date: Thu, 5 May 2011 18:58:06 -0700 (PDT) Subject: [FFmpeg-user] Unexpected decoded video output Message-ID: <130891.18214.qm@web46207.mail.sp1.yahoo.com> Hi, I encoded raw yuv video to h.264 format and decoded it back to yuv format in ffmpeg. But while watching decoded yuv video I noticed that user can see that frames are moving vertically and it is unexpected video output as compared to original video.I encoded the yuv video using following command - ffmpeg -y -s 720x480 -f rawvideo -r 30 -i akiyo_720x480.yuv -vcodec libx264 -y -b 256k -r 30 -preset fast? -g 10 -s 720x480 -coder 1 -flags +loop -cmp +chroma -f h264 -flags2 -fastpskip -trellis 1????? -partitions +parti4x4+partp8x8+partb8x8+parti8x8+partp4x4 -chromaoffset 0 -b_qfactor 0.45 -flags2 +wpred -subq 8 -flags2 +mixed_refs -flags2 +dct8x8 -me_range 16 -me_method umh -keyint_min 10 -sc_threshold 40 -i_qfactor 0.71 -qcomp 0.5 -rc_eq 'blurCplx^(1-qComp)' -level 40 -bf 16 -bframebias 0 -b_strategy 2 -bidir_refine 1 -refs 6 -cqp 28 -qmin 8 -qmax 48 -deblockalpha 0 -deblockbeta 0 -threads 0 -an output_akiyo.264 and decoded using following command- ffmpeg -y -i output_akiyo.264 -vcodec rawvideo -f yuv4mpegpipe out264_new.y4m I encoded same video with 2-pass too but getting same incorrect output. Let me know if I am missing or using some wrong parameter. I will really appreciate your help. Thanks. Regards, Hardik Sharma ? ? ? From dellwust at 163.com Fri May 6 08:25:16 2011 From: dellwust at 163.com (dellwust) Date: Thu, 5 May 2011 23:25:16 -0700 (PDT) Subject: [FFmpeg-user] meet problems when compiling FFMpeg Message-ID: <1304663116525-3501308.post@n4.nabble.com> Hello, EveryBody: When i compiling i meet a problem: YASM libavcodec/x86/vc1dsp_yasm.o libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands libavcodec/x86/vc1dsp_yasm.asm:329: invalid combination of opcode and operands any suggestions? -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/meet-problems-when-compiling-FFMpeg-tp3501308p3501308.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From yangtzj at hotmail.com Fri May 6 08:53:50 2011 From: yangtzj at hotmail.com (JiangJie) Date: Fri, 6 May 2011 14:53:50 +0800 Subject: [FFmpeg-user] Has ffserver's "av loose sync" problem been fixed? Message-ID: Hi all, I'm considering building a stream media server based on ffmpeg + ffserver. The video will come from a v4l2 device and the audio comes from an alsa device, i.e, /dev/dsp. I'd like to stream both the encoded audio and video simultaneously. However, from http://www.ffmpeg.org/ffserver.html, it says that "The audio and video loose sync after a while". Has this problem been fixed in the latest release (such as ffmpeg-0.6.3, or 0.7-rc1)? Is it possible to build a product-level stream media server based on ffmpeg + ffserver? Regards, Jie From sdbhabal at gmail.com Fri May 6 08:56:32 2011 From: sdbhabal at gmail.com (santosh bhabal) Date: Fri, 6 May 2011 12:26:32 +0530 Subject: [FFmpeg-user] ffmpeg on windows Message-ID: Hi All, I just wanted to know how to get H.264 as a video codec & AAC as a audio codec of my output stream. What will be the ffmpeg command & what parameter i need to use in ffserver.conf Any advice will be appreciated. Thanks in advance. Regards Santosh From funnylookinhat at gmail.com Fri May 6 15:07:53 2011 From: funnylookinhat at gmail.com (David Overcash) Date: Fri, 6 May 2011 07:07:53 -0600 Subject: [FFmpeg-user] ffmpeg and x264 In-Reply-To: References: <61386.97286.qm@web86408.mail.ird.yahoo.com> <237821.66460.qm@web86407.mail.ird.yahoo.com> Message-ID: > > > "The main issue is that I can get really great quality movie which is > larger and it seems not it's not working in all platforms (Mac, iPad, ...) > or the quality is really terrible." > > That probably didn't make sense to you because it doesn't make any sense at all... You mean you can't find a codec that works on all computers? -David From cs_palkar at yahoo.com Fri May 6 15:29:54 2011 From: cs_palkar at yahoo.com (Charu Palkar) Date: Fri, 6 May 2011 06:29:54 -0700 (PDT) Subject: [FFmpeg-user] Question In-Reply-To: References: Message-ID: <386740.64003.qm@web65610.mail.ac4.yahoo.com> How do I convert a .mp4 file with? x264 video and AAC audio encoding to? x264 video and AC3 audio encoding ? ? When I try it it gives me an error of wrong bit-rate. ? Thanx ? Charu Palkar -- From rodney.baker at iinet.net.au Fri May 6 15:59:25 2011 From: rodney.baker at iinet.net.au (Rodney Baker) Date: Fri, 6 May 2011 23:29:25 +0930 Subject: [FFmpeg-user] avi file "operation not permitted" In-Reply-To: References: Message-ID: <201105062329.25103.rodney.baker@iinet.net.au> On Fri, 6 May 2011 06:48:29 Jon Drukman wrote: > /usr/local/bin/ffmpeg -loglevel quiet -v 0 -i list.avi -s 480x352 -an -pass > 1 -vcodec libx264 -vpre fast_firstpass -b 300k -bt 300k -threads 0 > list.avi.mp4 > > ffmpeg version UNKNOWN, Copyright (c) 2000-2011 the FFmpeg developers > built on May 5 2011 11:01:27 with gcc 4.1.2 20080704 (Red Hat 4.1.2-48) > configuration: --enable-gpl --enable-libfaac --enable-libmp3lame > --enable-libx264 --enable-pthreads --enable-static --disable-shared > --disable-network --enable-nonfree > libavutil 51. 2. 0 / 51. 2. 0 > libavcodec 53. 3. 0 / 53. 3. 0 > libavformat 53. 0. 3 / 53. 0. 3 > libavdevice 53. 0. 0 / 53. 0. 0 > libavfilter 2. 4. 0 / 2. 4. 0 > libswscale 0. 14. 0 / 0. 14. 0 > list.avi: Operation not permitted > > What does it mean? I built ffmpeg from source this morning using last > night's tarball/snapshot. > It probably means you either don't have permission to read list.avi (or it doesn't exist in the directory where you're running ffmpeg) or perhaps to write list.avi.mp4. Make sure you're in the correct directory and that you have the appropriate file permissions. Also, make sure you didn't compile ffmpeg as root. Always configure and compile as a normal user, then switch to root to do "make install". -- =================================================== Rodney Baker VK5ZTV rodney.baker at iinet.net.au =================================================== From maarten at webersg.nl Fri May 6 16:30:19 2011 From: maarten at webersg.nl (Maarten Weber) Date: Fri, 6 May 2011 16:30:19 +0200 Subject: [FFmpeg-user] Trimming an FLV file Message-ID: <4E121F2089414EF5A2CAC8AD647C3437@MaartenPC> Hi all, I'm new to this mailinglist and I've got an issue on my hands concerning the trim function of FFmpeg. I must note that I'm using the FFmpeg version which is built in with Xuggler. The main problem is that I can use the -ss parameter to crop an FLV file from the beginning, but using the -t parameter gives no result. I get the same FLV size. As I'm writing this I'm wondering if the -acodec copy and vcodec copy can have something to do with the -t not taking any effect. Though this shouldn't prevent the -t from working does? So, I'm using ffmpeg as follows: ffmpeg -i test.flv -acodec copy -vcodec copy -t 2 test_trimmed.flv Does anyone have an idea as to why the -t isn't working? Am I doing something wrong here? Thanks in advance! Maarten From seandarcy2 at gmail.com Fri May 6 18:04:31 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Fri, 06 May 2011 12:04:31 -0400 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC335D8.5080504@gmail.com> References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> Message-ID: On 05/05/2011 07:42 PM, Baptiste Coudurier wrote: > On 05/05/2011 04:34 PM, sean darcy wrote: >> On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: >>> Hi, >>> >>> On 05/05/2011 01:19 PM, sean darcy wrote: >>>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>>> >>>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>>> >>>> If I just leave it like that, is out.mp4 interlaced or progressive? >>> >>> progressive. By default encoding is progressive. >>> >>>> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 >>>> >>>> Here I assume out.mp4 is progressive. The ffmpeg documentation says: >>>> >>>> "The alternative is to deinterlace the input stream with `-deinterlace', >>>> but deinterlacing introduces losses." >>> >>> Correct, it is progressive. Use -vf yadif instead of -deinterlace >>> Deinterlacing may be a bit destructive, especially if the input is _not_ >>> interlaced. >>> >>>> [...] >>>> >>>> Given this note about losses, am I right we should never deinterlace? >>>> Almost never? When is deinterlacing required/better? >>> >>> You have options: >>> if the receiving end playback interlaced (CRT tv): >>> encode interlaced >>> else if the receiver is going to deinterlace if the file is marked as >>> interlaced and you trust this deinterlacer, then you may encode >>> interlaced (deinterlacing will take cpu time) >>> otherwise you should deinterlace yourself using a good deinterlacer. >> >> Right. I knew that! Just passed right out of my mind, though. >>> >>> I suggest always deinterlace using -vf yadif if the source content is >>> interlaced >>> >> >> Well I found http://guru.multimedia.cx/deinterlacing-filters/ >> >> so I thought I'd try: >> >> -vf "yadif=3:0,mp=mcdeint=2:0:10" >> >> (dv is bottom-field first, right?) >> >> That generates a lot of perplexing output: >> >> [snow @ 0x230df40] pass:4mv changed:1384 >> [snow @ 0x230df40] pass:0 changed:1083 >> [snow @ 0x230df40] pass:1 changed:407 >> [snow @ 0x230df40] pass:2 changed:147 >> [snow @ 0x230df40] pass:3 changed:50 >> [snow @ 0x230df40] pass:4 changed:17 >> [snow @ 0x230df40] pass:5 changed:11 >> [snow @ 0x230df40] pass:6 changed:2 >> [snow @ 0x230df40] pass:7 changed:1 >> [snow @ 0x230df40] pass:8 changed:1 >> [snow @ 0x230df40] pass:9 changed:1 >> [snow @ 0x230df40] pass:10 changed:1 >> [snow @ 0x230df40] pass:11 changed:2 >> [snow @ 0x230df40] pass:12 changed:2 >> [snow @ 0x230df40] pass:13 changed:2 >> [snow @ 0x230df40] pass:14 changed:0 >> [snow @ 0x230df40] pass:4mv changed:1864 > > This is some debug messages, ignore them. > >> I realize the filter comparison is from five years ago, and yadif may >> have changed significantly since then. Does mcdeint still add anything >> to yadif? > > I think nothing has changed much since then :) > Now I've tried yadif=1:0. As I understand it, this is "bob" deinterlacing - field doubling (each field becomes a frame) - with spatial and temporal weaving. But the output is strange: [yadif @ 0xfbd9c0] mode:1 parity:0 ......... frame=38981 fps= 15 q=-1.0 Lsize= 624133kB time=1300.60 bitrate=3931.2kbits/s dup=0 drop=38979 There's a "drop" for each input frame. I'd understand this for yadif=0, where (as I understand it) 2 fields are combined into 1 frame. But yadif=0 shows _no_ drops. sean From belcampo at zonnet.nl Fri May 6 18:42:53 2011 From: belcampo at zonnet.nl (belcampo) Date: Fri, 06 May 2011 18:42:53 +0200 Subject: [FFmpeg-user] Question In-Reply-To: <386740.64003.qm@web65610.mail.ac4.yahoo.com> References: <386740.64003.qm@web65610.mail.ac4.yahoo.com> Message-ID: <4DC4250D.60602@zonnet.nl> On 05/06/11 15:29, Charu Palkar wrote: > How do I convert a .mp4 file with x264 video and AAC audio encoding to x264 > video and AC3 audio encoding ? > > When I try it it gives me an error of wrong bit-rate. As always give us the full uncut command and the resulting error message. So we don't have to guess what verion of ffmpeg you are using with which parameters etc etc. > > Thanx > > Charu Palkar > -- > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From cs_palkar at yahoo.com Fri May 6 18:45:05 2011 From: cs_palkar at yahoo.com (Charu Palkar) Date: Fri, 6 May 2011 09:45:05 -0700 (PDT) Subject: [FFmpeg-user] Question In-Reply-To: <4DC4250D.60602@zonnet.nl> References: <386740.64003.qm@web65610.mail.ac4.yahoo.com> <4DC4250D.60602@zonnet.nl> Message-ID: <889011.57950.qm@web65613.mail.ac4.yahoo.com> I will post the details once I get home. Sorry Charu -- ----- Original Message ---- From: belcampo To: FFmpeg user questions and RTFMs Sent: Fri, May 6, 2011 12:42:53 PM Subject: Re: [FFmpeg-user] Question On 05/06/11 15:29, Charu Palkar wrote: > How do I convert a .mp4 file with? x264 video and AAC audio encoding to? x264 > video and AC3 audio encoding ? > > When I try it it gives me an error of wrong bit-rate. As always give us the full uncut command and the resulting error message. So we don't have to guess what verion of ffmpeg you are using with which parameters etc etc. > > Thanx > > Charu Palkar > -- > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user _______________________________________________ ffmpeg-user mailing list ffmpeg-user at ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user From belcampo at zonnet.nl Fri May 6 18:49:09 2011 From: belcampo at zonnet.nl (belcampo) Date: Fri, 06 May 2011 18:49:09 +0200 Subject: [FFmpeg-user] Trimming an FLV file In-Reply-To: <4E121F2089414EF5A2CAC8AD647C3437@MaartenPC> References: <4E121F2089414EF5A2CAC8AD647C3437@MaartenPC> Message-ID: <4DC42685.3000004@zonnet.nl> On 05/06/11 16:30, Maarten Weber wrote: > Hi all, > > I'm new to this mailinglist and I've got an issue on my hands concerning the trim function of FFmpeg. I must note that I'm using the FFmpeg version which is built in with Xuggler. > > The main problem is that I can use the -ss parameter to crop an FLV file from the beginning, but using the -t parameter gives no result. I get the same FLV size. > > As I'm writing this I'm wondering if the -acodec copy and vcodec copy can have something to do with the -t not taking any effect. Though this shouldn't prevent the -t from working does? > > So, I'm using ffmpeg as follows: > > ffmpeg -i test.flv -acodec copy -vcodec copy -t 2 test_trimmed.flv Put the -t 2 before -i or as an alternative -vframes 50 if it is a 25fps source. > > > Does anyone have an idea as to why the -t isn't working? Am I doing something wrong here? > > Thanks in advance! > > Maarten > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From seandarcy2 at gmail.com Fri May 6 19:02:40 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Fri, 06 May 2011 13:02:40 -0400 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> Message-ID: On 05/06/2011 12:04 PM, sean darcy wrote: > On 05/05/2011 07:42 PM, Baptiste Coudurier wrote: >> On 05/05/2011 04:34 PM, sean darcy wrote: >>> On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: >>>> Hi, >>>> >>>> On 05/05/2011 01:19 PM, sean darcy wrote: >>>>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>>>> >>>>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>>>> >>>>> If I just leave it like that, is out.mp4 interlaced or progressive? >>>> >>>> progressive. By default encoding is progressive. >>>> >>>>> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 >>>>> >>>>> Here I assume out.mp4 is progressive. The ffmpeg documentation says: >>>>> >>>>> "The alternative is to deinterlace the input stream with >>>>> `-deinterlace', >>>>> but deinterlacing introduces losses." >>>> >>>> Correct, it is progressive. Use -vf yadif instead of -deinterlace >>>> Deinterlacing may be a bit destructive, especially if the input is >>>> _not_ >>>> interlaced. >>>> >>>>> [...] >>>>> >>>>> Given this note about losses, am I right we should never deinterlace? >>>>> Almost never? When is deinterlacing required/better? >>>> >>>> You have options: >>>> if the receiving end playback interlaced (CRT tv): >>>> encode interlaced >>>> else if the receiver is going to deinterlace if the file is marked as >>>> interlaced and you trust this deinterlacer, then you may encode >>>> interlaced (deinterlacing will take cpu time) >>>> otherwise you should deinterlace yourself using a good deinterlacer. >>> >>> Right. I knew that! Just passed right out of my mind, though. >>>> >>>> I suggest always deinterlace using -vf yadif if the source content is >>>> interlaced >>>> >>> >>> Well I found http://guru.multimedia.cx/deinterlacing-filters/ >>> >>> so I thought I'd try: >>> >>> -vf "yadif=3:0,mp=mcdeint=2:0:10" >>> >>> (dv is bottom-field first, right?) >>> >>> That generates a lot of perplexing output: >>> >>> [snow @ 0x230df40] pass:4mv changed:1384 >>> [snow @ 0x230df40] pass:0 changed:1083 >>> [snow @ 0x230df40] pass:1 changed:407 >>> [snow @ 0x230df40] pass:2 changed:147 >>> [snow @ 0x230df40] pass:3 changed:50 >>> [snow @ 0x230df40] pass:4 changed:17 >>> [snow @ 0x230df40] pass:5 changed:11 >>> [snow @ 0x230df40] pass:6 changed:2 >>> [snow @ 0x230df40] pass:7 changed:1 >>> [snow @ 0x230df40] pass:8 changed:1 >>> [snow @ 0x230df40] pass:9 changed:1 >>> [snow @ 0x230df40] pass:10 changed:1 >>> [snow @ 0x230df40] pass:11 changed:2 >>> [snow @ 0x230df40] pass:12 changed:2 >>> [snow @ 0x230df40] pass:13 changed:2 >>> [snow @ 0x230df40] pass:14 changed:0 >>> [snow @ 0x230df40] pass:4mv changed:1864 >> >> This is some debug messages, ignore them. >> >>> I realize the filter comparison is from five years ago, and yadif may >>> have changed significantly since then. Does mcdeint still add anything >>> to yadif? >> >> I think nothing has changed much since then :) >> > > Now I've tried yadif=1:0. As I understand it, this is "bob" > deinterlacing - field doubling (each field becomes a frame) - with > spatial and temporal weaving. > > But the output is strange: > > [yadif @ 0xfbd9c0] mode:1 parity:0 > ......... > frame=38981 fps= 15 q=-1.0 Lsize= 624133kB time=1300.60 > bitrate=3931.2kbits/s dup=0 drop=38979 > > There's a "drop" for each input frame. I'd understand this for yadif=0, > where (as I understand it) 2 fields are combined into 1 frame. But > yadif=0 shows _no_ drops. > > sean Ran it with yadif=0: [yadif @ 0x1d359c0] mode:0 parity:0 ........ frame=38980 fps= 16 q=-1.0 Lsize= 622885kB time=1300.57 bitrate=3923.4kbits/s s video:622275kB audio:0kB global headers:0kB muxing overhead 0.097974% frame I:166 Avg QP:16.67 size: 50398 No drops. And the resulting file size is approximately the same. But shouldn't the yadif=0 file be ~1/2 the size of the yadif=1 file? That is, 2 fields are becoming 1 frame, so 1/2 the number of frames. Or is x264 just compressing the related "bob" frames so effectively? Or am I misunderstanding this entirely? But I still don't get why yadif=1 drops a frame for each input frame. sean From jsd at cluttered.com Fri May 6 19:20:12 2011 From: jsd at cluttered.com (Jon Drukman) Date: Fri, 6 May 2011 17:20:12 +0000 (UTC) Subject: [FFmpeg-user] avi file "operation not permitted" References: <201105062329.25103.rodney.baker@iinet.net.au> Message-ID: Rodney Baker iinet.net.au> writes: > > On Fri, 6 May 2011 06:48:29 Jon Drukman wrote: > > /usr/local/bin/ffmpeg -loglevel quiet -v 0 -i list.avi -s 480x352 -an -pass > > 1 -vcodec libx264 -vpre fast_firstpass -b 300k -bt 300k -threads 0 > > list.avi.mp4 > > > > ffmpeg version UNKNOWN, Copyright (c) 2000-2011 the FFmpeg developers > > built on May 5 2011 11:01:27 with gcc 4.1.2 20080704 (Red Hat 4.1.2-48) > > configuration: --enable-gpl --enable-libfaac --enable-libmp3lame > > --enable-libx264 --enable-pthreads --enable-static --disable-shared > > --disable-network --enable-nonfree > > libavutil 51. 2. 0 / 51. 2. 0 > > libavcodec 53. 3. 0 / 53. 3. 0 > > libavformat 53. 0. 3 / 53. 0. 3 > > libavdevice 53. 0. 0 / 53. 0. 0 > > libavfilter 2. 4. 0 / 2. 4. 0 > > libswscale 0. 14. 0 / 0. 14. 0 > > list.avi: Operation not permitted > > > > What does it mean? I built ffmpeg from source this morning using last > > night's tarball/snapshot. > > > > It probably means you either don't have permission to read list.avi (or it > doesn't exist in the directory where you're running ffmpeg) or perhaps to > write list.avi.mp4. Make sure you're in the correct directory and that you > have the appropriate file permissions. > > Also, make sure you didn't compile ffmpeg as root. Always configure and > compile as a normal user, then switch to root to do "make install". It's not a permissions problem: [jsds-macbook:/tmp/foo] jsd% ls -l total 194440 -rw-r--r--@ 1 jsd staff 99548126 May 5 14:01 list.avi [jsds-macbook:/tmp/foo] jsd% touch list.avi [jsds-macbook:/tmp/foo] jsd% /usr/local/bin/ffmpeg -loglevel quiet -v 0 -i list.avi -s 480x352 -an -pass 1 -vcodec libx264 -vpre fast_firstpass -b 300k -bt 300k -threads 0 list.avi.mp4 ffmpeg version UNKNOWN, Copyright (c) 2000-2011 the FFmpeg developers built on Apr 25 2011 11:47:40 with gcc 4.2.1 (Apple Inc. build 5666) (dot 3) configuration: --enable-pthreads --enable-libfaac --enable-libmp3lame --enable-libx264 --enable-gpl --enable-nonfree --disable-network --enable-libxvid --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-version3 libavutil 51. 0. 0 / 51. 0. 0 libavcodec 53. 1. 0 / 53. 1. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 0. 0 / 2. 0. 0 libswscale 0. 13. 0 / 0. 13. 0 list.avi: Operation not permitted [jsds-macbook:/tmp/foo] jsd% touch list.avi.mp4 [jsds-macbook:/tmp/foo] jsd% ls -l total 194440 -rw-r--r--@ 1 jsd staff 99548126 May 6 10:16 list.avi -rw-r--r-- 1 jsd wheel 0 May 6 10:16 list.avi.mp4 [jsds-macbook:/tmp/foo] jsd% I compiled ffmpeg as jsd and installed with sudo, although I don't know why that would make a difference. I've been using unix for almost 20 years now and that's the first time I've ever heard someone say that. BTW, same result with apr 25 ffmpeg on mac os and yesterday's ffmpeg on centos linux... so it's not likely to be OS-related. From baptiste.coudurier at gmail.com Fri May 6 19:27:08 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Fri, 06 May 2011 10:27:08 -0700 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> Message-ID: <4DC42F6C.9060207@gmail.com> On 05/06/2011 10:02 AM, sean darcy wrote: > On 05/06/2011 12:04 PM, sean darcy wrote: >> On 05/05/2011 07:42 PM, Baptiste Coudurier wrote: >>> On 05/05/2011 04:34 PM, sean darcy wrote: >>>> On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: >>>>> Hi, >>>>> >>>>> On 05/05/2011 01:19 PM, sean darcy wrote: >>>>>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>>>>> >>>>>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>>>>> >>>>>> If I just leave it like that, is out.mp4 interlaced or progressive? >>>>> >>>>> progressive. By default encoding is progressive. >>>>> >>>>>> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 >>>>>> >>>>>> Here I assume out.mp4 is progressive. The ffmpeg documentation says: >>>>>> >>>>>> "The alternative is to deinterlace the input stream with >>>>>> `-deinterlace', >>>>>> but deinterlacing introduces losses." >>>>> >>>>> Correct, it is progressive. Use -vf yadif instead of -deinterlace >>>>> Deinterlacing may be a bit destructive, especially if the input is >>>>> _not_ >>>>> interlaced. >>>>> >>>>>> [...] >>>>>> >>>>>> Given this note about losses, am I right we should never deinterlace? >>>>>> Almost never? When is deinterlacing required/better? >>>>> >>>>> You have options: >>>>> if the receiving end playback interlaced (CRT tv): >>>>> encode interlaced >>>>> else if the receiver is going to deinterlace if the file is marked as >>>>> interlaced and you trust this deinterlacer, then you may encode >>>>> interlaced (deinterlacing will take cpu time) >>>>> otherwise you should deinterlace yourself using a good deinterlacer. >>>> >>>> Right. I knew that! Just passed right out of my mind, though. >>>>> >>>>> I suggest always deinterlace using -vf yadif if the source content is >>>>> interlaced >>>>> >>>> >>>> Well I found http://guru.multimedia.cx/deinterlacing-filters/ >>>> >>>> so I thought I'd try: >>>> >>>> -vf "yadif=3:0,mp=mcdeint=2:0:10" >>>> >>>> (dv is bottom-field first, right?) >>>> >>>> That generates a lot of perplexing output: >>>> >>>> [snow @ 0x230df40] pass:4mv changed:1384 >>>> [snow @ 0x230df40] pass:0 changed:1083 >>>> [snow @ 0x230df40] pass:1 changed:407 >>>> [snow @ 0x230df40] pass:2 changed:147 >>>> [snow @ 0x230df40] pass:3 changed:50 >>>> [snow @ 0x230df40] pass:4 changed:17 >>>> [snow @ 0x230df40] pass:5 changed:11 >>>> [snow @ 0x230df40] pass:6 changed:2 >>>> [snow @ 0x230df40] pass:7 changed:1 >>>> [snow @ 0x230df40] pass:8 changed:1 >>>> [snow @ 0x230df40] pass:9 changed:1 >>>> [snow @ 0x230df40] pass:10 changed:1 >>>> [snow @ 0x230df40] pass:11 changed:2 >>>> [snow @ 0x230df40] pass:12 changed:2 >>>> [snow @ 0x230df40] pass:13 changed:2 >>>> [snow @ 0x230df40] pass:14 changed:0 >>>> [snow @ 0x230df40] pass:4mv changed:1864 >>> >>> This is some debug messages, ignore them. >>> >>>> I realize the filter comparison is from five years ago, and yadif may >>>> have changed significantly since then. Does mcdeint still add anything >>>> to yadif? >>> >>> I think nothing has changed much since then :) >>> >> >> Now I've tried yadif=1:0. As I understand it, this is "bob" >> deinterlacing - field doubling (each field becomes a frame) - with >> spatial and temporal weaving. >> >> But the output is strange: >> >> [yadif @ 0xfbd9c0] mode:1 parity:0 >> ......... >> frame=38981 fps= 15 q=-1.0 Lsize= 624133kB time=1300.60 >> bitrate=3931.2kbits/s dup=0 drop=38979 >> >> There's a "drop" for each input frame. I'd understand this for yadif=0, >> where (as I understand it) 2 fields are combined into 1 frame. But >> yadif=0 shows _no_ drops. >> >> sean > > Ran it with yadif=0: > > [yadif @ 0x1d359c0] mode:0 parity:0 > ........ > frame=38980 fps= 16 q=-1.0 Lsize= 622885kB time=1300.57 > bitrate=3923.4kbits/s s > video:622275kB audio:0kB global headers:0kB muxing overhead 0.097974% > frame I:166 Avg QP:16.67 size: 50398 > > No drops. And the resulting file size is approximately the same. But > shouldn't the yadif=0 file be ~1/2 the size of the yadif=1 file? That > is, 2 fields are becoming 1 frame, so 1/2 the number of frames. Or is > x264 just compressing the related "bob" frames so effectively? Or am I > misunderstanding this entirely? > > But I still don't get why yadif=1 drops a frame for each input frame. No, when using mode 1, please read the documentation: * 1: send 1 frame for each field You are outputting 2 frames for one field. If you want no drop you need to double the frame rate. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From bahamutzero8825 at gmail.com Fri May 6 20:07:16 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Fri, 06 May 2011 13:07:16 -0500 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: <4DC30A66.3040006@gmail.com> Message-ID: <4DC438D4.10705@gmail.com> On 2011.05.05 06:34 PM, sean darcy wrote: > Does mcdeint still add anything > to yadif? Yes, but unless you have a really powerful CPU, it's not worth the extra time that it adds. It's very slow. In any case, using Yadif to deinterlace is almost always a very good idea. Interlaced displays are reaching their last days, and encoders do a much better job with progressive input. From h.reindl at thelounge.net Fri May 6 20:09:03 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Fri, 06 May 2011 20:09:03 +0200 Subject: [FFmpeg-user] ffmpeg 0.6.3 vs. g1caa412/20110331 Message-ID: <4DC4393F.1020000@thelounge.net> hi we are using snapshot 20110331 / git g1caa412 packed as rpm for fedora 13/14, it seems that this was one of the latest before the ABI-change to 0.7 now i found out that rpmfusion has 0.6.3 in updates-testing dated with 2011-04-26, but this version does not know the option "--enable-libfreetype" are the security-fixes from 0.6.3 included in g1caa412 or was there some security-backport after this snapshot for 0.6.3 and if there backports how would i include them in our snapshot? thanks and best wishes from austria -- Reindl Harald the lounge interactive design GmbH A-1060 Vienna, Hofm?hlgasse 17 CTO / software-development / cms-solutions p: +43 (1) 595 3999 33, m: +43 (676) 40 221 40 icq: 154546673, http://www.thelounge.net/ http://www.thelounge.net/signature.asc.what.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From seandarcy2 at gmail.com Fri May 6 20:35:03 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Fri, 06 May 2011 14:35:03 -0400 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC42F6C.9060207@gmail.com> References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> Message-ID: On 05/06/2011 01:27 PM, Baptiste Coudurier wrote: > On 05/06/2011 10:02 AM, sean darcy wrote: >> On 05/06/2011 12:04 PM, sean darcy wrote: >>> On 05/05/2011 07:42 PM, Baptiste Coudurier wrote: >>>> On 05/05/2011 04:34 PM, sean darcy wrote: >>>>> On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: >>>>>> Hi, >>>>>> >>>>>> On 05/05/2011 01:19 PM, sean darcy wrote: >>>>>>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>>>>>> >>>>>>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>>>>>> >>>>>>> If I just leave it like that, is out.mp4 interlaced or progressive? >>>>>> >>>>>> progressive. By default encoding is progressive. >>>>>> >>>>>>> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace out.mp4 >>>>>>> >>>>>>> Here I assume out.mp4 is progressive. The ffmpeg documentation says: >>>>>>> >>>>>>> "The alternative is to deinterlace the input stream with >>>>>>> `-deinterlace', >>>>>>> but deinterlacing introduces losses." >>>>>> >>>>>> Correct, it is progressive. Use -vf yadif instead of -deinterlace >>>>>> Deinterlacing may be a bit destructive, especially if the input is >>>>>> _not_ >>>>>> interlaced. >>>>>> >>>>>>> [...] >>>>>>> >>>>>>> Given this note about losses, am I right we should never deinterlace? >>>>>>> Almost never? When is deinterlacing required/better? >>>>>> >>>>>> You have options: >>>>>> if the receiving end playback interlaced (CRT tv): >>>>>> encode interlaced >>>>>> else if the receiver is going to deinterlace if the file is marked as >>>>>> interlaced and you trust this deinterlacer, then you may encode >>>>>> interlaced (deinterlacing will take cpu time) >>>>>> otherwise you should deinterlace yourself using a good deinterlacer. >>>>> >>>>> Right. I knew that! Just passed right out of my mind, though. >>>>>> >>>>>> I suggest always deinterlace using -vf yadif if the source content is >>>>>> interlaced >>>>>> >>>>> >>>>> Well I found http://guru.multimedia.cx/deinterlacing-filters/ >>>>> >>>>> so I thought I'd try: >>>>> >>>>> -vf "yadif=3:0,mp=mcdeint=2:0:10" >>>>> >>>>> (dv is bottom-field first, right?) >>>>> >>>>> That generates a lot of perplexing output: >>>>> >>>>> [snow @ 0x230df40] pass:4mv changed:1384 >>>>> [snow @ 0x230df40] pass:0 changed:1083 >>>>> [snow @ 0x230df40] pass:1 changed:407 >>>>> [snow @ 0x230df40] pass:2 changed:147 >>>>> [snow @ 0x230df40] pass:3 changed:50 >>>>> [snow @ 0x230df40] pass:4 changed:17 >>>>> [snow @ 0x230df40] pass:5 changed:11 >>>>> [snow @ 0x230df40] pass:6 changed:2 >>>>> [snow @ 0x230df40] pass:7 changed:1 >>>>> [snow @ 0x230df40] pass:8 changed:1 >>>>> [snow @ 0x230df40] pass:9 changed:1 >>>>> [snow @ 0x230df40] pass:10 changed:1 >>>>> [snow @ 0x230df40] pass:11 changed:2 >>>>> [snow @ 0x230df40] pass:12 changed:2 >>>>> [snow @ 0x230df40] pass:13 changed:2 >>>>> [snow @ 0x230df40] pass:14 changed:0 >>>>> [snow @ 0x230df40] pass:4mv changed:1864 >>>> >>>> This is some debug messages, ignore them. >>>> >>>>> I realize the filter comparison is from five years ago, and yadif may >>>>> have changed significantly since then. Does mcdeint still add anything >>>>> to yadif? >>>> >>>> I think nothing has changed much since then :) >>>> >>> >>> Now I've tried yadif=1:0. As I understand it, this is "bob" >>> deinterlacing - field doubling (each field becomes a frame) - with >>> spatial and temporal weaving. >>> >>> But the output is strange: >>> >>> [yadif @ 0xfbd9c0] mode:1 parity:0 >>> ......... >>> frame=38981 fps= 15 q=-1.0 Lsize= 624133kB time=1300.60 >>> bitrate=3931.2kbits/s dup=0 drop=38979 >>> >>> There's a "drop" for each input frame. I'd understand this for yadif=0, >>> where (as I understand it) 2 fields are combined into 1 frame. But >>> yadif=0 shows _no_ drops. >>> >>> sean >> >> Ran it with yadif=0: >> >> [yadif @ 0x1d359c0] mode:0 parity:0 >> ........ >> frame=38980 fps= 16 q=-1.0 Lsize= 622885kB time=1300.57 >> bitrate=3923.4kbits/s s >> video:622275kB audio:0kB global headers:0kB muxing overhead 0.097974% >> frame I:166 Avg QP:16.67 size: 50398 >> >> No drops. And the resulting file size is approximately the same. But >> shouldn't the yadif=0 file be ~1/2 the size of the yadif=1 file? That >> is, 2 fields are becoming 1 frame, so 1/2 the number of frames. Or is >> x264 just compressing the related "bob" frames so effectively? Or am I >> misunderstanding this entirely? >> >> But I still don't get why yadif=1 drops a frame for each input frame. > > No, when using mode 1, please read the documentation: > * 1: send 1 frame for each field > > You are outputting 2 frames for one field. If you want no drop you need > to double the frame rate. > Lost. let me go back to basics. I've got an interlaced input with 38980 "frames". But each of these frames is of 2 fields - each half the size of a progressive frame. And ~60 (59.94) fields are shown each second. For yadif=0, 2 fields are combined into 1 frame. So with my input, I should get the same number of "frames". The framerate would be to ~30 (29.97) frames per second. For yadif=1, each field is reconstructed into a frame. "send 1 frame for each field" . So I have twice the number of "frames", and each frame is a full size progressive frame. And the framerate should now be ~60?? So if I use yadif=1 with a standard 29.97 frame rate, half the frames are discarded. Which means there's no benefit to yadif=1! You'd need to set -r 59.94, and there'd be few if any players for your clip! Am I getting closer? Why would anyone ever use yadif=1 "bob" deinterlacing? sean From baptiste.coudurier at gmail.com Fri May 6 21:03:25 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Fri, 06 May 2011 12:03:25 -0700 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> Message-ID: <4DC445FD.8050003@gmail.com> On 05/06/2011 11:35 AM, sean darcy wrote: > On 05/06/2011 01:27 PM, Baptiste Coudurier wrote: >> On 05/06/2011 10:02 AM, sean darcy wrote: >>> On 05/06/2011 12:04 PM, sean darcy wrote: >>>> On 05/05/2011 07:42 PM, Baptiste Coudurier wrote: >>>>> On 05/05/2011 04:34 PM, sean darcy wrote: >>>>>> On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: >>>>>>> Hi, >>>>>>> >>>>>>> On 05/05/2011 01:19 PM, sean darcy wrote: >>>>>>>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>>>>>>> >>>>>>>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>>>>>>> >>>>>>>> If I just leave it like that, is out.mp4 interlaced or progressive? >>>>>>> >>>>>>> progressive. By default encoding is progressive. >>>>>>> >>>>>>>> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace >>>>>>>> out.mp4 >>>>>>>> >>>>>>>> Here I assume out.mp4 is progressive. The ffmpeg documentation >>>>>>>> says: >>>>>>>> >>>>>>>> "The alternative is to deinterlace the input stream with >>>>>>>> `-deinterlace', >>>>>>>> but deinterlacing introduces losses." >>>>>>> >>>>>>> Correct, it is progressive. Use -vf yadif instead of -deinterlace >>>>>>> Deinterlacing may be a bit destructive, especially if the input is >>>>>>> _not_ >>>>>>> interlaced. >>>>>>> >>>>>>>> [...] >>>>>>>> >>>>>>>> Given this note about losses, am I right we should never >>>>>>>> deinterlace? >>>>>>>> Almost never? When is deinterlacing required/better? >>>>>>> >>>>>>> You have options: >>>>>>> if the receiving end playback interlaced (CRT tv): >>>>>>> encode interlaced >>>>>>> else if the receiver is going to deinterlace if the file is >>>>>>> marked as >>>>>>> interlaced and you trust this deinterlacer, then you may encode >>>>>>> interlaced (deinterlacing will take cpu time) >>>>>>> otherwise you should deinterlace yourself using a good deinterlacer. >>>>>> >>>>>> Right. I knew that! Just passed right out of my mind, though. >>>>>>> >>>>>>> I suggest always deinterlace using -vf yadif if the source >>>>>>> content is >>>>>>> interlaced >>>>>>> >>>>>> >>>>>> Well I found http://guru.multimedia.cx/deinterlacing-filters/ >>>>>> >>>>>> so I thought I'd try: >>>>>> >>>>>> -vf "yadif=3:0,mp=mcdeint=2:0:10" >>>>>> >>>>>> (dv is bottom-field first, right?) >>>>>> >>>>>> That generates a lot of perplexing output: >>>>>> >>>>>> [snow @ 0x230df40] pass:4mv changed:1384 >>>>>> [snow @ 0x230df40] pass:0 changed:1083 >>>>>> [snow @ 0x230df40] pass:1 changed:407 >>>>>> [snow @ 0x230df40] pass:2 changed:147 >>>>>> [snow @ 0x230df40] pass:3 changed:50 >>>>>> [snow @ 0x230df40] pass:4 changed:17 >>>>>> [snow @ 0x230df40] pass:5 changed:11 >>>>>> [snow @ 0x230df40] pass:6 changed:2 >>>>>> [snow @ 0x230df40] pass:7 changed:1 >>>>>> [snow @ 0x230df40] pass:8 changed:1 >>>>>> [snow @ 0x230df40] pass:9 changed:1 >>>>>> [snow @ 0x230df40] pass:10 changed:1 >>>>>> [snow @ 0x230df40] pass:11 changed:2 >>>>>> [snow @ 0x230df40] pass:12 changed:2 >>>>>> [snow @ 0x230df40] pass:13 changed:2 >>>>>> [snow @ 0x230df40] pass:14 changed:0 >>>>>> [snow @ 0x230df40] pass:4mv changed:1864 >>>>> >>>>> This is some debug messages, ignore them. >>>>> >>>>>> I realize the filter comparison is from five years ago, and yadif may >>>>>> have changed significantly since then. Does mcdeint still add >>>>>> anything >>>>>> to yadif? >>>>> >>>>> I think nothing has changed much since then :) >>>>> >>>> >>>> Now I've tried yadif=1:0. As I understand it, this is "bob" >>>> deinterlacing - field doubling (each field becomes a frame) - with >>>> spatial and temporal weaving. >>>> >>>> But the output is strange: >>>> >>>> [yadif @ 0xfbd9c0] mode:1 parity:0 >>>> ......... >>>> frame=38981 fps= 15 q=-1.0 Lsize= 624133kB time=1300.60 >>>> bitrate=3931.2kbits/s dup=0 drop=38979 >>>> >>>> There's a "drop" for each input frame. I'd understand this for yadif=0, >>>> where (as I understand it) 2 fields are combined into 1 frame. But >>>> yadif=0 shows _no_ drops. >>>> >>>> sean >>> >>> Ran it with yadif=0: >>> >>> [yadif @ 0x1d359c0] mode:0 parity:0 >>> ........ >>> frame=38980 fps= 16 q=-1.0 Lsize= 622885kB time=1300.57 >>> bitrate=3923.4kbits/s s >>> video:622275kB audio:0kB global headers:0kB muxing overhead 0.097974% >>> frame I:166 Avg QP:16.67 size: 50398 >>> >>> No drops. And the resulting file size is approximately the same. But >>> shouldn't the yadif=0 file be ~1/2 the size of the yadif=1 file? That >>> is, 2 fields are becoming 1 frame, so 1/2 the number of frames. Or is >>> x264 just compressing the related "bob" frames so effectively? Or am I >>> misunderstanding this entirely? >>> >>> But I still don't get why yadif=1 drops a frame for each input frame. >> >> No, when using mode 1, please read the documentation: >> * 1: send 1 frame for each field >> >> You are outputting 2 frames for one field. If you want no drop you need >> to double the frame rate. >> > > Lost. let me go back to basics. I've got an interlaced input with 38980 > "frames". But each of these frames is of 2 fields - each half the size > of a progressive frame. And ~60 (59.94) fields are shown each second. > > For yadif=0, 2 fields are combined into 1 frame. So with my input, I > should get the same number of "frames". The framerate would be to ~30 > (29.97) frames per second. > > For yadif=1, each field is reconstructed into a frame. "send 1 frame for > each field" . So I have twice the number of "frames", and each frame is > a full size progressive frame. And the framerate should now be ~60?? > > So if I use yadif=1 with a standard 29.97 frame rate, half the frames > are discarded. Which means there's no benefit to yadif=1! > > You'd need to set -r 59.94, and there'd be few if any players for your > clip! > > Am I getting closer? > > Why would anyone ever use yadif=1 "bob" deinterlacing? You can do 1080i25 to 720p50 for example, but I'm sure there are other usage since the feature is there. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From hallucinet at free.fr Fri May 6 22:04:01 2011 From: hallucinet at free.fr (hallucinet at free.fr) Date: Fri, 6 May 2011 22:04:01 +0200 (CEST) Subject: [FFmpeg-user] Metadata problem from mp3 to ogg Message-ID: <67656660.2102481304712241281.JavaMail.root@zimbra6-e1.priv.proxad.net> Hi everybody, this is my first message. I searched the archive about my problem, actually found answers, but they somehow seem to not apply to my very problem : I cannot pass the metadata from my mp3 files to ogg files. The mp3s appear to have both Id3V1 and Id3V2 metadata, correctly tagged. When I lauch the ffmpeg command, it reads and displays what seem to be the ID3V2 tags of the mp3 file. But I cannot retrieve them in the ogg file. #id3v2 -l Elephant_Rouge-01-Splash.mp3 id3v1 tag info for Elephant_Rouge-01-Splash.mp3: Title : Splash Artist: Elephant Rouge Album : Splash Year: 2009, Genre: Avantgarde (90) Comment: Track: 1 id3v2 tag info for Elephant_Rouge-01-Splash.mp3: TIT2 (Title/songname/content description): Splash TPE1 (Lead performer(s)/Soloist(s)): Elephant Rouge TALB (Album/Movie/Show title): Splash TYER (Year): 2009 TCON (Content type): Avantgarde (90) TRCK (Track number/Position in set): 1 #ffmpeg -i Elephant_Rouge-01-Splash.mp3 -acodec libvorbis -ab 64k -map_meta_data Elephant_Rouge-01-Splash.ogg:Elephant_Rouge-01-Splash.mp3 Elephant_Rouge-01-Splash.ogg FFmpeg version SVN-r22960, Copyright (c) 2000-2010 the FFmpeg developers built on Apr 1 2011 19:24:52 with gcc 4.4.3 configuration: --prefix=/usr --enable-shared --libdir=/usr/lib --shlibdir=/usr/lib --incdir=/usr/include --disable-stripping --enable-postproc --enable-gpl --enable-pthreads --enable-libtheora --enable-libvorbis --disable-encoder=vorbis --enable-x11grab --enable-runtime-cpudetect --enable-libdc1394 --enable-libschroedinger --enable-libmp3lame --enable-libfaad --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-version3 --enable-libx264 libavutil 50.14. 0 / 50.14. 0 libavcodec 52.66. 0 / 52.66. 0 libavformat 52.61. 0 / 52.61. 0 libavdevice 52. 2. 0 / 52. 2. 0 libswscale 0.10. 0 / 0.10. 0 libpostproc 51. 2. 0 / 51. 2. 0 [mp3 @ 0x9709510]max_analyze_duration reached [mp3 @ 0x9709510]Estimating duration from bitrate, this may be inaccurate Input #0, mp3, from 'Elephant_Rouge-01-Splash.mp3': Metadata: TIT2 : Splash TPE1 : Elephant Rouge TALB : Splash TYER : 2009 TCON : Avantgarde TRCK : 1 Duration: 00:02:56.45, start: 0.000000, bitrate: 128 kb/s Stream #0.0: Audio: mp3, 44100 Hz, 2 channels, s16, 128 kb/s Output #0, ogg, to 'Elephant_Rouge-01-Splash.ogg': Metadata: title : Splash artist : Elephant Rouge album : Splash TYER : 2009 genre : Avantgarde TRACKNUMBER : 1 encoder : Lavf52.61.0 Stream #0.0: Audio: libvorbis, 44100 Hz, 2 channels, s16, 64 kb/s Stream mapping: Stream #0.0 -> #0.0 Press [q] to stop encoding [mp3 @ 0x970a6e0]Header missingrate= 76.7kbits/s Error while decoding stream #0.0 size= 1650kB time=176.20 bitrate= 76.7kbits/s video:0kB audio:1378kB global headers:3kB muxing overhead 19.444910% #id3v2 -l Elephant_Rouge-01-Splash.ogg Elephant_Rouge-01-Splash.ogg: No ID3 tag (the file plays fine, BTW) I tried a lot of outfile:infile combinations, only 0:0 produced no errors, but still no tags in the resulting files... :( What is crutial to me are the Song name and track # Any idea of what I'm doing wrong ? Thanks From blacktrash at gmx.net Fri May 6 22:21:53 2011 From: blacktrash at gmx.net (Christian Ebert) Date: Fri, 6 May 2011 22:21:53 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> Message-ID: <20110506202153.GC740@krille.blacktrash.org> * sean darcy on Friday, May 06, 2011 at 14:35:03 -0400 > Why would anyone ever use yadif=1 "bob" deinterlacing? To show a converted PAL source via a projector at 50 fps. The results from yadif=3,mcdeint=2:1:10 are brilliant, really. You just need a few days to transcode ;-) c -- \black\trash movie _SAME TIME SAME PLACE_ New York, in the summer of 2001 --->> http://www.blacktrash.org/underdogma/stsp.php From bahamutzero8825 at gmail.com Fri May 6 22:22:52 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Fri, 06 May 2011 15:22:52 -0500 Subject: [FFmpeg-user] Metadata problem from mp3 to ogg In-Reply-To: <67656660.2102481304712241281.JavaMail.root@zimbra6-e1.priv.proxad.net> References: <67656660.2102481304712241281.JavaMail.root@zimbra6-e1.priv.proxad.net> Message-ID: <4DC4589C.1000102@gmail.com> On 2011.05.06 03:04 PM, hallucinet at free.fr wrote: > Any idea of what I'm doing wrong ? Doesn't OGG store its own type of tags? It could be that FFmpeg stores the metadata in tags of a format other than ID3v2. If you look at the file with MediaInfo or similar, do you see tags? From lou at fakeoutdoorsman.com Fri May 6 23:01:55 2011 From: lou at fakeoutdoorsman.com (Lou) Date: Fri, 6 May 2011 13:01:55 -0800 Subject: [FFmpeg-user] Metadata problem from mp3 to ogg In-Reply-To: <67656660.2102481304712241281.JavaMail.root@zimbra6-e1.priv.proxad.net> References: <67656660.2102481304712241281.JavaMail.root@zimbra6-e1.priv.proxad.net> Message-ID: <20110506130155.4a26ec67@lrcd.com> On Fri, 6 May 2011 22:04:01 +0200 (CEST) hallucinet at free.fr wrote: > Hi everybody, this is my first message. I searched the archive about > my problem, actually found answers, but they somehow seem to not > apply to my very problem : > > I cannot pass the metadata from my mp3 files to ogg files. The mp3s > appear to have both Id3V1 and Id3V2 metadata, correctly tagged. When > I lauch the ffmpeg command, it reads and displays what seem to be the > ID3V2 tags of the mp3 file. But I cannot retrieve them in the ogg > file. > ... > > #id3v2 -l Elephant_Rouge-01-Splash.ogg > Elephant_Rouge-01-Splash.ogg: No ID3 tag > > (the file plays fine, BTW) > > I tried a lot of outfile:infile combinations, only 0:0 produced no > errors, but still no tags in the resulting files... :( > > What is crutial to me are the Song name and track # > > Any idea of what I'm doing wrong ? ID3 is MP3 specific, I think. You can use something like "vorbiscomment -l input.ogg" to view the metadata in your ogg output (or simply "ffmpeg -i input.ogg"). Also, you won't need to use -map_meta_data with a recent FFmpeg because it will automatically attempt to copy the metadata from input to output. From lou at fakeoutdoorsman.com Fri May 6 23:08:46 2011 From: lou at fakeoutdoorsman.com (Lou) Date: Fri, 6 May 2011 13:08:46 -0800 Subject: [FFmpeg-user] Trimming an FLV file In-Reply-To: <4E121F2089414EF5A2CAC8AD647C3437@MaartenPC> References: <4E121F2089414EF5A2CAC8AD647C3437@MaartenPC> Message-ID: <20110506130846.3dffa6bf@lrcd.com> On Fri, 6 May 2011 16:30:19 +0200 Maarten Weber wrote: > Hi all, > > I'm new to this mailinglist and I've got an issue on my hands > concerning the trim function of FFmpeg. I must note that I'm using > the FFmpeg version which is built in with Xuggler. > > The main problem is that I can use the -ss parameter to crop an FLV > file from the beginning, but using the -t parameter gives no result. > I get the same FLV size. > > As I'm writing this I'm wondering if the -acodec copy and vcodec copy > can have something to do with the -t not taking any effect. Though > this shouldn't prevent the -t from working does? > > So, I'm using ffmpeg as follows: > > ffmpeg -i test.flv -acodec copy -vcodec copy -t 2 test_trimmed.flv > > > Does anyone have an idea as to why the -t isn't working? Am I doing > something wrong here? > > Thanks in advance! > > Maarten You should also provide the complete FFmpeg terminal output. This will provide some useful info about your input and your FFmpeg version. There was a bug (issue1712) present in FFmpeg where -t didn't work with some flv and was fixed on Feb 4 2010. From bouke at editb.nl Fri May 6 23:30:24 2011 From: bouke at editb.nl (Bouke) Date: Fri, 6 May 2011 23:30:24 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? References: <4DC30A66.3040006@gmail.com> <4DC438D4.10705@gmail.com> Message-ID: <005d01cc0c34$d2927cb0$4301a8c0@hpkantoor> ----- Original Message ----- From: "Andrew Berg" To: "FFmpeg user questions and RTFMs" Sent: Friday, May 06, 2011 8:07 PM Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? > On 2011.05.05 06:34 PM, sean darcy wrote: >> Does mcdeint still add anything >> to yadif? > Yes, but unless you have a really powerful CPU, it's not worth the extra > time that it adds. It's very slow. > > In any case, using Yadif to deinterlace is almost always a very good > idea. Interlaced displays are reaching their last days, Excuse me? Where did you get this information about interlacing going away? Do you have anything to backup that what you are saying is actually meaningfull? Sorry if i sound harsh (Trying to be so now), but IMHO, this is just plain bullshit. But i love to be prooven wrong. Bouke > and encoders do > a much better job with progressive input. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From gialloporpora at gmail.com Sat May 7 01:25:31 2011 From: gialloporpora at gmail.com (gialloporpora) Date: Sat, 07 May 2011 01:25:31 +0200 Subject: [FFmpeg-user] Differences downloading RA files with FFMpeg and mplayer Message-ID: <4DC4836B.5030701@gmail.com> Dear all, I have noted some differences downloading ra files (via rtsp) using FFMpeg and mplayer, for example: fmpeg -i rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra output1.ra mplayer rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra -dumpstream if I listen the resulting files with FFPlay, they seems to be the same audio stream, but, if I open them with media player cassic (Real Alternative codecs), the file downloaded with FFMpeg have a bad quality. Someone knows the reason? Maybe, seeking problems, I think, but I am not sure. I use Windows XP, but I don't think it is the reason. Thanks Sandro PS: if you want I could upload the resulting output. ffmpeg -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 5153 bytes Desc: S/MIME Cryptographic Signature URL: From james.darnley at gmail.com Sat May 7 10:06:57 2011 From: james.darnley at gmail.com (James Darnley) Date: Sat, 7 May 2011 10:06:57 +0200 Subject: [FFmpeg-user] Metadata problem from mp3 to ogg In-Reply-To: <67656660.2102481304712241281.JavaMail.root@zimbra6-e1.priv.proxad.net> References: <67656660.2102481304712241281.JavaMail.root@zimbra6-e1.priv.proxad.net> Message-ID: On 06/05/2011, hallucinet at free.fr wrote: > FFmpeg version SVN-r22960, Copyright (c) 2000-2010 the FFmpeg developers Your ffmpeg is too old. From mark at mdsh.com Sat May 7 10:10:25 2011 From: mark at mdsh.com (Mark Himsley) Date: Sat, 07 May 2011 09:10:25 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC438D4.10705@gmail.com> References: <4DC30A66.3040006@gmail.com> <4DC438D4.10705@gmail.com> Message-ID: <4DC4FE71.30800@mdsh.com> On 06/05/2011 19:07, Andrew Berg wrote: > Interlaced displays are reaching their last days Not true. -- Mark From cfaf at hotmail.com Sat May 7 15:32:07 2011 From: cfaf at hotmail.com (christian fafard) Date: Sat, 7 May 2011 13:32:07 +0000 Subject: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? Message-ID: Any idea someone?? Baptiste, i know it's your field of expertise, do you have a procedure that works? Maybe the problem is the quicktime container, i don't know if Avid accept AVI for instance. I'm open to any suggestions, it's not vital for me but it's something i would like being able to do. Thanks Christian Fafard From bahamutzero8825 at gmail.com Sat May 7 19:19:10 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Sat, 07 May 2011 12:19:10 -0500 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <005d01cc0c34$d2927cb0$4301a8c0@hpkantoor> References: <4DC30A66.3040006@gmail.com> <4DC438D4.10705@gmail.com> <005d01cc0c34$d2927cb0$4301a8c0@hpkantoor> Message-ID: <4DC57F0E.4010308@gmail.com> On 2011.05.06 04:30 PM, Bouke wrote: > Excuse me? Where did you get this information about interlacing going away? > Do you have anything to backup that what you are saying is actually > meaningfull? > > Sorry if i sound harsh (Trying to be so now), but IMHO, this is just plain > bullshit. The only bullshit is you calling me out on something I didn't say. I never said interlacing is going away. In fact, I see it being around for a long time (most broadcasters want to present sports and other fast-motion content at 50 or 59.94Hz instead of 25 or 29.97Hz and still have a "full HD" picture, and no one seems to be willing to broadcast 1080p50/60). HDTVs (and software media players sending video to an LCD monitor) can deinterlace, so broadcasting interlaced material isn't a huge issue. Pure interlaced displays are going away in the near future, to be replaced by LCD monitors and HDTVs. I'm not saying everyone will have an HDTV in the next year, but they are being replaced. As always, there are going to be special situations where encoding interlaced is appropriate for a non-broadcaster, but in general, progressive is the way to go. From h.reindl at thelounge.net Sat May 7 20:40:56 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Sat, 07 May 2011 20:40:56 +0200 Subject: [FFmpeg-user] vlc-troubles with git-N-29534-g66b1f21 Message-ID: <4DC59238.4040901@thelounge.net> today (after some off-list) explanations (thanks again) i got a in the first moment working ffmpeg with "oldabi" git-N-29534-g66b1f21 encoding works wonderful with several formats but vlc-1.9.1 from rpmfusion (Fedora 14) says there are no decoders for all formats except .vob i can confirm that this ffmpeg-build is the problem because after downgrade to our latest working snapshot from 2011-03-31 the problem went away and also a lot of test-videos encoded with git-N-29534-g66b1f21 playing without any issue i wonder that even a "rpmbuild --rebuild vlc-rpmfusion.src.rpm" did not solve the problem, so is there any known issue? -- Mit besten Gr??en, Reindl Harald the lounge interactive design GmbH A-1060 Vienna, Hofm?hlgasse 17 CTO / software-development / cms-solutions p: +43 (1) 595 3999 33, m: +43 (676) 40 221 40 icq: 154546673, http://www.thelounge.net/ http://www.thelounge.net/signature.asc.what.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From mark at mdsh.com Sat May 7 20:55:18 2011 From: mark at mdsh.com (Mark Himsley) Date: Sat, 07 May 2011 19:55:18 +0100 Subject: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? In-Reply-To: References: Message-ID: <4DC59596.5090505@mdsh.com> On 07/05/2011 14:32, christian fafard wrote: > > Any idea someone?? > > Baptiste, i know it's your field of expertise, do you have a procedure that works? > > Maybe the problem is the quicktime container, i don't know if Avid accept AVI for instance. > > I'm open to any suggestions, it's not vital for me but it's something i would like being able to do. > > Thanks > Christian Fafard What do you mean by "accept"? Have you tried Baptiste's FFmbc: http://code.google.com/p/ffmbc/ -- Mark From PHILLIP.BARNETT at ITN.CO.UK Sat May 7 22:09:40 2011 From: PHILLIP.BARNETT at ITN.CO.UK (Barnett, Phillip) Date: Sat, 7 May 2011 20:09:40 +0000 Subject: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? In-Reply-To: References: Message-ID: <662D22E018CA064A9F0AEF3AC30520FA071E36@MBX1.ITN.LOCAL> Avid needs MXF files. Ffmbc will easily generate D10 files, but then you'll need to wrap them as MXF files, which writeavidmxf will do for you. See these instructions. http://mdsh.com/wiki/jsp/Wiki?writeavidmxf&highlight=d10 Cheers PHILLIP BARNETT SERVER MANAGER 200 GRAY'S INN ROAD LONDON WC1X 8XZ UNITED KINGDOM T +44 (0)20 7430 4474 F E PHILLIP.BARNETT at ITN.CO.UK WWW.ITN.CO.UK Please consider the environment. Do you really need to print this email? -----Original Message----- From: ffmpeg-user-bounces at ffmpeg.org [mailto:ffmpeg-user-bounces at ffmpeg.org] On Behalf Of christian fafard Sent: 07 May 2011 14:32 To: ffmpeg-user at ffmpeg.org Subject: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? Any idea someone?? Baptiste, i know it's your field of expertise, do you have a procedure that works? Maybe the problem is the quicktime container, i don't know if Avid accept AVI for instance. I'm open to any suggestions, it's not vital for me but it's something i would like being able to do. Thanks Christian Fafard _______________________________________________ ffmpeg-user mailing list ffmpeg-user at ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user Please Note: Any views or opinions are solely those of the author and do not necessarily represent those of Independent Television News Limited unless specifically stated. This email and any files attached are confidential and intended solely for the use of the individual or entity to which they are addressed. If you have received this email in error, please notify postmaster at itn.co.uk Please note that to ensure regulatory compliance and for the protection of our clients and business, we may monitor and read messages sent to and from our systems. Thank You. From blatwurst at digitalfish.com Sun May 8 01:16:43 2011 From: blatwurst at digitalfish.com (Blatwurst) Date: Sat, 7 May 2011 16:16:43 -0700 Subject: [FFmpeg-user] Fixing up Stephen Dranger's ffmpeg tutorial code Message-ID: Greetings, I've been working with the source code from Stephen Dranger's excellent FFmpeg tutorial. I've got all of the examples building under C++, and mostly running with the latest version of ffmpeg. In case anyone is interested in my efforts, I've summarized what I'm doing here: http://drangertuts.wordpress.com This page includes a link to a tarball with my versions of the sources. I'm posting this information for three purposes: 1. Allow others to benefit from my work 2. Allow others to tell me what I've done wrong 3. Allow others to help me to fix the remaining problems I'm finding with the code. If you find yourself being one of the "others" to which #2 or #3 apply, please email me at blatwurst -at- digitalfish -dot- com, or respond to this post, or post comments on my blog (the URL above). Have fun! Blat From cool.druker at gmail.com Sun May 8 01:53:46 2011 From: cool.druker at gmail.com (Cool Druker) Date: Sun, 8 May 2011 05:23:46 +0530 Subject: [FFmpeg-user] (no subject) Message-ID: http://adza.gameitis.com/wp-content/themes/motion/mylife.html From jswordtestem at yahoo.co.uk Sun May 8 02:08:10 2011 From: jswordtestem at yahoo.co.uk (phil curb) Date: Sun, 8 May 2011 01:08:10 +0100 (BST) Subject: [FFmpeg-user] problem converting and combining these FLV files Message-ID: <528545.17864.qm@web25905.mail.ukl.yahoo.com> I have this file, i'm trying to combine it with a copy of itself. http://www.sendspace.com/file/dkg3rz But the result while being twice as big, is the same duration. C:\blah>dir Karateka.flv (399,375) I want to create one big file that is 2 of them combined. So I convert the flv to mpg which one can combine with copy /B ffmpeg -i Karateka.flv -sameq Karateka.mpg this has created Karateka.mpg 753,664 I create 2 copies of this. So now 3 in total 753,664 Karateka.mpg 753,664 Karateka2.mpg 753,664 Karateka3.mpg Now I do Copy /B Karateka.mpg+Karateka2.mpg 1,507,328 Karateka.mpg 753,664 Karateka2.mpg 753,664 Karateka3.mpg The problem here is that the mpg that is twice the size, is exactly the same duration as the original. It hasn't combined. From seandarcy2 at gmail.com Sun May 8 03:43:41 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Sat, 07 May 2011 21:43:41 -0400 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC445FD.8050003@gmail.com> References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> <4DC445FD.8050003@gmail.com> Message-ID: On 05/06/2011 03:03 PM, Baptiste Coudurier wrote: > On 05/06/2011 11:35 AM, sean darcy wrote: >> On 05/06/2011 01:27 PM, Baptiste Coudurier wrote: >>> On 05/06/2011 10:02 AM, sean darcy wrote: >>>> On 05/06/2011 12:04 PM, sean darcy wrote: >>>>> On 05/05/2011 07:42 PM, Baptiste Coudurier wrote: >>>>>> On 05/05/2011 04:34 PM, sean darcy wrote: >>>>>>> On 05/05/2011 04:36 PM, Baptiste Coudurier wrote: >>>>>>>> Hi, >>>>>>>> >>>>>>>> On 05/05/2011 01:19 PM, sean darcy wrote: >>>>>>>>> I have an interlaced dv file. I'm transcoding it with x264 to mp4. >>>>>>>>> >>>>>>>>> 1. ffmpeg -i file.dv -an -vcodec libx264 -b out.mp4 >>>>>>>>> >>>>>>>>> If I just leave it like that, is out.mp4 interlaced or progressive? >>>>>>>> >>>>>>>> progressive. By default encoding is progressive. >>>>>>>> >>>>>>>>> 2. ffmpeg -i file.dv -an -vcodec libx264 -b -deinterlace >>>>>>>>> out.mp4 >>>>>>>>> >>>>>>>>> Here I assume out.mp4 is progressive. The ffmpeg documentation >>>>>>>>> says: >>>>>>>>> >>>>>>>>> "The alternative is to deinterlace the input stream with >>>>>>>>> `-deinterlace', >>>>>>>>> but deinterlacing introduces losses." >>>>>>>> >>>>>>>> Correct, it is progressive. Use -vf yadif instead of -deinterlace >>>>>>>> Deinterlacing may be a bit destructive, especially if the input is >>>>>>>> _not_ >>>>>>>> interlaced. >>>>>>>> >>>>>>>>> [...] >>>>>>>>> >>>>>>>>> Given this note about losses, am I right we should never >>>>>>>>> deinterlace? >>>>>>>>> Almost never? When is deinterlacing required/better? >>>>>>>> >>>>>>>> You have options: >>>>>>>> if the receiving end playback interlaced (CRT tv): >>>>>>>> encode interlaced >>>>>>>> else if the receiver is going to deinterlace if the file is >>>>>>>> marked as >>>>>>>> interlaced and you trust this deinterlacer, then you may encode >>>>>>>> interlaced (deinterlacing will take cpu time) >>>>>>>> otherwise you should deinterlace yourself using a good deinterlacer. >>>>>>> >>>>>>> Right. I knew that! Just passed right out of my mind, though. >>>>>>>> >>>>>>>> I suggest always deinterlace using -vf yadif if the source >>>>>>>> content is >>>>>>>> interlaced >>>>>>>> >>>>>>> >>>>>>> Well I found http://guru.multimedia.cx/deinterlacing-filters/ >>>>>>> >>>>>>> so I thought I'd try: >>>>>>> >>>>>>> -vf "yadif=3:0,mp=mcdeint=2:0:10" >>>>>>> >>>>>>> (dv is bottom-field first, right?) >>>>>>> >>>>>>> That generates a lot of perplexing output: >>>>>>> >>>>>>> [snow @ 0x230df40] pass:4mv changed:1384 >>>>>>> [snow @ 0x230df40] pass:0 changed:1083 >>>>>>> [snow @ 0x230df40] pass:1 changed:407 >>>>>>> [snow @ 0x230df40] pass:2 changed:147 >>>>>>> [snow @ 0x230df40] pass:3 changed:50 >>>>>>> [snow @ 0x230df40] pass:4 changed:17 >>>>>>> [snow @ 0x230df40] pass:5 changed:11 >>>>>>> [snow @ 0x230df40] pass:6 changed:2 >>>>>>> [snow @ 0x230df40] pass:7 changed:1 >>>>>>> [snow @ 0x230df40] pass:8 changed:1 >>>>>>> [snow @ 0x230df40] pass:9 changed:1 >>>>>>> [snow @ 0x230df40] pass:10 changed:1 >>>>>>> [snow @ 0x230df40] pass:11 changed:2 >>>>>>> [snow @ 0x230df40] pass:12 changed:2 >>>>>>> [snow @ 0x230df40] pass:13 changed:2 >>>>>>> [snow @ 0x230df40] pass:14 changed:0 >>>>>>> [snow @ 0x230df40] pass:4mv changed:1864 >>>>>> >>>>>> This is some debug messages, ignore them. >>>>>> >>>>>>> I realize the filter comparison is from five years ago, and yadif may >>>>>>> have changed significantly since then. Does mcdeint still add >>>>>>> anything >>>>>>> to yadif? >>>>>> >>>>>> I think nothing has changed much since then :) >>>>>> >>>>> >>>>> Now I've tried yadif=1:0. As I understand it, this is "bob" >>>>> deinterlacing - field doubling (each field becomes a frame) - with >>>>> spatial and temporal weaving. >>>>> >>>>> But the output is strange: >>>>> >>>>> [yadif @ 0xfbd9c0] mode:1 parity:0 >>>>> ......... >>>>> frame=38981 fps= 15 q=-1.0 Lsize= 624133kB time=1300.60 >>>>> bitrate=3931.2kbits/s dup=0 drop=38979 >>>>> >>>>> There's a "drop" for each input frame. I'd understand this for yadif=0, >>>>> where (as I understand it) 2 fields are combined into 1 frame. But >>>>> yadif=0 shows _no_ drops. >>>>> >>>>> sean >>>> >>>> Ran it with yadif=0: >>>> >>>> [yadif @ 0x1d359c0] mode:0 parity:0 >>>> ........ >>>> frame=38980 fps= 16 q=-1.0 Lsize= 622885kB time=1300.57 >>>> bitrate=3923.4kbits/s s >>>> video:622275kB audio:0kB global headers:0kB muxing overhead 0.097974% >>>> frame I:166 Avg QP:16.67 size: 50398 >>>> >>>> No drops. And the resulting file size is approximately the same. But >>>> shouldn't the yadif=0 file be ~1/2 the size of the yadif=1 file? That >>>> is, 2 fields are becoming 1 frame, so 1/2 the number of frames. Or is >>>> x264 just compressing the related "bob" frames so effectively? Or am I >>>> misunderstanding this entirely? >>>> >>>> But I still don't get why yadif=1 drops a frame for each input frame. >>> >>> No, when using mode 1, please read the documentation: >>> * 1: send 1 frame for each field >>> >>> You are outputting 2 frames for one field. If you want no drop you need >>> to double the frame rate. >>> >> >> Lost. let me go back to basics. I've got an interlaced input with 38980 >> "frames". But each of these frames is of 2 fields - each half the size >> of a progressive frame. And ~60 (59.94) fields are shown each second. >> >> For yadif=0, 2 fields are combined into 1 frame. So with my input, I >> should get the same number of "frames". The framerate would be to ~30 >> (29.97) frames per second. >> >> For yadif=1, each field is reconstructed into a frame. "send 1 frame for >> each field" . So I have twice the number of "frames", and each frame is >> a full size progressive frame. And the framerate should now be ~60?? >> >> So if I use yadif=1 with a standard 29.97 frame rate, half the frames >> are discarded. Which means there's no benefit to yadif=1! >> >> You'd need to set -r 59.94, and there'd be few if any players for your >> clip! >> >> Am I getting closer? >> >> Why would anyone ever use yadif=1 "bob" deinterlacing? > > You can do 1080i25 to 720p50 for example, but I'm sure there are other > usage since the feature is there. > Now here's an interesting result, at least to me! The output from: ffmpeg -i $INPUT -an -vcodec libx264 -preset slower -tune film \ -vf "yadif=0:0,hqdn3d" -r 29.97 -crf 17 -threads 0 $1.m4 680394195 May 6 12:47 JulyPlay-crop-yadif0-hqdn3d.mp4 And from: ffmpeg -i $INPUT -an -vcodec libx264 -preset slower -tune film \ -vf "yadif=1:0,hqdn3d" -r 59.94 -crf 17 -threads 0 $1.m4 697669397 May 7 20:57 JulyPlay-crop-yadif1-hqdn3d.mp4 These are 1990 vhs tapes. Since I'm uploading to blip.tv, I'd like the best I can do, even at the expense of encoding time. Therefore the cost (in terms of file size) of encoding with x264 a stream from yadif=1 (with twice the frames of yadif=0) is less than 3% (698meg v. 681meg). If blip.tv can deal with 59.94 framerates, I'm better off using yadif=1, -r=59.94. BTW, all the examples I see have denoising after deinterlacing. Shouldn't that be reversed: -vf="hqdn3d,yadif". Wouldn't it be better to lower the noise before you deinterlace? sean From cfaf at hotmail.com Sun May 8 07:04:01 2011 From: cfaf at hotmail.com (christian fafard) Date: Sun, 8 May 2011 05:04:01 +0000 Subject: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? In-Reply-To: <4DC59596.5090505@mdsh.com> References: , <4DC59596.5090505@mdsh.com> Message-ID: Hi Mark, You're right, the word 'accept' doesn't mean anything in a description of a technical problem. What i meant actually is that Avid import the video without complaining (IMX30 in a quicktime movie encoded with ffmpeg). But the video is a white frame all along. So i took a video that i sure is working, re-wrap to a new quicktime without re-encoding ( -vcodec copy) and it's still the same. So my conclusion is that it has to do with the quicktime header or something. And finally, no i didn't try Baptiste's ffmbc but that's exactly where i'm at right now. I found the source but is it possible to download windows binary somewhere? Thanks Christian > Date: Sat, 7 May 2011 19:55:18 +0100 > From: mark at mdsh.com > To: ffmpeg-user at ffmpeg.org > Subject: Re: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? > > On 07/05/2011 14:32, christian fafard wrote: > > > > Any idea someone?? > > > > Baptiste, i know it's your field of expertise, do you have a procedure that works? > > > > Maybe the problem is the quicktime container, i don't know if Avid accept AVI for instance. > > > > I'm open to any suggestions, it's not vital for me but it's something i would like being able to do. > > > > Thanks > > Christian Fafard > > What do you mean by "accept"? > > Have you tried Baptiste's FFmbc: http://code.google.com/p/ffmbc/ > > -- > Mark > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From bouke at editb.nl Sun May 8 10:43:33 2011 From: bouke at editb.nl (Bouke) Date: Sun, 8 May 2011 10:43:33 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? References: <4DC30A66.3040006@gmail.com> <4DC438D4.10705@gmail.com><005d01cc0c34$d2927cb0$4301a8c0@hpkantoor> <4DC57F0E.4010308@gmail.com> Message-ID: <002401cc0d5c$06d64600$4301a8c0@hpkantoor> ----- Original Message ----- From: "Andrew Berg" > On 2011.05.06 04:30 PM, Bouke wrote: >> Excuse me? Where did you get this information about interlacing going >> away? >> Do you have anything to backup that what you are saying is actually >> meaningfull? >> >> Sorry if i sound harsh (Trying to be so now), but IMHO, this is just >> plain >> bullshit. > The only bullshit is you calling me out on something I didn't say. I > never said interlacing is going away. Andrew, No, you did not, but you did say: > idea. Interlaced displays are reaching their last days, and And this is not the way i see it.. > In fact, I see it being around for > a long time (most broadcasters want to present sports and other > fast-motion content at 50 or 59.94Hz instead of 25 or 29.97Hz and still > have a "full HD" picture, and no one seems to be willing to broadcast > 1080p50/60). HDTVs (and software media players sending video to an LCD > monitor) can deinterlace, so broadcasting interlaced material isn't a > huge issue. Pure interlaced displays are going away in the near future, Yes, if you mean tubes. But an LCD / plasma capable of displaying interlaced video is an interlaced device IMHO. So then, let's agree to disagree. It's way off topic from the original question aynway. Bouke From phil_rhodes at rocketmail.com Sun May 8 11:17:24 2011 From: phil_rhodes at rocketmail.com (Phil Rhodes) Date: Sun, 08 May 2011 10:17:24 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <002401cc0d5c$06d64600$4301a8c0@hpkantoor> References: <4DC30A66.3040006@gmail.com> <4DC438D4.10705@gmail.com> <005d01cc0c34$d2927cb0$4301a8c0@hpkantoor> <4DC57F0E.4010308@gmail.com> <002401cc0d5c$06d64600$4301a8c0@hpkantoor> Message-ID: > Interlaced displays are reaching their last days I haven't been following the thread closely, but be aware: this is not so. Much as everyone from camera crews to software engineers would like things to be otherwise, interlaced material is going nowhere fast. It's often required by broadcasters, and the entire broadcast chain is still interlaced in almost all cases. Once again - this is not about what we would -like- to be true. We would like there never to have been a fractional-framerate NTSC and we would like there to be no interlacing. Unfortunately, out her in reality-world, there are stupid fractional framerates and there is drop-frame timecode and there is interlacing and we just have to deal with it. Video-related software that omits interlacing is simply incomplete. P From betonpfeiler at googlemail.com Sun May 8 12:45:04 2011 From: betonpfeiler at googlemail.com (betonpfeiler) Date: Sun, 08 May 2011 12:45:04 +0200 Subject: [FFmpeg-user] problem converting and combining these FLV files In-Reply-To: <528545.17864.qm@web25905.mail.ukl.yahoo.com> References: <528545.17864.qm@web25905.mail.ukl.yahoo.com> Message-ID: <4DC67430.8010004@googlemail.com> Am 08.05.2011 02:08, schrieb phil curb: > I have this file, i'm trying to combine it with a copy of itself. > http://www.sendspace.com/file/dkg3rz > > But the result while being twice as big, is the same duration. > > C:\blah>dir > Karateka.flv (399,375) > > I want to create one big file that is 2 of them combined. > > So I convert the flv to mpg which one can combine with copy /B > > ffmpeg -i Karateka.flv -sameq Karateka.mpg > Hello, I did the same on a linux system and while just copying I had similar issues. In my case it helped to do a remux with ffmpeg of the combined file with this command: |cat video1.mxf video2.mxf | ffmpeg -i - -acodec copy -vcodec mpeg2video -b 5M whole.mpg |||This is how it worked on Ubuntu, on a windows set-up you might replace 'cat' with you copy-routine. The '-' as input tells ffmpeg to use the cat-merged file as input. This command is explained on ffmbc.wordpress.com greetings, b ||| | From sheen.andy at googlemail.com Sun May 8 13:36:49 2011 From: sheen.andy at googlemail.com (Andy Sheen) Date: Sun, 08 May 2011 12:36:49 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: <4DC30A66.3040006@gmail.com> <4DC438D4.10705@gmail.com> <005d01cc0c34$d2927cb0$4301a8c0@hpkantoor> <4DC57F0E.4010308@gmail.com> <002401cc0d5c$06d64600$4301a8c0@hpkantoor> Message-ID: <4DC68051.3000100@googlemail.com> Phil Rhodes wrote on Sun 08 May at 10:17 UK time >> Interlaced displays are reaching their last days > > I haven't been following the thread closely, but be aware: this is not so. > > Much as everyone from camera crews to software engineers would like > things to be otherwise, interlaced material is going nowhere fast. It's > often required by broadcasters, and the entire broadcast chain is still > interlaced in almost all cases. There is a difference between interlaced displays and interlaced material. All flat panels are progressive display devices that will accept both interlaced and progressive material. The flat panel has to deinterlace and display progressively and how good this is depends entirely on the underlying TV hardware (some do a good job, others don't). Personally, all the stuff I recode from off air gets deinterlaced when recoded because the display chain it will be played on is capable of progressive. From seandarcy2 at gmail.com Sun May 8 15:20:16 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Sun, 08 May 2011 09:20:16 -0400 Subject: [FFmpeg-user] Error on 2-pass x264: timebase mismatch with 1st pass (50/2997 vs 1001/30000) Message-ID: Pass 1 works. Pass 2 dies with timebase mismatch: ffmpeg -i play.dv -an -pass 1 -vf yadif=1:0,hqdn3d -vcodec libx264 -preset slower -tune film -r 59.94 -b 4000k -threads 0 -f null -y /dev/null which runs fine, no errors: ffmpeg version git-N-29568-g3b4621a, Copyright (c) 2000-2011 the FFmpeg developers built on May 2 2011 13:50:23 with gcc 4.6.0 20110428 (Red Hat 4.6.0-6) configuration: --prefix=/usr --bindir=/usr/bin --datadir=/usr/share/ffmpeg --libdir=/usr/lib64 --mandir=/usr/share/man --shlibdir=/usr/lib64 --extra-cflags='-O2 -march=native -mtune=native -fopenmp -fomit-frame-pointer -pipe' --enable-static --enable-shared --enable-gpl --enable-nonfree --enable-version3 --enable-postproc --enable-avfilter --enable-pthreads --enable-x11grab --enable-gray --enable-vaapi --enable-hardcoded-tables --enable-frei0r --enable-libdirac --disable-decoder=libdirac --enable-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-librtmp --enable-libschroedinger --disable-encoder=libschroedinger --enable-libspeex --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libxvid --enable-zlib --disable-debug --cpu=amdfam10 --arch=x86_64 --enable-pic --enable-libopencv libavutil 51. 2. 0 / 51. 2. 0 libavcodec 53. 3. 0 / 53. 3. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 4. 0 / 2. 4. 0 libswscale 0. 14. 0 / 0. 14. 0 libpostproc 51. 2. 0 / 51. 2. 0 [dv @ 0x2797f00] Estimating duration from bitrate, this may be inaccurate Input #0, dv, from 'play.dv': Duration: 00:51:08.43, start: 0.000000, bitrate: 28771 kb/s Stream #0.0: Video: dvvideo, yuv411p, 720x480, 28771 kb/s, PAR 8:9 DAR 4:3, 29.97 tbr, 29.97 tbn, 29.97 tbc Stream #0.1: Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s Incompatible pixel format 'yuv411p' for codec 'libx264', auto-selecting format 'yuv420p' [buffer @ 0x279e0a0] w:720 h:480 pixfmt:yuv411p [yadif @ 0x279ebc0] mode:1 parity:0 [hqdn3d @ 0x279f220] ls:4.000000 cs:3.000000 lt:6.000000 ct:4.500000 [hqdn3d @ 0x279f220] auto-inserting filter 'auto-inserted scaler 0' between the filter 'Parsed filter 0 yadif' and the filter 'Parsed filter 1 hqdn3d' [scale @ 0x279f8e0] w:720 h:480 fmt:yuv411p -> w:720 h:480 fmt:yuv420p flags:0xa0000004 [libx264 @ 0x279afc0] using SAR=8/9 [libx264 @ 0x279afc0] using cpu capabilities: MMX2 SSE2Fast FastShuffle SSEMisalign LZCNT [libx264 @ 0x279afc0] profile Main, level 3.1 Output #0, null, to '/dev/null': Metadata: encoder : Lavf53.0.3 Stream #0.0: Video: libx264, yuv420p, 720x480 [PAR 8:9 DAR 4:3], q=2-31, pass 1, 4000 kb/s, 90k tbn, 59.94 tbc Stream mapping: Stream #0.0 -> #0.0 Press [q] to stop encoding frame=183920 fps=121 q=0.0 Lsize= -0kB time=3068.37 bitrate= -0.0kbits/s video:0kB audio:0kB global headers:0kB muxing overhead -inf% final ratefactor: 19.69 But pass 2 dies: ffmpeg -i play.dv -an -pass 2 -vf yadif=1:0,hqdn3d -vcodec libx264 -preset slower -tune film -r 59.94 -timestamp now -b 4000k -threads 0 Play-yadif1-4000k.m4v .......... [dv @ 0x22f6f00] Estimating duration from bitrate, this may be inaccurate Input #0, dv, from 'play.dv': Duration: 00:51:08.43, start: 0.000000, bitrate: 28771 kb/s Stream #0.0: Video: dvvideo, yuv411p, 720x480, 28771 kb/s, PAR 8:9 DAR 4:3, 29.97 tbr, 29.97 tbn, 29.97 tbc Stream #0.1: Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s Incompatible pixel format 'yuv411p' for codec 'libx264', auto-selecting format 'yuv420p' [buffer @ 0x22fd260] w:720 h:480 pixfmt:yuv411p [yadif @ 0x22fdd20] mode:1 parity:0 [hqdn3d @ 0x22fe380] ls:4.000000 cs:3.000000 lt:6.000000 ct:4.500000 [hqdn3d @ 0x22fe380] auto-inserting filter 'auto-inserted scaler 0' between the filter 'Parsed filter 0 yadif' and the filter 'Parsed filter 1 hqdn3d' [scale @ 0x22fea00] w:720 h:480 fmt:yuv411p -> w:720 h:480 fmt:yuv420p flags:0xa0000004 [libx264 @ 0x22f9fc0] using SAR=8/9 [libx264 @ 0x22f9fc0] using cpu capabilities: MMX2 SSE2Fast FastShuffle SSEMisalign LZCNT [libx264 @ 0x22f9fc0] timebase mismatch with 1st pass (50/2997 vs 1001/30000) Output #0, ipod, to 'Play-yadif1-4000k.m4v': Stream #0.0: Video: libx264, yuv420p, 720x480 [PAR 8:9 DAR 4:3], q=2-31, pass 2, 4000 kb/s, 90k tbn, 59.94 tbc Stream mapping: Stream #0.0 -> #0.0 Error while opening encoder for output stream #0.0 - maybe incorrect parameters such as bit_rate, rate, width or height I also tried this without putting a framerate in pass 1. Same result. Also used -r 1001/60000, again same result: [libx264 @ 0xdd10c0] timebase mismatch with 1st pass (1001/60000 vs 1001/30000) How do you ( can you) do a 2 pass encoding while changing the frame rate? sean From jswordtestem at yahoo.co.uk Sun May 8 17:24:32 2011 From: jswordtestem at yahoo.co.uk (phil curb) Date: Sun, 8 May 2011 16:24:32 +0100 (BST) Subject: [FFmpeg-user] problem converting and combining these FLV files Message-ID: <483950.39980.qm@web25905.mail.ukl.yahoo.com> Thanks, so copy /B Karateka.mpg+Karateka2.mpg ffmpeg -i Karateka.mpg -acodec copy -vcodec mpeg2video -sameq whole.mpg That did it. But Howcome this one combines no problem, with copy /b or cat. Without any remux necessary http://www.berkut13.com/videos/brakeup.mpg It suggests to me that Karateka.flv is not being correctly converted to mpg, so the mpegs themselves are a bit odd hence not combining normally. But what's up with them? Am 08.05.2011 02:08, schrieb phil curb: > I have this file, i'm trying to combine it with a copy of itself. > http://www.sendspace.com/file/dkg3rz > > But the result while being twice as big, is the same duration. > > C:\blah>dir > Karateka.flv (399,375) > > I want to create one big file that is 2 of them combined. > > So I convert the flv to mpg which one can combine with copy /B > > ffmpeg -i Karateka.flv -sameq Karateka.mpg > Hello, I did the same on a linux system and while just copying I had similar issues. In my case it helped to do a remux with ffmpeg of the combined file with this command: |cat video1.mxf video2.mxf | ffmpeg -i - -acodec copy -vcodec mpeg2video -b 5M whole.mpg |||This is how it worked on Ubuntu, on a windows set-up you might replace 'cat' with you copy-routine. The '-' as input tells ffmpeg to use the cat-merged file as input. This command is explained on ffmbc.wordpress.com greetings, b ||| | _______________ From mark at mdsh.com Sun May 8 18:25:42 2011 From: mark at mdsh.com (Mark Himsley) Date: Sun, 08 May 2011 17:25:42 +0100 Subject: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? In-Reply-To: References: , <4DC59596.5090505@mdsh.com> Message-ID: <4DC6C406.4050206@mdsh.com> On 08/05/2011 06:04, christian fafard wrote: > > Hi Mark, > Hi Christian, Please can you not top-post on the ffmpeg email lists. http://en.wikipedia.org/wiki/Posting_style I have taken the liberty of re-ordering your response below: >> Date: Sat, 7 May 2011 19:55:18 +0100 >> From: mark at mdsh.com >> To: ffmpeg-user at ffmpeg.org >> Subject: Re: [FFmpeg-user] How to trick AVID to accept ffmpeg encoded D10 video? >> >> On 07/05/2011 14:32, christian fafard wrote: >>> >>> Any idea someone?? >>> >>> Baptiste, i know it's your field of expertise, do you have a procedure that works? >>> >>> Maybe the problem is the quicktime container, i don't know if Avid accept AVI for instance. >>> >>> I'm open to any suggestions, it's not vital for me but it's something i would like being able to do. >>> >>> Thanks >>> Christian Fafard >> >> What do you mean by "accept"? >> >> Have you tried Baptiste's FFmbc: http://code.google.com/p/ffmbc/ >> >> -- >> Mark >> > Hi Mark, > > You're right, > > the word 'accept' doesn't mean anything in a description of a technical problem. > > What i meant actually is that Avid import the video without complaining (IMX30 in a quicktime movie encoded with ffmpeg). > But the video is a white frame all along. > > So i took a video that i sure is working, re-wrap to a new quicktime without re-encoding ( -vcodec copy) and it's still the same. > So my conclusion is that it has to do with the quicktime header or something. If I recall correctly, Avid used QuickTime to do the import of MOV files, and QuickTime is very picky about flagging the video correctly. I expect you didn't include "-vbsf imxdump -vtag mx3p" in your command line. But, since you have not posted your command line this is just a pure guess! > And finally, no i didn't try Baptiste's ffmbc but that's exactly where i'm at right now. > I found the source but is it possible to download windows binary somewhere? I have sent you a private email off list. -- Mark From koxaniy at mail.ru Sun May 8 20:16:20 2011 From: koxaniy at mail.ru (Tuuls) Date: Sun, 8 May 2011 11:16:20 -0700 (PDT) Subject: [FFmpeg-user] LXF decoder ! Need help ! Message-ID: <1304878580857-3507549.post@n4.nabble.com> Hello. Someone can help with unpacking LXF container? The fact that the LXF files from the new servers can not be preprocessed in ffmpeg. The case is probably not in the files themselves, as they are exported from the old systems are handled beautifully, but the problem is probably in the file header. And can we somehow add the module metadata extraction? http://refile.net/f/?6j Here is a small file that I exported from Harris server. It contains the metadata and the video format DV?PRO25 . when I try to open it in ffmpeg, I get an error [lxf @ 000000000035A920] checksum error [lxf @ 000000000035A920] expected 120 B size header, got 0 Thanks in advance ! -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/LXF-decoder-Need-help-tp3507549p3507549.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From bahamutzero8825 at gmail.com Sun May 8 22:01:48 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Sun, 08 May 2011 15:01:48 -0500 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <002401cc0d5c$06d64600$4301a8c0@hpkantoor> References: <4DC30A66.3040006@gmail.com> <4DC438D4.10705@gmail.com><005d01cc0c34$d2927cb0$4301a8c0@hpkantoor> <4DC57F0E.4010308@gmail.com> <002401cc0d5c$06d64600$4301a8c0@hpkantoor> Message-ID: <4DC6F6AC.4070109@gmail.com> On 2011.05.08 03:43 AM, Bouke wrote: > Yes, if you mean tubes. But an LCD / plasma capable of displaying interlaced > video is an interlaced device IMHO. But they're not really capable of displaying interlaced video correctly like a tube TV. They have to deinterlace in order to display it correctly (as Andy Sheen said, this depends on other hardware in the TV). They're still progressive displays, whether they have some way of deinterlacing the source video before displaying it or not. From jswordtestem at yahoo.co.uk Mon May 9 07:27:27 2011 From: jswordtestem at yahoo.co.uk (phil curb) Date: Mon, 9 May 2011 06:27:27 +0100 (BST) Subject: [FFmpeg-user] problem converting and combining these FLV files Message-ID: <930698.62402.qm@web25908.mail.ukl.yahoo.com> I just tried it on another one, combining them, then fixing it with remux ffmpeg -i file.mpg -acodec copy -vcodec mpeg2video file2.mpg and in the middle of file2.mpg, the audio sped up went high and then finished so the video was silent. The audio just lost sync at some point. I don't have a link to the file where that happened, but perhaps somebody knows of why. ---- Howcome this one combines no problem, with copy /b or cat. Without any remux necessary http://www.berkut13.com/videos/brakeup.mpg It suggests to me that Karateka.flv is not being correctly converted to mpg, so the mpegs themselves are a bit odd hence not combining normally. But what's up with them? From tim.nicholson at bbc.co.uk Mon May 9 09:57:57 2011 From: tim.nicholson at bbc.co.uk (Tim Nicholson) Date: Mon, 09 May 2011 08:57:57 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC445FD.8050003@gmail.com> References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> <4DC445FD.8050003@gmail.com> Message-ID: <4DC79E85.5050704@bbc.co.uk> On 06/05/11 20:03, Baptiste Coudurier wrote: > On 05/06/2011 11:35 AM, sean darcy wrote: [...] >> Why would anyone ever use yadif=1 "bob" deinterlacing? > > You can do 1080i25 to 720p50 for example, but I'm sure there are other > usage since the feature is there. > But what I would really like is to go from 720p50 to 576i25 (SD) ;) -- Tim http://www.bbc.co.uk/ This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. If you have received it in error, please delete it from your system. Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. Please note that the BBC monitors e-mails sent or received. Further communication will signify your consent to this. From bouke at editb.nl Mon May 9 10:02:13 2011 From: bouke at editb.nl (Bouke) Date: Mon, 9 May 2011 10:02:13 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> <4DC445FD.8050003@gmail.com> <4DC79E85.5050704@bbc.co.uk> Message-ID: <004901cc0e1f$6b2c1d50$4301a8c0@hpkantoor> ---- Original Message ----- From: "Tim Nicholson" To: "FFmpeg user questions and RTFMs" Sent: Monday, May 09, 2011 9:57 AM Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? > On 06/05/11 20:03, Baptiste Coudurier wrote: >> On 05/06/2011 11:35 AM, sean darcy wrote: > [...] > >>> Why would anyone ever use yadif=1 "bob" deinterlacing? >> >> You can do 1080i25 to 720p50 for example, but I'm sure there are other >> usage since the feature is there. >> > > But what I would really like is to go from 720p50 to 576i25 (SD) ;) Tim, You know that AE can do this? (Perhaps you don't like the downscaler in AE, but it is a handy tool to split fields to frames and vice versa, and it can be scripted) Or, otherwise a trip to AviSynth Forrest? Bouke > -- > Tim > > http://www.bbc.co.uk/ > This e-mail (and any attachments) is confidential and may contain personal > views which are not the views of the BBC unless specifically stated. > If you have received it in error, please delete it from your system. > Do not use, copy or disclose the information in any way nor act in > reliance on it and notify the sender immediately. > Please note that the BBC monitors e-mails sent or received. > Further communication will signify your consent to this. > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From tim.nicholson at bbc.co.uk Mon May 9 10:29:35 2011 From: tim.nicholson at bbc.co.uk (Tim Nicholson) Date: Mon, 09 May 2011 09:29:35 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <004901cc0e1f$6b2c1d50$4301a8c0@hpkantoor> References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> <4DC445FD.8050003@gmail.com> <4DC79E85.5050704@bbc.co.uk> <004901cc0e1f$6b2c1d50$4301a8c0@hpkantoor> Message-ID: <4DC7A5EF.5010608@bbc.co.uk> On 09/05/11 09:02, Bouke wrote: > > ---- Original Message ----- > From: "Tim Nicholson" > To: "FFmpeg user questions and RTFMs" > Sent: Monday, May 09, 2011 9:57 AM > Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? > > >> On 06/05/11 20:03, Baptiste Coudurier wrote: >>> On 05/06/2011 11:35 AM, sean darcy wrote: >> [...] >> >>>> Why would anyone ever use yadif=1 "bob" deinterlacing? >>> >>> You can do 1080i25 to 720p50 for example, but I'm sure there are other >>> usage since the feature is there. >>> >> >> But what I would really like is to go from 720p50 to 576i25 (SD) ;) > > Tim, > You know that AE can do this? > (Perhaps you don't like the downscaler in AE, but it is a handy tool to > split fields to frames and vice versa, and it can be scripted) > But difficult to justify the cost of AE on a media mangling box just to do this.... > Or, otherwise a trip to AviSynth Forrest? Possibly but OS dependant and everything else is X of some flavour.. -- Tim http://www.bbc.co.uk/ This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. If you have received it in error, please delete it from your system. Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. Please note that the BBC monitors e-mails sent or received. Further communication will signify your consent to this. From baptiste.coudurier at gmail.com Mon May 9 11:01:01 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Mon, 09 May 2011 02:01:01 -0700 Subject: [FFmpeg-user] Error on 2-pass x264: timebase mismatch with 1st pass (50/2997 vs 1001/30000) In-Reply-To: References: Message-ID: <4DC7AD4D.7010000@gmail.com> On 5/8/11 6:20 AM, sean darcy wrote: > ffmpeg -i play.dv -an -pass 1 -vf yadif=1:0,hqdn3d -vcodec libx264 > -preset slower -tune film -r 59.94 -b 4000k -threads 0 -f null -y /dev/null Use -f mp4 /dev/null -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From bouke at editb.nl Mon May 9 11:14:56 2011 From: bouke at editb.nl (bouke) Date: Mon, 9 May 2011 11:14:56 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> <4DC445FD.8050003@gmail.com> <4DC79E85.5050704@bbc.co.uk><004901cc0e1f$6b2c1d50$4301a8c0@hpkantoor> <4DC7A5EF.5010608@bbc.co.uk> Message-ID: <00c201cc0e29$93bb85d0$4301a8c0@hpkantoor> ----- Original Message ----- From: "Tim Nicholson" To: "FFmpeg user questions and RTFMs" Sent: Monday, May 09, 2011 10:29 AM Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? > On 09/05/11 09:02, Bouke wrote: >> >> ---- Original Message ----- >> From: "Tim Nicholson" >> To: "FFmpeg user questions and RTFMs" >> Sent: Monday, May 09, 2011 9:57 AM >> Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? >> >> >>> On 06/05/11 20:03, Baptiste Coudurier wrote: >>>> On 05/06/2011 11:35 AM, sean darcy wrote: >>> [...] >>> >>>>> Why would anyone ever use yadif=1 "bob" deinterlacing? >>>> >>>> You can do 1080i25 to 720p50 for example, but I'm sure there are other >>>> usage since the feature is there. >>>> >>> >>> But what I would really like is to go from 720p50 to 576i25 (SD) ;) >> >> Tim, >> You know that AE can do this? >> (Perhaps you don't like the downscaler in AE, but it is a handy tool to >> split fields to frames and vice versa, and it can be scripted) >> > > But difficult to justify the cost of AE on a media mangling box just to do > this.... > >> Or, otherwise a trip to AviSynth Forrest? > > Possibly but OS dependant and everything else is X of some flavour.. Ok, you want cheap, you get cheap. Found a way with QT to do it. (Old trick to send two images over one interlaced pipe, used in the old days of expensive video conferencing) Export your clip as 25 fps and get rid of the last frame. >From the original, remove the first frame and export again. QT will omit one frame every frame if you go from 50 to 25 Pick one of the two and apply a interlaced key signal as mask. (One line black / white alternating) Paste that clip over the other, and you got an interlaced composition. Reverse the mask if you have the wrong field dominance, or apply the mask to the other clip. Bouke > -- > Tim > > http://www.bbc.co.uk/ > This e-mail (and any attachments) is confidential and may contain personal > views which are not the views of the BBC unless specifically stated. > If you have received it in error, please delete it from your system. > Do not use, copy or disclose the information in any way nor act in > reliance on it and notify the sender immediately. > Please note that the BBC monitors e-mails sent or received. > Further communication will signify your consent to this. > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From mailer.tovis at freemail.hu Mon May 9 12:54:53 2011 From: mailer.tovis at freemail.hu (tovis) Date: Mon, 9 May 2011 12:54:53 +0200 (CEST) Subject: [FFmpeg-user] ffmpeg and multichannel USB grabber Message-ID: <1966513538b922c0d831142b0eaec780.squirrel@nasi> Dir List! I have using a cheap, multichannel USB grabber dongle, with EasyCAP DC60 Driver from sourceforge net http://sourceforge.net/projects/easycapdc60/ It's have 6 video inputs (listed by v4l-info) named "CVBS0", "CVBS1" ... "CVBS4" and "S-VIDEO". Now I have use command for grabbing: ffmpeg -s 640x480 -f video4linux2 -i /dev/video0 http://localhost:8090/feed1.ffm It's working well, camera is pluged in to input wire number "1". QUESTIONS: - I can not found what syntax to use if I want to use input 2,3 or 4? - I have trying to put timestamp, on to stream, adding "-timestamp now" switch at command line, but I have nothing :( What is the right syntax? Could some one give me an example? Any help would be highly appreciated! Sincerely tovis From stefan at konink.de Mon May 9 13:51:52 2011 From: stefan at konink.de (Stefan de Konink) Date: Mon, 9 May 2011 13:51:52 +0200 (CEST) Subject: [FFmpeg-user] ffmpeg and multichannel USB grabber In-Reply-To: <1966513538b922c0d831142b0eaec780.squirrel@nasi> References: <1966513538b922c0d831142b0eaec780.squirrel@nasi> Message-ID: On Mon, 9 May 2011, tovis wrote: > I have using a cheap, multichannel USB grabber dongle, with EasyCAP DC60 > Driver from sourceforge net http://sourceforge.net/projects/easycapdc60/ > It's have 6 video inputs (listed by v4l-info) named "CVBS0", "CVBS1" ... > "CVBS4" and "S-VIDEO". Is your grabber really 6 channel, or is it just one chip with 6 possible connections? Cheap and 6 channel doesn't really make sense. If the chip is able to do > 120fps on one channel, then maybe it is possible. But usually you should get different /dev/video[0-9] for those. Stefan From stefan at konink.de Mon May 9 13:51:52 2011 From: stefan at konink.de (Stefan de Konink) Date: Mon, 9 May 2011 13:51:52 +0200 (CEST) Subject: [FFmpeg-user] ffmpeg and multichannel USB grabber In-Reply-To: <1966513538b922c0d831142b0eaec780.squirrel@nasi> References: <1966513538b922c0d831142b0eaec780.squirrel@nasi> Message-ID: On Mon, 9 May 2011, tovis wrote: > I have using a cheap, multichannel USB grabber dongle, with EasyCAP DC60 > Driver from sourceforge net http://sourceforge.net/projects/easycapdc60/ > It's have 6 video inputs (listed by v4l-info) named "CVBS0", "CVBS1" ... > "CVBS4" and "S-VIDEO". Is your grabber really 6 channel, or is it just one chip with 6 possible connections? Cheap and 6 channel doesn't really make sense. If the chip is able to do > 120fps on one channel, then maybe it is possible. But usually you should get different /dev/video[0-9] for those. Stefan From adi235 at gmail.com Mon May 9 14:01:51 2011 From: adi235 at gmail.com (Aditya) Date: Mon, 9 May 2011 12:01:51 +0000 (UTC) Subject: [FFmpeg-user] =?utf-8?q?Segmentation_Faults_-_avcodec=5Fencode=5F?= =?utf-8?q?video_Please_Help?= Message-ID: I am writing a simple C program to accept video input decode it encode it into flv and then save output video file. I have looked at danger's tutorials, output-example.c,api-example.c and such. But i seem to be missing something very essential. And always end up encountering , segmentation faults, ( sometimes buffer under flows or warning max analysis time reached. The below is the program. I would very much appreciate for a clear solution to this problem. I have spend to many days trying to figure this out already. [CODE] #include #include #include #include #include #include static void SaveFrame(AVFrame *pFrame, int width, int height, int iFrame); AVFormatContext *pFormatCtx; static int i, videoStream; AVCodecContext *pCodecCtx; AVCodec *pCodec; AVFrame *pFrame; AVFrame *pFrameRGB; AVPacket packet; static int frameFinished; static int numBytes; static uint8_t *buffer; AVCodec *codec; AVCodecContext *c = NULL; int out_size, size, outbuf_size; uint8_t *outbuf; FILE *f; int init_video_file(char *inputFileName) { // Register all formats and codecs av_register_all(); // Open video file if (av_open_input_file(&pFormatCtx, inputFileName, NULL, 0, NULL) != 0) return -1; // Couldn't open file // Retrieve stream information if (av_find_stream_info(pFormatCtx) < 0) return -1; // Couldn't find stream information // Dump information about file onto standard error dump_format(pFormatCtx, 0, inputFileName, false); // Find the first video stream videoStream = -1; for (i = 0; i < pFormatCtx->nb_streams; i++) if (pFormatCtx->streams[i]->codec->codec_type == CODEC_TYPE_VIDEO) { videoStream = i; break; } if (videoStream == -1) return -1; // Didn't find a video stream return 0; } int init_video_codec() { // Get a pointer to the codec context for the video stream pCodecCtx = pFormatCtx->streams[videoStream]->codec; // Find the decoder for the video stream pCodec = avcodec_find_decoder(pCodecCtx->codec_id); if (pCodec == NULL) return -1; // Codec not found // Open codec if (avcodec_open(pCodecCtx, pCodec) < 0) return -1; // Could not open codec return 0; } int set_frame_data() { // Hack to correct wrong frame rates that seem to be generated by some codecs if (pCodecCtx->time_base.num > 1000 && pCodecCtx->time_base.den == 1) pCodecCtx->time_base.den = 1000; // Allocate video frame pFrame = avcodec_alloc_frame(); // Allocate an AVFrame structure pFrameRGB = avcodec_alloc_frame(); if (pFrameRGB == NULL) return -1; // Determine required buffer size and allocate buffer numBytes = avpicture_get_size(PIX_FMT_RGB24, pCodecCtx->width, pCodecCtx->height); buffer = malloc(numBytes); // Assign appropriate parts of buffer to image planes in pFrameRGB avpicture_fill((AVPicture *) pFrameRGB, buffer, PIX_FMT_RGB24, pCodecCtx->width, pCodecCtx->height); return 0; } void process_packet_data(char *outFileName) { // Read frames and save first five frames to disk i = 0; while (av_read_frame(pFormatCtx, &packet) >= 0) { // Is this a packet from the video stream? if (packet.stream_index == videoStream) { // Decode video frame avcodec_decode_video(pCodecCtx, pFrame, &frameFinished, packet.data, packet.size); // Did we get a video frame? if (frameFinished) { //printf("\n packet size= %d \t pframe number=n%d ", packet.size, // pFrame->coded_picture_number); // Insane segmentation faults when encoding the got frame.. !!Help. //out_size = avcodec_encode_video(c, outbuf, outbuf_size, pFrame); save_frame_as_image(); } } // Free the packet that was allocated by av_read_frame av_free_packet(&packet); } } void set_encoder_for_output(const char *filename) { uint8_t *outbuf; printf("Video encoding\n"); /* find the mpeg1 video encoder */ codec = avcodec_find_encoder(CODEC_ID_FLASHSV); if (!codec) { fprintf(stderr, "codec not found\n"); exit(1); } c = avcodec_alloc_context(); pFrame = avcodec_alloc_frame(); /* put sample parameters */ c->bit_rate = 128000; /* resolution must be a multiple of two */ c->width = 320; c->height = 240; /* frames per second */ c->time_base = (AVRational) {1,25}; c->gop_size = 10; /* emit one intra frame every ten frames */ c->max_b_frames=1; c->pix_fmt = PIX_FMT_YUV420P; /* open it */ if (avcodec_open(c, codec) < 0) { fprintf(stderr, "could not open codec\n"); exit(1); } f = fopen(filename, "wb"); if (!f) { fprintf(stderr, "could not open ______::: %s\n", filename); exit(1); } /* alloc image and output buffer */ outbuf_size = 230400; outbuf = malloc(outbuf_size); size = c->width * c->height; //out_size = avcodec_encode_video(c, outbuf, outbuf_size, picture); } void cleanup() { // Free the RGB image' free(outbuf); avcodec_close(c); av_free(c); printf("\n"); free(buffer); av_free(pFrameRGB); // Free the YUV frame av_free(pFrame); // Close the codec avcodec_close(pCodecCtx); // Close the video file av_close_input_file(pFormatCtx); } void encode_frames_to_video() { } void save_frame_as_image() { static struct SwsContext *img_convert_ctx; // Convert the image into YUV format that SDL uses if (img_convert_ctx == NULL) { int w = pCodecCtx->width; int h = pCodecCtx->height; img_convert_ctx = sws_getContext(w, h, pCodecCtx->pix_fmt, w, h, PIX_FMT_RGB24, SWS_BICUBIC, NULL, NULL, NULL); if (img_convert_ctx == NULL) { fprintf(stderr, "Cannot initialize the conversion context!\n"); exit(1); } } int ret = sws_scale(img_convert_ctx, pFrame->data, pFrame->linesize, 0, pCodecCtx->height, pFrameRGB->data, pFrameRGB->linesize); // Save the frame to disk if (i++ <= 5) SaveFrame(pFrameRGB, pCodecCtx->width, pCodecCtx->height, i); } int main(int argc, char * argv[]) { init_video_file(argv[1]); init_video_codec(); set_frame_data(); set_encoder_for_output(argv[2]); process_packet_data(argv[2]); cleanup(); return 0; } static void SaveFrame(AVFrame *pFrame, int width, int height, int iFrame) { FILE *pFile; char szFilename[32]; int y; // Open file sprintf(szFilename, "frame%d.ppm", iFrame); pFile = fopen(szFilename, "wb"); if (pFile == NULL) return; // Write header fprintf(pFile, "P6\n%d %d\n255\n", width, height); // Write pixel data for (y = 0; y < height; y++) fwrite(pFrame->data[0] + y * pFrame->linesize[0], 1, width * 3, pFile); // Close file fclose(pFile); } [/CODE] [b] Kindly help in resolving the segmentation faults in the above program [/b] I have been unable to find any straight forward example on how to do this. Thanks From Cecil at Decebal.nl Mon May 9 15:17:58 2011 From: Cecil at Decebal.nl (Cecil Westerhof) Date: Mon, 09 May 2011 15:17:58 +0200 Subject: [FFmpeg-user] Converting to h264 Message-ID: <1304947078.17848.5.camel@Aspire.decebal.comp> I was told to make mp4 files. I did, but now at the client I find out that the format is wrong. My files have the following format: Stream #0.0(und): Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 1085 kb/s, 15 fps, 15 tbr, 15 tbn, 15 tbc but what they need is: Stream #0.0(eng): Video: h264 (High), yuv420p, 640x480 [PAR 1:1 DAR 4:3], 30 kb/s, 10 fps, 10 tbr, 10k tbn, 20 tbc I tried several things, but until now with no avail. I tried for example: ffmpeg -i afspraak.mp4 -vcodec copy -vbsf h264_mp4toannexb -an of.h264 Which gives: Output #0, h264, to 'of.h264': Metadata: major_brand : isom minor_version : 512 compatible_brands: isomiso2mp41 creation_time : 1970-01-01 00:00:00 encoder : Lavf52.108.0 Stream #0.0(und): Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], q=2-31, 1085 kb/s, 90k tbn, 15 tbc Metadata: creation_time : 1970-01-01 00:00:00 Stream mapping: Stream #0.0 -> #0.0 Press [q] to stop encoding [NULL @ 0x80981c0] Warning: SPS NALU missing or invalid. The resulting stream may not play. [NULL @ 0x80981c0] Warning: PPS NALU missing or invalid. The resulting stream may not play. h264_mp4toannexb failed for stream 0, codec copy: Invalid data found when processing input frame= 2384 fps= 0 q=-1.0 Lsize= 20947kB time=158.93 bitrate=1079.7kbits/s video:21068kB audio:0kB global headers:0kB muxing overhead -0.571794% Does anyone have a pointer for me where to find the info I need? -- Cecil Westerhof From dave at avpreserve.com Mon May 9 15:28:40 2011 From: dave at avpreserve.com (Dave Rice) Date: Mon, 9 May 2011 09:28:40 -0400 Subject: [FFmpeg-user] LXF decoder ! Need help ! In-Reply-To: <1304878580857-3507549.post@n4.nabble.com> References: <1304878580857-3507549.post@n4.nabble.com> Message-ID: <8278B7DB-4B1A-4C40-B246-2D6D4E9FB9A6@avpreserve.com> On May 8, 2011, at 2:16 PM, Tuuls wrote: > Hello. Someone can help with unpacking LXF container? Sorry for an un-ffmpeg answer, but you could try to investigate the header using mediainfo, which added lxf support recently. Try `mediainfo --inform='Details;1' file.lxf`. Dave Rice avpreserve.com > The fact that the LXF files from the new servers can not be preprocessed in > ffmpeg. The case is probably not in the files themselves, as they are > exported from the old systems are handled beautifully, but the problem is > probably in the file header. > And can we somehow add the module metadata extraction? > > http://refile.net/f/?6j > > Here is a small file that I exported from Harris server. It contains the > metadata and the video format DV?PRO25 . > > when I try to open it in ffmpeg, I get an error > [lxf @ 000000000035A920] checksum error > [lxf @ 000000000035A920] expected 120 B size header, got 0 > > Thanks in advance ! > > -- > View this message in context: http://ffmpeg-users.933282.n4.nabble.com/LXF-decoder-Need-help-tp3507549p3507549.html > Sent from the FFmpeg-users mailing list archive at Nabble.com. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From edward.guo at hotmail.com Mon May 9 16:39:03 2011 From: edward.guo at hotmail.com (=?gb2312?B?zPogxL3I3Q==?=) Date: Mon, 9 May 2011 22:39:03 +0800 Subject: [FFmpeg-user] I Need Some Help In-Reply-To: References: , Message-ID: Dear Sir or Madam Many thanks for reading this email. I am a new learner of FFMPEG. When I was reading your code recently, come to a confused problem which appears in and line 327. As I consider the pointer "s->idct_put[0]" points to the "dsp.idct_put" which is initialized in line 4180. Then I looked through the code found the function "ff_simple_idct_put" (line 4252) and found it appears in line 371 also . Logically, I used it to decode DVCPRO 625i50 file but found the result is not exactly correct. Subsequently I compared the address of "c->idct_put= ff_simple_idct_put" in line 4252 and the address in when directly call the "ff_simple_idct_put" in . The interesting thing is they are different functions even though they have the same name. To be more sure, I modified the code "mb->idct_put = s->idct_put[dct_mode && log2_blocksize == 3];" line 521 into "mb->idct_put = s->idct_put[0]" and print a message in the function "ff_simple_idct_put" of line 371. As you can imagine, the code never go into there. So I wonder, 1. which and how you use the idct method when decoding the DVCPRO 625i50 frame? 2. After the initilization of "dsp.idct_put" where maybe modify it? 3. where can I get the correct "ff_simple_idct_put" method except the , when decoding DVCPRO 625i50? 4. what's the pointer "dsp.idct_put" really point to? Please help me. Kind Wishes Yours Chao.Guo From tim.nicholson at bbc.co.uk Mon May 9 17:44:32 2011 From: tim.nicholson at bbc.co.uk (Tim Nicholson) Date: Mon, 09 May 2011 16:44:32 +0100 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <00c201cc0e29$93bb85d0$4301a8c0@hpkantoor> References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> <4DC445FD.8050003@gmail.com> <4DC79E85.5050704@bbc.co.uk><004901cc0e1f$6b2c1d50$4301a8c0@hpkantoor> <4DC7A5EF.5010608@bbc.co.uk> <00c201cc0e29$93bb85d0$4301a8c0@hpkantoor> Message-ID: <4DC80BE0.1060109@bbc.co.uk> On 09/05/11 10:14, bouke wrote: > > ----- Original Message ----- > From: "Tim Nicholson" > To: "FFmpeg user questions and RTFMs" > Sent: Monday, May 09, 2011 10:29 AM > Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? > > >> On 09/05/11 09:02, Bouke wrote: >>> >>> ---- Original Message ----- >>> From: "Tim Nicholson" >>> To: "FFmpeg user questions and RTFMs" >>> Sent: Monday, May 09, 2011 9:57 AM >>> Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? >>>> [..] >>>> But what I would really like is to go from 720p50 to 576i25 (SD) ;) >>> >>> Tim, >>> You know that AE can do this? >>> (Perhaps you don't like the downscaler in AE, but it is a handy tool to >>> split fields to frames and vice versa, and it can be scripted) >>> >>[...] > > Ok, you want cheap, you get cheap. Not so much that as not using a sledgehammer. It has been pointed out to me that the tinterlace filter in mplayer can make a useful nutcracker in this case (with a tweaked version in ffmbc). > Found a way with QT to do it. > [...] > Pick one of the two and apply a interlaced key signal as mask. > (One line black / white alternating) >...] Now that is a goldie oldie. Somewhere I still have a field test image with field one lines all red and field two all blue. Great for spotting dominance changes but not so good for the eyes if you freeze frame instead of freeze field...:( -- Tim http://www.bbc.co.uk/ This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. If you have received it in error, please delete it from your system. Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. Please note that the BBC monitors e-mails sent or received. Further communication will signify your consent to this. From mike.scheutzow at alcatel-lucent.com Mon May 9 18:14:50 2011 From: mike.scheutzow at alcatel-lucent.com (Mike Scheutzow) Date: Mon, 09 May 2011 12:14:50 -0400 Subject: [FFmpeg-user] Converting to h264 In-Reply-To: <1304947078.17848.5.camel@Aspire.decebal.comp> References: <1304947078.17848.5.camel@Aspire.decebal.comp> Message-ID: <4DC812FA.9020403@alcatel-lucent.com> Cecil Westerhof wrote: > I was told to make mp4 files. I did, but now at the client I find out > that the format is wrong. This confusion happens a lot, because client's don't always specify whether they are referring to the container or to the codec. You have to ask. > My files have the following format: > Stream #0.0(und): Video: mpeg4, yuv420p, 1028x795 [PAR 1:1 DAR 1028:795], 1085 kb/s, 15 fps, 15 tbr, 15 tbn, 15 tbc > but what they need is: > Stream #0.0(eng): Video: h264 (High), yuv420p, 640x480 [PAR 1:1 DAR 4:3], 30 kb/s, 10 fps, 10 tbr, 10k tbn, 20 tbc > > I tried several things, but until now with no avail. > > I tried for example: > ffmpeg -i afspraak.mp4 -vcodec copy -vbsf h264_mp4toannexb -an of.h264 You can not use '-vcodec copy' to convert from 'Video: mp4' to 'Video: h264'. These video types are different and have incompatible binary formats. You need to do a full re-encode, typically using '-vcodec libx264', along with the additional command line arguments that codec requires. Either mp4 video or h264 video can be stored in a .mp4 container. Mike Scheutzow From lord_nenad at yahoo.com Fri May 6 20:19:35 2011 From: lord_nenad at yahoo.com (Nenad Lukic) Date: Fri, 6 May 2011 11:19:35 -0700 (PDT) Subject: [FFmpeg-user] ffmpeg -acodec problem Message-ID: <287439.57502.qm@web125701.mail.ne1.yahoo.com> Hello, I'm Nenad Lukic, php developer, I recently start dealing with ffmpeg, and I have serious problems with it... So if you are a expert, and I'm just a beginner I will really appreciate your help. Here is the problem: I'm trying to make conversions for html5 videos, converting to mp4, webm, ogg/ogv formats. First I had version 0.6.0 that didn't had webm installed, so I ask operators to update that version, and they did, and after that, none of the conversions didn't work... Then I hire someone, to make that... Then that man delete ffmpeg and install new copy... BUT, in php information page version is 0.6.3, and when I run ffmpeg -version I get 0.7rc-1 version... Here is the real problem now: mp4 and webm are converting okay, because their default codecs are ones used for html5, but with ogg/ogv default audio format is flac, that can't be viewed in browsers. That why I have videos without sound. But, problem is that when I try to use sound codecs -acodec no matter what codec I try to use, it doesn't work. I tried with 10 different audio codecs, and only 3 of them ware creating files, and 0 bytes files. So I can't do conversions, and I don't know why is this... Can someone help me? Tell me what to do, how this can be handled, tell me what is the problem, anything... Thank you! I will really appreciate your help! Best regards, Nenad Lukic From h.reindl at thelounge.net Mon May 9 19:46:23 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Mon, 09 May 2011 19:46:23 +0200 Subject: [FFmpeg-user] ffmpeg -acodec problem In-Reply-To: <287439.57502.qm@web125701.mail.ne1.yahoo.com> References: <287439.57502.qm@web125701.mail.ne1.yahoo.com> Message-ID: <4DC8286F.80304@thelounge.net> Am 06.05.2011 20:19, schrieb Nenad Lukic: > Hello, > > I'm Nenad Lukic, php developer, I recently start dealing with ffmpeg, and I have serious problems with it... > So if you are a expert, and I'm just a beginner I will really appreciate your help. > > Here is the problem: > > I'm trying to make conversions for html5 videos, converting to mp4, webm, ogg/ogv formats. > First I had version 0.6.0 that didn't had webm installed, so I ask > operators to update that version, and they did, and after that, none of > the conversions didn't work... > Then I hire someone, to make that... Then that man delete ffmpeg and > install new copy... BUT, in php information page version is 0.6.3, and > when I run ffmpeg -version I get 0.7rc-1 version... > > Here is the real problem now: > > mp4 and webm are converting okay, because their default codecs are ones > used for html5 not really - HTML5 needs H264-Video not only MP4 > but with ogg/ogv default audio format is flac, that > can't be viewed in browsers that why I have videos without sound. vcodec: libtheora acodec: libvorbis are you sure that you specify the codecs AFTER the input-file? -- Mit besten Gr??en, Reindl Harald the lounge interactive design GmbH A-1060 Vienna, Hofm?hlgasse 17 CTO / software-development / cms-solutions p: +43 (1) 595 3999 33, m: +43 (676) 40 221 40 icq: 154546673, http://www.thelounge.net/ http://www.thelounge.net/signature.asc.what.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From Umair.Khan at uni-klu.ac.at Mon May 9 19:56:45 2011 From: Umair.Khan at uni-klu.ac.at (Umair.Khan at uni-klu.ac.at) Date: Mon, 09 May 2011 19:56:45 +0200 Subject: [FFmpeg-user] MJPEG decoding issue Message-ID: <4DC846FD020000E000010D20@gwx1.uni-klu.ac.at> Hi, I am trying to get a RTSP stream from an IP camera in MJPEG format. But FFMPEG has some issue with MJPEG format. I get the error messages: mjpeg: unsupported coding type What can be the problem? I will really appreciate any help in this regard. Regards, Dipl.-Ing. Umair Ali Khan Research Staff Member, Pervasive Computing Group Institute of Networked and Embedded Systems Klagenfurt University, Austria -------------------------------------------------------- Lakeside Park L.2.1.33, 9020 Klagenfurt Voice: +43(0)463 2700 3872 Fax: +43(0)463 2700 3679 -------------------------------------------------------- www.pervasive.uni-klu.ac.at From kosta.brazzers at gmail.com Mon May 9 20:40:23 2011 From: kosta.brazzers at gmail.com (Kosta Vlotis) Date: Mon, 9 May 2011 14:40:23 -0400 Subject: [FFmpeg-user] cloud encoding Message-ID: Hello, i was wondering. is there a way that i can get some thing like this: http://blog.picloud.com/tag/ffmpeg/ set up locally? i'd like to have a server rack set up locally and have a script that sends tasks to my "cloud". this solution seems idea for my needs but for the amount of videos that we encode, paying someone for computer time and bandwidth might mean the cost will be too high. so does anyone know of a way that i can set this up locally? anyone have any suggestions? From h.reindl at thelounge.net Mon May 9 20:45:26 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Mon, 09 May 2011 20:45:26 +0200 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: <4DC83646.9020402@thelounge.net> Am 09.05.2011 20:40, schrieb Kosta Vlotis: > Hello, > > i was wondering. is there a way that i can get some thing like this: > http://blog.picloud.com/tag/ffmpeg/ set up locally? > > i'd like to have a server rack set up locally and have a script that > sends tasks to my "cloud". this solution seems idea for my needs but > for the amount of videos that we encode, paying someone for computer > time and bandwidth might mean the cost will be too high. > > so does anyone know of a way that i can set this up locally? anyone > have any suggestions? take a mysql-database, a nfs-storage and mount the storage on all nodes the mysql-db is for tasklist, status (started, canceld, failed) and contains the path. yes you will need some logic for how to spread the load but it should not be too different with cronjobs every minute or even a sysv service with some scirpt in an endless loop to let fetch every instance some videos to encode and marke in the table as "in progress" whith this way and a little time there are nice things possible since you can add additional infos in the database, make some webinterfaces to control the ffmpeg-params and store them with the record -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From houndeyex at gmail.com Mon May 9 20:46:20 2011 From: houndeyex at gmail.com (James O.) Date: Mon, 9 May 2011 14:46:20 -0400 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: I have a setup here that involves Python scripts and a CRON job. I didn't architect the whole thing, but it leverages the super compute cluster at the university. Basically files are sent via SFTP to a folder watched by a CRON job that runs every minute. CRON will move files and kick off a transcode, and then sends out a notification when it's done. There is some other automation in place, but it may not fit for your circumstance. On Mon, May 9, 2011 at 2:40 PM, Kosta Vlotis wrote: > Hello, > > i was wondering. is there a way that i can get some thing like this: > http://blog.picloud.com/tag/ffmpeg/ set up locally? > > i'd like to have a server rack set up locally and have a script that > sends tasks to my "cloud". this solution seems idea for my needs but > for the amount of videos that we encode, paying someone for computer > time and bandwidth might mean the cost will be too high. > > so does anyone know of a way that i can set this up locally? anyone > have any suggestions? > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From koxaniy at mail.ru Mon May 9 21:36:09 2011 From: koxaniy at mail.ru (Tuuls) Date: Mon, 9 May 2011 12:36:09 -0700 (PDT) Subject: [FFmpeg-user] LXF decoder ! Need help ! In-Reply-To: <8278B7DB-4B1A-4C40-B246-2D6D4E9FB9A6@avpreserve.com> References: <1304878580857-3507549.post@n4.nabble.com> <8278B7DB-4B1A-4C40-B246-2D6D4E9FB9A6@avpreserve.com> Message-ID: <1304969769243-3510144.post@n4.nabble.com> MediaInfo well read this lxf , and well send metadata : General Complete name : D:\AT110416_7349.lxf Format : LXF File size : 114 MiB Duration : 29s 880ms Overall bit rate : 31.9 Mbps Track name : ???????????? ???????. ?????? < its a well right metadata Writing library : NX3000VESX-16C/VR1 Service name : aeg Audio #1 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) Audio #2 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) Audio #3 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) Audio #4 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/LXF-decoder-Need-help-tp3507549p3510144.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From koxaniy at mail.ru Mon May 9 21:36:09 2011 From: koxaniy at mail.ru (Tuuls) Date: Mon, 9 May 2011 12:36:09 -0700 (PDT) Subject: [FFmpeg-user] LXF decoder ! Need help ! In-Reply-To: <8278B7DB-4B1A-4C40-B246-2D6D4E9FB9A6@avpreserve.com> References: <1304878580857-3507549.post@n4.nabble.com> <8278B7DB-4B1A-4C40-B246-2D6D4E9FB9A6@avpreserve.com> Message-ID: <1304969769300-3510145.post@n4.nabble.com> MediaInfo well read this lxf , and well send metadata : General Complete name : D:\AT110416_7349.lxf Format : LXF File size : 114 MiB Duration : 29s 880ms Overall bit rate : 31.9 Mbps Track name : ???????????? ???????. ?????? < its a well right metadata Writing library : NX3000VESX-16C/VR1 Service name : aeg Audio #1 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) Audio #2 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) Audio #3 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) Audio #4 Format : PCM Duration : 29s 880ms Bit rate mode : Constant Bit rate : 768 Kbps Channel(s) : 1 channel Sampling rate : 48.0 KHz Bit depth : 16 bits Stream size : 2.74 MiB (2%) -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/LXF-decoder-Need-help-tp3507549p3510145.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From phil_rhodes at rocketmail.com Mon May 9 22:02:13 2011 From: phil_rhodes at rocketmail.com (Phil Rhodes) Date: Mon, 09 May 2011 21:02:13 +0100 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: > i'd like to have a server rack set up locally and have a script that > sends tasks to my "cloud". this solution seems idea for my needs but > for the amount of videos that we encode, paying someone for computer > time and bandwidth might mean the cost will be too high. In my day we used to call this a render farm, amazing how these concepts can be reinvigorated by renaming them! P From kosta.brazzers at gmail.com Mon May 9 22:58:49 2011 From: kosta.brazzers at gmail.com (Kosta Vlotis) Date: Mon, 9 May 2011 16:58:49 -0400 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: On Mon, May 9, 2011 at 4:02 PM, Phil Rhodes wrote: >> i'd like to have a server rack set up locally and have a script that >> sends tasks to my "cloud". this solution seems idea for my needs but >> for the amount of videos that we encode, paying someone for computer >> time and bandwidth might mean the cost will be too high. > > In my day we used to call this a render farm, amazing how these concepts can > be reinvigorated by renaming them! > > P > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > lol.. right now im using a stack of about 20 PCs to run all the encoding using windows batch scripts. it gets the job done but i need something more scalable and/or faster. i can't just keep stacking pcs on the floor.. From belcampo at zonnet.nl Mon May 9 23:01:39 2011 From: belcampo at zonnet.nl (belcampo) Date: Mon, 09 May 2011 23:01:39 +0200 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: <4DC85633.5090501@zonnet.nl> On 05/09/11 20:40, Kosta Vlotis wrote: > Hello, > > i was wondering. is there a way that i can get some thing like this: > http://blog.picloud.com/tag/ffmpeg/ set up locally? > > i'd like to have a server rack set up locally and have a script that > sends tasks to my "cloud". this solution seems idea for my needs but > for the amount of videos that we encode, paying someone for computer > time and bandwidth might mean the cost will be too high. > > so does anyone know of a way that i can set this up locally? anyone > have any suggestions? I do have a local setup like that, more or less. I use a script found here: http://code.google.com/p/ppss/ I use it for putting 5 PC's with a total of 12 Cores to work at transcoding with ffmpeg. 12 single-threaded instances to get the highest transcoded fps. NOT the fastest way to get 1 file encoded though. There are solutions to do that also, but then you have to evenly divide your source in parts equal to the amount of available cores. After encoding you'll have to mux the encoded parts together again. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From Cecil at decebal.nl Mon May 9 23:17:14 2011 From: Cecil at decebal.nl (Cecil Westerhof) Date: Mon, 09 May 2011 23:17:14 +0200 Subject: [FFmpeg-user] Converting to h264 In-Reply-To: <4DC812FA.9020403@alcatel-lucent.com> (Mike Scheutzow's message of "Mon, 09 May 2011 12:14:50 -0400") References: <1304947078.17848.5.camel@Aspire.decebal.comp> <4DC812FA.9020403@alcatel-lucent.com> Message-ID: <87liyfa811.fsf@Compaq.site> Op maandag 9 mei 2011 18:14 CEST schreef Mike Scheutzow: > You can not use '-vcodec copy' to convert from 'Video: mp4' to 'Video: > h264'. These video types are different and have incompatible binary > formats. > > You need to do a full re-encode, typically using '-vcodec libx264', > along with the additional command line arguments that codec requires. I found something that works: ffmpeg -y -i huiswerkIQCoaches02.avi -acodec libmp3lame -ar 44100 -ab 96k -vcodec libx264 -vpre slow -level 41 -crf 20 -bufsize 20000k -maxrate 25000k -g 250 -r 20 -s vga -coder 1 -flags +loop -cmp +chroma -partitions +parti4x4+partp8x8+partb8x8 -subq 7 -me_range 16 -keyint_min 25 -sc_threshold 40 -i_qfactor 0.71 -rc_eq 'blurCplx^(1-qComp)' -bf 16 -b_strategy 1 -bidir_refine 1 -refs 6 -deblockalpha 0 -deblockbeta 0 output.mp4 I do not understand it yet, but it looks likes it works. I have to check it tomorrow at the client. -- Cecil Westerhof Senior Software Engineer LinkedIn: http://www.linkedin.com/in/cecilwesterhof From jerome at mediaarea.net Mon May 9 23:18:17 2011 From: jerome at mediaarea.net (Jerome Martinez) Date: Mon, 9 May 2011 14:18:17 -0700 (PDT) Subject: [FFmpeg-user] LXF decoder ! Need help ! In-Reply-To: <1304969769243-3510144.post@n4.nabble.com> References: <1304878580857-3507549.post@n4.nabble.com> <8278B7DB-4B1A-4C40-B246-2D6D4E9FB9A6@avpreserve.com> <1304969769243-3510144.post@n4.nabble.com> Message-ID: <1304975897622-3510358.post@n4.nabble.com> Tuuls wrote: > > MediaInfo well read this lxf , and well send metadata : I apologize if I am out of topic (not focused on ffmpeg). Actually, video part is missing (I adapted my code so DV PAL is now displayed) + Windows version of MediaInfo has "trace" feature disabled for binary size optimization (Dave is using the Linux/Mac version, with trace feature activated, his command has not the same result than your test). Trace of your file with the command from Dave + MediaInfo compiled with trace feature is available at this URL: http://pastebin.com/Z6wKx8ik Example: 00000000 Header (355 bytes) 00000000 Header (72 bytes) 00000000 Signature: LEITCH 00000008 Version: 1 (0x00000001) 0000000C Header size: 72 (0x00000048) 00000010 Type: 2 (0x2) 00000014 Stream ID: 0 (0x00000000) 00000018 ? (Always 0x00000000): 0 (0x0) 00000020 ?: 21513600 (0x1484580) 00000028 ? (Always 0x00000001): 1 (0x00000001) 0000002C Block size: 120 (0x78) 00000030 Block size: 163 (0xA3) 00000034 ? (Always 0x00000000): 0 (0x00000000) 00000038 ? (Always 0x00000000): 0 (0x00000000) 0000003C ? (Always 0x00000000): 0 (0x00000000) 00000040 Reverse TimeStamp?: 2859346826 (0xAA6E2B8A) - 3971315.036 ms 00000048 Info? (120 bytes) 00000048 Unknown: (120 bytes) 000000C0 Tags? (163 bytes) 000000C0 Size: 8 (0x08) 000000C1 ? (8 opaque bytes): (8 bytes) 000000C9 Size: 18 (0x12) 000000CA Library?: NX3000VESX-16C/VR1 000000DC Size: 0 (0x00) 000000DD Size: 0 (0x00) 000000DE Size: 16 (0x10) 000000DF 0x00000008: 0 (0x00000000) 000000E3 0x00000000: 4 (0x00000004) 000000E7 0x00000000: 0 (0x00000000) 000000EB 0x00000092: 146 (0x00000092) 000000EF Size: 0 (0x00) 000000F0 Size: 0 (0x00) 000000F1 Size: 6 (0x06) 000000F2 Channel?: aeg 000000F8 Size: 0 (0x00) 000000F9 Size: 0 (0x00) 000000FA Size: 0 (0x00) 000000FB Size: 0 (0x00) 000000FC Size: 56 (0x38) 000000FD Title: ???????????? ???????. ?????? 00000135 Size: 14 (0x0E) 00000136 ? (in UTF-16): Arh New 00000144 Size: 8 (0x08) 00000145 Data: (8 bytes) 0000014D Size: 8 (0x08) 0000014E Data: (8 bytes) 00000156 Size: 0 (0x00) 00000157 Size: 0 (0x00) 00000158 Size: 0 (0x00) 00000159 Size: 0 (0x00) 0000015A Size: 8 (0x08) 0000015B ? (8 opaque bytes): (8 bytes) 00000163 Video (72 bytes) 00000163 Header (72 bytes) 00000163 Signature: LEITCH 0000016B Version: 1 (0x00000001) 0000016F Header size: 72 (0x00000048) 00000173 Type: 0 (0x0) 00000177 Stream ID: 0 (0x00000000) 0000017B TimeStamp: 0 (0x0) - 0.000 ms 00000183 Duration: 28800 (0x7080) - 40.000 ms 0000018B Format: 5 (0x05) - DV 0000018B GOP (N): 1 (0x01) 0000018C GOP (M): 1 (0x01) 0000018C Bit rate: 25 (0x19) - 25000000 bps 0000018D Picture type: 0 (0x00) - I 0000018E Reserved: 0 (0x00) 0000018F Video data size: 144000 (0x23280) 00000197 VBI data size: 0 (0x0) 0000019F ? (Always 0x00000000): 0 (0x00000000) 000001A3 Reverse TimeStamp: 2880276243 (0xABAD8713) - 4000383.671 ms 000001AB Stream (144000 bytes) ... With all of this, you can adapt ffmpeg to handle your file (this should not be complex, video stream is a basic DV PAL) You can also check MediaInfo source code for LXF: http://mediainfo.svn.sourceforge.net/viewvc/mediainfo/MediaInfoLib/trunk/Source/MediaInfo/Multiple/File_Lxf.cpp?view=markup Note: LXF parsing in MediaInfo is 100% made by reverse engineering, so a lot of bytes are still opaque and I may have done some mistakes, this is only some help for you with unpacking LXF container. Good luck for implementing a better ffmpeg LXF parser! Note 2: your error is at offset [lxf @ 000000000035A920], this is inside the DV data (I intentionaly removed the DV trace in my dump, but I also have it), so maybe your problem is inside DV parsing, not at the LXF level. Jerome, developer of MediaInfo -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/LXF-decoder-Need-help-tp3507549p3510358.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From funnylookinhat at gmail.com Mon May 9 23:31:09 2011 From: funnylookinhat at gmail.com (David Overcash) Date: Mon, 9 May 2011 15:31:09 -0600 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: On Mon, May 9, 2011 at 2:58 PM, Kosta Vlotis wrote: > On Mon, May 9, 2011 at 4:02 PM, Phil Rhodes > wrote: > >> i'd like to have a server rack set up locally and have a script that > >> sends tasks to my "cloud". this solution seems idea for my needs but > >> for the amount of videos that we encode, paying someone for computer > >> time and bandwidth might mean the cost will be too high. > Have you considered simply using an external service or writing your own setup? I've created ( and currently manage ) two video encoding platforms that scale per the necessary requirements of my current encoding queue, but it was a lot of work, etc. I think you're looking for a silver bullet that will magically turn a bunch of computers into an encoding farm for you with little or no effort on your end, and you might as well just use encoding.comor a similar service if that's the case. Realistically though, if you end up going this alone, you might as well follow the instructions of the earlier post by using a CRON script to invoke the encoding job, and then simply manage your queue by sending files via NFS to your various boxes, etc. If you're beginning to think an external service is your best bet, send me an email (not to the list) and I can make a suggestion or two. Cheers, David From kosta.brazzers at gmail.com Mon May 9 23:53:35 2011 From: kosta.brazzers at gmail.com (Kosta Vlotis) Date: Mon, 9 May 2011 17:53:35 -0400 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: On Mon, May 9, 2011 at 5:31 PM, David Overcash wrote: > On Mon, May 9, 2011 at 2:58 PM, Kosta Vlotis wrote: > >> On Mon, May 9, 2011 at 4:02 PM, Phil Rhodes >> wrote: >> >> i'd like to have a server rack set up locally and have a script that >> >> sends tasks to my "cloud". this solution seems idea for my needs but >> >> for the amount of videos that we encode, paying someone for computer >> >> time and bandwidth might mean the cost will be too high. >> > > Have you considered simply using an external service or writing your own > setup? ?I've created ( and currently manage ) two video encoding platforms > that scale per the necessary requirements of my current encoding queue, but > it was a lot of work, etc. ?I think you're looking for a silver bullet that > will magically turn a bunch of computers into an encoding farm for you with > little or no effort on your end, and you might as well just use > encoding.comor a similar service if that's the case. > > Realistically though, if you end up going this alone, you might as well > follow the instructions of the earlier post by using a CRON script to invoke > the encoding job, and then simply manage your queue by sending files via NFS > to your various boxes, etc. > > If you're beginning to think an external service is your best bet, send me > an email (not to the list) and I can make a suggestion or two. > > Cheers, > David > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > im definately not looking for a magic solution. but this looks ideal. i just need to figure out the monthly cost of this: http://blog.picloud.com/tag/ffmpeg/. the advantage of an external service is i don't need to worry about buying/setting up/maintaining the servers. the down side is obviously cost.. building a solution is also on the table (probably what my employer would prefer). but i have very little knowledge of programming and scripting outside of a windows batch environment. but i am not opposed to learning as i go (i'm quite good at that :D ). i just need a place to start. From houndeyex at gmail.com Tue May 10 01:52:07 2011 From: houndeyex at gmail.com (James O.) Date: Mon, 9 May 2011 19:52:07 -0400 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: If you are looking for something totally headache free, I have heard good things about Panda from a start-up in my area. http://www.pandastream.com On Mon, May 9, 2011 at 5:53 PM, Kosta Vlotis wrote: > On Mon, May 9, 2011 at 5:31 PM, David Overcash > wrote: > > On Mon, May 9, 2011 at 2:58 PM, Kosta Vlotis >wrote: > > > >> On Mon, May 9, 2011 at 4:02 PM, Phil Rhodes > > >> wrote: > >> >> i'd like to have a server rack set up locally and have a script that > >> >> sends tasks to my "cloud". this solution seems idea for my needs but > >> >> for the amount of videos that we encode, paying someone for computer > >> >> time and bandwidth might mean the cost will be too high. > >> > > > > Have you considered simply using an external service or writing your own > > setup? I've created ( and currently manage ) two video encoding > platforms > > that scale per the necessary requirements of my current encoding queue, > but > > it was a lot of work, etc. I think you're looking for a silver bullet > that > > will magically turn a bunch of computers into an encoding farm for you > with > > little or no effort on your end, and you might as well just use > > encoding.comor a similar service if that's the case. > > > > Realistically though, if you end up going this alone, you might as well > > follow the instructions of the earlier post by using a CRON script to > invoke > > the encoding job, and then simply manage your queue by sending files via > NFS > > to your various boxes, etc. > > > > If you're beginning to think an external service is your best bet, send > me > > an email (not to the list) and I can make a suggestion or two. > > > > Cheers, > > David > > _______________________________________________ > > ffmpeg-user mailing list > > ffmpeg-user at ffmpeg.org > > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > > > im definately not looking for a magic solution. but this looks ideal. > i just need to figure out the monthly cost of this: > http://blog.picloud.com/tag/ffmpeg/. the advantage of an external > service is i don't need to worry about buying/setting up/maintaining > the servers. the down side is obviously cost.. > building a solution is also on the table (probably what my employer > would prefer). but i have very little knowledge of programming and > scripting outside of a windows batch environment. but i am not opposed > to learning as i go (i'm quite good at that :D ). i just need a place > to start. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From phamsyquybk at gmail.com Tue May 10 11:42:41 2011 From: phamsyquybk at gmail.com (Quy Pham Sy) Date: Tue, 10 May 2011 18:42:41 +0900 Subject: [FFmpeg-user] output_example.c compile error on Mac OSX Message-ID: Hi, I try to compile output_example.c on mac osx, and i got the error ouput: gcc -o output-example.o output-example.c -I/usr/local/include In file included from /usr/include/machine/_types.h:34, from /usr/include/sys/_types.h:33, from /usr/include/_types.h:27, from /usr/include/stdlib.h:63, from output-example.c:25: /usr/include/i386/_types.h:37: error: expected ?=?, ?,?, ?;?, ?asm? or ?__attribute__? before ?typedef? output-example.c: In function ?write_video_frame?: output-example.c:347: warning: ?sws_getContext? is deprecated (declared at /usr/local/include/libswscale/swscale.h:194) output-example.c:359: warning: passing argument 2 of ?sws_scale? from incompatible pointer type output-example.c: In function ?main?: output-example.c:495: warning: ?url_fopen? is deprecated (declared at /usr/local/include/libavformat/avio.h:424) output-example.c:548: warning: ?url_fclose? is deprecated (declared at /usr/local/include/libavformat/avio.h:425) can anyone help me with this? Thanks in advance, Quy From joolzg at btinternet.com Tue May 10 12:07:47 2011 From: joolzg at btinternet.com (JULIAN GARDNER) Date: Tue, 10 May 2011 11:07:47 +0100 (BST) Subject: [FFmpeg-user] ffmpeg and x264 In-Reply-To: References: <61386.97286.qm@web86408.mail.ird.yahoo.com><237821.66460.qm@web86407.mail.ird.yahoo.com> Message-ID: <401438.62782.qm@web86405.mail.ird.yahoo.com> OK tried this ffmpeg -i udp://230.10.0.104:1234 -vcodec libx264 -preset medium -profile baseline -b 750 -bn 125 -acodec libfaac -ab 96k -ar 32k -g 75 -async 1 udp://231.10.0.104:1234?pkt_size=1316 and i get this x264 [error]: invalid preset 'xcEum' now both ffmpeg and x264 are latest git pulls and builds Also if i remove the -preset i get x264 [error]: invalid profile 'ejd????? ' So me thinks a bug? or am i doing something wrong joolz ----- Original Message ----- > From: Dave Pope > To: FFmpeg user questions and RTFMs > Cc: > Sent: Saturday, 30 April 2011, 16:39 > Subject: Re: [FFmpeg-user] ffmpeg and x264 > >T hanks!? The change makes perfect sense, I just had to dig around for while > before figuring it out.? Do you know if there's a patch in the works to make > 'ffmpeg? --help' report something about these?? It seems like it might > require building some "handshaking" between them that might not > already exist in general form. > > ________________________________ > > From: ffmpeg-user-bounces at ffmpeg.org on behalf of Robert Kr?ger > Sent: Sat 4/30/2011 4:04 AM > To: FFmpeg user questions and RTFMs > Subject: Re: [FFmpeg-user] ffmpeg and x264 > > > > > On Apr 29, 2011, at 17:37 , Dave Pope wrote: > >> Yeah, I hit this one too.? Use "-preset" and "-profile" > switches instead >> of -vpre.? "-preset fast -profile baseline" works OK for me.? > Annoying >> that it's not reflected in the --help output; I don't see a way to > list >> the available values yet either. >> > > Check x264 docs. As far as I understood the purpose of the change was not to > duplicate efforts from the x264 project and that is documented quite well. > > try x264 --fullhelp > > > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > > > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From phamsyquybk at gmail.com Tue May 10 12:15:47 2011 From: phamsyquybk at gmail.com (Quy Pham Sy) Date: Tue, 10 May 2011 19:15:47 +0900 Subject: [FFmpeg-user] output_example.c compile error on Mac OSX In-Reply-To: References: Message-ID: I figured it out, some of my accident modification caused it. On Tue, May 10, 2011 at 6:42 PM, Quy Pham Sy wrote: > Hi, > > I try to compile output_example.c on mac osx, and i got the error ouput: > > gcc -o output-example.o output-example.c -I/usr/local/include > In file included from /usr/include/machine/_types.h:34, > from /usr/include/sys/_types.h:33, > from /usr/include/_types.h:27, > from /usr/include/stdlib.h:63, > from output-example.c:25: > /usr/include/i386/_types.h:37: error: expected ?=?, ?,?, ?;?, ?asm? or > ?__attribute__? before ?typedef? > output-example.c: In function ?write_video_frame?: > output-example.c:347: warning: ?sws_getContext? is deprecated (declared at > /usr/local/include/libswscale/swscale.h:194) > output-example.c:359: warning: passing argument 2 of ?sws_scale? from > incompatible pointer type > output-example.c: In function ?main?: > output-example.c:495: warning: ?url_fopen? is deprecated (declared at > /usr/local/include/libavformat/avio.h:424) > output-example.c:548: warning: ?url_fclose? is deprecated (declared at > /usr/local/include/libavformat/avio.h:425) > > > can anyone help me with this? > > Thanks in advance, > Quy > From adi235 at gmail.com Tue May 10 12:56:05 2011 From: adi235 at gmail.com (Aditya) Date: Tue, 10 May 2011 10:56:05 +0000 (UTC) Subject: [FFmpeg-user] useful link for output-example.c on osx Message-ID: http://www.cantgetnosleep.com/wordpress/?p=111#compiled1 From kosta.brazzers at gmail.com Tue May 10 16:16:46 2011 From: kosta.brazzers at gmail.com (Kosta Vlotis) Date: Tue, 10 May 2011 10:16:46 -0400 Subject: [FFmpeg-user] cloud encoding In-Reply-To: References: Message-ID: On Mon, May 9, 2011 at 7:52 PM, James O. wrote: > If you are looking for something totally headache free, I have heard good > things about Panda from a start-up in my area. http://www.pandastream.com > > On Mon, May 9, 2011 at 5:53 PM, Kosta Vlotis wrote: > >> On Mon, May 9, 2011 at 5:31 PM, David Overcash >> wrote: >> > On Mon, May 9, 2011 at 2:58 PM, Kosta Vlotis > >wrote: >> > >> >> On Mon, May 9, 2011 at 4:02 PM, Phil Rhodes > > >> >> wrote: >> >> >> i'd like to have a server rack set up locally and have a script that >> >> >> sends tasks to my "cloud". this solution seems idea for my needs but >> >> >> for the amount of videos that we encode, paying someone for computer >> >> >> time and bandwidth might mean the cost will be too high. >> >> >> > >> > Have you considered simply using an external service or writing your own >> > setup? ?I've created ( and currently manage ) two video encoding >> platforms >> > that scale per the necessary requirements of my current encoding queue, >> but >> > it was a lot of work, etc. ?I think you're looking for a silver bullet >> that >> > will magically turn a bunch of computers into an encoding farm for you >> with >> > little or no effort on your end, and you might as well just use >> > encoding.comor a similar service if that's the case. >> > >> > Realistically though, if you end up going this alone, you might as well >> > follow the instructions of the earlier post by using a CRON script to >> invoke >> > the encoding job, and then simply manage your queue by sending files via >> NFS >> > to your various boxes, etc. >> > >> > If you're beginning to think an external service is your best bet, send >> me >> > an email (not to the list) and I can make a suggestion or two. >> > >> > Cheers, >> > David >> > _______________________________________________ >> > ffmpeg-user mailing list >> > ffmpeg-user at ffmpeg.org >> > http://ffmpeg.org/mailman/listinfo/ffmpeg-user >> > >> >> im definately not looking for a magic solution. but this looks ideal. >> i just need to figure out the monthly cost of this: >> http://blog.picloud.com/tag/ffmpeg/. the advantage of an external >> service is i don't need to worry about buying/setting up/maintaining >> the servers. the down side is obviously cost.. >> building a solution is also on the table (probably what my employer >> would prefer). but i have very little knowledge of programming and >> scripting outside of a windows batch environment. but i am not opposed >> to learning as i go (i'm quite good at that :D ). i just need a place >> to start. >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user >> > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > hmmm.. thanks for the suggestion.. this panda site doesn't say if they use ffmpeg though.. so far every service\product i've tested has given me inferior results to ffmpeg From koxaniy at mail.ru Tue May 10 19:48:38 2011 From: koxaniy at mail.ru (Tuuls) Date: Tue, 10 May 2011 10:48:38 -0700 (PDT) Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: References: Message-ID: <1305049718405-3512566.post@n4.nabble.com> Okay, and how to make MP4 clip with the bottom fields first? -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/dv-mp4-deinterlace-or-not-and-how-tp3499389p3512566.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From bahamutzero8825 at gmail.com Tue May 10 19:57:09 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Tue, 10 May 2011 12:57:09 -0500 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <1305049718405-3512566.post@n4.nabble.com> References: <1305049718405-3512566.post@n4.nabble.com> Message-ID: <4DC97C75.1070902@gmail.com> On 2011.05.10 12:48 PM, Tuuls wrote: > Okay, and how to make MP4 clip with the bottom fields first? I don't think field order can be specified in the container. The video stream holds that information. If you need to convert from TFF or progressive to BFF, there's likely a filter for that. If you need to override the flag in the video stream, you'll have to do so with a setting in whatever media player you use. From koxaniy at mail.ru Tue May 10 20:05:35 2011 From: koxaniy at mail.ru (Tuuls) Date: Tue, 10 May 2011 11:05:35 -0700 (PDT) Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC97C75.1070902@gmail.com> References: <1305049718405-3512566.post@n4.nabble.com> <4DC97C75.1070902@gmail.com> Message-ID: <1305050735825-3512605.post@n4.nabble.com> I do not need to change the field order. I need to keep the bottom field first from DV source clips. And x264 set top field first :( In the setting of the X264 is an option that allows you to specify the priority fields, but due to the introduction of its presets can not be fully applied. In page of x264 i read : pic-struct Default: Not Set Force sending pic_struct in Picture Timing SEI. Implied when you use --pulldown or --tff/--bff. Recommendation: Default http://mewiki.project357.com/wiki/X264_Settings#pic-struct how use this ? -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/dv-mp4-deinterlace-or-not-and-how-tp3499389p3512605.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From koxaniy at mail.ru Tue May 10 20:47:10 2011 From: koxaniy at mail.ru (Tuuls) Date: Tue, 10 May 2011 11:47:10 -0700 (PDT) Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <1305050735825-3512605.post@n4.nabble.com> References: <1305049718405-3512566.post@n4.nabble.com> <4DC97C75.1070902@gmail.com> <1305050735825-3512605.post@n4.nabble.com> Message-ID: <1305053230649-3512692.post@n4.nabble.com> I so so sorry. Now tested encoded with next parameters : -y -threads 2 -f mp4 -r 25 -vcodec libx264 -preset veryfast -profile main -top 0 -s 720x576 -b 1000k -aspect 5:4 from DV sourse with BFF , and ffmpeg create mp4 file with BFF fields first order ! Sorry :) -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/dv-mp4-deinterlace-or-not-and-how-tp3499389p3512692.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From Umair.Khan at uni-klu.ac.at Tue May 10 21:10:23 2011 From: Umair.Khan at uni-klu.ac.at (Umair.Khan at uni-klu.ac.at) Date: Tue, 10 May 2011 21:10:23 +0200 Subject: [FFmpeg-user] mjpeg decoding problem Message-ID: <4DC9A9BF020000E000010DB0@gwx1.uni-klu.ac.at> Hi, I am streaming MJPEG stream from an IP camera over RTSP. The stream runs fine in VLC but when I try to stream with FFMPEG, it gives following error: "mjpeg: unsupported coding type" I am using following command to record a video from the camera: ffmpeg -i rtsp://192.168.1.168:8555/PSIA/Streaming/channels/0?videoCodecType=MJPEG -vcodec mjpeg sample.avi Can anyone please tell me the solution of this issue? I have spent many days to resolve it, but to no avail. I will really appreciate any help in this regard. Regards, Dipl.-Ing. Umair Ali Khan Research Staff Member, Pervasive Computing Group Institute of Networked and Embedded Systems Klagenfurt University, Austria -------------------------------------------------------- Lakeside Park L.2.1.33, 9020 Klagenfurt Voice: +43(0)463 2700 3872 Fax: +43(0)463 2700 3679 -------------------------------------------------------- www.pervasive.uni-klu.ac.at From ethan at veetle.com Wed May 11 03:32:42 2011 From: ethan at veetle.com (Ethan Wang) Date: Tue, 10 May 2011 18:32:42 -0700 Subject: [FFmpeg-user] ffmpeg crashes using trying to convert Wirecast stream to flv Message-ID: <4DC9E73A.5020708@veetle.com> ffmpeg, when given a stream originally produced from a set up using Wirecast dies with the following error: [flv @ 0x90c2100]st:0 error, non monotone timestamps 10708119 >= 10708119 (This is the error message from ffmpeg-0.6.1, I tried ffmpeg-0.7 rc1 as well, which produces a similar error, worded slightly differently) Our exact set up is this: Wirecast -->QuickTime (h264 video, AAC audio) --> VLC (ASF container re-enveloping, still h264 video, AAC audio unmodified) --> ffmpeg (transcoding to FLV, h264 video at lower bitrate, mp3 audio) I would assume the re-enveloping from QuickTime container to ASF would not affect the DTS/PTS of the elementary streams within so it must then be some issue with the h264 video stream produced by Wirecast that ffmpeg does not like. Would it be advisable to hack it locally here so when ffmpeg encounters an error like this, it simply keep going, by either discarding the frame with the duplicate timestamp, or ignoring it altogether? Ethan. -- Live HD to millions for free. http://veetle.com From mailer.tovis at freemail.hu Wed May 11 09:33:03 2011 From: mailer.tovis at freemail.hu (tovis) Date: Wed, 11 May 2011 09:33:03 +0200 (CEST) Subject: [FFmpeg-user] ffmpeg video4linux2 grabber input selection Message-ID: Dear list! I have using a cheap 4 channel USB grabber, using driver EasyCap60 from sourceforge. First channel is working well, using command: $ffmpeg -s 640x480 -f video4linux2 -i /dev/video0 http://localhost:1090/feed1.ffm What syntax I should have use for other channels? Sincerely tovis From mailer.tovis at freemail.hu Wed May 11 09:47:21 2011 From: mailer.tovis at freemail.hu (tovis) Date: Wed, 11 May 2011 09:47:21 +0200 (CEST) Subject: [FFmpeg-user] Streaming delay Message-ID: <2343327bbbccbc2a84acd5578c896de0.squirrel@nasi> Dear list! I'm trying to get live streaming, from a camera attached to PC using cheap, 4 channel USB grabber. I have using sample configuration for ffserver (http://ffmpeg.org/sample.html) and appropriate command for ffmpeg : $ffmpeg -s 640x480 -f video4linux2 -i /dev/video0 http://localhost:10090/feed1.ffm To view the stream, on client I have using flowplayer. It's working well, exclude that the delay between real contents and stream is more then 10 sec on 100 MBIT ethernet LAN. What part (ffserver.conf or ffmpeg command, may be both) should I have "tune" to lower this delay? What parameter could lower this delay? Sincerely tovis From weikun0905 at gmail.com Wed May 11 10:56:03 2011 From: weikun0905 at gmail.com (Vincent,Wei) Date: Wed, 11 May 2011 16:56:03 +0800 Subject: [FFmpeg-user] mjpeg decoding problem In-Reply-To: <4DC9A9BF020000E000010DB0@gwx1.uni-klu.ac.at> References: <4DC9A9BF020000E000010DB0@gwx1.uni-klu.ac.at> Message-ID: Hi, As I know, the avi file format is use the CODEC_ID_MPEG4 as the codec, so this may not support the MJPEG, you may try the MPEG4 stream , or you can edit ffmpeg.c to change the codec type ,and rebuild the ffmpeg. AVOutputFormat avi_muxer = { "avi", "avi format", "video/x-msvideo", "avi", sizeof(AVIContext), CODEC_ID_MP2, CODEC_ID_MPEG4, avi_write_header, avi_write_packet, avi_write_trailer, .codec_tag= (const AVCodecTag*[]){codec_bmp_tags, codec_wav_tags, 0}, }; 2011/5/11 > Hi, > > > I am streaming MJPEG stream from an IP camera over RTSP. The stream runs > fine in VLC but when I try to stream with FFMPEG, it gives following error: > > > "mjpeg: unsupported coding type" > > > I am using following command to record a video from the camera: > > > ffmpeg -i rtsp:// > 192.168.1.168:8555/PSIA/Streaming/channels/0?videoCodecType=MJPEG -vcodec > mjpeg sample.avi > > > Can anyone please tell me the solution of this issue? I have spent many > days to resolve it, but to no avail. I will really appreciate any help in > this regard. > > > Regards, > > > Dipl.-Ing. Umair Ali Khan > Research Staff Member, > Pervasive Computing Group > Institute of Networked and Embedded Systems > Klagenfurt University, Austria > -------------------------------------------------------- > Lakeside Park L.2.1.33, 9020 Klagenfurt > Voice: +43(0)463 2700 3872 > Fax: +43(0)463 2700 3679 > -------------------------------------------------------- > www.pervasive.uni-klu.ac.at > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From weikun0905 at gmail.com Wed May 11 10:56:03 2011 From: weikun0905 at gmail.com (Vincent,Wei) Date: Wed, 11 May 2011 16:56:03 +0800 Subject: [FFmpeg-user] mjpeg decoding problem In-Reply-To: <4DC9A9BF020000E000010DB0@gwx1.uni-klu.ac.at> References: <4DC9A9BF020000E000010DB0@gwx1.uni-klu.ac.at> Message-ID: Hi, As I know, the avi file format is use the CODEC_ID_MPEG4 as the codec, so this may not support the MJPEG, you may try the MPEG4 stream , or you can edit ffmpeg.c to change the codec type ,and rebuild the ffmpeg. AVOutputFormat avi_muxer = { "avi", "avi format", "video/x-msvideo", "avi", sizeof(AVIContext), CODEC_ID_MP2, CODEC_ID_MPEG4, avi_write_header, avi_write_packet, avi_write_trailer, .codec_tag= (const AVCodecTag*[]){codec_bmp_tags, codec_wav_tags, 0}, }; 2011/5/11 > Hi, > > > I am streaming MJPEG stream from an IP camera over RTSP. The stream runs > fine in VLC but when I try to stream with FFMPEG, it gives following error: > > > "mjpeg: unsupported coding type" > > > I am using following command to record a video from the camera: > > > ffmpeg -i rtsp:// > 192.168.1.168:8555/PSIA/Streaming/channels/0?videoCodecType=MJPEG -vcodec > mjpeg sample.avi > > > Can anyone please tell me the solution of this issue? I have spent many > days to resolve it, but to no avail. I will really appreciate any help in > this regard. > > > Regards, > > > Dipl.-Ing. Umair Ali Khan > Research Staff Member, > Pervasive Computing Group > Institute of Networked and Embedded Systems > Klagenfurt University, Austria > -------------------------------------------------------- > Lakeside Park L.2.1.33, 9020 Klagenfurt > Voice: +43(0)463 2700 3872 > Fax: +43(0)463 2700 3679 > -------------------------------------------------------- > www.pervasive.uni-klu.ac.at > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From Umair.Khan at uni-klu.ac.at Wed May 11 11:32:09 2011 From: Umair.Khan at uni-klu.ac.at (Umair.Khan at uni-klu.ac.at) Date: Wed, 11 May 2011 11:32:09 +0200 Subject: [FFmpeg-user] mjpeg decoding problem Message-ID: <4DCA73BA020000E000010DBD@gwx1.uni-klu.ac.at> Hi, My aim is not just to convert into avi format. I want to get the MJPEG streaming and save the images into JPEG. In every case, ffmpeg shows the error "mjpeg: unsupported codec type". When I looked into ffmpeg supported codecs, I could only find MKTAG ('M', 'J', 'P', 'G') and no MJPEG. Offcourse this is mjpeg codec, but my camera streams with the codec ID: MJPEG. Anyone please suggest me what should I do? Regards, Dipl.-Ing. Umair Ali Khan Research Staff Member, Pervasive Computing Group Institute of Networked and Embedded Systems Klagenfurt University, Austria -------------------------------------------------------- Lakeside Park L.2.1.33, 9020 Klagenfurt Voice: +43(0)463 2700 3872 Fax: +43(0)463 2700 3679 -------------------------------------------------------- www.pervasive.uni-klu.ac.at >>> "Vincent,Wei" 05/11/11 10:56 AM >>> Hi, As I know, the avi file format is use the CODEC_ID_MPEG4 as the codec, so this may not support the MJPEG, you may try the MPEG4 stream , or you can edit ffmpeg.c to change the codec type ,and rebuild the ffmpeg. AVOutputFormat avi_muxer = { "avi", "avi format", "video/x-msvideo", "avi", sizeof(AVIContext), CODEC_ID_MP2, CODEC_ID_MPEG4, avi_write_header, avi_write_packet, avi_write_trailer, .codec_tag= (const AVCodecTag*[]){codec_bmp_tags, codec_wav_tags, 0}, }; 2011/5/11 > Hi, > > > I am streaming MJPEG stream from an IP camera over RTSP. The stream runs > fine in VLC but when I try to stream with FFMPEG, it gives following error: > > > "mjpeg: unsupported coding type" > > > I am using following command to record a video from the camera: > > > ffmpeg -i rtsp:// > 192.168.1.168:8555/PSIA/Streaming/channels/0?videoCodecType=MJPEG -vcodec > mjpeg sample.avi > > > Can anyone please tell me the solution of this issue? I have spent many > days to resolve it, but to no avail. I will really appreciate any help in > this regard. > > > Regards, > > > Dipl.-Ing. Umair Ali Khan > Research Staff Member, > Pervasive Computing Group > Institute of Networked and Embedded Systems > Klagenfurt University, Austria > -------------------------------------------------------- > Lakeside Park L.2.1.33, 9020 Klagenfurt > Voice: +43(0)463 2700 3872 > Fax: +43(0)463 2700 3679 > -------------------------------------------------------- > www.pervasive.uni-klu.ac.at > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > _______________________________________________ ffmpeg-user mailing list ffmpeg-user at ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user From andrew at designwildwest.com Wed May 11 11:42:31 2011 From: andrew at designwildwest.com (Andrew Pettican) Date: Wed, 11 May 2011 10:42:31 +0100 Subject: [FFmpeg-user] -vf pad and -aspect problem Message-ID: <76617E67-6AB4-4EDE-85BA-AFF4CDBA7F41@designwildwest.com> Hi, I'm having a little trouble with the -vf pad and -aspect options. The basic story is that I have a source video (176x144, approx 5:4) which should be scaled up to 384x216 (16:9) adding pillerbox padding where necessary. I know scaling up should be avoided, but this is for a website which expects all videos to have identical dimensions. On my local PC I use the following command to achieve this, and it appears to work fine: c:\ffmpeg\bin\ffmpeg -i c:\ffmpeg\original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red c:\ffmpeg\test-output-aspect.16.9.webm i.e. scale from 176x144 to 264x216 then pad to 384x216 (60px each side) The returned video is 384x216 and in a 16:9 aspect ratio: Stream #0.0(eng): Video: libvpx, yuv420p, 384x216 [PAR 1:1 DAR 16:9], q=2-31, 400 kb/s, 1k tbn, 15 tbc ------ The problem lies with my web server where this will actually be running when live. The same command appears to stretch the video, making it extra wide. The returned video is still 384x216 but it no longer has a 1:1 pixel aspect ratio so the video ends up with a 256:99 (16:6.19) aspect ratio: Stream #0.0(eng): Video: libvpx, yuv420p, 384x216 [PAR 16:11 DAR 256:99], q=2-31, 400 kb/s, 1k tbn, 15 tbc Examining frames from both of these videos in VLC: - the PC has 60px of padding, 264px of image, and 60px of padding = 384px total width. This is what I expect. - the server has approx 87px of padding, 384px of image, and 87px of padding = 558px total width - Both videos are exactly 216px high as expected. It seems to me that the server applies the 16:9 aspect and ignores the padding that was applied. This stretches the 5:4 video to a 16:9 aspect and also stretches the 60px of padding to 87px on either side. I've also tried running the commands on both machines without the -aspect option. In this case both machines generate the same size video: - Stream #0.0(eng): Video: libvpx, yuv420p, 384x216 [PAR 193:176 DAR 193:99], q=2-31, 400 kb/s, 1k tbn, 15 tbc (pc) - Stream #0.0(eng): Video: libvpx, yuv420p, 384x216 [PAR 193:176 DAR 193:99], q=2-31, 400 kb/s, 1k tbn, 15 tbc (server) Note: - The version of ffmpeg on my PC was obtained from http://hawkeye.arrozcru.org/ - "FFmpeg git-95f163b 32-bit Static (Latest)" - The version of ffmpeg on my server was compiled from source by my server administrator. Any help/suggestions you have would be greatly appreciated. Is there something wrong with my command, or anything I can do to make it clear to ffmpeg what I want? The full ffmpeg output for each command is as follows: PC: ---------------------------------- C:\Users\Andrew>c:\ffmpeg\bin\ffmpeg -i c:\ffmpeg\original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red c:\ffmpeg\test-output-aspect.16.9.webm ffmpeg version git-N-29638-g95f163b, Copyright (c) 2000-2011 the FFmpeg developers built on May 6 2011 12:50:01 with gcc 4.5.3 configuration: --enable-gpl --enable-version3 --enable-memalign-hack --enable-runtime-cpudetect --enable-avisynth --enable-bzlib --enable-frei0r --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libfreetype --enable-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-librtmp --enable-libschroedinger --enable-libspeex --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libxavs --enable-libxvid --enable-zlib --pkg-config=pkg-config libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 3. 0 / 53. 3. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 4. 0 / 2. 4. 0 libswscale 0. 14. 0 / 0. 14. 0 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'c:\ffmpeg\original.mp4': Metadata: major_brand : mp42 minor_version : 0 compatible_brands: mp42isom creation_time : 2010-04-09 18:16:16 encoder : mp4creator 1.4.4 Duration: 00:02:33.93, start: 0.000000, bitrate: 197 kb/s Stream #0.0(eng): Video: mpeg4, yuv420p, 176x144 [PAR 193:176 DAR 193:144], 133 kb/s, 15 fps, 15 tbr, 90k tbn, 15 tbc Metadata: creation_time : 2010-04-09 18:16:16 Stream #0.1(eng): Audio: aac, 11025 Hz, stereo, s16, 63 kb/s Metadata: creation_time : 2010-04-09 18:16:17 [buffer @ 029964E0] w:176 h:144 pixfmt:yuv420p [scale @ 029968A0] w:176 h:144 fmt:yuv420p -> w:264 h:216 fmt:yuv420p flags:0x4 [pad @ 02996D00] w:264 h:216 -> w:384 h:216 x:60 y:0 color:0x515AF0FF[yuva] [libvpx @ 0029F6C0] v0.9.6 Output #0, webm, to 'c:\ffmpeg\test-output-aspect.16.9.webm': Metadata: major_brand : mp42 minor_version : 0 compatible_brands: mp42isom creation_time : 2010-04-09 18:16:16 encoder : Lavf53.0.3 Stream #0.0(eng): Video: libvpx, yuv420p, 384x216 [PAR 1:1 DAR 16:9], q=2-31, 400 kb/s, 1k tbn, 15 tbc Metadata: creation_time : 2010-04-09 18:16:16 Stream #0.1(eng): Audio: libvorbis, 22050 Hz, stereo, s16, 64 kb/s Metadata: creation_time : 2010-04-09 18:16:17 Stream mapping: Stream #0.0 -> #0.0 Stream #0.1 -> #0.1 Press [q] to stop encoding frame= 2309 fps= 42 q=0.0 Lsize= 8308kB time=153.72 bitrate= 442.8kbits/s video:7515kB audio:726kB global headers:4kB muxing overhead 0.772855% Server: ---------------------------------- root at WildWildWest [~]> nice -n 19 /usr/bin/ffmpeg -i /path/to/video/original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red /path/to/video/test-output-aspect.16.9.webm FFmpeg version SVN-r26400, Copyright (c) 2000-2011 the FFmpeg developers built on May 10 2011 16:39:45 with gcc 4.1.2 20080704 (Red Hat 4.1.2-50) configuration: --prefix=/usr --libdir=/usr/lib64 --shlibdir=/usr/lib64 --mandir=/usr/share/man --incdir=/usr/include --disable-avisynth --extra-cflags='-O2 -g -m64 -mtune=generic -fPIC' --enable-avfilter --enable-libdirac --enable-libfaac --enable-libgsm --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libvorbis --enable-libvpx --enable-libx264 --enable-gpl --enable-nonfree --enable-postproc --enable-pthreads --enable-shared --enable-swscale --enable-vdpau --enable-version3 --enable-x11grab libavutil 50.36. 0 / 50.36. 0 libavcore 0.16. 1 / 0.16. 1 libavcodec 52.108. 0 / 52.108. 0 libavformat 52.93. 0 / 52.93. 0 libavdevice 52. 2. 3 / 52. 2. 3 libavfilter 1.74. 0 / 1.74. 0 libswscale 0.12. 0 / 0.12. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from '/path/to/video/original.mp4': Metadata: major_brand : mp42 minor_version : 0 compatible_brands: mp42isom creation_time : 2010-04-09 18:16:16 encoder : mp4creator 1.4.4 Duration: 00:02:33.93, start: 0.000000, bitrate: 197 kb/s Stream #0.0(eng): Video: mpeg4, yuv420p, 176x144 [PAR 193:176 DAR 193:144], 133 kb/s, 15 fps, 15 tbr, 90k tbn, 15 tbc Metadata: creation_time : 2010-04-09 18:16:16 Stream #0.1(eng): Audio: aac, 11025 Hz, stereo, s16, 63 kb/s Metadata: creation_time : 2010-04-09 18:16:17 [buffer @ 0xeb1e1b0] w:176 h:144 pixfmt:yuv420p [scale @ 0xeb1e580] w:176 h:144 fmt:yuv420p -> w:264 h:216 fmt:yuv420p flags:0xa0000004 [pad @ 0xeb1e8b0] w:264 h:216 -> w:384 h:216 x:60 y:0 color:0x515AF0FF[yuva] [libvpx @ 0xeb1aa90] v0.9.6 Output #0, webm, to '/path/to/video/test-output-aspect.16.9.webm': Metadata: major_brand : mp42 minor_version : 0 compatible_brands: mp42isom creation_time : 2010-04-09 18:16:16 encoder : Lavf52.93.0 Stream #0.0(eng): Video: libvpx, yuv420p, 384x216 [PAR 16:11 DAR 256:99], q=2-31, 400 kb/s, 1k tbn, 15 tbc Metadata: creation_time : 2010-04-09 18:16:16 Stream #0.1(eng): Audio: libvorbis, 22050 Hz, stereo, s16, 64 kb/s Metadata: creation_time : 2010-04-09 18:16:17 Stream mapping: Stream #0.0 -> #0.0 Stream #0.1 -> #0.1 Press [q] to stop encoding frame= 2309 fps= 60 q=0.0 Lsize= 8327kB time=153.72 bitrate= 443.8kbits/s video:7516kB audio:744kB global headers:4kB muxing overhead 0.772233% From etienne.buira.lists at free.fr Wed May 11 12:22:59 2011 From: etienne.buira.lists at free.fr (Etienne Buira) Date: Wed, 11 May 2011 12:22:59 +0200 Subject: [FFmpeg-user] -vf pad and -aspect problem In-Reply-To: <76617E67-6AB4-4EDE-85BA-AFF4CDBA7F41@designwildwest.com> References: <76617E67-6AB4-4EDE-85BA-AFF4CDBA7F41@designwildwest.com> Message-ID: <20110511102258.GA28878@epicure.lazyet.homelinux.net> On Wed, May 11, 2011 at 10:42:31AM +0100, Andrew Pettican wrote: > Hi, Hi. ../.. > C:\Users\Andrew>c:\ffmpeg\bin\ffmpeg -i c:\ffmpeg\original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red c:\ffmpeg\test-output-aspect.16.9.webm > ffmpeg version git-N-29638-g95f163b, Copyright (c) 2000-2011 the FFmpeg developers ../.. > Server: > ---------------------------------- > root at WildWildWest [~]> nice -n 19 /usr/bin/ffmpeg -i /path/to/video/original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red /path/to/video/test-output-aspect.16.9.webm > FFmpeg version SVN-r26400, Copyright (c) 2000-2011 the FFmpeg developers ../.. Some issues with filters/old style -s -aspect... have been adressed by Baptiste Coudurier, after the switch to git IIRC, so try with a newer version. From Umair.Khan at uni-klu.ac.at Wed May 11 16:08:34 2011 From: Umair.Khan at uni-klu.ac.at (Umair.Khan at uni-klu.ac.at) Date: Wed, 11 May 2011 16:08:34 +0200 Subject: [FFmpeg-user] mjpeg decoding problem Message-ID: <4DCAB482020000E000010DD8@gwx1.uni-klu.ac.at> I also tried output-example.c in libavcodec. When I try to run it with the following arguments: ./output-example rtsp://192.168.1.168:8555/PSIA/Streaming/channels/0?videoCodecType=MJPEG I get following errors: [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cd) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c8) [mjpeg @ 0x9846750] Can not process SOS before SOF, skipping [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (ce) [mjpeg @ 0x9846750] dqt: 16bit precision [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cd) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (ca) [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c6) [mjpeg @ 0x9846750] invalid id 61 [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cd) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] [IMGUTILS @ 0xbfb7f138] Picture size 51117x18481 is invalid [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cf) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c7) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c5) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cb) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cd) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cd) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c7) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c5) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] invalid id 198 [mjpeg @ 0x9846750] dqt: 16bit precision [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c6) [mjpeg @ 0x9846750] [IMGUTILS @ 0xbfb7f138] Picture size 27136x19751 is invalid [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c5) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (ca) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c5) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cd) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cf) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c7) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (ce) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c6) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (cd) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] Can not process SOS before SOF, skipping [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c6) [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] Can not process SOS before SOF, skipping [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c6) [mjpeg @ 0x9846750] Can not process SOS before SOF, skipping [mjpeg @ 0x9846750] Found EOI before any SOF, ignoring [mjpeg @ 0x9846750] mjpeg: unsupported coding type (ca) [mjpeg @ 0x9846750] mjpeg: unsupported coding type (c7) [mjpeg @ 0x9846750] only 8 bits/component accepted [mjpeg @ 0x9846750] dqt: 16bit precision [mjpeg @ 0x9846750] No JPEG data found in image Please anyone help me what is this problem? Regards, Dipl.-Ing. Umair Ali Khan Research Staff Member, Pervasive Computing Group Institute of Networked and Embedded Systems Klagenfurt University, Austria -------------------------------------------------------- Lakeside Park L.2.1.33, 9020 Klagenfurt Voice: +43(0)463 2700 3872 Fax: +43(0)463 2700 3679 -------------------------------------------------------- www.pervasive.uni-klu.ac.at >>> "Vincent,Wei" 05/11/11 10:56 AM >>> Hi, As I know, the avi file format is use the CODEC_ID_MPEG4 as the codec, so this may not support the MJPEG, you may try the MPEG4 stream , or you can edit ffmpeg.c to change the codec type ,and rebuild the ffmpeg. AVOutputFormat avi_muxer = { "avi", "avi format", "video/x-msvideo", "avi", sizeof(AVIContext), CODEC_ID_MP2, CODEC_ID_MPEG4, avi_write_header, avi_write_packet, avi_write_trailer, .codec_tag= (const AVCodecTag*[]){codec_bmp_tags, codec_wav_tags, 0}, }; 2011/5/11 > Hi, > > > I am streaming MJPEG stream from an IP camera over RTSP. The stream runs > fine in VLC but when I try to stream with FFMPEG, it gives following error: > > > "mjpeg: unsupported coding type" > > > I am using following command to record a video from the camera: > > > ffmpeg -i rtsp:// > 192.168.1.168:8555/PSIA/Streaming/channels/0?videoCodecType=MJPEG -vcodec > mjpeg sample.avi > > > Can anyone please tell me the solution of this issue? I have spent many > days to resolve it, but to no avail. I will really appreciate any help in > this regard. > > > Regards, > > > Dipl.-Ing. Umair Ali Khan > Research Staff Member, > Pervasive Computing Group > Institute of Networked and Embedded Systems > Klagenfurt University, Austria > -------------------------------------------------------- > Lakeside Park L.2.1.33, 9020 Klagenfurt > Voice: +43(0)463 2700 3872 > Fax: +43(0)463 2700 3679 > -------------------------------------------------------- > www.pervasive.uni-klu.ac.at > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > _______________________________________________ ffmpeg-user mailing list ffmpeg-user at ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-user From andrew at designwildwest.com Wed May 11 16:15:01 2011 From: andrew at designwildwest.com (Andrew Pettican) Date: Wed, 11 May 2011 15:15:01 +0100 Subject: [FFmpeg-user] -vf pad and -aspect problem In-Reply-To: <20110511102258.GA28878@epicure.lazyet.homelinux.net> References: <76617E67-6AB4-4EDE-85BA-AFF4CDBA7F41@designwildwest.com> <20110511102258.GA28878@epicure.lazyet.homelinux.net> Message-ID: <0705BDE8-33F2-4E0E-A23A-AF1FE5E1EFCA@designwildwest.com> Thanks for your reply. I'm sure that will solve the problem. I've asked my server admin to get the latest version of ffmpeg from git, however he is unsure of how to get the libavfilter filters. Specifically he has said to me: I was following instruction about avfilters, which is to use checkout.sh script that downloads ffmpeg version from svn and compile it after checkout. Can these avfilters be obtained from git? thanks Andrew On 11 May 2011, at 11:22, Etienne Buira wrote: > On Wed, May 11, 2011 at 10:42:31AM +0100, Andrew Pettican wrote: >> Hi, > > Hi. > > ../.. > >> C:\Users\Andrew>c:\ffmpeg\bin\ffmpeg -i c:\ffmpeg\original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red c:\ffmpeg\test-output-aspect.16.9.webm >> ffmpeg version git-N-29638-g95f163b, Copyright (c) 2000-2011 the FFmpeg developers > > ../.. > >> Server: >> ---------------------------------- >> root at WildWildWest [~]> nice -n 19 /usr/bin/ffmpeg -i /path/to/video/original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red /path/to/video/test-output-aspect.16.9.webm >> FFmpeg version SVN-r26400, Copyright (c) 2000-2011 the FFmpeg developers > > ../.. > > Some issues with filters/old style -s -aspect... have been adressed by > Baptiste Coudurier, after the switch to git IIRC, so try with a newer > version. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From etienne.buira.lists at free.fr Wed May 11 16:30:00 2011 From: etienne.buira.lists at free.fr (Etienne Buira) Date: Wed, 11 May 2011 16:30:00 +0200 Subject: [FFmpeg-user] -vf pad and -aspect problem In-Reply-To: <0705BDE8-33F2-4E0E-A23A-AF1FE5E1EFCA@designwildwest.com> References: <76617E67-6AB4-4EDE-85BA-AFF4CDBA7F41@designwildwest.com> <20110511102258.GA28878@epicure.lazyet.homelinux.net> <0705BDE8-33F2-4E0E-A23A-AF1FE5E1EFCA@designwildwest.com> Message-ID: <20110511143000.GB28878@epicure.lazyet.homelinux.net> On Wed, May 11, 2011 at 03:15:01PM +0100, Andrew Pettican wrote: > Thanks for your reply. I'm sure that will solve the problem. > > I've asked my server admin to get the latest version of ffmpeg from git, however he is unsure of how to get the libavfilter filters. > > Specifically he has said to me: I was following instruction about avfilters, which is to use checkout.sh script that downloads ffmpeg version from svn and compile it after checkout. > > Can these avfilters be obtained from git? > > thanks > Andrew Don't know everything about avfilter's history, but at least some are available using main git repo (including pad), which you could experiment with the version on your host. So I think you can forget about checkout.sh and avfilter repo. > > > On 11 May 2011, at 11:22, Etienne Buira wrote: > > > On Wed, May 11, 2011 at 10:42:31AM +0100, Andrew Pettican wrote: > >> Hi, > > > > Hi. > > > > ../.. > > > >> C:\Users\Andrew>c:\ffmpeg\bin\ffmpeg -i c:\ffmpeg\original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red c:\ffmpeg\test-output-aspect.16.9.webm > >> ffmpeg version git-N-29638-g95f163b, Copyright (c) 2000-2011 the FFmpeg developers > > > > ../.. > > > >> Server: > >> ---------------------------------- > >> root at WildWildWest [~]> nice -n 19 /usr/bin/ffmpeg -i /path/to/video/original.mp4 -vcodec libvpx -acodec libvorbis -ac 2 -ar 22050 -ab 64k -b 400k -r 15 -f webm -y -s 264x216 -aspect 16:9 -vf pad=384:216:60:0:red /path/to/video/test-output-aspect.16.9.webm > >> FFmpeg version SVN-r26400, Copyright (c) 2000-2011 the FFmpeg developers > > > > ../.. > > > > Some issues with filters/old style -s -aspect... have been adressed by > > Baptiste Coudurier, after the switch to git IIRC, so try with a newer > > version. > > _______________________________________________ > > ffmpeg-user mailing list > > ffmpeg-user at ffmpeg.org > > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From jshupert at pps-inc.com Wed May 11 18:55:23 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Wed, 11 May 2011 12:55:23 -0400 Subject: [FFmpeg-user] possible to do mp3 @ 48kHz Message-ID: <4DCABF7B.2040305@pps-inc.com> friends I am successful transcoding files with ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 44100 -b 8000000 -s 720x480 -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 44100 -b 1500000 -r 30 -s 720x480 -f flv D:\path\vid\my.flv but the problem is i would very much enjoy having the audio sampling rate of 48kHz ; -ar 48000 and libmp3lame only will do 44100 , 22050 , 11025 so if i really wanted to have mp3 48 kHz audio --- is there any solution? thanks From james.darnley at gmail.com Wed May 11 18:59:04 2011 From: james.darnley at gmail.com (James Darnley) Date: Wed, 11 May 2011 18:59:04 +0200 Subject: [FFmpeg-user] possible to do mp3 @ 48kHz In-Reply-To: <4DCABF7B.2040305@pps-inc.com> References: <4DCABF7B.2040305@pps-inc.com> Message-ID: On 11/05/2011, Jim Shupert wrote: > friends > > I am successful transcoding files with > > ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 44100 -b 8000000 -s 720x480 > -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi > ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 44100 -b 1500000 -r 30 -s > 720x480 -f flv D:\path\vid\my.flv > > but the problem is i would very much enjoy having the audio sampling > rate of 48kHz ; -ar 48000 > > and libmp3lame only will do 44100 , 22050 , 11025 > > so if i really wanted to have mp3 48 kHz audio --- is there any solution? lame happily supports 48 KHz. FLV does not. Stop using FLV and all will be fine. From lazarusportugalwebmasters at gmail.com Wed May 11 19:06:33 2011 From: lazarusportugalwebmasters at gmail.com (lazarusportugalwebmasters) Date: Wed, 11 May 2011 18:06:33 +0100 Subject: [FFmpeg-user] Record the screen with ffmpeg Message-ID: <4DCAC219.60505@gmail.com> Hello. I want record my own screen in video but I don't know how do it with ffmpeg. Anyone can help me with that. -- #lazarus-br em irc.freenode.org #lazarusportugal em irc.ptnet.org(Cloud IRC em lazarusportugal.org/doku.php/irc) Em caso de d?vida sobre o cliente a usar clique em http://lazarusportugal.org/doku.php/duvidas_irc Juntem-se a estes fant?sticos IRC's sobre Lazarus From cfaf at hotmail.com Wed May 11 19:14:09 2011 From: cfaf at hotmail.com (christian fafard) Date: Wed, 11 May 2011 17:14:09 +0000 Subject: [FFmpeg-user] Does ffmbc related questions belongs to this forum? Message-ID: If not then all my excuses... I'm trying to encode to IMX30 using the built-in target in ffmbc but i have an error message about smpte170m color primairies What am i doing wrong? D:\>ffmbc -i test.mpg -target imx30 test1.mov FFmpeg version FFmbc-0.6-rc4, Copyright (c) 2000-2011 the FFmpeg developers built on May 6 2011 12:04:48 with gcc 4.2.1-sjlj (mingw32-2) configuration: --cpu=i686 --arch=i686 --prefix=/usr/local/i586-mingw32msvc --t arget-os=mingw32 --enable-memalign-hack --cross-prefix=i586-mingw32msvc- --extra -cflags='--static -I $PREFIX/include -D_WIN32_WINNT=0x0501' --extra-ldflags='-st atic -L $PREFIX/lib' --enable-gpl --enable-version3 --enable-nonfree --enable-pt hreads --enable-static --disable-shared --enable-libdirac --enable-libfaac --ena ble-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-libspeex --enable-l ibtheora --enable-libvorbis --enable-libxvid --enable-libschroedinger --enable-l ibx264 --enable-libvpx libavutil 50. 38. 0 / 50. 38. 0 libavcodec 52.112. 1 / 52.112. 1 libavformat 52. 99. 0 / 52. 99. 0 libavdevice 52. 2. 3 / 52. 2. 3 libavfilter 1. 76. 0 / 1. 76. 0 libswscale 0. 12. 0 / 0. 12. 0 Input #0, mpeg, from 'test.mpg': Duration: 00:01:09.56, start: 0.213367, bitrate: 4253 kb/s Stream #0.0[0x1e0](und): Video: mpeg2video, yuv420p, 720x480i bff [PAR 8:9 D AR 4:3], 8000 kb/s, 29.97 fps Stream #0.1[0x1c0](und): Audio: mp2, 44100 Hz, stereo, s16, 224 kb/s [swscaler @ 0x2e268e0] [Eval @ 0x22ec20] Undefined constant or missing '(' in 's mpte170m' [swscaler @ 0x2e268e0] Unable to parse option value "smpte170m" Invalid value 'smpte170m' for option 'color_primaries' If i'm not using the target and instead enter the following command line, then it encode but the video, once imported into avid, is a white frame all way through. It's maybe the 'zig-zag' vs 'alternate' scan method issue? ffmbc -i test.mpg -vcodec mpeg2video -r 29.97 -pix_fmt yuv422p -minrate 30000k -maxrate 30000k -b 30000k -intra -flags +ildct+low_delay -dc 10 -flags2 +ivlc+non_linear_q -ps 1 -qmin 1 -qmax 8 -top 1 -bufsize 1200000 -rc_init_occupancy 1200000 test3.mov thanks Christian From lazarusportugalwebmasters at gmail.com Wed May 11 19:21:29 2011 From: lazarusportugalwebmasters at gmail.com (lazarusportugalwebmasters) Date: Wed, 11 May 2011 18:21:29 +0100 Subject: [FFmpeg-user] Record the screen with ffmpeg Message-ID: <4DCAC599.1030803@gmail.com> Hello. I use this comand and it giveme the bug : ffmpeg -f alsa -i hw:1 -f x11grab -s 800x600 -r 24 -b 100k -bf 2 -g 300 -i :0.0 -ar 22050 -ab 64k -acodec libmp3lame outputvid.mpeg It give-me the bug Unknown input format: 'alsa'. -- #lazarus-br em irc.freenode.org #lazarusportugal em irc.ptnet.org(Cloud IRC em lazarusportugal.org/doku.php/irc) Em caso de d?vida sobre o cliente a usar clique em http://lazarusportugal.org/doku.php/duvidas_irc Juntem-se a estes fant?sticos IRC's sobre Lazarus From lazarusportugalwebmasters at gmail.com Wed May 11 19:30:36 2011 From: lazarusportugalwebmasters at gmail.com (lazarusportugalwebmasters) Date: Wed, 11 May 2011 18:30:36 +0100 Subject: [FFmpeg-user] Record the screen with ffmpeg Message-ID: <4DCAC7BC.7030600@gmail.com> I'm using windows to do it. -- #lazarus-br em irc.freenode.org #lazarusportugal em irc.ptnet.org(Cloud IRC em lazarusportugal.org/doku.php/irc) Em caso de d?vida sobre o cliente a usar clique em http://lazarusportugal.org/doku.php/duvidas_irc Juntem-se a estes fant?sticos IRC's sobre Lazarus From ffmpeg at neoprimitive.net Wed May 11 20:19:56 2011 From: ffmpeg at neoprimitive.net (jlucas) Date: Wed, 11 May 2011 14:19:56 -0400 Subject: [FFmpeg-user] Record the screen with ffmpeg In-Reply-To: <4DCAC7BC.7030600@gmail.com> References: <4DCAC7BC.7030600@gmail.com> Message-ID: <20110511181955.GZ5996@neoprimitive.net> On 11/05/11 18:30 +0100, lazarusportugalwebmasters wrote: > I'm using windows to do it. ALSA stands for Advanced Linux Sound Architecture, so the fact you're running Windows explains why using '-f alsa' isn't working out for you. From houndeyex at gmail.com Wed May 11 21:11:33 2011 From: houndeyex at gmail.com (James O.) Date: Wed, 11 May 2011 15:11:33 -0400 Subject: [FFmpeg-user] Record the screen with ffmpeg In-Reply-To: <20110511181955.GZ5996@neoprimitive.net> References: <4DCAC7BC.7030600@gmail.com> <20110511181955.GZ5996@neoprimitive.net> Message-ID: If you're on Windows you might try BB FlashBack Express. The express edition is free to use. http://www.bbsoftware.co.uk/BBFlashBack_FreePlayer.aspx On Wed, May 11, 2011 at 2:19 PM, jlucas wrote: > On 11/05/11 18:30 +0100, lazarusportugalwebmasters wrote: > > I'm using windows to do it. > > ALSA stands for Advanced Linux Sound Architecture, so the fact you're > running Windows explains why using '-f alsa' isn't working out for > you. > > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From cfaf at hotmail.com Wed May 11 21:18:10 2011 From: cfaf at hotmail.com (christian fafard) Date: Wed, 11 May 2011 19:18:10 +0000 Subject: [FFmpeg-user] Does ffmbc related questions belongs to this forum? In-Reply-To: References: Message-ID: > From: cfaf at hotmail.com > To: ffmpeg-user at ffmpeg.org > Date: Wed, 11 May 2011 17:14:09 +0000 > Subject: [FFmpeg-user] Does ffmbc related questions belongs to this forum? > > > If not then all my excuses... > > I'm trying to encode to IMX30 using the built-in target in ffmbc but i have an error message about smpte170m color primairies > What am i doing wrong? > > D:\>ffmbc -i test.mpg -target imx30 test1.mov > FFmpeg version FFmbc-0.6-rc4, Copyright (c) 2000-2011 the FFmpeg developers > built on May 6 2011 12:04:48 with gcc 4.2.1-sjlj (mingw32-2) > configuration: --cpu=i686 --arch=i686 --prefix=/usr/local/i586-mingw32msvc --t > arget-os=mingw32 --enable-memalign-hack --cross-prefix=i586-mingw32msvc- --extra > -cflags='--static -I $PREFIX/include -D_WIN32_WINNT=0x0501' --extra-ldflags='-st > atic -L $PREFIX/lib' --enable-gpl --enable-version3 --enable-nonfree --enable-pt > hreads --enable-static --disable-shared --enable-libdirac --enable-libfaac --ena > ble-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-libspeex --enable-l > ibtheora --enable-libvorbis --enable-libxvid --enable-libschroedinger --enable-l > ibx264 --enable-libvpx > libavutil 50. 38. 0 / 50. 38. 0 > libavcodec 52.112. 1 / 52.112. 1 > libavformat 52. 99. 0 / 52. 99. 0 > libavdevice 52. 2. 3 / 52. 2. 3 > libavfilter 1. 76. 0 / 1. 76. 0 > libswscale 0. 12. 0 / 0. 12. 0 > Input #0, mpeg, from 'test.mpg': > Duration: 00:01:09.56, start: 0.213367, bitrate: 4253 kb/s > Stream #0.0[0x1e0](und): Video: mpeg2video, yuv420p, 720x480i bff [PAR 8:9 D > AR 4:3], 8000 kb/s, 29.97 fps > Stream #0.1[0x1c0](und): Audio: mp2, 44100 Hz, stereo, s16, 224 kb/s > [swscaler @ 0x2e268e0] [Eval @ 0x22ec20] Undefined constant or missing '(' in 's > mpte170m' > [swscaler @ 0x2e268e0] Unable to parse option value "smpte170m" > Invalid value 'smpte170m' for option 'color_primaries' > > > > If i'm not using the target and instead enter the following command line, then it encode but the video, once imported into avid, is a white frame all way through. > It's maybe the 'zig-zag' vs 'alternate' scan method issue? > > ffmbc -i test.mpg -vcodec mpeg2video -r 29.97 -pix_fmt yuv422p -minrate 30000k -maxrate 30000k -b 30000k -intra -flags +ildct+low_delay -dc 10 -flags2 +ivlc+non_linear_q -ps 1 -qmin 1 -qmax 8 -top 1 -bufsize 1200000 -rc_init_occupancy 1200000 test3.mov > > thanks > Christian > _______________________________________________ One more thing, Where can i get more info about the bistream filters? U:\>ffmbc -bsfs FFmpeg version FFmbc-0.6-rc4, Copyright (c) 2000-2011 the FFmpeg developers built on May 6 2011 12:04:48 with gcc 4.2.1-sjlj (mingw32-2) configuration: --cpu=i686 --arch=i686 --prefix=/usr/local/i586-mingw32msvc - arget-os=mingw32 --enable-memalign-hack --cross-prefix=i586-mingw32msvc- --ext -cflags='--static -I $PREFIX/include -D_WIN32_WINNT=0x0501' --extra-ldflags='- atic -L $PREFIX/lib' --enable-gpl --enable-version3 --enable-nonfree --enable- hreads --enable-static --disable-shared --enable-libdirac --enable-libfaac --e ble-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-libspeex --enable ibtheora --enable-libvorbis --enable-libxvid --enable-libschroedinger --enable ibx264 --enable-libvpx libavutil 50. 38. 0 / 50. 38. 0 libavcodec 52.112. 1 / 52.112. 1 libavformat 52. 99. 0 / 52. 99. 0 libavdevice 52. 2. 3 / 52. 2. 3 libavfilter 1. 76. 0 / 1. 76. 0 libswscale 0. 12. 0 / 0. 12. 0 Bitstream filters: text2movsub remove_extra noise mov2textsub mpeg2setdar mpeg2seqdump mp3decomp mp3comp mjpegadump mjpeg2jpeg imxremoveklv imxdump h264_mp4toannexb dump_extra chomp aac_adtstoasc The 'imxremoveklv' and 'imxdump' caught my attention and i'm wondering if they could be of any use in my problem. Thanks Christian From jshupert at pps-inc.com Wed May 11 21:22:22 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Wed, 11 May 2011 15:22:22 -0400 Subject: [FFmpeg-user] possible to do mp3 @ 48kHz In-Reply-To: References: <4DCABF7B.2040305@pps-inc.com> Message-ID: <4DCAE1EE.8020004@pps-inc.com> On 5/11/2011 12:59 PM, James Darnley wrote: > On 11/05/2011, Jim Shupert wrote: >> friends >> >> I am successful transcoding files with >> >> ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 44100 -b 8000000 -s 720x480 >> -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi >> ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 44100 -b 1500000 -r 30 -s >> 720x480 -f flv D:\path\vid\my.flv >> >> but the problem is i would very much enjoy having the audio sampling >> rate of 48kHz ; -ar 48000 >> >> and libmp3lame only will do 44100 , 22050 , 11025 >> >> so if i really wanted to have mp3 48 kHz audio --- is there any solution? > lame happily supports 48 KHz. FLV does not. Stop using FLV and all > will be fine. > _______________________________________________ > Yes, thank you -- you are exactly correct. I am have just made 48kHz mp3 in the avi.. it seems (from what i have now read) that FLV supports only 4 rates 5.5 kHz, 11 kHz, 22.05 kHz, 44.1 kHz But I now have a second Q of the same encodes if I look at my result in mediainfo i see that the video datarate is not what i used in my command mediainfo reports the 2 files at 2 diffrent rates Bit rate : 4557 Kbps Bit rate : 2983 Kbps my command is ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 48000 -b 8000000 -s 720x480 -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi so insted of -b 8000000 [ 8 mbps ] i am getting a vbr ? is that true and how do i make it cbr @ 8mbps thanks (again) js From jshupert at pps-inc.com Wed May 11 21:34:49 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Wed, 11 May 2011 15:34:49 -0400 Subject: [FFmpeg-user] Record the screen with ffmpeg In-Reply-To: References: <4DCAC7BC.7030600@gmail.com> <20110511181955.GZ5996@neoprimitive.net> Message-ID: <4DCAE4D9.10303@pps-inc.com> On 5/11/2011 3:11 PM, James O. wrote: > If you're on Windows you might try BB FlashBack Express. The express edition > is free to use. > > http://www.bbsoftware.co.uk/BBFlashBack_FreePlayer.aspx > > On Wed, May 11, 2011 at 2:19 PM, jlucas wrote: > >> On 11/05/11 18:30 +0100, lazarusportugalwebmasters wrote: >>> I'm using windows to do it. >> ALSA stands for Advanced Linux Sound Architecture, so the fact you're >> running Windows explains why using '-f alsa' isn't working out for >> you. since you are on windows - you could use the windows media encoder screencapture codec -- google the: windows media encoder sdk or....just get a linux box :) From denis.muraviev at mac.com Wed May 11 23:19:50 2011 From: denis.muraviev at mac.com (Denis Muraviev) Date: Thu, 12 May 2011 01:19:50 +0400 Subject: [FFmpeg-user] ac3 6ch -> 6 mono wav files Message-ID: <7C2FDC84-1149-40B0-9EA4-86EC431A6B10@mac.com> Hi, I have .ac3 file with 6 channels, and I want to decode it to 6 different mono .wav files (every channel need to be in separate wav file). I found that there is `-ac channels' to set the number of audio channels, but how can I select the channel I need (not the first one). Thanks, Denis. From baptiste.coudurier at gmail.com Wed May 11 23:42:03 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Wed, 11 May 2011 14:42:03 -0700 Subject: [FFmpeg-user] Does ffmbc related questions belongs to this forum? In-Reply-To: References: Message-ID: <4DCB02AB.2070206@gmail.com> Hi, On 05/11/2011 10:14 AM, christian fafard wrote: > > If not then all my excuses... No. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From lclemens at gmail.com Thu May 12 01:05:44 2011 From: lclemens at gmail.com (Ph0t0n) Date: Wed, 11 May 2011 16:05:44 -0700 (PDT) Subject: [FFmpeg-user] Flip in sws_scale In-Reply-To: <7049fdd50901081108j147e55a6m5b65f71c260e1572@mail.gmail.com> References: <7049fdd50901081108j147e55a6m5b65f71c260e1572@mail.gmail.com> Message-ID: <1305155144405-3516095.post@n4.nabble.com> A filter would work, but if you're doing a conversion/scaling anyway with sw_scale(), theoretically it's faster to flip at the same time. I saw this posted somewhere else... int nDivisor, nMaxLineSize = 0; // find max linesize for (int i = 0; i < 4; i++) { if (pic->linesize[i] > nMaxLineSize) { nMaxLineSize = pic->linesize[i]; } } if (pic->linesize[0]) { nDivisor = (nMaxLineSize / pic->linesize[0]); if (!nDivisor) { nDivisor = 1; } pic->data[0] += (pic->linesize[0] * ((nHeight/nDivisor) - 1)); } if (pic->linesize[1]) { nDivisor = (nMaxLineSize / pic->linesize[1]); if (!nDivisor) { nDivisor = 1; } pic->data[1] += (pic->linesize[1] * ((nHeight/nDivisor) - 1)); } if (pic->linesize[2]) { nDivisor = (nMaxLineSize / pic->linesize[2]); if (!nDivisor) { nDivisor = 1; } pic->data[2] += (pic->linesize[2] * ((nHeight/nDivisor) - 1)); } if (pic->linesize[3]) { nDivisor = (nMaxLineSize / pic->linesize[3]); if (!nDivisor) { nDivisor = 1; } pic->data[3] += (pic->linesize[3] * ((nHeight/nDivisor) - 1)); } pic->linesize[0] *= -1; pic->linesize[1] *= -1; pic->linesize[2] *= -1; pic->linesize[3] *= -1; -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/Flip-in-sws-scale-tp939665p3516095.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From hardik.sharma22 at yahoo.com Thu May 12 01:20:39 2011 From: hardik.sharma22 at yahoo.com (Hardik Sharma) Date: Wed, 11 May 2011 16:20:39 -0700 (PDT) Subject: [FFmpeg-user] Can we change NALU size in X264 encoding Message-ID: <898221.89422.qm@web46206.mail.sp1.yahoo.com> Hi guys, Till this time I was using JM for X264 encoding and decoding but now I want to compare my results with FFmpeg keeping same parameters. My question is can we specify NALU size in ffmpeg for X264 codec? It's really important for me to specify NALU size in ffmpeg.? Thanks, Hardik Sharma ? ?? From frederiksunne at gmail.com Thu May 12 09:03:38 2011 From: frederiksunne at gmail.com (Frederik Dam Sunne) Date: Thu, 12 May 2011 09:03:38 +0200 Subject: [FFmpeg-user] -vf pad and -aspect problem In-Reply-To: <0705BDE8-33F2-4E0E-A23A-AF1FE5E1EFCA@designwildwest.com> References: <76617E67-6AB4-4EDE-85BA-AFF4CDBA7F41@designwildwest.com> <20110511102258.GA28878@epicure.lazyet.homelinux.net> <0705BDE8-33F2-4E0E-A23A-AF1FE5E1EFCA@designwildwest.com> Message-ID: - Specifically he has said to me: I was following instruction about avfilters, which is to use checkout.sh script that downloads ffmpeg version from svn and compile it after checkout. I think that dates back half a year ago when not all the filters was part of the FFmpeg codebase. This is no longer an issue, so all you need to do is to checkout from git and compile... Regards, Frederik From lazarusportugalwebmasters at gmail.com Thu May 12 10:04:54 2011 From: lazarusportugalwebmasters at gmail.com (lazarusportugalwebmasters) Date: Thu, 12 May 2011 09:04:54 +0100 Subject: [FFmpeg-user] Record the screen with ffmpeg In-Reply-To: <4DCAE4D9.10303@pps-inc.com> References: <4DCAC7BC.7030600@gmail.com> <20110511181955.GZ5996@neoprimitive.net> <4DCAE4D9.10303@pps-inc.com> Message-ID: <4DCB94A6.4000403@gmail.com> I'm writing a programa that use ffmpeg, and any software that you're say at here are't Opensource (like my) and cross platform(i want that my program be opensource). :) -- #lazarus-br em irc.freenode.org #lazarusportugal em irc.ptnet.org(Cloud IRC em lazarusportugal.org/doku.php/irc) Em caso de d?vida sobre o cliente a usar clique em http://lazarusportugal.org/doku.php/duvidas_irc Juntem-se a estes fant?sticos IRC's sobre Lazarus From mroper at kinect.co.nz Thu May 12 10:25:31 2011 From: mroper at kinect.co.nz (mroper) Date: Thu, 12 May 2011 20:25:31 +1200 Subject: [FFmpeg-user] trying to work out ffmpeg and avi conversion Message-ID: <4DCB997B.3050406@kinect.co.nz> hi, hoping someone can help, i'm fairly new to ffmpeg, trying to work out how to convert an HD 1280x720 AVI file to mpeg2 so my tv via linux media tomb can play it. When i try doing.... ffmpeg -i "$filename" -target pal-dvd -b 25000k /destpath/$filename i get .....Error while opening encoder for output stream #0.0- maybe incorrect parameters such as bit_rate, rate, width or height i've tried searching on the error and drawn a blank, lots of information but nothing I can relate to it. likewise the man pages don't give me a clue. reason why I have the -b 25000k is I'm trying to maintain the quality of the orginal avi, if I don't set this it works fine. likewise if I set a lower rate it also works fine (but at lower quality). The orginal avi when doing a ffmpeg -i filename shows up Stream #0.0: Video: mjpeg, yuvj422p, 1280x720, 30 tbr, 30 tbn, 30 tbc Metadata: strn : FUJIFILM AVI STREAM 0100 Stream #0.1: Audio: pcm_s16le, 44100 Hz, 2 channels, s16, 1411 kb/s Where as the best I can get is via ffmpeg -i "$filename" -target pal-dvd -s hd720 /destpath/$filename and produces Duration: 00:00:02.00, start: 0.500000, bitrate: 7176 kb/s Stream #0.0[0x1e0]: Video: mpeg2video, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 9000 kb/s, 25 fps, 25 tbr, 90k tbn, 50 tbc Stream #0.1[0x80]: Audio: ac3, 48000 Hz, stereo, s16, 448 kb/s Which has a much lower quality/bitrate (you can see the difference on screen). Any ideas? Thanks Miles From belcampo at zonnet.nl Thu May 12 12:58:52 2011 From: belcampo at zonnet.nl (belcampo) Date: Thu, 12 May 2011 12:58:52 +0200 Subject: [FFmpeg-user] trying to work out ffmpeg and avi conversion In-Reply-To: <4DCB997B.3050406@kinect.co.nz> References: <4DCB997B.3050406@kinect.co.nz> Message-ID: <4DCBBD6C.5070400@zonnet.nl> On 05/12/11 10:25, mroper wrote: > hi, > > > > hoping someone can help, i'm fairly new to ffmpeg, trying to > work out how to convert an HD 1280x720 AVI file to mpeg2 so my > tv via linux media tomb can play it. > > > > When i try doing.... > > > > ffmpeg -i "$filename" -target pal-dvd -b 25000k > /destpath/$filename > > > > i get .....Error while opening encoder for output stream #0.0- maybe > incorrect parameters such > as bit_rate, rate, width or height > > > i've tried searching on the error and drawn a blank, lots of information > but nothing I can relate to it. likewise the man pages don't give me a > clue. reason why I have the -b 25000k is I'm trying to maintain the > quality of the orginal avi, if I don't set this it works fine. likewise > if I set a lower rate it also works fine (but at lower quality). The > orginal avi when doing a ffmpeg -i filename shows up > > Stream #0.0: Video: mjpeg, yuvj422p, 1280x720, 30 tbr, 30 tbn, 30 tbc > Metadata: > strn : FUJIFILM AVI STREAM 0100 > Stream #0.1: Audio: pcm_s16le, 44100 Hz, 2 channels, s16, 1411 kb/s > > Where as the best I can get is via > > ffmpeg -i "$filename" -target pal-dvd -s hd720 /destpath/$filename > > and produces > > Duration: 00:00:02.00, start: 0.500000, bitrate: 7176 kb/s > Stream #0.0[0x1e0]: Video: mpeg2video, yuv420p, 1280x720 [PAR 1:1 DAR > 16:9], 9000 kb/s, 25 fps, 25 tbr, 90k tbn, 50 tbc > Stream #0.1[0x80]: Audio: ac3, 48000 Hz, stereo, s16, 448 kb/s > > Which has a much lower quality/bitrate (you can see the difference on > screen). target pal-dvd is limited to 9000 kb/s AFAIK, the dvd-spec doesn't allow more. You'll have to remove 'target pal-dvd' and specify the parameters your-self like vcodec mpeg2 -b 25000k and then it should work. You also have to specify -acodec to something your TV 'understands'. > > Any ideas? > > Thanks > > Miles > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From koxaniy at mail.ru Thu May 12 13:11:48 2011 From: koxaniy at mail.ru (Tuuls) Date: Thu, 12 May 2011 04:11:48 -0700 (PDT) Subject: [FFmpeg-user] ac3 6ch -> 6 mono wav files In-Reply-To: <7C2FDC84-1149-40B0-9EA4-86EC431A6B10@mac.com> References: <7C2FDC84-1149-40B0-9EA4-86EC431A6B10@mac.com> Message-ID: <1305198708158-3517039.post@n4.nabble.com> use SOX ! The best of sounds program of command line. -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/ac3-6ch-6-mono-wav-files-tp3515894p3517039.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From wernam at hotmail.com Thu May 12 13:53:37 2011 From: wernam at hotmail.com (Wernam Wer) Date: Thu, 12 May 2011 13:53:37 +0200 Subject: [FFmpeg-user] Building for ARM Message-ID: Hi all, I'm building ffmpeg for an ARM with Linux headers 2.6 to send video frames over RTP, ffmpeg only have to packet these frames and send it. I need to optimize the space so I use the minimal configuration. I have compile the revision 26400 with this configuration and works fine: ./configure --enable-cross-compile --cross-prefix=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu- --target-os=linux --cc=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu-gcc --host-cc=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu-gcc --cpu=armv5te --arch=arm --enable-static --disable-asm --enable-armv5te --disable-stripping --disable-debug --disable-encoders --disable-decoders --disable-parsers --disable-ffplay --disable-ffserver --disable-devices --disable-bsfs --disable-muxers --enable-parser=mpeg4video --enable-muxer=rtp --enable-parser=aac --disable-filters --disable-demuxers --enable-demuxer=m4v --extra-libs=-static --extra-cflags=--static --disable-ffprobe I use with this command (with this output): # ffmpeg -vcodec copy -i pipe -an -f rtp rtp://224.52.52.22:7004 -sameq -v 0 < /dev/null & [3] 178 # FFmpeg version SVN-r26400, Copyright (c) 2000-2011 the FFmpeg developers built on Jan 18 2011 09:48:17 with gcc 4.1.2 configuration: --enable-cross-compile --cross-prefix=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu- --target-os=linux --cc=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu-gcc --host-cc=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu-gcc --cpu=armv5te --arch=arm --enable-static --disable-asm --enable-armv5te --disable-stripping --disable-debug --disable-encoders --disable-decoders --disable-parsers --disable-ffplay --disable-ffserver --disable-devices --disable-bsfs --disable-muxers --enable-parser=mpeg4video --enable-muxer=rtp --enable-parser=aac --disable-filters --disable-demuxers --enable-demuxer=m4v --extra-libs=-static --extra-cflags=--static libavutil 50.36. 0 / 50.36. 0 libavcore 0.16. 1 / 0.16. 1 libavcodec 52.108. 0 / 52.108. 0 libavformat 52.93. 0 / 52.93. 0 libavdevice 52. 2. 3 / 52. 2. 3 libavfilter 1.74. 0 / 1.74. 0 libswscale 0.12. 0 / 0.12. 0 puntero a frame (fis): 0xc2780004 [m4v @ 0x3538a0] max_analyze_duration reached [m4v @ 0x3538a0] Estimating duration from bitrate, this may be inaccurate Input #0, m4v, from 'pipe': Duration: N/A, bitrate: N/A Stream #0.0: Video: [0][0][0][0] / 0x0000, 352x288 [PAR 1:1 DAR 11:9], 25 fps, 25 tbr, 1200k tbn, 25 tbc Output #0, rtp, to 'rtp://224.52.52.22:7004': Metadata: encoder : Lavf52.93.0 Stream #0.0: Video: [0][0][0][0] / 0x0000, 352x288 [PAR 1:1 DAR 11:9], q=2-31, 90k tbn, 25 tbc Stream mapping: Stream #0.0 -> #0.0 SDP: v=0 o=- 0 0 IN IP4 127.0.0.1 s=No Name c=IN IP4 224.52.52.22 t=0 0 a=tool:libavformat 52.93.0 m=video 7004 RTP/AVP 96 a=rtpmap:96 MP4V-ES/90000 a=fmtp:96 profile-level-id=1; config=000001B003000001B50900000100000001200086C400668582120A31 Press [q] to stop encoding If I try to use the version 0.6.3, I can compile it fine, but it doesn't work, i get this output: # ./ffmpeg.last -vcodec copy -i pipe -an -f rtp rtp://224.52.52.22:7004 -sameq -re FFmpeg version 0.6.3, Copyright (c) 2000-2010 the FFmpeg developers built on May 12 2011 13:00:49 with gcc 4.1.2 configuration: --enable-cross-compile --cross-prefix=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu- --target-os=linux --cc=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu-gcc --host-cc=../../Cross/arm-softfloat-linux-gnu/bin/arm-softfloat-linux-gnu-gcc --cpu=armv5te --arch=arm --enable-static --disable-asm --enable-armv5te --disable-stripping --disable-debug --disable-encoders --disable-decoders --disable-parsers --disable-ffplay --disable-ffserver --disable-devices --disable-bsfs --disable-muxers --enable-parser=mpeg4video --enable-muxer=rtp --enable-parser=aac --disable-filters --disable-demuxers --enable-demuxer=m4v --extra-libs=-static --extra-cflags=--static --disable-ffprobe libavutil 50.15. 1 / 50.15. 1 libavcodec 52.72. 2 / 52.72. 2 libavformat 52.64. 2 / 52.64. 2 libavdevice 52. 2. 0 / 52. 2. 0 libswscale 0.11. 0 / 0.11. 0 [m4v @ 0x30e360]max_analyze_duration reached [m4v @ 0x30e360]Estimating duration from bitrate, this may be inaccurate Input #0, m4v, from 'pipe': Duration: N/A, bitrate: N/A Stream #0.0: Video: 0x0000, 352x288 [PAR 1:1 DAR 11:9], 25 fps, 25 tbr, 1200k tbn, 25 tbc And doesn't sent anything. My question is I how can I debug it or if I'm doing anything wrong. I also try to use the last version or the 0.7.rc1 but both doesn't compile, I get this error: /home/john/ffmpeg/ffmpeg-0.7-rc1/libavformat/libavformat.a(utils.o): In function `.L2670': utils.c:(.text+0x928c): undefined reference to `ff_find_pix_fmt' collect2: ld returned 1 exit status make: *** [ffmpeg_g] Error 1 Thanks for the answers. Wernam From betonpfeiler at googlemail.com Thu May 12 14:12:23 2011 From: betonpfeiler at googlemail.com (betonpfeiler) Date: Thu, 12 May 2011 14:12:23 +0200 Subject: [FFmpeg-user] ac3 6ch -> 6 mono wav files In-Reply-To: <7C2FDC84-1149-40B0-9EA4-86EC431A6B10@mac.com> References: <7C2FDC84-1149-40B0-9EA4-86EC431A6B10@mac.com> Message-ID: <4DCBCEA7.5060706@googlemail.com> Am 11.05.2011 23:19, schrieb Denis Muraviev: > Hi, > > I have .ac3 file with 6 channels, and I want to decode it to 6 different mono .wav files (every channel need to be in separate wav file). > I found that there is `-ac channels' to set the number of audio channels, but how can I select the channel I need (not the first one). > > Thanks, > Denis. Hey! If you don't want to use SOX, you can also achieve this with the -map_audio_channel xi:yi:zi:xo:yo:zo command (x=file, y=stream, z=channel; i=in, o=out) you can find a more detailed explanation under ffmbc.wordpress.com From mark at mdsh.com Thu May 12 14:13:07 2011 From: mark at mdsh.com (Mark Himsley) Date: Thu, 12 May 2011 13:13:07 +0100 Subject: [FFmpeg-user] flag YUV input as interlaced Message-ID: <4DCBCED3.9050103@mdsh.com> Imagine I am using FFmpeg to read a raw YUV file. How do I flag that input file as interlaced so that interlaced aware filters process the stream correctly? For instance: I have an SD PAL sized yvyu422 file that I know is top-field-first I want to output that as PAL DV 25, which is bottom-field-first I expect to use a command like like this: ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i input.yvyu422 -vf fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov But, since the input file is not flagged as interlaced the fieldorder filter cannot do it's job. Thanks. -- Mark From bouke at editb.nl Thu May 12 15:11:36 2011 From: bouke at editb.nl (bouke) Date: Thu, 12 May 2011 15:11:36 +0200 Subject: [FFmpeg-user] flag YUV input as interlaced References: <4DCBCED3.9050103@mdsh.com> Message-ID: <017f01cc10a6$228be1f0$4301a8c0@hpkantoor> ----- Original Message ----- From: "Mark Himsley" To: "FFmpeg user questions and RTFMs" Sent: Thursday, May 12, 2011 2:13 PM Subject: [FFmpeg-user] flag YUV input as interlaced > Imagine I am using FFmpeg to read a raw YUV file. How do I flag that input > file as interlaced so that interlaced aware filters process the stream > correctly? > > For instance: > I have an SD PAL sized yvyu422 file that I know is top-field-first > I want to output that as PAL DV 25, which is bottom-field-first > > I expect to use a command like like this: > > ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i input.yvyu422 -vf > fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov Normally you put the known input specs before the input file, as in your example. So did you test ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -vf eldorder=tff -i input.yvyu422 -vf fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov ? Bouke > But, since the input file is not flagged as interlaced the fieldorder > filter cannot do it's job. > > Thanks. > > -- > Mark > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From Michael.Feurstein at wu.ac.at Thu May 12 16:04:05 2011 From: Michael.Feurstein at wu.ac.at (Feurstein, Michael) Date: Thu, 12 May 2011 16:04:05 +0200 Subject: [FFmpeg-user] unicast input to ffmpeg Message-ID: Hi, I'm trying to to merge a video and audio stream with ffmpeg. I can successfully work with a multicast input for ffmpeg as in: ffmpeg -i udp://@224.0.0.1:1234 -vcodec copy -i http://mp3stream1.apasf.apa.at:8000/ -acodec copy /muxed.mov However I'd like to use an unicast stream as input for ffmpeg. The unicast is created with vlc directed at 192.168.1.1 ([...] standard{access=udp, mux=ts, dst=192.168.1.1, port=5353} [...]) and can be opened with vlc on the machine its directed at (192.168.1.1) by simply opening udp:// Now when I try to open udp:// with ffmpeg as in: ffmpeg -i udp:// -vcodec copy -i http://mp3stream1.apasf.apa.at:8000/ -acodec copy /muxed.mov ffmpeg just sits there and does nothing. I already thought about piping the stream with openRTSP but that would be my fallback option. Has anyone encountered this same behavior or does someone know what I am doing wrong? I'd be glad for any input on this topic. Thanks Best Regards Michael Feurstein From mark at mdsh.com Thu May 12 16:51:20 2011 From: mark at mdsh.com (Mark Himsley) Date: Thu, 12 May 2011 15:51:20 +0100 Subject: [FFmpeg-user] flag YUV input as interlaced In-Reply-To: <017f01cc10a6$228be1f0$4301a8c0@hpkantoor> References: <4DCBCED3.9050103@mdsh.com> <017f01cc10a6$228be1f0$4301a8c0@hpkantoor> Message-ID: <4DCBF3E8.3050700@mdsh.com> On 12/05/11 14:11, bouke wrote: > > ----- Original Message ----- > From: "Mark Himsley" > To: "FFmpeg user questions and RTFMs" > Sent: Thursday, May 12, 2011 2:13 PM > Subject: [FFmpeg-user] flag YUV input as interlaced > > >> Imagine I am using FFmpeg to read a raw YUV file. How do I flag that input >> file as interlaced so that interlaced aware filters process the stream >> correctly? >> >> For instance: >> I have an SD PAL sized yvyu422 file that I know is top-field-first >> I want to output that as PAL DV 25, which is bottom-field-first >> >> I expect to use a command like like this: >> >> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i input.yvyu422 -vf >> fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov > > Normally you put the known input specs before the input file, as in your > example. > So did you test > ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -vf > eldorder=tff -i input.yvyu422 -vf fieldorder=bff -vcodec dvvideo -pix_fmt > yuv420p -y output.mov > > ? Hi Bouke, The -vf command is defining a filter, not setting switches on a file, so that command does not work. Thanks though. > Bouke > >> But, since the input file is not flagged as interlaced the fieldorder >> filter cannot do it's job. >> >> Thanks. >> >> -- >> Mark From bouke at editb.nl Thu May 12 17:03:24 2011 From: bouke at editb.nl (bouke) Date: Thu, 12 May 2011 17:03:24 +0200 Subject: [FFmpeg-user] flag YUV input as interlaced References: <4DCBCED3.9050103@mdsh.com><017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com> Message-ID: <020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> ----- Original Message ----- From: "Mark Himsley" To: "FFmpeg user questions and RTFMs" Sent: Thursday, May 12, 2011 4:51 PM Subject: Re: [FFmpeg-user] flag YUV input as interlaced > On 12/05/11 14:11, bouke wrote: >> >> ----- Original Message ----- >> From: "Mark Himsley" >> To: "FFmpeg user questions and RTFMs" >> Sent: Thursday, May 12, 2011 2:13 PM >> Subject: [FFmpeg-user] flag YUV input as interlaced >> >> >>> Imagine I am using FFmpeg to read a raw YUV file. How do I flag that >>> input >>> file as interlaced so that interlaced aware filters process the stream >>> correctly? >>> >>> For instance: >>> I have an SD PAL sized yvyu422 file that I know is top-field-first >>> I want to output that as PAL DV 25, which is bottom-field-first >>> >>> I expect to use a command like like this: >>> >>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i >>> input.yvyu422 -vf >>> fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov >> >> Normally you put the known input specs before the input file, as in your >> example. >> So did you test >> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -vf >> eldorder=tff -i input.yvyu422 -vf fieldorder=bff -vcodec >> dvvideo -pix_fmt >> yuv420p -y output.mov >> >> ? > > Hi Bouke, > > The -vf command is defining a filter, not setting switches on a file, so > that command does not work. Doh! Note to self, first think, then post.... Dirty trick, if you can spare a line, you could crop off one line at the top, and pad one line at the bottom. That switches field dominance as well. Bouke > Thanks though. > >> Bouke >> >>> But, since the input file is not flagged as interlaced the fieldorder >>> filter cannot do it's job. >>> >>> Thanks. >>> >>> -- >>> Mark > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From mark at mdsh.com Thu May 12 17:17:43 2011 From: mark at mdsh.com (Mark Himsley) Date: Thu, 12 May 2011 16:17:43 +0100 Subject: [FFmpeg-user] flag YUV input as interlaced In-Reply-To: <020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> References: <4DCBCED3.9050103@mdsh.com><017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com> <020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> Message-ID: <4DCBFA17.2070303@mdsh.com> On 12/05/11 16:03, bouke wrote: > ----- Original Message ----- > From: "Mark Himsley" > To: "FFmpeg user questions and RTFMs" > Sent: Thursday, May 12, 2011 4:51 PM > Subject: Re: [FFmpeg-user] flag YUV input as interlaced > > >> On 12/05/11 14:11, bouke wrote: >>> >>> ----- Original Message ----- >>> From: "Mark Himsley" >>> To: "FFmpeg user questions and RTFMs" >>> Sent: Thursday, May 12, 2011 2:13 PM >>> Subject: [FFmpeg-user] flag YUV input as interlaced >>> >>> >>>> Imagine I am using FFmpeg to read a raw YUV file. How do I flag that >>>> input >>>> file as interlaced so that interlaced aware filters process the stream >>>> correctly? >>>> >>>> For instance: >>>> I have an SD PAL sized yvyu422 file that I know is top-field-first >>>> I want to output that as PAL DV 25, which is bottom-field-first >>>> >>>> I expect to use a command like like this: >>>> >>>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i >>>> input.yvyu422 -vf >>>> fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov >>> >>> Normally you put the known input specs before the input file, as in your >>> example. >>> So did you test >>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -vf >>> eldorder=tff -i input.yvyu422 -vf fieldorder=bff -vcodec >>> dvvideo -pix_fmt >>> yuv420p -y output.mov >>> >>> ? >> >> Hi Bouke, >> >> The -vf command is defining a filter, not setting switches on a file, so >> that command does not work. > > Doh! > Note to self, first think, then post.... > Dirty trick, if you can spare a line, you could crop off one line at the > top, and pad one line at the bottom. > That switches field dominance as well. Hi Bouke, That is true, but I wrote the fieldorder filter to do a _slightly_ better job then that (it fills the blank line created by the pad with roughly the right video data and only does the shift up/down it its required). What I want to do, though, is to set a flag like -s, -r and -pix_fmt that sets the interlaced flag and tff flag on -f rawvideo media before the video hits the filter chain. I just assumed I'd missed something. Perhaps I need to patch ffmpeg instead. Thanks :-) > Bouke > > >> Thanks though. >> >>> Bouke >>> >>>> But, since the input file is not flagged as interlaced the fieldorder >>>> filter cannot do it's job. >>>> >>>> Thanks. >>>> >>>> -- >>>> Mark From bouke at editb.nl Thu May 12 17:37:44 2011 From: bouke at editb.nl (bouke) Date: Thu, 12 May 2011 17:37:44 +0200 Subject: [FFmpeg-user] flag YUV input as interlaced References: <4DCBCED3.9050103@mdsh.com><017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com><020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> <4DCBFA17.2070303@mdsh.com> Message-ID: <024601cc10ba$8d331b40$4301a8c0@hpkantoor> ---- Original Message ----- From: "Mark Himsley" To: "FFmpeg user questions and RTFMs" Sent: Thursday, May 12, 2011 5:17 PM Subject: Re: [FFmpeg-user] flag YUV input as interlaced > On 12/05/11 16:03, bouke wrote: >> ----- Original Message ----- >> From: "Mark Himsley" >> To: "FFmpeg user questions and RTFMs" >> Sent: Thursday, May 12, 2011 4:51 PM >> Subject: Re: [FFmpeg-user] flag YUV input as interlaced >> >> >>> On 12/05/11 14:11, bouke wrote: >>>> >>>> ----- Original Message ----- >>>> From: "Mark Himsley" >>>> To: "FFmpeg user questions and RTFMs" >>>> Sent: Thursday, May 12, 2011 2:13 PM >>>> Subject: [FFmpeg-user] flag YUV input as interlaced >>>> >>>> >>>>> Imagine I am using FFmpeg to read a raw YUV file. How do I flag that >>>>> input >>>>> file as interlaced so that interlaced aware filters process the stream >>>>> correctly? >>>>> >>>>> For instance: >>>>> I have an SD PAL sized yvyu422 file that I know is top-field-first >>>>> I want to output that as PAL DV 25, which is bottom-field-first >>>>> >>>>> I expect to use a command like like this: >>>>> >>>>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i >>>>> input.yvyu422 -vf >>>>> fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov >>>> >>>> Normally you put the known input specs before the input file, as in >>>> your >>>> example. >>>> So did you test >>>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -vf >>>> eldorder=tff -i input.yvyu422 -vf fieldorder=bff -vcodec >>>> dvvideo -pix_fmt >>>> yuv420p -y output.mov >>>> >>>> ? >>> >>> Hi Bouke, >>> >>> The -vf command is defining a filter, not setting switches on a file, so >>> that command does not work. >> >> Doh! >> Note to self, first think, then post.... >> Dirty trick, if you can spare a line, you could crop off one line at the >> top, and pad one line at the bottom. >> That switches field dominance as well. > > Hi Bouke, > > That is true, but I wrote the fieldorder filter to do a _slightly_ better > job then that (it fills the blank line created by the pad with roughly the > right video data and only does the shift up/down it its required). Good for HD, but SD is an overscanned format that you have to crop for computer use anyways, or you have incorrect framing. Lot's of mics hanging in the top of the shot, matte boxes, and overall incorrect framing as any cam op that shot SD knew that the entire frame was never to be shown. And on top of that blanking on analogue sources... So, loosing one line is not that bad... Bouke > What I want to do, though, is to set a flag like -s, -r and -pix_fmt that > sets the interlaced flag and tff flag on -f rawvideo media before the > video hits the filter chain. > > I just assumed I'd missed something. Perhaps I need to patch ffmpeg > instead. > > Thanks :-) > > >> Bouke >> >> >>> Thanks though. >>> >>>> Bouke >>>> >>>>> But, since the input file is not flagged as interlaced the fieldorder >>>>> filter cannot do it's job. >>>>> >>>>> Thanks. >>>>> >>>>> -- >>>>> Mark > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From mark at mdsh.com Thu May 12 18:07:09 2011 From: mark at mdsh.com (Mark Himsley) Date: Thu, 12 May 2011 17:07:09 +0100 Subject: [FFmpeg-user] flag YUV input as interlaced In-Reply-To: <024601cc10ba$8d331b40$4301a8c0@hpkantoor> References: <4DCBCED3.9050103@mdsh.com><017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com><020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> <4DCBFA17.2070303@mdsh.com> <024601cc10ba$8d331b40$4301a8c0@hpkantoor> Message-ID: <4DCC05AD.70102@mdsh.com> On 12/05/11 16:37, bouke wrote: > > ---- Original Message ----- > From: "Mark Himsley" > To: "FFmpeg user questions and RTFMs" > Sent: Thursday, May 12, 2011 5:17 PM > Subject: Re: [FFmpeg-user] flag YUV input as interlaced > > >> On 12/05/11 16:03, bouke wrote: >>> ----- Original Message ----- >>> From: "Mark Himsley" >>> To: "FFmpeg user questions and RTFMs" >>> Sent: Thursday, May 12, 2011 4:51 PM >>> Subject: Re: [FFmpeg-user] flag YUV input as interlaced >>> >>> >>>> On 12/05/11 14:11, bouke wrote: >>>>> >>>>> ----- Original Message ----- >>>>> From: "Mark Himsley" >>>>> To: "FFmpeg user questions and RTFMs" >>>>> Sent: Thursday, May 12, 2011 2:13 PM >>>>> Subject: [FFmpeg-user] flag YUV input as interlaced >>>>> >>>>> >>>>>> Imagine I am using FFmpeg to read a raw YUV file. How do I flag that >>>>>> input >>>>>> file as interlaced so that interlaced aware filters process the stream >>>>>> correctly? >>>>>> >>>>>> For instance: >>>>>> I have an SD PAL sized yvyu422 file that I know is top-field-first >>>>>> I want to output that as PAL DV 25, which is bottom-field-first >>>>>> >>>>>> I expect to use a command like like this: >>>>>> >>>>>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i >>>>>> input.yvyu422 -vf >>>>>> fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov >>>>> >>>>> Normally you put the known input specs before the input file, as in >>>>> your >>>>> example. >>>>> So did you test >>>>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -vf >>>>> eldorder=tff -i input.yvyu422 -vf fieldorder=bff -vcodec >>>>> dvvideo -pix_fmt >>>>> yuv420p -y output.mov >>>>> >>>>> ? >>>> >>>> Hi Bouke, >>>> >>>> The -vf command is defining a filter, not setting switches on a file, so >>>> that command does not work. >>> >>> Doh! >>> Note to self, first think, then post.... >>> Dirty trick, if you can spare a line, you could crop off one line at the >>> top, and pad one line at the bottom. >>> That switches field dominance as well. >> >> Hi Bouke, >> >> That is true, but I wrote the fieldorder filter to do a _slightly_ better >> job then that (it fills the blank line created by the pad with roughly the >> right video data and only does the shift up/down it its required). > > Good for HD, but SD is an overscanned format that you have to crop for > computer use anyways, or you have incorrect framing. > Lot's of mics hanging in the top of the shot, matte boxes, and overall > incorrect framing as any cam op that shot SD knew that the entire frame was > never to be shown. > And on top of that blanking on analogue sources... > So, loosing one line is not that bad... Hi Bouke, This is a digression from my question, so I hope the original question is not lost. But I think I must have been very bad at explaining in my previous emails. I'm only using fieldorder as an example. But, FYI, the field order filter DOES drop (or raise) the picture by one line when doing the tff <-> bff conversion. It ALSO adds pertinent picture data into the new line created by the crop+pad, AND it only does the tff <-> bff conversion if the original file is not in the interlaced format that has been requested. Anyway... What I need is a way to flag rawvideo media when its used as an input into ffmpeg. Anyone? Thanks you your enthusiastic responses though :-) > Bouke > >> What I want to do, though, is to set a flag like -s, -r and -pix_fmt that >> sets the interlaced flag and tff flag on -f rawvideo media before the >> video hits the filter chain. >> >> I just assumed I'd missed something. Perhaps I need to patch ffmpeg >> instead. >> >> Thanks :-) >> >> >>> Bouke >>> >>> >>>> Thanks though. >>>> >>>>> Bouke >>>>> >>>>>> But, since the input file is not flagged as interlaced the fieldorder >>>>>> filter cannot do it's job. >>>>>> >>>>>> Thanks. >>>>>> >>>>>> -- >>>>>> Mark From andrew at designwildwest.com Thu May 12 18:10:42 2011 From: andrew at designwildwest.com (Andrew Pettican) Date: Thu, 12 May 2011 17:10:42 +0100 Subject: [FFmpeg-user] -vf pad and -aspect problem In-Reply-To: References: <76617E67-6AB4-4EDE-85BA-AFF4CDBA7F41@designwildwest.com> <20110511102258.GA28878@epicure.lazyet.homelinux.net> <0705BDE8-33F2-4E0E-A23A-AF1FE5E1EFCA@designwildwest.com> Message-ID: Thank you both Frederik & Etienne. I've got everything working now. Andrew On 12 May 2011, at 08:03, Frederik Dam Sunne wrote: > - Specifically he has said to me: I was following instruction about > avfilters, which is to use checkout.sh script that downloads ffmpeg version > from svn and compile it after checkout. > > I think that dates back half a year ago when not all the filters was part of > the FFmpeg codebase. This is > no longer an issue, so all you need to do is to checkout from git and > compile... > > Regards, > > Frederik > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From jshupert at pps-inc.com Thu May 12 23:50:12 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Thu, 12 May 2011 17:50:12 -0400 Subject: [FFmpeg-user] xVid and version Q Message-ID: <4DCC5614.1060608@pps-inc.com> Friends, I made a file with ffmpeg [ -vcodec libxvid ] ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 48000 -b 8000000 -s 720x480 -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi ( and file looks fine - plays in VLC , wonderful ) when I analyze a file with Gspot it repots the file video codec to be xvid ISO MPEG-4. ISO I think this is xvid 1.3.1 . when i examine with media info or ffprobe it simply says Codec:xvid. What are the major differences between xvid 1.1.2 and xvid ISO MPEG-4. ISO? If i wish to provide a person Xvid 1.1.2 can I get that older codec version or would it be suggested to have the person update thier player codecs. They are asking for Xvid 1.1.2 is this a reasonable request isn't it just older - don't ya think they could just update their codec? Simply said is the file I made with ffmpeg [-vcodec libxvid ] simply a newer version of xvid 1.1.2 or is something fundamentally different. Thanks much j From adi235 at gmail.com Fri May 13 10:01:42 2011 From: adi235 at gmail.com (Aditya) Date: Fri, 13 May 2011 08:01:42 +0000 (UTC) Subject: [FFmpeg-user] libav-api Understanding Error Logs when decoding and encoding Message-ID: [mpeg @ 19140050] max_analyze_duration reached //--?? Input #0, mpeg, from 'tv5.mpg': Duration: 00:02:58.72, start: 0.500000, bitrate: 4162 kb/s Stream #0.0[0x1c0]: Audio: mp2, 48000 Hz, stereo, s16, 64 kb/s Stream #0.1[0x1e0]: Video: mpeg2video (Main), yuv420p, 720x576 [PAR 16:15 DA R 4:3], 15000 kb/s, 25 fps, 25 tbr, 90k tbn, 50 tbc Output #0, avi, to 'output.avi': Stream #0.0: Audio: mp2, 48000 Hz, 2 channels, 64 kb/s Stream #0.1: Video: mpeg4, yuv420p, 720x576 [PAR 16:15 DAR 4:3], q=2-31, 150 00 kb/s, 90k tbn, 50 tbc [mpeg2video @ 1915FE30] ac-tex damaged at 33 22 //--?? what is ac-tex how to /-- prevent its damage [mpeg2video @ 1915FE30] Warning MVs not available /--?? what are mv's [mpeg2video @ 1915FE30] concealing 90 DC, 90 AC, 90 MV errors //-- what is happening here [mpeg2video @ 1915FE30] Warning MVs not available [mpeg2video @ 1915FE30] concealing 30 DC, 30 AC, 30 MV errors [mpeg2video @ 1915FE30] 00 motion_type at 1 2 [mpeg2video @ 1915FE30] Warning MVs not available [mpeg2video @ 1915FE30] concealing 90 DC, 90 AC, 90 MV errors From ratheendran.s at gmail.com Fri May 13 13:46:21 2011 From: ratheendran.s at gmail.com (Ratheendran R) Date: Fri, 13 May 2011 17:16:21 +0530 Subject: [FFmpeg-user] help needed to understand the features Message-ID: Dear All, I am exploring the features of FFMPEG as a streaming server. my requirement is 1. I want to record the video only for specific period,say I want to store the video of the last 2 hours from a WEBCAM/IP CAMERA/HDMI CAMAERA. how this can be achieved. I mean the parameters need to be configured. Regards, Ratheendran From jshupert at pps-inc.com Fri May 13 15:57:23 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Fri, 13 May 2011 09:57:23 -0400 Subject: [FFmpeg-user] xVid and version Q In-Reply-To: <4DCC5614.1060608@pps-inc.com> References: <4DCC5614.1060608@pps-inc.com> Message-ID: <4DCD38C3.7010002@pps-inc.com> On 5/12/2011 5:50 PM, Jim Shupert wrote: > Friends, > > I made a file with ffmpeg [ -vcodec libxvid ] > > ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 48000 -b 8000000 -s 720x480 > -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi > ( and file looks fine - plays in VLC , wonderful ) > > when I analyze a file with Gspot it repots the file video codec to be > xvid ISO MPEG-4. ISO > I think this is xvid 1.3.1 . > when i examine with media info or ffprobe it simply says Codec:xvid. > > What are the major differences between > xvid 1.1.2 and xvid ISO MPEG-4. ISO? > > > If i wish to provide a person Xvid 1.1.2 can I get that older codec > version or would it be suggested to have the person update thier > player codecs. > > They are asking for Xvid 1.1.2 is this a reasonable request isn't it > just older - don't ya think they could just update their codec? > > Simply said is the file I made with ffmpeg [-vcodec libxvid ] simply a > newer version of xvid 1.1.2 or is something fundamentally different. > > Thanks much > > is there any reason to think that the file made with -vcodec libxvid { that i think is Xvid 1.3.1 isn't that true ? } would not be backward compatible to xvid 1.1.2 ? meaning play ok in any player that can play xvid1.1.2 i did some googling -- and i current think -vcodec libxvid = = XviD 1.3.1 = = xvid ISO MPEG-4. ISO -------is that true? and those files should play in anything that plays xvid 1.1.2 any wisdom would be appreciated - From anil_jangam at persistent.co.in Fri May 13 16:35:41 2011 From: anil_jangam at persistent.co.in (Anil Jangam) Date: Fri, 13 May 2011 20:05:41 +0530 Subject: [FFmpeg-user] Converted file size is bigger than source. Message-ID: Team, We have observed a scenario in which the size of the output transcoded file (~500kb) is greater than the size of the original video file (~140kb). Upon checking with media info tool (and with ffmpeg), we found that the video codec is h263 and audio codec used is amr (see below). Duration: 00:00:13.20, start: 0.000000, bitrate: 86 kb/s Stream #0.0(und): Video: h263, yuv420p, 176x144 [PAR 12:11 DAR 4:3], 29.97 tbr, 30k tbn, 29.97 tbc Stream #0.1(und): Audio: libopencore_amrnb, 8000 Hz, mono, s16 Upon checking the properties of the output file, it shows that video codec is 'mpeg4_sp' and audio codec is 'aac_lc'. I am not sure if 'mpeg4_sp' is a low compression ratio codec as compare to the 'h263'. In that case, conversion of 'h263' coded video into 'mpeg4_sp' codec video will obviously result in the higher size of the transcoded output. Can you please confirm this? /anil. DISCLAIMER ========== This e-mail may contain privileged and confidential information which is the property of Persistent Systems Ltd. It is intended only for the use of the individual or entity to which it is addressed. If you are not the intended recipient, you are not authorized to read, retain, copy, print, distribute or use this message. If you have received this communication in error, please notify the sender and delete all copies of this message. Persistent Systems Ltd. does not accept any liability for virus infected mails. From h.reindl at thelounge.net Fri May 13 17:03:12 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Fri, 13 May 2011 17:03:12 +0200 Subject: [FFmpeg-user] Converted file size is bigger than source. In-Reply-To: References: Message-ID: <4DCD4830.5030000@thelounge.net> Am 13.05.2011 16:35, schrieb Anil Jangam: > Team, > > We have observed a scenario in which the size of the output transcoded file (~500kb) is greater than the size of the original video file (~140kb). Upon checking with media info tool (and with ffmpeg), we found that the video codec is h263 and audio codec used is amr (see below). > > Duration: 00:00:13.20, start: 0.000000, bitrate: 86 kb/s > Stream #0.0(und): Video: h263, yuv420p, 176x144 [PAR 12:11 DAR 4:3], 29.97 tbr, 30k tbn, 29.97 tbc > Stream #0.1(und): Audio: libopencore_amrnb, 8000 Hz, mono, s16 > > Upon checking the properties of the output file, it shows that video codec is 'mpeg4_sp' and audio codec is 'aac_lc'. > > I am not sure if 'mpeg4_sp' is a low compression ratio codec as compare to the 'h263'. In that case, conversion of 'h263' coded video into 'mpeg4_sp' codec video will obviously result in the higher size of the transcoded output. > > Can you please confirm this? Depends on codec and bitrates 86 kb/s input file is a very low bitrate if your output is higher there is no other option as getting bigger with H264/AAC-Audio and a low bitrate it should get even smaller but anyways: post your full cli-call, without it is impossible to say anything away from guessing -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From etienne.buira.lists at free.fr Fri May 13 18:57:43 2011 From: etienne.buira.lists at free.fr (Etienne Buira) Date: Fri, 13 May 2011 18:57:43 +0200 Subject: [FFmpeg-user] xVid and version Q In-Reply-To: <4DCD38C3.7010002@pps-inc.com> References: <4DCC5614.1060608@pps-inc.com> <4DCD38C3.7010002@pps-inc.com> Message-ID: <20110513165742.GC28878@epicure.lazyet.homelinux.net> On Fri, May 13, 2011 at 09:57:23AM -0400, Jim Shupert wrote: > > > On 5/12/2011 5:50 PM, Jim Shupert wrote: > > Friends, > > > > I made a file with ffmpeg [ -vcodec libxvid ] > > > > ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 48000 -b 8000000 -s 720x480 > > -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi > > ( and file looks fine - plays in VLC , wonderful ) > > > > when I analyze a file with Gspot it repots the file video codec to be > > xvid ISO MPEG-4. ISO > > I think this is xvid 1.3.1 . > > when i examine with media info or ffprobe it simply says Codec:xvid. > > > > What are the major differences between > > xvid 1.1.2 and xvid ISO MPEG-4. ISO? > > > > > > If i wish to provide a person Xvid 1.1.2 can I get that older codec > > version or would it be suggested to have the person update thier > > player codecs. > > > > They are asking for Xvid 1.1.2 is this a reasonable request isn't it > > just older - don't ya think they could just update their codec? > > > > Simply said is the file I made with ffmpeg [-vcodec libxvid ] simply a > > newer version of xvid 1.1.2 or is something fundamentally different. > > > > Thanks much > > > > > is there any reason to think that the file made with -vcodec libxvid > { that i think is Xvid 1.3.1 isn't that true ? } > would not be backward compatible to xvid 1.1.2 ? > meaning play ok in any player that can play xvid1.1.2 > > i did some googling -- and i current think > -vcodec libxvid = = XviD 1.3.1 = = xvid ISO MPEG-4. ISO -------is that > true? > and > those files should play in anything that plays xvid 1.1.2 > > any wisdom would be appreciated - Hi. Actually, you have to distinguish two things: 1. The bitstream format, here you are talking about mpeg 4 part 2 2. The software that outputs a bitstream, taking raw input and compressing it the best it can (in parameters limits) to compress it using the bitstream format constructs. xvid stands here. So, it should be as simple as "use divx/xvid/whatever" to encode, and use whatever mpeg4part2 decoder to play. This is often true. From jshupert at pps-inc.com Fri May 13 19:21:15 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Fri, 13 May 2011 13:21:15 -0400 Subject: [FFmpeg-user] xVid and version Q In-Reply-To: <20110513165742.GC28878@epicure.lazyet.homelinux.net> References: <4DCC5614.1060608@pps-inc.com> <4DCD38C3.7010002@pps-inc.com> <20110513165742.GC28878@epicure.lazyet.homelinux.net> Message-ID: <4DCD688B.1000302@pps-inc.com> On 5/13/2011 12:57 PM, Etienne Buira wrote: > On Fri, May 13, 2011 at 09:57:23AM -0400, Jim Shupert wrote: >> >> On 5/12/2011 5:50 PM, Jim Shupert wrote: >>> Friends, >>> >>> I made a file with ffmpeg [ -vcodec libxvid ] >>> >>> ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 48000 -b 8000000 -s 720x480 >>> -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi >>> ( and file looks fine - plays in VLC , wonderful ) >>> >>> when I analyze a file with Gspot it repots the file video codec to be >>> xvid ISO MPEG-4. ISO >>> I think this is xvid 1.3.1 . >>> when i examine with media info or ffprobe it simply says Codec:xvid. >>> >>> What are the major differences between >>> xvid 1.1.2 and xvid ISO MPEG-4. ISO? >>> >>> >>> If i wish to provide a person Xvid 1.1.2 can I get that older codec >>> version or would it be suggested to have the person update thier >>> player codecs. >>> >>> They are asking for Xvid 1.1.2 is this a reasonable request isn't it >>> just older - don't ya think they could just update their codec? >>> >>> Simply said is the file I made with ffmpeg [-vcodec libxvid ] simply a >>> newer version of xvid 1.1.2 or is something fundamentally different. >>> >>> Thanks much >>> >>> >> is there any reason to think that the file made with -vcodec libxvid >> { that i think is Xvid 1.3.1 isn't that true ? } >> would not be backward compatible to xvid 1.1.2 ? >> meaning play ok in any player that can play xvid1.1.2 >> >> i did some googling -- and i current think >> -vcodec libxvid = = XviD 1.3.1 = = xvid ISO MPEG-4. ISO -------is that >> true? >> and >> those files should play in anything that plays xvid 1.1.2 >> >> any wisdom would be appreciated - > Hi. > > Actually, you have to distinguish two things: > 1. The bitstream format, here you are talking about mpeg 4 part 2 > 2. The software that outputs a bitstream, taking raw input and > compressing it the best it can (in parameters limits) to compress it > using the bitstream format constructs. xvid stands here. > > So, it should be as simple as "use divx/xvid/whatever" to encode, and > use whatever mpeg4part2 decoder to play. This is often true. > _______________________________________________ > Thank you for your reply -- i do appreciate it. and I follow what you state here. simply said here is my challenge. I have made files for a fellow who looks at them in gSpot and sees xvid ISO MPEG-4. ISO not xvid 1.1.2 I contend that what i made is xvid 1.3.1 and that is even better. Am i incorrect about that ? isn't the files made via -vcodec libxvid using the codec known as xvid 1.3.1 and isn't that backwards compatible with xvid1.1.2. actually even better? Thanks again! From etienne.buira.lists at free.fr Fri May 13 19:41:27 2011 From: etienne.buira.lists at free.fr (Etienne Buira) Date: Fri, 13 May 2011 19:41:27 +0200 Subject: [FFmpeg-user] xVid and version Q In-Reply-To: <4DCD688B.1000302@pps-inc.com> References: <4DCC5614.1060608@pps-inc.com> <4DCD38C3.7010002@pps-inc.com> <20110513165742.GC28878@epicure.lazyet.homelinux.net> <4DCD688B.1000302@pps-inc.com> Message-ID: <20110513174127.GD28878@epicure.lazyet.homelinux.net> On Fri, May 13, 2011 at 01:21:15PM -0400, Jim Shupert wrote: > > > On 5/13/2011 12:57 PM, Etienne Buira wrote: > > On Fri, May 13, 2011 at 09:57:23AM -0400, Jim Shupert wrote: > >> > >> On 5/12/2011 5:50 PM, Jim Shupert wrote: > >>> Friends, > >>> > >>> I made a file with ffmpeg [ -vcodec libxvid ] > >>> > >>> ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 48000 -b 8000000 -s 720x480 > >>> -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi > >>> ( and file looks fine - plays in VLC , wonderful ) > >>> > >>> when I analyze a file with Gspot it repots the file video codec to be > >>> xvid ISO MPEG-4. ISO > >>> I think this is xvid 1.3.1 . > >>> when i examine with media info or ffprobe it simply says Codec:xvid. > >>> > >>> What are the major differences between > >>> xvid 1.1.2 and xvid ISO MPEG-4. ISO? > >>> > >>> > >>> If i wish to provide a person Xvid 1.1.2 can I get that older codec > >>> version or would it be suggested to have the person update thier > >>> player codecs. > >>> > >>> They are asking for Xvid 1.1.2 is this a reasonable request isn't it > >>> just older - don't ya think they could just update their codec? > >>> > >>> Simply said is the file I made with ffmpeg [-vcodec libxvid ] simply a > >>> newer version of xvid 1.1.2 or is something fundamentally different. > >>> > >>> Thanks much > >>> > >>> > >> is there any reason to think that the file made with -vcodec libxvid > >> { that i think is Xvid 1.3.1 isn't that true ? } > >> would not be backward compatible to xvid 1.1.2 ? > >> meaning play ok in any player that can play xvid1.1.2 > >> > >> i did some googling -- and i current think > >> -vcodec libxvid = = XviD 1.3.1 = = xvid ISO MPEG-4. ISO -------is that > >> true? > >> and > >> those files should play in anything that plays xvid 1.1.2 > >> > >> any wisdom would be appreciated - > > Hi. > > > > Actually, you have to distinguish two things: > > 1. The bitstream format, here you are talking about mpeg 4 part 2 > > 2. The software that outputs a bitstream, taking raw input and > > compressing it the best it can (in parameters limits) to compress it > > using the bitstream format constructs. xvid stands here. > > > > So, it should be as simple as "use divx/xvid/whatever" to encode, and > > use whatever mpeg4part2 decoder to play. This is often true. > > _______________________________________________ > > > Thank you for your reply -- i do appreciate it. > and I follow what you state here. > > simply said here is my challenge. > I have made files for a fellow who looks at them in gSpot and sees > > xvid ISO MPEG-4. ISO not xvid 1.1.2 I'd then say that gspot have a strange way to report formats, but I understand gspot's output as "mpeg4 part2". > I contend that what i made is xvid 1.3.1 and that is even better. Newer is not always better, but that's another story. > Am i incorrect about that ? isn't the files made via -vcodec libxvid using the codec known as xvid 1.3.1 and isn't that backwards compatible with xvid1.1.2. > actually even better? -vcodec libxvid will encode to mpeg4 part2, whatever version you are using (and you use the version that were found when compiling ffmpeg). But I don't see where is your trouble, making sure your files will play flawlessly on all players/platforms? Target a particular player that says it plays xvid version.no.no? Something else? > > Thanks again! > > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From anil_jangam at persistent.co.in Fri May 13 20:48:25 2011 From: anil_jangam at persistent.co.in (Anil Jangam) Date: Sat, 14 May 2011 00:18:25 +0530 Subject: [FFmpeg-user] Converted file size is bigger than source. In-Reply-To: <4DCD4830.5030000@thelounge.net> References: <4DCD4830.5030000@thelounge.net> Message-ID: Inline comments. > -----Original Message----- > From: ffmpeg-user-bounces at ffmpeg.org [mailto:ffmpeg-user- > bounces at ffmpeg.org] On Behalf Of Reindl Harald > Sent: Friday, May 13, 2011 8:33 PM > To: ffmpeg-user at ffmpeg.org > Subject: Re: [FFmpeg-user] Converted file size is bigger than source. > > > > Am 13.05.2011 16:35, schrieb Anil Jangam: > > Team, > > > > We have observed a scenario in which the size of the output > transcoded file (~500kb) is greater than the size of the original video > file (~140kb). Upon checking with media info tool (and with ffmpeg), we > found that the video codec is h263 and audio codec used is amr (see > below). > > > > Duration: 00:00:13.20, start: 0.000000, bitrate: 86 kb/s Stream > > #0.0(und): Video: h263, yuv420p, 176x144 [PAR 12:11 DAR 4:3], 29.97 > > tbr, 30k tbn, 29.97 tbc Stream #0.1(und): Audio: libopencore_amrnb, > > 8000 Hz, mono, s16 > > > > Upon checking the properties of the output file, it shows that video > codec is 'mpeg4_sp' and audio codec is 'aac_lc'. > > > > I am not sure if 'mpeg4_sp' is a low compression ratio codec as > compare to the 'h263'. In that case, conversion of 'h263' coded video > into 'mpeg4_sp' codec video will obviously result in the higher size of > the transcoded output. > > > > Can you please confirm this? > > Depends on codec and bitrates > > 86 kb/s input file is a very low bitrate if your output is higher there > is no other option as getting bigger with H264/AAC-Audio and a low > bitrate it should get even smaller > > but anyways: post your full cli-call, without it is impossible to say > anything away from guessing [[Anil Jangam]] Thanks Reindl. The command line that we used is below. ffmpeg -i video_dc.3gp -y -vcodec mpeg4 -acodec libfaac -ar 16000 -ab 96000 -async 1 out-video.3gp What could be wrong in this case? What should we change here to keep the size of output video in control? /anil. DISCLAIMER ========== This e-mail may contain privileged and confidential information which is the property of Persistent Systems Ltd. It is intended only for the use of the individual or entity to which it is addressed. If you are not the intended recipient, you are not authorized to read, retain, copy, print, distribute or use this message. If you have received this communication in error, please notify the sender and delete all copies of this message. Persistent Systems Ltd. does not accept any liability for virus infected mails. From jshupert at pps-inc.com Fri May 13 20:54:06 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Fri, 13 May 2011 14:54:06 -0400 Subject: [FFmpeg-user] xVid and version Q In-Reply-To: <20110513174127.GD28878@epicure.lazyet.homelinux.net> References: <4DCC5614.1060608@pps-inc.com> <4DCD38C3.7010002@pps-inc.com> <20110513165742.GC28878@epicure.lazyet.homelinux.net> <4DCD688B.1000302@pps-inc.com> <20110513174127.GD28878@epicure.lazyet.homelinux.net> Message-ID: <4DCD7E4E.4000802@pps-inc.com> On 5/13/2011 1:41 PM, Etienne Buira wrote: > On Fri, May 13, 2011 at 01:21:15PM -0400, Jim Shupert wrote: >> >> On 5/13/2011 12:57 PM, Etienne Buira wrote: >>> On Fri, May 13, 2011 at 09:57:23AM -0400, Jim Shupert wrote: >>>> On 5/12/2011 5:50 PM, Jim Shupert wrote: >>>>> Friends, >>>>> >>>>> I made a file with ffmpeg [ -vcodec libxvid ] >>>>> >>>>> ffmpeg -i D:\path\vid\my.mpg -ab 192000 -ar 48000 -b 8000000 -s 720x480 >>>>> -vcodec libxvid -acodec libmp3lame D:\path\vid\my.avi >>>>> ( and file looks fine - plays in VLC , wonderful ) >>>>> >>>>> when I analyze a file with Gspot it repots the file video codec to be >>>>> xvid ISO MPEG-4. ISO >>>>> I think this is xvid 1.3.1 . >>>>> when i examine with media info or ffprobe it simply says Codec:xvid. >>>>> >>>>> What are the major differences between >>>>> xvid 1.1.2 and xvid ISO MPEG-4. ISO? >>>>> >>>>> >>>>> If i wish to provide a person Xvid 1.1.2 can I get that older codec >>>>> version or would it be suggested to have the person update thier >>>>> player codecs. >>>>> >>>>> They are asking for Xvid 1.1.2 is this a reasonable request isn't it >>>>> just older - don't ya think they could just update their codec? >>>>> >>>>> Simply said is the file I made with ffmpeg [-vcodec libxvid ] simply a >>>>> newer version of xvid 1.1.2 or is something fundamentally different. >>>>> >>>>> Thanks much >>>>> >>>>> >>>> is there any reason to think that the file made with -vcodec libxvid >>>> { that i think is Xvid 1.3.1 isn't that true ? } >>>> would not be backward compatible to xvid 1.1.2 ? >>>> meaning play ok in any player that can play xvid1.1.2 >>>> >>>> i did some googling -- and i current think >>>> -vcodec libxvid = = XviD 1.3.1 = = xvid ISO MPEG-4. ISO -------is that >>>> true? >>>> and >>>> those files should play in anything that plays xvid 1.1.2 >>>> >>>> any wisdom would be appreciated - >>> Hi. >>> >>> Actually, you have to distinguish two things: >>> 1. The bitstream format, here you are talking about mpeg 4 part 2 >>> 2. The software that outputs a bitstream, taking raw input and >>> compressing it the best it can (in parameters limits) to compress it >>> using the bitstream format constructs. xvid stands here. >>> >>> So, it should be as simple as "use divx/xvid/whatever" to encode, and >>> use whatever mpeg4part2 decoder to play. This is often true. >>> _______________________________________________ >>> >> Thank you for your reply -- i do appreciate it. >> and I follow what you state here. >> >> simply said here is my challenge. >> I have made files for a fellow who looks at them in gSpot and sees >> >> xvid ISO MPEG-4. ISO not xvid 1.1.2 > I'd then say that gspot have a strange way to report formats, but I > understand gspot's output as "mpeg4 part2". > >> I contend that what i made is xvid 1.3.1 and that is even better. > Newer is not always better, but that's another story. > yes , i know what you mean....it can be ..... checkered >> Am i incorrect about that ? isn't the files made via -vcodec libxvid using the codec known as xvid 1.3.1 and isn't that backwards compatible with xvid1.1.2. >> actually even better? > -vcodec libxvid will encode to mpeg4 part2, whatever version you are > using (and you use the version that were found when compiling ffmpeg). > > But I don't see where is your trouble, making sure your files will play > flawlessly on all players/platforms? Target a particular player that > says it plays xvid version.no.no? yes , exactly this --- targeting a player that says it plays xvid 1.1.2 -thanks > Something else? > >> Thanks again! >> >> >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > From h.reindl at thelounge.net Fri May 13 21:12:23 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Fri, 13 May 2011 21:12:23 +0200 Subject: [FFmpeg-user] Converted file size is bigger than source. In-Reply-To: References: <4DCD4830.5030000@thelounge.net> Message-ID: <4DCD8297.70401@thelounge.net> Am 13.05.2011 20:48, schrieb Anil Jangam: > Duration: 00:00:13.20, start: 0.000000, bitrate: 86 kb/s Stream > ffmpeg -i video_dc.3gp -y -vcodec mpeg4 -acodec libfaac -ar 16000 -ab 96000 -async 1 out-video.3gp > > What could be wrong in this case? What should we change here to keep the size of output video in control? What about define a -vb? you input file has 86 kb/s you define only audio with 96 kb/s and you wonder why it get's bigger? -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From etienne.buira.lists at free.fr Fri May 13 21:16:21 2011 From: etienne.buira.lists at free.fr (Etienne Buira) Date: Fri, 13 May 2011 21:16:21 +0200 Subject: [FFmpeg-user] xVid and version Q In-Reply-To: <4DCD7E4E.4000802@pps-inc.com> References: <4DCC5614.1060608@pps-inc.com> <4DCD38C3.7010002@pps-inc.com> <20110513165742.GC28878@epicure.lazyet.homelinux.net> <4DCD688B.1000302@pps-inc.com> <20110513174127.GD28878@epicure.lazyet.homelinux.net> <4DCD7E4E.4000802@pps-inc.com> Message-ID: <20110513191621.GE28878@epicure.lazyet.homelinux.net> On Fri, May 13, 2011 at 02:54:06PM -0400, Jim Shupert wrote: > yes , exactly this --- targeting a player that says it plays xvid 1.1.2 So there should be no trouble. From anil_jangam at persistent.co.in Fri May 13 22:11:49 2011 From: anil_jangam at persistent.co.in (Anil Jangam) Date: Sat, 14 May 2011 01:41:49 +0530 Subject: [FFmpeg-user] Converted file size is bigger than source. In-Reply-To: <4DCD8297.70401@thelounge.net> References: <4DCD4830.5030000@thelounge.net> <4DCD8297.70401@thelounge.net> Message-ID: > -----Original Message----- > From: ffmpeg-user-bounces at ffmpeg.org [mailto:ffmpeg-user- > bounces at ffmpeg.org] On Behalf Of Reindl Harald > Sent: Saturday, May 14, 2011 12:42 AM > To: ffmpeg-user at ffmpeg.org > Subject: Re: [FFmpeg-user] Converted file size is bigger than source. > > > > Am 13.05.2011 20:48, schrieb Anil Jangam: > > Duration: 00:00:13.20, start: 0.000000, bitrate: 86 kb/s Stream > ffmpeg > > -i video_dc.3gp -y -vcodec mpeg4 -acodec libfaac -ar 16000 -ab 96000 > > -async 1 out-video.3gp > > > > What could be wrong in this case? What should we change here to keep > the size of output video in control? > > What about define a -vb? [[Anil Jangam]] You mean video bit rate right? The option is -b actually. Option -vb does not exist. I will try this and test. Thanks. > > you input file has 86 kb/s [[Anil Jangam]] Is this 86kb/s bit rate shown by ffmpeg is video bit rate or audio bit rate? Sorry if it too obvious question. > you define only audio with 96 kb/s > > and you wonder why it get's bigger? DISCLAIMER ========== This e-mail may contain privileged and confidential information which is the property of Persistent Systems Ltd. It is intended only for the use of the individual or entity to which it is addressed. If you are not the intended recipient, you are not authorized to read, retain, copy, print, distribute or use this message. If you have received this communication in error, please notify the sender and delete all copies of this message. Persistent Systems Ltd. does not accept any liability for virus infected mails. From jshupert at pps-inc.com Fri May 13 22:10:17 2011 From: jshupert at pps-inc.com (Jim Shupert) Date: Fri, 13 May 2011 16:10:17 -0400 Subject: [FFmpeg-user] xVid and version Q In-Reply-To: <20110513191621.GE28878@epicure.lazyet.homelinux.net> References: <4DCC5614.1060608@pps-inc.com> <4DCD38C3.7010002@pps-inc.com> <20110513165742.GC28878@epicure.lazyet.homelinux.net> <4DCD688B.1000302@pps-inc.com> <20110513174127.GD28878@epicure.lazyet.homelinux.net> <4DCD7E4E.4000802@pps-inc.com> <20110513191621.GE28878@epicure.lazyet.homelinux.net> Message-ID: <4DCD9029.5050205@pps-inc.com> On 5/13/2011 3:16 PM, Etienne Buira wrote: > On Fri, May 13, 2011 at 02:54:06PM -0400, Jim Shupert wrote: >> yes , exactly this --- targeting a player that says it plays xvid 1.1.2 > So there should be no trouble. > _______________________________________________ Thanks -- i owe you one , brother have a great day! From herve.flores at free.fr Fri May 13 22:22:46 2011 From: herve.flores at free.fr (Herve Flores) Date: Fri, 13 May 2011 22:22:46 +0200 Subject: [FFmpeg-user] flag YUV input as interlaced In-Reply-To: <4DCC05AD.70102@mdsh.com> References: <4DCBCED3.9050103@mdsh.com><017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com><020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> <4DCBFA17.2070303@mdsh.com> <024601cc10ba$8d331b40$4301a8c0@hpkantoor> <4DCC05AD.70102@mdsh.com> Message-ID: <6EEA57F7-01F2-4FB3-8030-9F3568E37C16@free.fr> Le 12 mai 2011 ? 18:07, Mark Himsley a ?crit : > On 12/05/11 16:37, bouke wrote: >> >> ---- Original Message ----- >> From: "Mark Himsley" >> To: "FFmpeg user questions and RTFMs" >> Sent: Thursday, May 12, 2011 5:17 PM >> Subject: Re: [FFmpeg-user] flag YUV input as interlaced >> >> >>> On 12/05/11 16:03, bouke wrote: >>>> ----- Original Message ----- >>>> From: "Mark Himsley" >>>> To: "FFmpeg user questions and RTFMs" >>>> Sent: Thursday, May 12, 2011 4:51 PM >>>> Subject: Re: [FFmpeg-user] flag YUV input as interlaced >>>> >>>> >>>>> On 12/05/11 14:11, bouke wrote: >>>>>> >>>>>> ----- Original Message ----- >>>>>> From: "Mark Himsley" >>>>>> To: "FFmpeg user questions and RTFMs" >>>>>> Sent: Thursday, May 12, 2011 2:13 PM >>>>>> Subject: [FFmpeg-user] flag YUV input as interlaced >>>>>> >>>>>> >>>>>>> Imagine I am using FFmpeg to read a raw YUV file. How do I flag that >>>>>>> input >>>>>>> file as interlaced so that interlaced aware filters process the stream >>>>>>> correctly? >>>>>>> >>>>>>> For instance: >>>>>>> I have an SD PAL sized yvyu422 file that I know is top-field-first >>>>>>> I want to output that as PAL DV 25, which is bottom-field-first >>>>>>> >>>>>>> I expect to use a command like like this: >>>>>>> >>>>>>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -i >>>>>>> input.yvyu422 -vf >>>>>>> fieldorder=bff -vcodec dvvideo -pix_fmt yuv420p -y output.mov >>>>>> >>>>>> Normally you put the known input specs before the input file, as in >>>>>> your >>>>>> example. >>>>>> So did you test >>>>>> ffmpeg -f rawvideo -r 25 -s 720x576 -pix_fmt yuyv422 -vf >>>>>> eldorder=tff -i input.yvyu422 -vf fieldorder=bff -vcodec >>>>>> dvvideo -pix_fmt >>>>>> yuv420p -y output.mov >>>>>> >>>>>> ? [?] > I'm only using fieldorder as an example. > > But, FYI, the field order filter DOES drop (or raise) the picture by one line when doing the tff <-> bff conversion. It ALSO adds pertinent picture data into the new line created by the crop+pad, AND it only does the tff <-> bff conversion if the original file is not in the interlaced format that has been requested. > > Anyway... > > What I need is a way to flag rawvideo media when its used as an input into ffmpeg. Anyone? Hi first, sorry this is not the answer you searched, I searched too, but I didn't find any way to do :-( I have the same concern with converted 1080i(ttf) to DV: some QuickTime versions does a good job during this conversion (good resample, preserve interlacing) BUT does not change field order, so my DV file contents is ttf (but flaged/readed as bff by ffmpeg) So your cool filter has no effect another eg: my file has an interlaced -ttf- contents, but is -badly- encoded as progressive, your filter cannot work here too (From memories about the cvs-log-list) I'm not sure that a automatic filter was the best choice. Is is possible to modify your filter to force fieldorder inside it? eg: "-vf fieldorder=ttf,bff" (or "-vf fieldorder=0,1" to keep the usual ffmpeg syntax) ?and a syntax like "-vf fieldorder=-1,1" will produce the same behavior than the actual filter ;-) (automatic detection of input fieldorder > change -or not- fieldorder according to 2nd parameter) was just my 2 cents as user good continuation bye Herv? From h.reindl at thelounge.net Fri May 13 23:37:38 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Fri, 13 May 2011 23:37:38 +0200 Subject: [FFmpeg-user] Converted file size is bigger than source. In-Reply-To: References: <4DCD4830.5030000@thelounge.net> <4DCD8297.70401@thelounge.net> Message-ID: <4DCDA4A2.1050009@thelounge.net> Am 13.05.2011 22:11, schrieb Anil Jangam: >> What about define a -vb? > > [[Anil Jangam]] You mean video bit rate right? The option is -b actually. Option -vb does not exist. > I will try this and test. Thanks. consider a newer ffmpeg-version -vb exists since a very long time Video options: -b bitrate set bitrate (in bits/s) -vb bitrate set bitrate (in bits/s) -vframes number set the number of video frames to record -r rate set frame rate (Hz value, fraction or abbreviation) -s size set frame size (WxH or abbreviation) -aspect aspect set aspect ratio (4:3, 16:9 or 1.3333, 1.7777) -vn disable video -vcodec codec force video codec ('copy' to copy stream) -sameq use same video quality as source (implies VBR) -pass n select the pass number (1 or 2) -passlogfile prefix select two pass log file name prefix -newvideo add a new video stream to the current output stream -vlang code set the ISO 639 language code (3 letters) of the current video stream >> you input file has 86 kb/s > > [[Anil Jangam]] Is this 86kb/s bit rate shown by ffmpeg is video bit rate or audio bit rate? Sorry if it too obvious question. Not 100% sure about the output but since it is the first line with duration i guess both combined But anyways, even if it is vbr only your audio-bitrate is higher alone Newer versions have better output Duration: 01:58:39.04, start: 0.000000, bitrate: 1937 kb/s Stream #0.0: Video: msmpeg4, yuv420p, 720x576, 25 tbr, 25 tbn, 25 tbc Stream #0.1: Audio: mp3, 48000 Hz, stereo, s16, 128 kb/s -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From johnbrown105 at hotmail.com Sat May 14 06:52:27 2011 From: johnbrown105 at hotmail.com (John Brown) Date: Sat, 14 May 2011 04:52:27 +0000 (UTC) Subject: [FFmpeg-user] New options -preset and -tune for x264 encoding References: <20110424093945.GA31290@mars> Message-ID: Klaus Kudielka writes: > > Hi, > > ... there is one big difference > that makes ffmpeg less usable for x264 encoding: > > (a) x264 & ffmpeg -vpre => user options override presets > (b) ffmpeg -preset => presets override user options > > Personally I would STRONGLY vote that ffmpeg -preset also conforms to > behaviour (a). > Me too. I need to override -nr (default=0), but now it has no effect. Regards, Alias John Brown. From herve.flores at free.fr Sat May 14 10:26:04 2011 From: herve.flores at free.fr (Herve Flores) Date: Sat, 14 May 2011 10:26:04 +0200 Subject: [FFmpeg-user] dv => mp4: deinterlace or not, and how? In-Reply-To: <4DC80BE0.1060109@bbc.co.uk> References: <4DC30A66.3040006@gmail.com> <4DC335D8.5080504@gmail.com> <4DC42F6C.9060207@gmail.com> <4DC445FD.8050003@gmail.com> <4DC79E85.5050704@bbc.co.uk><004901cc0e1f$6b2c1d50$4301a8c0@hpkantoor> <4DC7A5EF.5010608@bbc.co.uk> <00c201cc0e29$93bb85d0$4301a8c0@hpkantoor> <4DC80BE0.1060109@bbc.co.uk> Message-ID: <8B7C8DEC-735B-4731-B423-9A715336D1B5@free.fr> Le 9 mai 2011 ? 17:44, Tim Nicholson a ?crit : > On 09/05/11 10:14, bouke wrote: >> >> ----- Original Message ----- >> From: "Tim Nicholson" >> To: "FFmpeg user questions and RTFMs" >> Sent: Monday, May 09, 2011 10:29 AM >> Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? >> >> >>> On 09/05/11 09:02, Bouke wrote: >>>> >>>> ---- Original Message ----- >>>> From: "Tim Nicholson" >>>> To: "FFmpeg user questions and RTFMs" >>>> Sent: Monday, May 09, 2011 9:57 AM >>>> Subject: Re: [FFmpeg-user] dv => mp4: deinterlace or not, and how? >>>>> [..] >>>>> But what I would really like is to go from 720p50 to 576i25 (SD) ;) >>>> >>>> Tim, >>>> You know that AE can do this? >>>> (Perhaps you don't like the downscaler in AE, but it is a handy tool to >>>> split fields to frames and vice versa, and it can be scripted) >>>> >>> [...] >> >> Ok, you want cheap, you get cheap. > > Not so much that as not using a sledgehammer. > > It has been pointed out to me that the tinterlace filter in mplayer can make a useful nutcracker in this case (with a tweaked version in ffmbc). > > >> Found a way with QT to do it. > > [...] >> Pick one of the two and apply a interlaced key signal as mask. >> (One line black / white alternating) > >...] > > Now that is a goldie oldie. Somewhere I still have a field test image with field one lines all red and field two all blue. Great for spotting dominance changes but not so good for the eyes if you freeze frame instead of freeze field...:( ?was maybe mine ;-) the aim of this stupid video was to test how each flat screen TV handles interlacing, and what they do with chroma (eg: if the TV displays a violet screen, it does a bad job with chroma ;-)) I made a "stupid DVD-VIDEO" for friends, it's still available here PS: I'm french, so the caption onto the video means: "What color do you see?" bye Herv? From stefano.sabatini-lala at poste.it Sat May 14 11:26:03 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Sat, 14 May 2011 11:26:03 +0200 Subject: [FFmpeg-user] flag YUV input as interlaced In-Reply-To: <4DCC05AD.70102@mdsh.com> References: <4DCBCED3.9050103@mdsh.com> <017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com> <020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> <4DCBFA17.2070303@mdsh.com> <024601cc10ba$8d331b40$4301a8c0@hpkantoor> <4DCC05AD.70102@mdsh.com> Message-ID: <20110514092603.GC13180@geppetto> On date Thursday 2011-05-12 17:07:09 +0100, Mark Himsley encoded: [...] > Anyway... > > What I need is a way to flag rawvideo media when its used as an > input into ffmpeg. Anyone? Mark, did you try with -top? -- ffmpeg-user random tip #13 Have you ever *seen* ffmpeg? find ~/src/ffmpeg/ -type f | xargs cat | ffmpeg -s qcif \ -f rawvideo -i - -y ff.mpeg From dashing.meng at gmail.com Sat May 14 12:21:57 2011 From: dashing.meng at gmail.com (littlebat) Date: Sat, 14 May 2011 18:21:57 +0800 Subject: [FFmpeg-user] Is it possible to run multiple ffmpeg intances? Message-ID: <20110514182157.600cd069.dashing.meng@gmail.com> Hi, I try to convert serveral videos using multiple ffmpeg intances concurrently, but it failed, only get 44 bytes result file. Is it possible to run multiple ffmpeg intances concurrently? or, there are some things wrong in my command as below? ffmpeg -y -i /home/mdx/pipetest/t1.mkv \ -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ -acodec libopencore_amrnb \ -ac 1 -ar 8000 -ab 10.2k -f 3gp 1.3gp & \ ffmpeg -y -i /home/mdx/pipetest/t2.mkv \ -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ -acodec libopencore_amrnb \ -ac 1 -ar 8000 -ab 10.2k -f 3gp 2.3gp & \ ffmpeg -y -i /home/mdx/pipetest/t3.mkv \ -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ -acodec libopencore_amrnb \ -ac 1 -ar 8000 -ab 10.2k -f 3gp 3.3gp & thanks. littlebat From h.reindl at thelounge.net Sat May 14 12:54:10 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Sat, 14 May 2011 12:54:10 +0200 Subject: [FFmpeg-user] Is it possible to run multiple ffmpeg intances? In-Reply-To: <20110514182157.600cd069.dashing.meng@gmail.com> References: <20110514182157.600cd069.dashing.meng@gmail.com> Message-ID: <4DCE5F52.4020207@thelounge.net> are you starting this as one-liner? start each command with the & at the end in a own line multiple ffmpeg-instances can not be a problem because this affects the operating system, not ffmpeg as long each instance has its own target - will mean: they do not know that they are running not alno Am 14.05.2011 12:21, schrieb littlebat: > Hi, > > I try to convert serveral videos using multiple ffmpeg intances > concurrently, but it failed, only get 44 bytes result > file. Is it possible to run multiple ffmpeg intances concurrently? or, > there are some things wrong in my command as below? > > > ffmpeg -y -i /home/mdx/pipetest/t1.mkv \ > -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ > -acodec libopencore_amrnb \ > -ac 1 -ar 8000 -ab 10.2k -f 3gp 1.3gp & \ > ffmpeg -y -i /home/mdx/pipetest/t2.mkv \ > -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ > -acodec libopencore_amrnb \ > -ac 1 -ar 8000 -ab 10.2k -f 3gp 2.3gp & \ > ffmpeg -y -i /home/mdx/pipetest/t3.mkv \ > -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ > -acodec libopencore_amrnb \ > -ac 1 -ar 8000 -ab 10.2k -f 3gp 3.3gp & > > > thanks. > > littlebat > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user -- Mit besten Gr??en, Reindl Harald the lounge interactive design GmbH A-1060 Vienna, Hofm?hlgasse 17 CTO / software-development / cms-solutions p: +43 (1) 595 3999 33, m: +43 (676) 40 221 40 icq: 154546673, http://www.thelounge.net/ http://www.thelounge.net/signature.asc.what.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From etienne.buira.lists at free.fr Sat May 14 12:55:28 2011 From: etienne.buira.lists at free.fr (Etienne Buira) Date: Sat, 14 May 2011 12:55:28 +0200 Subject: [FFmpeg-user] Is it possible to run multiple ffmpeg intances? In-Reply-To: <20110514182157.600cd069.dashing.meng@gmail.com> References: <20110514182157.600cd069.dashing.meng@gmail.com> Message-ID: <20110514105527.GF28878@epicure.lazyet.homelinux.net> On Sat, May 14, 2011 at 06:21:57PM +0800, littlebat wrote: > Hi, > > I try to convert serveral videos using multiple ffmpeg intances > concurrently, but it failed, only get 44 bytes result > file. Is it possible to run multiple ffmpeg intances concurrently? or, > there are some things wrong in my command as below? > > > ffmpeg -y -i /home/mdx/pipetest/t1.mkv \ > -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ > -acodec libopencore_amrnb \ > -ac 1 -ar 8000 -ab 10.2k -f 3gp 1.3gp & \ > ffmpeg -y -i /home/mdx/pipetest/t2.mkv \ > -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ > -acodec libopencore_amrnb \ > -ac 1 -ar 8000 -ab 10.2k -f 3gp 2.3gp & \ > ffmpeg -y -i /home/mdx/pipetest/t3.mkv \ > -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ > -acodec libopencore_amrnb \ > -ac 1 -ar 8000 -ab 10.2k -f 3gp 3.3gp & > > > thanks. > > littlebat Hi. There should be no problem, however, with this kind of construct, you will miss error messages (or at least get them mixed together). You can redirect stdin/out/err, or run each instance in a shell (I use screen http://www.gnu.org/software/screen/ for that). From dashing.meng at gmail.com Sat May 14 16:41:49 2011 From: dashing.meng at gmail.com (littlebat) Date: Sat, 14 May 2011 22:41:49 +0800 Subject: [FFmpeg-user] Can anyone explain why need append " Hi, I try to convert video using ffmpeg in background, but it failed, only get 44 bytes result file. The command as below: ffmpeg -y -i /home/mdx/pipetest/t1.mkv \ -vcodec h263 -b 200k -r 15 -s 176x144 -aspect 11:9 \ -acodec libopencore_amrnb \ -ac 1 -ar 8000 -ab 10.2k -f 3gp 1.3gp & But, when I replace "&" with " References: <4DCBCED3.9050103@mdsh.com> <017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com> <020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> <4DCBFA17.2070303@mdsh.com> <024601cc10ba$8d331b40$4301a8c0@hpkantoor> <4DCC05AD.70102@mdsh.com> <20110514092603.GC13180@geppetto> Message-ID: <4DCEB9C2.8040103@mdsh.com> On 14/05/2011 10:26, Stefano Sabatini wrote: > On date Thursday 2011-05-12 17:07:09 +0100, Mark Himsley encoded: > [...] >> Anyway... >> >> What I need is a way to flag rawvideo media when its used as an >> input into ffmpeg. Anyone? > > Mark, did you try with -top? Hi Stefano, Yes. Sorry, I should have said that -top was my first guess, but your showinfo filter says the rawvideo file is flagged as progressive when I have -top 1 in the input files settings. ./ffmpeg -f rawvideo -s 720x576 -r 25 -pix_fmt uyvy422 -top 1 -i ~/Videos/Clock10tone.uyvy422 -vf showinfo -vcodec dvvideo -pix_fmt yuv420p ~/Videos/junk.mov ffmpeg version git-N-29944-gc0b1cae, Copyright (c) 2000-2011 the FFmpeg developers built on May 14 2011 17:01:50 with gcc 4.5.2 configuration: --prefix=/usr/local --enable-ffplay --enable-ffprobe --enable-ffserver --enable-gpl --enable-nonfree --enable-pthreads --enable-runtime-cpudetect --enable-libdirac --enable-libfaac --enable-libgsm --enable-libmp3lame --enable-libopenjpeg --enable-libschroedinger --enable-libspeex --enable-libtheora --enable-libvorbis --enable-libxvid --enable-libx264 --enable-x11grab libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 5. 0 / 53. 5. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 5. 0 / 2. 5. 0 libswscale 0. 14. 0 / 0. 14. 0 libpostproc 51. 2. 0 / 51. 2. 0 [rawvideo @ 0x97f02c0] Estimating duration from bitrate, this may be inaccurate Input #0, rawvideo, from '/home/mdsh/Videos/Clock10tone.uyvy422': Duration: N/A, start: 0.000000, bitrate: N/A Stream #0.0: Video: rawvideo, uyvy422, 720x576, 25 tbr, 25 tbn, 25 tbc [buffer @ 0x97f0180] w:720 h:576 pixfmt:uyvy422 tb:1/1000000 sar:0/1 [showinfo @ 0x97dfe00] auto-inserting filter 'auto-inserted scaler 0' between the filter 'src' and the filter 'Parsed filter 0 showinfo' [scale @ 0x97e04a0] w:720 h:576 fmt:uyvy422 -> w:720 h:576 fmt:yuv420p flags:0xa2000004 Output #0, mov, to '/home/mdsh/Videos/junk.mov': Metadata: encoder : Lavf53.0.3 Stream #0.0: Video: dvvideo, yuv420p, 720x576, q=2-31, 200 kb/s, 25 tbn, 25 tbc Stream mapping: Stream #0.0 -> #0.0 Press [q] to stop encoding [showinfo @ 0x97dfe00] n:0 pts:0 pts_time:0.000000 pos:0 fmt:yuv420p sar:0/1 s:720x576 i:P iskey:1 type:? crc:154113476 plane_crc:[1658531253 4049777664 4049777664 0] [showinfo @ 0x97dfe00] n:1 pts:40000 pts_time:0.040000 pos:829440 fmt:yuv420p sar:0/1 s:720x576 i:P iskey:1 type:? crc:154113476 plane_crc:[1658531253 4049777664 4049777664 0] [...] [showinfo @ 0x97dfe00] n:249 pts:9960000 pts_time:9.960000 pos:206530560 fmt:yuv420p sar:0/1 s:720x576 i:P iskey:1 type:? crc:154113476 plane_crc:[1658531253 4049777664 4049777664 0] frame= 250 fps= 50 q=0.0 Lsize= 35158kB time=10.00 bitrate=28801.3kbits/s video:35156kB audio:0kB global headers:0kB muxing overhead 0.004647% -- Mark From stefano.sabatini-lala at poste.it Sat May 14 19:46:54 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Sat, 14 May 2011 19:46:54 +0200 Subject: [FFmpeg-user] flag YUV input as interlaced In-Reply-To: <4DCEB9C2.8040103@mdsh.com> References: <4DCBCED3.9050103@mdsh.com> <017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com> <020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> <4DCBFA17.2070303@mdsh.com> <024601cc10ba$8d331b40$4301a8c0@hpkantoor> <4DCC05AD.70102@mdsh.com> <20110514092603.GC13180@geppetto> <4DCEB9C2.8040103@mdsh.com> Message-ID: <20110514174654.GA18009@geppetto> On date Saturday 2011-05-14 18:20:02 +0100, Mark Himsley encoded: > On 14/05/2011 10:26, Stefano Sabatini wrote: > > On date Thursday 2011-05-12 17:07:09 +0100, Mark Himsley encoded: > > [...] > >> Anyway... > >> > >> What I need is a way to flag rawvideo media when its used as an > >> input into ffmpeg. Anyone? > > > > Mark, did you try with -top? > > Hi Stefano, > > Yes. Sorry, I should have said that -top was my first guess, but your > showinfo filter says the rawvideo file is flagged as progressive when I > have -top 1 in the input files settings. Yes, check ffmpeg.c:do_video_out, top_field_first is used for setting the field type *just before encoding*. I wonder what could be a better approach, maybe we could implement a setprops filter for changing the properties of the video in the filterchain, and drop -top from ffmpeg.c. What do you think of this idea? -- ffmpeg-user random tip #12 One minute of video noise with ffmpeg: ffmpeg -t 60 -s qcif -f rawvideo -pix_fmt rgb24 -r 25 -i /dev/urandom \ -y noise.mpeg From mark at mdsh.com Sun May 15 20:48:11 2011 From: mark at mdsh.com (Mark Himsley) Date: Sun, 15 May 2011 19:48:11 +0100 Subject: [FFmpeg-user] flag YUV input as interlaced In-Reply-To: <20110514174654.GA18009@geppetto> References: <4DCBCED3.9050103@mdsh.com> <017f01cc10a6$228be1f0$4301a8c0@hpkantoor> <4DCBF3E8.3050700@mdsh.com> <020d01cc10b5$c15bb6c0$4301a8c0@hpkantoor> <4DCBFA17.2070303@mdsh.com> <024601cc10ba$8d331b40$4301a8c0@hpkantoor> <4DCC05AD.70102@mdsh.com> <20110514092603.GC13180@geppetto> <4DCEB9C2.8040103@mdsh.com> <20110514174654.GA18009@geppetto> Message-ID: <4DD01FEB.9090507@mdsh.com> On 14/05/2011 18:46, Stefano Sabatini wrote: > On date Saturday 2011-05-14 18:20:02 +0100, Mark Himsley encoded: >> On 14/05/2011 10:26, Stefano Sabatini wrote: >>> On date Thursday 2011-05-12 17:07:09 +0100, Mark Himsley encoded: >>> [...] >>>> Anyway... >>>> >>>> What I need is a way to flag rawvideo media when its used as an >>>> input into ffmpeg. Anyone? >>> >>> Mark, did you try with -top? >> >> Hi Stefano, >> >> Yes. Sorry, I should have said that -top was my first guess, but your >> showinfo filter says the rawvideo file is flagged as progressive when I >> have -top 1 in the input files settings. > > Yes, check ffmpeg.c:do_video_out, top_field_first is used for setting > the field type *just before encoding*. > > I wonder what could be a better approach, maybe we could implement a > setprops filter for changing the properties of the video in the > filterchain, and drop -top from ffmpeg.c. > > What do you think of this idea? Replied in FFmpeg-devel. -- Mark From shop at open-t.co.uk Mon May 16 00:11:36 2011 From: shop at open-t.co.uk (Sebastian Arcus) Date: Sun, 15 May 2011 23:11:36 +0100 Subject: [FFmpeg-user] Drawtext filter with date overlay is not working in git-N-29946-g27614b1 Message-ID: <4DD04F98.9090305@open-t.co.uk> Just downloaded latest ffmpeg from git and tried the drawtext filter with date overlay. No errors on the command line, but the date on the video is completely static - it picks up the date the moment ffmpeg is started, and it doesn't change any more. The command I use is: ffmpeg -fflags +genpts -t 600 -f mjpeg -r 8 -s 640x480 \ -i http://localhost:8080/?action=stream -vcodec mpeg4 \ -vf drawtext="fontfile=/usr/share/fonts/TTF/mitra.ttf:x=70:y=455: \ text='\%H\:\%M\:\%S | \%a \%d/\%b/\%Y | S500ATV | camera 0': \ fontcolor=0xFFFFFFFF:fontsize=18: \ shadowcolor=0x000000EE:shadowx=1:shadowy=1" \ -b 1500000 -r 8 video_file.avi This used to work just fine only few weeks ago. The input stream comes from mjpg-streamer, which generates it from a usb webcam - but I don't think that makes any difference. Anybody any suggestions? Any recent changes that broke the drawtext filter? Sebastian From seandarcy2 at gmail.com Mon May 16 00:29:35 2011 From: seandarcy2 at gmail.com (sean darcy) Date: Sun, 15 May 2011 18:29:35 -0400 Subject: [FFmpeg-user] 420p out from ffmpeg becomes 420i in to x264 Message-ID: I'm trying to use ffmpeg to pipe a stream to x264. Both ffmpeg and x264 from git today. I'm deinterlacing the input with yadif, piped output is yuv4mpeg. And ffmpeg thinks is generating yuv420p. But x264 is seeing the piped stream as interlaced. I thought yuv4mpeg had a header describing the stream. Is that so? If so, is the problem here that ffmpeg is not writing the header correctly, x264 is not reading it correctly, or have I screwed up something? sean + ffmpeg -i JulyPlay1990.avi -an -vf yadif=1:0,hqdn3d -r 60000/1001 -pix_fmt yuv420p -f yuv4mpegpipe - + x264 --demuxer y4m --preset slower --tune film --bitrate 1200 -o out3.m4v - ffmpeg version git-N-29946-g27614b1, Copyright (c) 2000-2011 the FFmpeg developers ...... Input #0, avi, from 'JulyPlay1990-interlaced.avi': Metadata: encoder : Lavf53.0.3 Duration: 00:21:40.66, start: 0.000000, bitrate: 30312 kb/s Stream #0.0: Video: dvvideo, yuv411p, 720x480 [PAR 8:9 DAR 4:3], 29.97 tbr, 29.97 tbn, 29.97 tbc Stream #0.1: Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s [buffer @ 0xe12400] w:720 h:480 pixfmt:yuv411p [yadif @ 0xe12ec0] mode:1 parity:0 [hqdn3d @ 0xe0db40] ls:4.000000 cs:3.000000 lt:6.000000 ct:4.500000 [hqdn3d @ 0xe0db40] auto-inserting filter 'auto-inserted scaler 0' between the filter 'Parsed filter 0 yadif' and the filter 'Parsed filter 1 hqdn3d' [scale @ 0xe0e380] w:720 h:480 fmt:yuv411p -> w:720 h:480 fmt:yuv420p flags:0xa0000004 Output #0, yuv4mpegpipe, to 'pipe:': Metadata: encoder : Lavf53.0.3 Stream #0.0: Video: rawvideo, yuv420p, 720x480 [PAR 8:9 DAR 4:3], q=2-31, 200 kb/s, 90k tbn, 59.94 tbc Stream mapping: Stream #0.0 -> #0.0 Press [q] to stop encoding y4m [info]: 720x480i 8:9 @ 60000/1001 fps (cfr) x264 [warning]: input appears to be interlaced, enabling bff interlaced mode. If you want otherwise, use --no-interlaced or --tff From rickcorteza at gmail.com Mon May 16 06:20:15 2011 From: rickcorteza at gmail.com (Rick C.) Date: Mon, 16 May 2011 12:20:15 +0800 Subject: [FFmpeg-user] understanding copy flag Message-ID: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> Hello, I was making a few tests using the following syntax: ffmpeg -i original.avi -acodec copy -vcodec copy output.avi To my surprise the results were not very good (mainly just the audio side I believe). Am I wrong in thinking this should work as a passthrough? I tried the same with .mp4 and .mpg and the results were what I expected. I didn't notice anything significant from the output log so I didn't post anything. Any advice would be great. Thanks! rc From hustmobile at gmail.com Mon May 16 07:02:48 2011 From: hustmobile at gmail.com (HUST Mobile) Date: Mon, 16 May 2011 13:02:48 +0800 Subject: [FFmpeg-user] Is there any plans to integrate Opencore AAC decoder to ffmpeg? Message-ID: Is there any plans to integrate Opencore AAC decoder to ffmpeg? There are already 2 AAC decoders for ffmpeg: libfaad and aac decoder of ffmpeg. For AAC-LC , the aac decoder of ffmpeg decoding performance is good. However, for EAAC+, it is rather slow. So, adding the Opencore AAC decoder to ffmpeg as a modules seems useful, especially for those mobile platform. How to start integrating Opencore AAC decoder to ffmpeg, just as the libfaad? And Opencore AAC decoder is Apache licensed, is this good for take it as a ffmpeg external module? From Cecil at decebal.nl Mon May 16 11:02:56 2011 From: Cecil at decebal.nl (Cecil Westerhof) Date: Mon, 16 May 2011 11:02:56 +0200 Subject: [FFmpeg-user] How to get the good type of mp4 Message-ID: <87k4draugv.fsf@Compaq.site> I made a mp4 file for someone. A file they can play is: Input #0, mov,mp4,m4a,3gp,3g2,mj2, from '/WinTransfer/blog.mp4': Metadata: major_brand : mp42 minor_version : 1 compatible_brands: M4V mp42isom creation_time : 2010-10-27 22:48:19 Duration: 00:01:17.27, start: 0.000000, bitrate: 86 kb/s Stream #0.0(eng): Video: h264 (High), yuv420p, 640x480 [PAR 1:1 DAR 4:3], 30 kb/s, 10 fps, 10 tbr, 10k tbn, 20 tbc Metadata: creation_time : 2010-10-27 22:48:19 Stream #0.1(eng): Audio: aac, 22050 Hz, stereo, s16, 54 kb/s Metadata: creation_time : 2010-10-27 22:48:19 I managed to make a file like: Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'huiswerkIQCoaches02.mp4': Metadata: major_brand : isom minor_version : 512 compatible_brands: isomiso2avc1mp41 creation_time : 1970-01-01 00:00:00 encoder : Lavf52.108.0 Duration: 00:12:36.04, start: 0.000000, bitrate: 78 kb/s Stream #0.0(und): Video: h264 (High), yuv420p, 640x480 [PAR 1:1 DAR 4:3], 29 kb/s, 20 fps, 20 tbr, 20 tbn, 40 tbc Metadata: creation_time : 1970-01-01 00:00:00 Stream #0.1(und): Audio: aac, 22050 Hz, stereo, s16, 45 kb/s Metadata: creation_time : 1970-01-01 00:00:00 They told me it worked. But now it turns out they only tested it in Windows Media player. They need to give a demo of there website with the video's integrated and now they find they do not play in there website. (Why they did not test this is beyond me.) When looking at it, I think the problem is the Metadata. I tried working with vtag and metadata, but to no avail. What would be a way to get the video playing? For the curious, these are the parameters I use: -strict experimental -acodec aac -ar 22050 -ab 64k -vcodec libx264 -vpre slow -level 41 -vtag mp42 -crf 20 -bufsize 20000k -maxrate 25000k -g 250 -r 20 -s vga -metadata major_brand="mp42" -ss 0 -t 30 -coder 1 -flags +loop -cmp +chroma -partitions +parti4x4+partp8x8+partb8x8 -subq 7 -me_range 16 -keyint_min 25 -sc_threshold 40 -i_qfactor 0.71 -rc_eq 'blurCplx^(1-qComp)' -bf 16 -b_strategy 1 -bidir_refine 1 -refs 6 -deblockalpha 0 -deblockbeta 0 By the way: why is creation_time not set? -- Cecil Westerhof Senior Software Engineer LinkedIn: http://www.linkedin.com/in/cecilwesterhof From Cecil at decebal.nl Mon May 16 11:44:43 2011 From: Cecil at decebal.nl (Cecil Westerhof) Date: Mon, 16 May 2011 11:44:43 +0200 Subject: [FFmpeg-user] How to get the good type of mp4 In-Reply-To: <87k4draugv.fsf@Compaq.site> (Cecil Westerhof's message of "Mon, 16 May 2011 11:02:56 +0200") References: <87k4draugv.fsf@Compaq.site> Message-ID: <877h9rasj8.fsf@Compaq.site> It was not the metadata. It was fps. When changing it to 10 it worked. But is there a way to set the metadata? Op maandag 16 mei 2011 11:02 CEST schreef Cecil Westerhof: > I made a mp4 file for someone. A file they can play is: > Input #0, mov,mp4,m4a,3gp,3g2,mj2, from '/WinTransfer/blog.mp4': > Metadata: > major_brand : mp42 > minor_version : 1 > compatible_brands: M4V mp42isom > creation_time : 2010-10-27 22:48:19 > Duration: 00:01:17.27, start: 0.000000, bitrate: 86 kb/s > Stream #0.0(eng): Video: h264 (High), yuv420p, 640x480 [PAR 1:1 DAR 4:3], 30 > kb/s, 10 fps, 10 tbr, 10k tbn, 20 tbc > Metadata: > creation_time : 2010-10-27 22:48:19 > Stream #0.1(eng): Audio: aac, 22050 Hz, stereo, s16, 54 kb/s > Metadata: > creation_time : 2010-10-27 22:48:19 > > I managed to make a file like: > Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'huiswerkIQCoaches02.mp4': > Metadata: > major_brand : isom > minor_version : 512 > compatible_brands: isomiso2avc1mp41 > creation_time : 1970-01-01 00:00:00 > encoder : Lavf52.108.0 > Duration: 00:12:36.04, start: 0.000000, bitrate: 78 kb/s > Stream #0.0(und): Video: h264 (High), yuv420p, 640x480 [PAR 1:1 DAR 4:3], 29 > kb/s, 20 fps, 20 tbr, 20 tbn, 40 tbc > Metadata: > creation_time : 1970-01-01 00:00:00 > Stream #0.1(und): Audio: aac, 22050 Hz, stereo, s16, 45 kb/s > Metadata: > creation_time : 1970-01-01 00:00:00 > > They told me it worked. But now it turns out they only tested it in > Windows Media player. They need to give a demo of there website with > the video's integrated and now they find they do not play in there > website. (Why they did not test this is beyond me.) > > When looking at it, I think the problem is the Metadata. I tried > working with vtag and metadata, but to no avail. What would be a way > to get the video playing? -- Cecil Westerhof Senior Software Engineer LinkedIn: http://www.linkedin.com/in/cecilwesterhof From mroper at kinect.co.nz Mon May 16 12:25:04 2011 From: mroper at kinect.co.nz (mroper) Date: Mon, 16 May 2011 22:25:04 +1200 Subject: [FFmpeg-user] trying to work out ffmpeg and avi conversion In-Reply-To: <4DCBBD6C.5070400@zonnet.nl> References: <4DCB997B.3050406@kinect.co.nz> <4DCBBD6C.5070400@zonnet.nl> Message-ID: <4DD0FB80.9040307@kinect.co.nz> thanks for the reply. yes, i've tried that and it does work, however it doesn't play on my panasonic tv, i have tried going through and trying to match settings which do not work but there must be something I am missing. any clues you can give me based on the differences below (I have tried changing the 1280x720 to a lower size, with no joy, however would like to keep in HD which the TV supports) Not Working Duration: 00:00:02.02, start: 0.000000, bitrate: 12358 kb/s Stream #0.0: Video: mpeg2video, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 104857 kb/s, 30 fps, 30 tbr, 30 tbn, 60 tbc Metadata: strn : FUJIFILM AVI STREAM 0100 Stream #0.1: Audio: ac3, 44100 Hz, stereo, s16, 64 kb/s Working Duration: 00:00:09.00, start: 0.500000, bitrate: 9624 kb/s Stream #0.0[0x1e0]: Video: mpeg2video, yuv420p, 720x576 [PAR 1:1 DAR 5:4], 9000 kb/s, 25 fps, 25 tbr, 90k tbn, 50 tbc Stream #0.1[0x80]: Audio: ac3, 48000 Hz, mono, s16, 448 kb/s Both recorded on same camera, one at HD, one at SD. On 12/05/11 22:58, belcampo wrote: > On 05/12/11 10:25, mroper wrote: >> hi, >> >> >> >> hoping someone can help, i'm fairly new to ffmpeg, trying to >> work out how to convert an HD 1280x720 AVI file to mpeg2 so my >> tv via linux media tomb can play it. >> >> >> >> When i try doing.... >> >> >> >> ffmpeg -i "$filename" -target pal-dvd -b 25000k >> /destpath/$filename >> >> >> >> i get .....Error while opening encoder for output stream #0.0- maybe >> incorrect parameters such >> as bit_rate, rate, width or height >> >> >> i've tried searching on the error and drawn a blank, lots of information >> but nothing I can relate to it. likewise the man pages don't give me a >> clue. reason why I have the -b 25000k is I'm trying to maintain the >> quality of the orginal avi, if I don't set this it works fine. likewise >> if I set a lower rate it also works fine (but at lower quality). The >> orginal avi when doing a ffmpeg -i filename shows up >> >> Stream #0.0: Video: mjpeg, yuvj422p, 1280x720, 30 tbr, 30 tbn, 30 tbc >> Metadata: >> strn : FUJIFILM AVI STREAM 0100 >> Stream #0.1: Audio: pcm_s16le, 44100 Hz, 2 channels, s16, 1411 kb/s >> >> Where as the best I can get is via >> >> ffmpeg -i "$filename" -target pal-dvd -s hd720 /destpath/$filename >> >> and produces >> >> Duration: 00:00:02.00, start: 0.500000, bitrate: 7176 kb/s >> Stream #0.0[0x1e0]: Video: mpeg2video, yuv420p, 1280x720 [PAR 1:1 DAR >> 16:9], 9000 kb/s, 25 fps, 25 tbr, 90k tbn, 50 tbc >> Stream #0.1[0x80]: Audio: ac3, 48000 Hz, stereo, s16, 448 kb/s >> >> Which has a much lower quality/bitrate (you can see the difference on >> screen). > target pal-dvd is limited to 9000 kb/s AFAIK, the dvd-spec doesn't allow more. > You'll have to remove 'target pal-dvd' and specify the parameters your-self like vcodec mpeg2 -b > 25000k and then it should work. You also have to specify -acodec to something your TV 'understands'. >> >> Any ideas? >> >> Thanks >> >> Miles >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From shop at open-t.co.uk Mon May 16 13:32:16 2011 From: shop at open-t.co.uk (Sebastian Arcus) Date: Mon, 16 May 2011 12:32:16 +0100 Subject: [FFmpeg-user] Drawtext filter with date overlay is not working in git-N-29946-g27614b1 In-Reply-To: <4DD04F98.9090305@open-t.co.uk> References: <4DD04F98.9090305@open-t.co.uk> Message-ID: <4DD10B40.8000000@open-t.co.uk> I've just noticed that ffmpeg is using 98%-100% of one of the processor cores. It used to take about 28%-38% on the same machine doing the same work using an ffmpeg build from 8th March 2011. Sebastian On 05/15/2011 11:11 PM, Sebastian Arcus wrote: > Just downloaded latest ffmpeg from git and tried the drawtext filter > with date overlay. No errors on the command line, but the date on the > video is completely static - it picks up the date the moment ffmpeg is > started, and it doesn't change any more. The command I use is: > > ffmpeg -fflags +genpts -t 600 -f mjpeg -r 8 -s 640x480 \ > -i http://localhost:8080/?action=stream -vcodec mpeg4 \ > -vf drawtext="fontfile=/usr/share/fonts/TTF/mitra.ttf:x=70:y=455: \ > text='\%H\:\%M\:\%S | \%a \%d/\%b/\%Y | S500ATV | camera 0': \ > fontcolor=0xFFFFFFFF:fontsize=18: \ > shadowcolor=0x000000EE:shadowx=1:shadowy=1" \ > -b 1500000 -r 8 video_file.avi > > This used to work just fine only few weeks ago. > > The input stream comes from mjpg-streamer, which generates it from a usb > webcam - but I don't think that makes any difference. > > Anybody any suggestions? Any recent changes that broke the drawtext filter? > > Sebastian > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From shop at open-t.co.uk Mon May 16 13:37:00 2011 From: shop at open-t.co.uk (Sebastian Arcus) Date: Mon, 16 May 2011 12:37:00 +0100 Subject: [FFmpeg-user] Piping realtime data to drawtext filter? Message-ID: <4DD10C5C.60901@open-t.co.uk> I would like to overlay on footage recorded from a usb webcam data acquired in realtime from a GPS receiver. I'm mainly interested in current vehicle speed. Drawtext takes static text on the command line - the only bit of realtime data it can display is the date. Is there any way to feed it constantly updating data, such as data coming from my gps receiver through gpsd? Sebastian From funnylookinhat at gmail.com Mon May 16 15:05:16 2011 From: funnylookinhat at gmail.com (David Overcash) Date: Mon, 16 May 2011 07:05:16 -0600 Subject: [FFmpeg-user] understanding copy flag In-Reply-To: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> References: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> Message-ID: Some codecs don't play nicely with copy... Also - why would you essentially run "cp original.avi output.avi" through FFMPEG? Is there something else you're trying to accomplish? If so then we might be able to help a bit more... :) On Sun, May 15, 2011 at 10:20 PM, Rick C. wrote: > Hello, > > I was making a few tests using the following syntax: > > ffmpeg -i original.avi -acodec copy -vcodec copy output.avi > > To my surprise the results were not very good (mainly just the audio side I > believe). Am I wrong in thinking this should work as a passthrough? I > tried the same with .mp4 and .mpg and the results were what I expected. I > didn't notice anything significant from the output log so I didn't post > anything. Any advice would be great. Thanks! > > rc > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From stefano.sabatini-lala at poste.it Mon May 16 16:16:14 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Mon, 16 May 2011 16:16:14 +0200 Subject: [FFmpeg-user] Drawtext filter with date overlay is not working in git-N-29946-g27614b1 In-Reply-To: <4DD10B40.8000000@open-t.co.uk> References: <4DD04F98.9090305@open-t.co.uk> <4DD10B40.8000000@open-t.co.uk> Message-ID: <20110516141614.GA17118@geppetto> On date Monday 2011-05-16 12:32:16 +0100, Sebastian Arcus encoded: > I've just noticed that ffmpeg is using 98%-100% of one of the > processor cores. It used to take about 28%-38% on the same machine > doing the same work using an ffmpeg build from 8th March 2011. > > Sebastian > > > On 05/15/2011 11:11 PM, Sebastian Arcus wrote: > >Just downloaded latest ffmpeg from git and tried the drawtext filter > >with date overlay. No errors on the command line, but the date on the > >video is completely static - it picks up the date the moment ffmpeg is > >started, and it doesn't change any more. The command I use is: > > > >ffmpeg -fflags +genpts -t 600 -f mjpeg -r 8 -s 640x480 \ > >-i http://localhost:8080/?action=stream -vcodec mpeg4 \ > >-vf drawtext="fontfile=/usr/share/fonts/TTF/mitra.ttf:x=70:y=455: \ > >text='\%H\:\%M\:\%S | \%a \%d/\%b/\%Y | S500ATV | camera 0': \ > >fontcolor=0xFFFFFFFF:fontsize=18: \ > >shadowcolor=0x000000EE:shadowx=1:shadowy=1" \ > >-b 1500000 -r 8 video_file.avi > > > >This used to work just fine only few weeks ago. Yes it was borken by a recent commit, I'll fix it soon. > >The input stream comes from mjpg-streamer, which generates it from a usb > >webcam - but I don't think that makes any difference. > >Anybody any suggestions? Wait... > Any recent changes that broke the drawtext filter? -- ffmpeg-user random tip #5 FFmpeg documentation: http://www.ffmpeg.org/documentation.html From stefano.sabatini-lala at poste.it Mon May 16 16:18:58 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Mon, 16 May 2011 16:18:58 +0200 Subject: [FFmpeg-user] Piping realtime data to drawtext filter? In-Reply-To: <4DD10C5C.60901@open-t.co.uk> References: <4DD10C5C.60901@open-t.co.uk> Message-ID: <20110516141858.GB17118@geppetto> On date Monday 2011-05-16 12:37:00 +0100, Sebastian Arcus encoded: > I would like to overlay on footage recorded from a usb webcam data > acquired in realtime from a GPS receiver. I'm mainly interested in > current vehicle speed. Drawtext takes static text on the command > line - the only bit of realtime data it can display is the date. > > Is there any way to feed it constantly updating data, such as data > coming from my gps receiver through gpsd? You may need a separate filter for that, and build a specific text overlay on top of that. Even better, the filter may inject metadata in frame and let the drawtext filter access it (but this involves a lot of changes in the code). -- ffmpeg-user random tip #8 Multimedia related stuff web site promo: http://wiki.multimedia.cx/ From andycivil at gmail.com Mon May 16 16:49:19 2011 From: andycivil at gmail.com (Andy Civil) Date: Mon, 16 May 2011 10:49:19 -0400 Subject: [FFmpeg-user] understanding copy flag In-Reply-To: References: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> Message-ID: <4DD1396F.3060101@gmail.com> On 2011-05-16 9:05 AM, David Overcash wrote: > Some codecs don't play nicely with copy... How can that be, when the codec shouldn't even be involved? > Also - why would you essentially > run "cp original.avi output.avi" through FFMPEG? I would have thought it's pretty obvious that he's trying to debug something, (for example, why his output file is so big) and he's trying to start from an obvious "base", ready to tweak things and see what breaks it - and he's upset because even his obvious base seems broken. Clearly copying is not his objective, but one step in a debugging path. P.S. I'm just guessing, but perhaps the original file has frames missing (e.g. it's declared as 60fps but only 30fps are present) and the 'copy' function is obediently duplicating frames (and inserting them into the output) to tidy up the mess. -- Andy From shop at open-t.co.uk Mon May 16 16:58:07 2011 From: shop at open-t.co.uk (Sebastian Arcus) Date: Mon, 16 May 2011 15:58:07 +0100 Subject: [FFmpeg-user] Drawtext filter with date overlay is not working in git-N-29946-g27614b1 In-Reply-To: <20110516141614.GA17118@geppetto> References: <4DD04F98.9090305@open-t.co.uk> <4DD10B40.8000000@open-t.co.uk> <20110516141614.GA17118@geppetto> Message-ID: <4DD13B7F.20001@open-t.co.uk> Thanks Stefano. I'll wait until you find some time to fix it. Thanks again, Sebastian On 05/16/2011 03:16 PM, Stefano Sabatini wrote: > On date Monday 2011-05-16 12:32:16 +0100, Sebastian Arcus encoded: >> I've just noticed that ffmpeg is using 98%-100% of one of the >> processor cores. It used to take about 28%-38% on the same machine >> doing the same work using an ffmpeg build from 8th March 2011. >> >> Sebastian >> >> >> On 05/15/2011 11:11 PM, Sebastian Arcus wrote: >>> Just downloaded latest ffmpeg from git and tried the drawtext filter >>> with date overlay. No errors on the command line, but the date on the >>> video is completely static - it picks up the date the moment ffmpeg is >>> started, and it doesn't change any more. The command I use is: >>> >>> ffmpeg -fflags +genpts -t 600 -f mjpeg -r 8 -s 640x480 \ >>> -i http://localhost:8080/?action=stream -vcodec mpeg4 \ >>> -vf drawtext="fontfile=/usr/share/fonts/TTF/mitra.ttf:x=70:y=455: \ >>> text='\%H\:\%M\:\%S | \%a \%d/\%b/\%Y | S500ATV | camera 0': \ >>> fontcolor=0xFFFFFFFF:fontsize=18: \ >>> shadowcolor=0x000000EE:shadowx=1:shadowy=1" \ >>> -b 1500000 -r 8 video_file.avi >>> >>> This used to work just fine only few weeks ago. > > Yes it was borken by a recent commit, I'll fix it soon. > >>> The input stream comes from mjpg-streamer, which generates it from a usb >>> webcam - but I don't think that makes any difference. > >>> Anybody any suggestions? > > Wait... > >> Any recent changes that broke the drawtext filter? From funnylookinhat at gmail.com Mon May 16 17:49:24 2011 From: funnylookinhat at gmail.com (David Overcash) Date: Mon, 16 May 2011 09:49:24 -0600 Subject: [FFmpeg-user] understanding copy flag In-Reply-To: <4DD1396F.3060101@gmail.com> References: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> <4DD1396F.3060101@gmail.com> Message-ID: > > P.S. I'm just guessing, but perhaps the original file has frames missing > (e.g. it's declared as 60fps but only 30fps are present) and the 'copy' > function is obediently duplicating frames (and inserting them into the > output) to tidy up the mess. > > Right - but the weird part is that his issues are with Audio - not Video... at least, if I'm reading correctly - and I know of no instances where the compressed audio would not simply be remuxed into the new container.... or am I not understanding how that would work? -David From mike.scheutzow at alcatel-lucent.com Mon May 16 18:09:22 2011 From: mike.scheutzow at alcatel-lucent.com (Mike Scheutzow) Date: Mon, 16 May 2011 12:09:22 -0400 Subject: [FFmpeg-user] understanding copy flag In-Reply-To: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> References: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> Message-ID: <4DD14C32.6020603@alcatel-lucent.com> Rick C. wrote: > Hello, > > I was making a few tests using the following syntax: > > ffmpeg -i original.avi -acodec copy -vcodec copy output.avi > > To my surprise the results were not very good (mainly just the audio side I believe). Am I wrong in thinking this should work as a passthrough? I tried the same with .mp4 and .mpg and the results were what I expected. I didn't notice anything significant from the output log so I didn't post anything. Any advice would be great. Thanks! If you want our help, you need to be more descriptive about the problem you have. "not very good" is too vague. Did you hear clicks? Is lip-sync bad? Also, show us the *complete* output from 'ffmpeg -i original.avi'. Mike Scheutzow From jswordtestem at yahoo.co.uk Tue May 17 00:35:45 2011 From: jswordtestem at yahoo.co.uk (phil curb) Date: Mon, 16 May 2011 23:35:45 +0100 (BST) Subject: [FFmpeg-user] how do I correct a differing frame rate? Message-ID: <910275.52305.qm@web25903.mail.ukl.yahoo.com> When I run ffmpeg -i on this FLV file, it says Seems stream 0 codec frame rate differs from container frame rate: 59.94 (2997/5 0) -> 29.92 (359/12) How do I fix this file so that it doesn't say that? http://www.sendspace.com/file/a4hzyc From h.reindl at thelounge.net Tue May 17 00:49:19 2011 From: h.reindl at thelounge.net (Reindl Harald) Date: Tue, 17 May 2011 00:49:19 +0200 Subject: [FFmpeg-user] Solved: Fwd: vlc-troubles with git-N-29534-g66b1f21 Message-ID: <4DD1A9EF.3030007@thelounge.net> fyi: this builf from today, oldabi works well on fedora 13/14 without any issues ecnoding and/or vlc, the first build since 2010-03-31 and i am happy again :-) ffmpeg version git-N-29954-g33651e3, Copyright (c) 2000-2011 the FFmpeg developers built on May 17 2011 00:31:33 with gcc 4.5.1 20100924 (Red Hat 4.5.1-4) -------- Original-Nachricht -------- Betreff: [FFmpeg-user] vlc-troubles with git-N-29534-g66b1f21 Datum: Sat, 07 May 2011 20:40:56 +0200 Von: Reindl Harald Antwort an: FFmpeg user questions and RTFMs Organisation: the lounge interactive design An: Mailing-List ffmpeg today (after some off-list) explanations (thanks again) i got a in the first moment working ffmpeg with "oldabi" git-N-29534-g66b1f21 encoding works wonderful with several formats but vlc-1.9.1 from rpmfusion (Fedora 14) says there are no decoders for all formats except .vob i can confirm that this ffmpeg-build is the problem because after downgrade to our latest working snapshot from 2011-03-31 the problem went away and also a lot of test-videos encoded with git-N-29534-g66b1f21 playing without any issue i wonder that even a "rpmbuild --rebuild vlc-rpmfusion.src.rpm" did not solve the problem, so is there any known issue? -- Mit besten Gr??en, Reindl Harald the lounge interactive design GmbH A-1060 Vienna, Hofm?hlgasse 17 CTO / software-development / cms-solutions p: +43 (1) 595 3999 33, m: +43 (676) 40 221 40 icq: 154546673, http://www.thelounge.net/ http://www.thelounge.net/signature.asc.what.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 261 bytes Desc: OpenPGP digital signature URL: From hardik.sharma22 at yahoo.com Tue May 17 01:40:30 2011 From: hardik.sharma22 at yahoo.com (Hardik Sharma) Date: Mon, 16 May 2011 16:40:30 -0700 (PDT) Subject: [FFmpeg-user] slice-max-size for h.264 in ffmpeg Message-ID: <337831.5679.qm@web46205.mail.sp1.yahoo.com> Hi, Does ffmpeg support an option like ?slice-max-size? for H.264 ? I know that the x264 encoder supports a parameter called ?slice-max-size?, running ?x264 ?fullhelp? reveals the following: --slice-max-size Limit the size of each slice in bytes Is it the case that while ffmpeg does not support any such option as yet, it will/may support such an option. Is there any way by which I can control NALU size? Thanks. Regards, Hardik Sharma? From rickcorteza at gmail.com Tue May 17 02:36:01 2011 From: rickcorteza at gmail.com (Rick C.) Date: Tue, 17 May 2011 08:36:01 +0800 Subject: [FFmpeg-user] understanding copy flag In-Reply-To: <4DD14C32.6020603@alcatel-lucent.com> References: <6BC1716A-CE07-40F4-A532-5C76A514D799@gmail.com> <4DD14C32.6020603@alcatel-lucent.com> Message-ID: <9553646B-C58A-43B4-9743-E81168445A22@gmail.com> On May 17, 2011, at 12:09 AM, Mike Scheutzow wrote: > Rick C. wrote: >> Hello, >> I was making a few tests using the following syntax: >> ffmpeg -i original.avi -acodec copy -vcodec copy output.avi >> To my surprise the results were not very good (mainly just the audio side I believe). Am I wrong in thinking this should work as a passthrough? I tried the same with .mp4 and .mpg and the results were what I expected. I didn't notice anything significant from the output log so I didn't post anything. Any advice would be great. Thanks! > > > If you want our help, you need to be more descriptive about the problem you have. "not very good" is too vague. Did you hear clicks? Is lip-sync bad? > > Also, show us the *complete* output from 'ffmpeg -i original.avi'. Thanks for all the replies. Ultimately I would like to use it in a more practical way such as if original.avi had mp3 sound if I decided to convert it to output.mp4 I could use -acodec copy for example. In my original tests I just went from same container to same container since I thought it would just copy over (and it does), but the audio issues keep popping up in .avi container for me. And to describe the issue there is mostly no sound but it cuts in from time to time in short 1 second bursts. I couldn't say if it's actually in sync? Here is the full output of a test I just did on a video: Last login: Mon May 16 12:25:54 on ttys000 mahalkos-MacBook:~ mahalko$ /Users/mahalko/Desktop/ffmpeg -i /Users/mahalko/Desktop/original.avi -acodec copy -vcodec copy /Users/mahalko/Desktop/output.avi FFmpeg version git-N-28698-g621f4c9, Copyright (c) 2000-2011 the FFmpeg developers built on Mar 30 2011 13:29:37 with gcc 4.2.1 (Apple Inc. build 5666) (dot 3) configuration: --prefix=/Volumes/Ramdisk/sw --enable-gpl --enable-pthreads --enable-version3 --enable-libspeex --enable-libvpx --disable-decoder=libvpx --enable-libmp3lame --enable-libtheora --enable-libvorbis --enable-libx264 --enable-avfilter --enable-libopencore_amrwb --enable-libopencore_amrnb --enable-filters --arch=x86_64 --enable-runtime-cpudetect libavutil 50. 40. 0 / 50. 40. 0 libavcodec 52.116. 0 / 52.116. 0 libavformat 52.104. 0 / 52.104. 0 libavdevice 52. 4. 0 / 52. 4. 0 libavfilter 1. 76. 0 / 1. 76. 0 libswscale 0. 13. 0 / 0. 13. 0 Seems stream 0 codec frame rate differs from container frame rate: 30000.00 (30000/1) -> 29.97 (30000/1001) Input #0, avi, from '/Users/mahalko/Desktop/original.avi': Metadata: artist : the Black Eyed Peas date : 2011 title : Just Can't Get Enough encoder : Lavf52.103.0 Duration: 00:03:55.94, start: 0.000000, bitrate: 2147 kb/s Stream #0.0: Video: mpeg4, yuv420p, 720x406 [PAR 1:1 DAR 360:203], 29.97 fps, 29.97 tbr, 29.97 tbn, 30k tbc Stream #0.1: Audio: mp3, 48000 Hz, stereo, s16, 192 kb/s Output #0, avi, to '/Users/mahalko/Desktop/output.avi': Metadata: IART : the Black Eyed Peas ICRD : 2011 INAM : Just Can't Get Enough ISFT : Lavf52.104.0 Stream #0.0: Video: mpeg4, yuv420p, 720x406 [PAR 1:1 DAR 360:203], q=2-31, 29.97 tbn, 29.97 tbc Stream #0.1: Audio: libmp3lame, 48000 Hz, stereo, 192 kb/s Stream mapping: Stream #0.0 -> #0.0 Stream #0.1 -> #0.1 Press ctrl-c to stop encoding frame= 2606 fps=2284 q=-1.0 size= 29154kB time=108.65 bitrate=2198.2kbits/s frame= 3655 fps=2164 q=-1.0 size= 40965kB time=152.42 bitrate=2201.7kbits/s frame= 4794 fps=2107 q=-1.0 size= 53747kB time=199.92 bitrate=2202.4kbits/s frame= 5657 fps=2150 q=-1.0 Lsize= 61854kB time=235.90 bitrate=2147.9kbits/s video:55915kB audio:5530kB global headers:0kB muxing overhead 0.665017% mahalkos-MacBook:~ mahalko$ rc From bahamutzero8825 at gmail.com Tue May 17 03:13:28 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Mon, 16 May 2011 20:13:28 -0500 Subject: [FFmpeg-user] how do I correct a differing frame rate? In-Reply-To: <910275.52305.qm@web25903.mail.ukl.yahoo.com> References: <910275.52305.qm@web25903.mail.ukl.yahoo.com> Message-ID: <4DD1CBB8.1070702@gmail.com> On 2011.05.16 05:35 PM, phil curb wrote: > When I run ffmpeg -i on this FLV file, it says > > Seems stream 0 codec frame rate differs from container frame rate: 59.94 (2997/5 > 0) -> 29.92 (359/12) > > How do I fix this file so that it doesn't say that? Remux? I've found those messages to be quite harmless, but then again, I've never really worked with FLV. Are you actually having any problems with the file? From grant.smith at envent-tech.com Tue May 17 04:09:36 2011 From: grant.smith at envent-tech.com (Grant Smith) Date: Mon, 16 May 2011 22:09:36 -0400 Subject: [FFmpeg-user] Installation Problems In-Reply-To: References: <20110501130356.3714c1d2.dashing.meng@gmail.com> Message-ID: I tried your tip and I still encountered the same error: ERROR: libmp3lame >= 3.98.3 not found If you think configure made a mistake, make sure you are using the latest version from Git. If the latest version fails, report the problem to the ffmpeg-user at ffmpeg.org mailing list or IRC #ffmpeg on irc.freenode.net. Include the log file "config.log" produced by configure as this will help solving the problem. Does anyone else have any ideas? Thanks, Grant Smith A+, Network+, MCP x 2, BSIT/VC, MIS Phone: +1.317.560.4457 Fax: +1.208.567.9281 On Mon, May 2, 2011 at 8:44 AM, Grant Smith wrote: > Interesting.... I don't think my LAME build is failing but I will give this > a try and let you know. > > Thanks, > > > Grant Smith > A+, Network+, MCP x 2, BSIT/VC, MIS > > Phone: +1.317.560.4457 > > > On Sun, May 1, 2011 at 1:03 AM, littlebat wrote: > >> >> On Wed, 27 Apr 2011 22:17:16 -0400 >> Grant Smith wrote: >> >> > I'm trying to install ffmpeg on Ubuntu 10.04 using the tutorial found >> > @ http://ubuntuforums.org/showpost.php?p=9868359&postcount=1289 >> >> I tested tutorial for installing a static linked ffmpeg with the >> style of Ubuntu's deb package. >> >> I found I must execute "sudo mkdir -p /usr/local/share/doc/lame" before >> "sudo checkinstall ..." when install LAME, or it will fail. >> _______________________________________________ >> ffmpeg-user mailing list >> ffmpeg-user at ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-user >> > > From dashing.meng at gmail.com Tue May 17 04:34:27 2011 From: dashing.meng at gmail.com (littlebat) Date: Tue, 17 May 2011 10:34:27 +0800 Subject: [FFmpeg-user] how to maximize the volume without distorting the sound using ffmpeg? Message-ID: <20110517103427.e0b2cacc.dashing.meng@gmail.com> Hi, Is there a way to maximize the volume without distorting the sound using Ffmpeg? just like the "-af volnorm" filter in Mencoder? littlebat From jswordtestem at yahoo.co.uk Tue May 17 04:50:59 2011 From: jswordtestem at yahoo.co.uk (phil curb) Date: Tue, 17 May 2011 03:50:59 +0100 (BST) Subject: [FFmpeg-user] how do I correct a differing frame rate? Message-ID: <433025.95987.qm@web25906.mail.ukl.yahoo.com> no but i'd still like to be able to remux the file so ffmpeg -i doesn't give that error about the frame rate. I just don't know the line to do it.. ---------------------------------------- On 2011.05.16 05:35 PM, phil curb wrote: > When I run ffmpeg -i on this FLV file, it says > > Seems stream 0 codec frame rate differs from container frame rate: 59.94 (2997/5 > 0) -> 29.92 (359/12) > > How do I fix this file so that it doesn't say that? >http://www.sendspace.com/file/a4hzyc Remux? I've found those messages to be quite harmless, but then again, I've never really worked with FLV. Are you actually having any problems with the file? From baptiste.coudurier at gmail.com Tue May 17 04:54:12 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Mon, 16 May 2011 19:54:12 -0700 Subject: [FFmpeg-user] slice-max-size for h.264 in ffmpeg In-Reply-To: <337831.5679.qm@web46205.mail.sp1.yahoo.com> References: <337831.5679.qm@web46205.mail.sp1.yahoo.com> Message-ID: <4DD1E354.20705@gmail.com> Hi, On 5/16/11 4:40 PM, Hardik Sharma wrote: > Hi, > > Does ffmpeg support an option like ?slice-max-size? for H.264 ? > > I know that the x264 encoder supports a parameter called > ?slice-max-size?, running ?x264 ?fullhelp? reveals the following: > --slice-max-size Limit the size of each slice in bytes > > Is it the case that while ffmpeg does not support any such option as > yet, it will/may support such an option. > > Is there any way by which I can control NALU size? Thanks. It doesn't currently, but it should be easy to add. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From baptiste.coudurier at gmail.com Tue May 17 04:57:46 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Mon, 16 May 2011 19:57:46 -0700 Subject: [FFmpeg-user] 420p out from ffmpeg becomes 420i in to x264 In-Reply-To: References: Message-ID: <4DD1E42A.5030807@gmail.com> Hi, On 5/15/11 3:29 PM, sean darcy wrote: > I'm trying to use ffmpeg to pipe a stream to x264. Both ffmpeg and x264 > from git today. > > I'm deinterlacing the input with yadif, piped output is yuv4mpeg. And > ffmpeg thinks is generating yuv420p. > > But x264 is seeing the piped stream as interlaced. I thought yuv4mpeg > had a header describing the stream. Is that so? If so, is the problem > here that ffmpeg is not writing the header correctly, x264 is not > reading it correctly, or have I screwed up something? > > sean > > + ffmpeg -i JulyPlay1990.avi -an -vf yadif=1:0,hqdn3d -r 60000/1001 > -pix_fmt yuv420p -f yuv4mpegpipe - > + x264 --demuxer y4m --preset slower --tune film --bitrate 1200 -o > out3.m4v - I suggest you use ffmpeg directly instead of piping. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From dashing.meng at gmail.com Tue May 17 05:05:30 2011 From: dashing.meng at gmail.com (littlebat) Date: Tue, 17 May 2011 11:05:30 +0800 Subject: [FFmpeg-user] What does "-vol" parameter mean in ffmpeg? Message-ID: <20110517110530.c7c0023d.dashing.meng@gmail.com> Hi, I can't understand how to use "-vol" parameter to increase volume in ffmpeg. I know "-vol 256" mean keep volume unchanged. But I don't know how to choose a proper number for a video file to increase volume. For example, does "-vol 512" mean volume intensity is twice as original file? ( volume intensity: http://en.wikipedia.org/wiki/Decibel#Power_quantities ) thanks. littlebat From hardik.sharma22 at yahoo.com Tue May 17 06:03:11 2011 From: hardik.sharma22 at yahoo.com (Hardik Sharma) Date: Mon, 16 May 2011 21:03:11 -0700 (PDT) Subject: [FFmpeg-user] slice-max-size for h.264 in ffmpeg Message-ID: <730529.61847.qm@web46215.mail.sp1.yahoo.com> Can you help me. Please let me know how I can add the NAL size control as I am new to ffmpeg and trying to do some experiments with it. Thanks Regards, Hardik Sharma On 5/16/11 4:40 PM, Hardik Sharma wrote: >Hi, >>Does ffmpeg support an option like ?slice-max-size? for H.264 ? >>I know that the x264 encoder supports a parameter called >?slice-max-size?, running ?x264 ?fullhelp? reveals the following: >--slice-max-size Limit the size of each slice in bytes >>Is it the case that while ffmpeg does not support any such option as >yet, it will/may support such an option. >>Is there any way by which I can control NALU size? Thanks. It doesn't currently, but it should be easy to add. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From adi235 at gmail.com Tue May 17 09:14:39 2011 From: adi235 at gmail.com (Aditya) Date: Tue, 17 May 2011 07:14:39 +0000 (UTC) Subject: [FFmpeg-user] Need Help in Stripping Down ffmpeg.c transcode Message-ID: for a simple video format conversion. 1 If i input a video file. the parse_options function takes the arguments translates and assigns the video encode values 2. the parse_option function calls and initializes the video output - do_video_out 3. the parse_options function calls and initializes the audio output -do_audio_out 4. the transcoding function is called which takes in the arguments got by the parsed_options function. 5. the trans coding function does a whole lot of processing with error checks ---- some of which might not be needed for basic transcode. 6. output_packet is called at /* at the end of stream, we must flush the decoder buffers */ 7. from the out_packet the data is written to the file. The the areas where i see the reduction of code is the input options. if i preset them and initialize the output/codec/streams the transcoding if limited to just video transcoding - not considering the subtitles or error checks.. How do i simplify the ffmpeg.c can any one help me do this? From oleber at gmail.com Tue May 17 09:20:55 2011 From: oleber at gmail.com (marcos rebelo) Date: Tue, 17 May 2011 09:20:55 +0200 Subject: [FFmpeg-user] Convert some hundreds files to a H264 Message-ID: Hi all I was giving some hundreds of files (more will come) with different formats (div, avi, mov, ...) and different sizes to be converted to H264 with: a maximum width 720, -b 500k -bt 500k -acodec libfaac Is there any simple configuration, configurable script or something like that, that is able of processing this different files without the human interaction. I'm a Perl developer, I may program something if is simple. CPU and time isn't a problem, so there is no problem if I need (I would prefer) to pass 2 times. Thanks for any help Best Regards Marcos Rebelo -- Marcos Rebelo http://www.oleber.com/ Milan Perl Mongers leader https://sites.google.com/site/milanperlmongers/ Webmaster of http://perl5notebook.oleber.com From betonpfeiler at googlemail.com Tue May 17 10:42:51 2011 From: betonpfeiler at googlemail.com (betonpfeiler) Date: Tue, 17 May 2011 10:42:51 +0200 Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: References: Message-ID: <4DD2350B.9030607@googlemail.com> Hello, Am 17.05.2011 09:20, schrieb marcos rebelo: > Hi all > > > I was giving some hundreds of files (more will come) with different > formats (div, avi, mov, ...) and different sizes to be converted to > H264 with: > a maximum width 720, > -b 500k > -bt 500k > -acodec libfaac > > > Is there any simple configuration, configurable script or something > like that, that is able of processing this different files without the > human interaction. I'm a Perl developer, I may program something if is > simple. > > CPU and time isn't a problem, so there is no problem if I need (I > would prefer) to pass 2 times. > > > Thanks for any help > > Best Regards > Marcos Rebelo > > you could use a simple batchscript to create a watchfolder. For linux you might use something like this: for f in *.avi; do ffmpeg -i "$f"/[insert your commands here]/"${f%.avi}.mp4"; done [You should be able to use *.* to execute all files in the folder, but in this case remeber to save the output to another folder... :)] Save this text to a scriptfile (with a simple texteditor) name.sh and make it executable sudo chmod +x name.sh Then just start the script in the terminal. On a Windows-Setup you'll have to use a batch-script. I hope this helps. From oleber at gmail.com Tue May 17 11:16:14 2011 From: oleber at gmail.com (marcos rebelo) Date: Tue, 17 May 2011 11:16:14 +0200 Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: <4DD2350B.9030607@googlemail.com> References: <4DD2350B.9030607@googlemail.com> Message-ID: My problem is to do the [insert your commands here], In my opinion, it should be easy to give some commands defining the final format of the video and ignore the original format, a smart script should analyse the video and it should create the [insert your commands here]. Note that my files are correctly visible on mplayer. Thanks for all help Best Regards Marcos Rebelo -- Marcos Rebelo http://www.oleber.com/ Milan Perl Mongers leader https://sites.google.com/site/milanperlmongers/ Webmaster of http://perl5notebook.oleber.com From belcampo at zonnet.nl Tue May 17 12:18:14 2011 From: belcampo at zonnet.nl (belcampo) Date: Tue, 17 May 2011 12:18:14 +0200 Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: References: <4DD2350B.9030607@googlemail.com> Message-ID: <4DD24B66.3010800@zonnet.nl> On 05/17/11 11:16, marcos rebelo wrote: > My problem is to do the [insert your commands here], > > In my opinion, it should be easy to give some commands defining the > final format of the video and ignore the original format, a smart > script should analyse the video and it should create the [insert your > commands here]. > > Note that my files are correctly visible on mplayer. If mplayer detects/analyses your files correct, so will ffmpeg, compiled with the same codecs, do the same. Nothing special to do AFAIK. > > Thanks for all help > > > Best Regards > Marcos Rebelo > > From smouli60 at gmail.com Tue May 17 13:05:21 2011 From: smouli60 at gmail.com (mouli s) Date: Tue, 17 May 2011 16:35:21 +0530 Subject: [FFmpeg-user] convesion to mkv fails Message-ID: Hello, Wat is the best way for the conversion of any files to mkv ??? i tried all the following... wat happened is the converted file cannot be opened.... "C:\Program Files\BullsHit Converter\Lib\FFmpeg\ffmpeg.exe" -i "C:\Documents and Settings\Desktop\Samples\video\Wild.wmv" -y "C:\Documents and Settings\broov\Desktop\ConversionSamples\video\Wild.mkv" also tried using some values... "C:\Program Files\BullsHit Converter\Lib\FFmpeg\ffmpeg.exe" -i "C:\Documents and Settings\Desktop\Samples\video.mp4" -y -vcodec libx264 -b 1250kb -acodec libmp3lame -ar 44100 -ab 160kb -ac 2 "C:\Documents and Settings\broov\Desktop\Samples\video.mkv" the converted video file cannot be opened.... wats the prob in this convertion ????? thanks for any replies... From smouli60 at gmail.com Tue May 17 13:05:21 2011 From: smouli60 at gmail.com (mouli s) Date: Tue, 17 May 2011 16:35:21 +0530 Subject: [FFmpeg-user] convesion to mkv fails Message-ID: Hello, Wat is the best way for the conversion of any files to mkv ??? i tried all the following... wat happened is the converted file cannot be opened.... "C:\Program Files\BullsHit Converter\Lib\FFmpeg\ffmpeg.exe" -i "C:\Documents and Settings\Desktop\Samples\video\Wild.wmv" -y "C:\Documents and Settings\broov\Desktop\ConversionSamples\video\Wild.mkv" also tried using some values... "C:\Program Files\BullsHit Converter\Lib\FFmpeg\ffmpeg.exe" -i "C:\Documents and Settings\Desktop\Samples\video.mp4" -y -vcodec libx264 -b 1250kb -acodec libmp3lame -ar 44100 -ab 160kb -ac 2 "C:\Documents and Settings\broov\Desktop\Samples\video.mkv" the converted video file cannot be opened.... wats the prob in this convertion ????? thanks for any replies... From mblainml at earthlink.net Tue May 17 18:24:08 2011 From: mblainml at earthlink.net (Mailing List) Date: Tue, 17 May 2011 12:24:08 -0400 Subject: [FFmpeg-user] Convert some hundreds files to a H264 Message-ID: marcos rebelo wrote in news:BANLkTi=43-RGdG9h4LPx+mi2AcxdZyv-zQ at mail.gmail.com: > I was giving some hundreds of files (more will come) with > different formats (div, avi, mov, ...) and different sizes to be > converted to H264 with: > a maximum width 720, > -b 500k > -bt 500k > -acodec libfaac As others have suggested, using a simple loop to process the files is not particularly difficult. The only issue is your "maximum width 720". Using -vf "scale=720:-1" is unreliable because it throws errors for odd-numbered heights, instead of rounding to a usable value as mencoder can. You could parse the output of either "ffprobe" or "ffmpeg -i" for the display aspect ratio and use it to calculate scaling dimensions, but it makes the script considerably more complex. My own crude solution is in Windows batch, which wouldn't help you with Perl. From belcampo at zonnet.nl Tue May 17 22:22:27 2011 From: belcampo at zonnet.nl (belcampo) Date: Tue, 17 May 2011 22:22:27 +0200 Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: References: Message-ID: <4DD2D903.2020407@zonnet.nl> On 05/17/11 18:24, Mailing List wrote: > marcos rebelo wrote in > news:BANLkTi=43-RGdG9h4LPx+mi2AcxdZyv-zQ at mail.gmail.com: > >> I was giving some hundreds of files (more will come) with >> different formats (div, avi, mov, ...) and different sizes to be >> converted to H264 with: >> a maximum width 720, >> -b 500k >> -bt 500k >> -acodec libfaac > > As others have suggested, using a simple loop to process the files is not > particularly difficult. The only issue is your "maximum width 720". Using > -vf "scale=720:-1" is unreliable because it throws errors for odd-numbered > heights, instead of rounding to a usable value as mencoder can. It throws error-messages for not-optimal, in the sense of not-optimal in file-size, encodings but it will encode nevertheless without trouble AFAIK. > > You could parse the output of either "ffprobe" or "ffmpeg -i" for the > display aspect ratio and use it to calculate scaling dimensions, but it > makes the script considerably more complex. My own crude solution is in > Windows batch, which wouldn't help you with Perl. > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From mblainml at earthlink.net Tue May 17 23:26:28 2011 From: mblainml at earthlink.net (Mark Blain) Date: Tue, 17 May 2011 17:26:28 -0400 Subject: [FFmpeg-user] Convert some hundreds files to a H264 References: <4DD2D903.2020407@zonnet.nl> Message-ID: belcampo wrote: >It throws error-messages for not-optimal, in the sense of not-optimal in >file-size, encodings but it will encode nevertheless without trouble AFAIK. When scaling with "-vf scale=640:-1" you can get a message like: [libx264 @ 0x33e68e0] width or height not divisible by 2 (640x361) if the calculated scaled height happens to be an odd number. From belcampo at zonnet.nl Tue May 17 23:33:41 2011 From: belcampo at zonnet.nl (belcampo) Date: Tue, 17 May 2011 23:33:41 +0200 Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: References: <4DD2D903.2020407@zonnet.nl> Message-ID: <4DD2E9B5.1030208@zonnet.nl> On 05/17/11 23:26, Mark Blain wrote: > belcampo wrote: > >> It throws error-messages for not-optimal, in the sense of not-optimal in >> file-size, encodings but it will encode nevertheless without trouble AFAIK. > > When scaling with "-vf scale=640:-1" you can get a message like: > [libx264 @ 0x33e68e0] width or height not divisible by 2 (640x361) > if the calculated scaled height happens to be an odd number. Yes I know, that's what I tried to say, you get this message, but that doesn't stop encoding as expected. > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From mblainml at earthlink.net Wed May 18 00:25:03 2011 From: mblainml at earthlink.net (Mark Blain) Date: Tue, 17 May 2011 18:25:03 -0400 Subject: [FFmpeg-user] Convert some hundreds files to a H264 References: <4DD2D903.2020407@zonnet.nl> <4DD2E9B5.1030208@zonnet.nl> Message-ID: belcampo wrote: > >On 05/17/11 23:26, Mark Blain wrote: >> belcampo wrote: >> >>> It throws error-messages for not-optimal, in the sense of not-optimal in >>> file-size, encodings but it will encode nevertheless without trouble AFAIK. >> >> When scaling with "-vf scale=640:-1" you can get a message like: >> [libx264 @ 0x33e68e0] width or height not divisible by 2 (640x361) >> if the calculated scaled height happens to be an odd number. >Yes I know, that's what I tried to say, you get this message, but that >doesn't stop encoding as expected. Odd, it does for me. Full output below. ffmpeg -i test.avi -vf "scale=640:-1" -vcodec libx264 -preset slow -crf 26 -an -f mp4 test.mp4 ffmpeg version git-N-29970-ge280a4d, Copyright (c) 2000-2011 the FFmpeg developers built on May 17 2011 02:28:56 with gcc 4.5.0 20100414 (Fedora MinGW 4.5.0-1.fc14) configuration: --prefix=/var/www/users/research/ffmpeg/snapshots/build --arch=x86 --target-os=mingw32 --cross-prefix=i686-pc-mingw 32- --cc='ccache i686-pc-mingw32-gcc' --enable-w32threads --enable-memalign-hack --enable-runtime-cpudetect --enable-cross-compile - -enable-static --disable-shared --extra-libs='-lws2_32 -lwinmm' --extra-cflags='--static -I/var/www/users/research/ffmpeg/snapshots/ build/include' --extra-ldflags='-static -L/var/www/users/research/ffmpeg/snapshots/build/lib' --enable-bzlib --enable-zlib --enable- gpl --enable-version3 --enable-nonfree --enable-libx264 --enable-libspeex --enable-libtheora --enable-libvorbis --enable-libfaac --e nable-libxvid --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libmp3lame --enable-libvpx --disable-decoder=libvpx libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 6. 0 / 53. 6. 0 libavformat 53. 1. 0 / 53. 1. 0 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 5. 0 / 2. 5. 0 libswscale 0. 14. 0 / 0. 14. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, avi, from 'test.avi': Duration: 00:06:00.12, start: 0.000000, bitrate: 938 kb/s Stream #0.0: Video: mpeg4, yuv420p, 624x352 [PAR 1:1 DAR 39:22], 29.97 tbr, 29.97 tbn, 29.97 tbc Stream #0.1: Audio: mp3, 48000 Hz, stereo, s16, 128 kb/s [buffer @ 0x190aea0] w:624 h:352 pixfmt:yuv420p tb:1/1000000 sar:1/1 [scale @ 0x39e73e0] w:624 h:352 fmt:yuv420p -> w:640 h:361 fmt:yuv420p flags:0xe2000004 [libx264 @ 0x1909e20] width or height not divisible by 2 (640x361) Output #0, mp4, to 'test.mp4': Stream #0.0: Video: libx264, yuv420p, 640x361 [PAR 1:1 DAR 640:361], q=2-31, 200 kb/s, 90k tbn, 29.97 tbc Stream mapping: Stream #0.0 -> #0.0 Error while opening encoder for output stream #0.0 - maybe incorrect parameters such as bit_rate, rate, width or height From mblainml at earthlink.net Wed May 18 00:25:45 2011 From: mblainml at earthlink.net (Mark Blain) Date: Tue, 17 May 2011 18:25:45 -0400 Subject: [FFmpeg-user] Convert some hundreds files to a H264 References: <4DD2D903.2020407@zonnet.nl> <4DD2E9B5.1030208@zonnet.nl> Message-ID: belcampo wrote: > >On 05/17/11 23:26, Mark Blain wrote: >> belcampo wrote: >> >>> It throws error-messages for not-optimal, in the sense of not-optimal in >>> file-size, encodings but it will encode nevertheless without trouble AFAIK. >> >> When scaling with "-vf scale=640:-1" you can get a message like: >> [libx264 @ 0x33e68e0] width or height not divisible by 2 (640x361) >> if the calculated scaled height happens to be an odd number. >Yes I know, that's what I tried to say, you get this message, but that >doesn't stop encoding as expected. Odd, it does for me. Full output below. ffmpeg -i test.avi -vf "scale=640:-1" -vcodec libx264 -preset slow -crf 26 -an -f mp4 test.mp4 ffmpeg version git-N-29970-ge280a4d, Copyright (c) 2000-2011 the FFmpeg developers built on May 17 2011 02:28:56 with gcc 4.5.0 20100414 (Fedora MinGW 4.5.0-1.fc14) configuration: --prefix=/var/www/users/research/ffmpeg/snapshots/build --arch=x86 --target-os=mingw32 --cross-prefix=i686-pc-mingw 32- --cc='ccache i686-pc-mingw32-gcc' --enable-w32threads --enable-memalign-hack --enable-runtime-cpudetect --enable-cross-compile - -enable-static --disable-shared --extra-libs='-lws2_32 -lwinmm' --extra-cflags='--static -I/var/www/users/research/ffmpeg/snapshots/ build/include' --extra-ldflags='-static -L/var/www/users/research/ffmpeg/snapshots/build/lib' --enable-bzlib --enable-zlib --enable- gpl --enable-version3 --enable-nonfree --enable-libx264 --enable-libspeex --enable-libtheora --enable-libvorbis --enable-libfaac --e nable-libxvid --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libmp3lame --enable-libvpx --disable-decoder=libvpx libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 6. 0 / 53. 6. 0 libavformat 53. 1. 0 / 53. 1. 0 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 5. 0 / 2. 5. 0 libswscale 0. 14. 0 / 0. 14. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, avi, from 'test.avi': Duration: 00:06:00.12, start: 0.000000, bitrate: 938 kb/s Stream #0.0: Video: mpeg4, yuv420p, 624x352 [PAR 1:1 DAR 39:22], 29.97 tbr, 29.97 tbn, 29.97 tbc Stream #0.1: Audio: mp3, 48000 Hz, stereo, s16, 128 kb/s [buffer @ 0x190aea0] w:624 h:352 pixfmt:yuv420p tb:1/1000000 sar:1/1 [scale @ 0x39e73e0] w:624 h:352 fmt:yuv420p -> w:640 h:361 fmt:yuv420p flags:0xe2000004 [libx264 @ 0x1909e20] width or height not divisible by 2 (640x361) Output #0, mp4, to 'test.mp4': Stream #0.0: Video: libx264, yuv420p, 640x361 [PAR 1:1 DAR 640:361], q=2-31, 200 kb/s, 90k tbn, 29.97 tbc Stream mapping: Stream #0.0 -> #0.0 Error while opening encoder for output stream #0.0 - maybe incorrect parameters such as bit_rate, rate, width or height From sheen.andy at googlemail.com Wed May 18 00:31:56 2011 From: sheen.andy at googlemail.com (Andy Sheen) Date: Tue, 17 May 2011 23:31:56 +0100 Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: References: <4DD2D903.2020407@zonnet.nl> <4DD2E9B5.1030208@zonnet.nl> Message-ID: <4DD2F75C.9080604@googlemail.com> Mark Blain wrote on Tue 17 May at 23:25 UK time > ffmpeg -i test.avi -vf "scale=640:-1" -vcodec libx264 -preset slow -crf 26 > -an -f mp4 test.mp4 > > ffmpeg version git-N-29970-ge280a4d, Copyright (c) 2000-2011 the FFmpeg > developers > built on May 17 2011 02:28:56 with gcc 4.5.0 20100414 (Fedora MinGW > 4.5.0-1.fc14) > configuration: --prefix=/var/www/users/research/ffmpeg/snapshots/build > --arch=x86 --target-os=mingw32 --cross-prefix=i686-pc-mingw > 32- --cc='ccache i686-pc-mingw32-gcc' --enable-w32threads > --enable-memalign-hack --enable-runtime-cpudetect --enable-cross-compile - > -enable-static --disable-shared --extra-libs='-lws2_32 -lwinmm' > --extra-cflags='--static -I/var/www/users/research/ffmpeg/snapshots/ > build/include' --extra-ldflags='-static > -L/var/www/users/research/ffmpeg/snapshots/build/lib' --enable-bzlib > --enable-zlib --enable- > gpl --enable-version3 --enable-nonfree --enable-libx264 --enable-libspeex > --enable-libtheora --enable-libvorbis --enable-libfaac --e > nable-libxvid --enable-libopencore-amrnb --enable-libopencore-amrwb > --enable-libmp3lame --enable-libvpx --disable-decoder=libvpx > libavutil 51. 2. 1 / 51. 2. 1 > libavcodec 53. 6. 0 / 53. 6. 0 > libavformat 53. 1. 0 / 53. 1. 0 > libavdevice 53. 0. 0 / 53. 0. 0 > libavfilter 2. 5. 0 / 2. 5. 0 > libswscale 0. 14. 0 / 0. 14. 0 > libpostproc 51. 2. 0 / 51. 2. 0 > Input #0, avi, from 'test.avi': > Duration: 00:06:00.12, start: 0.000000, bitrate: 938 kb/s > Stream #0.0: Video: mpeg4, yuv420p, 624x352 [PAR 1:1 DAR 39:22], 29.97 > tbr, 29.97 tbn, 29.97 tbc > Stream #0.1: Audio: mp3, 48000 Hz, stereo, s16, 128 kb/s > [buffer @ 0x190aea0] w:624 h:352 pixfmt:yuv420p tb:1/1000000 sar:1/1 > [scale @ 0x39e73e0] w:624 h:352 fmt:yuv420p -> w:640 h:361 fmt:yuv420p > flags:0xe2000004 > [libx264 @ 0x1909e20] width or height not divisible by 2 (640x361) > Output #0, mp4, to 'test.mp4': > Stream #0.0: Video: libx264, yuv420p, 640x361 [PAR 1:1 DAR 640:361], > q=2-31, 200 kb/s, 90k tbn, 29.97 tbc > Stream mapping: > Stream #0.0 -> #0.0 > Error while opening encoder for output stream #0.0 - maybe incorrect > parameters such as bit_rate, rate, width or height I have to say it doesn't look like a scaling problem but the 90k tbn looks suspicious to me. No idea why it thinks that, but the value should probably be 29.97... From mblainml at earthlink.net Wed May 18 01:08:24 2011 From: mblainml at earthlink.net (Mark Blain) Date: Tue, 17 May 2011 19:08:24 -0400 Subject: [FFmpeg-user] Convert some hundreds files to a H264 References: <4DD2D903.2020407@zonnet.nl> <4DD2E9B5.1030208@zonnet.nl> <4DD2F75C.9080604@googlemail.com> Message-ID: Andy Sheen wrote: >I have to say it doesn't look like a scaling problem but the 90k tbn >looks suspicious to me. No idea why it thinks that, but the value should >probably be 29.97... Good observation. I think the output tbn shown is just a side-effect. If I use -vf "scale=640:360" or just remove the scale filter, the output shows 30k tbn and encoding begins. See below. With either -vf "scale=640:-1" or -vf "scale=640:361" (just for testing, of course) it shows 90k tbn and the program exits. ffmpeg -i test.avi -vcodec libx264 -preset slow -crf 26 -an -f mp4 test.mp4 ffmpeg version git-N-29970-ge280a4d, Copyright (c) 2000-2011 the FFmpeg developers built on May 17 2011 02:28:56 with gcc 4.5.0 20100414 (Fedora MinGW 4.5.0-1.fc14) configuration: --prefix=/var/www/users/research/ffmpeg/snapshots/build --arch=x86 --target-os=mingw32 --cross-prefix=i686-pc-mingw 32- --cc='ccache i686-pc-mingw32-gcc' --enable-w32threads --enable-memalign-hack --enable-runtime-cpudetect --enable-cross-compile - -enable-static --disable-shared --extra-libs='-lws2_32 -lwinmm' --extra-cflags='--static -I/var/www/users/research/ffmpeg/snapshots/ build/include' --extra-ldflags='-static -L/var/www/users/research/ffmpeg/snapshots/build/lib' --enable-bzlib --enable-zlib --enable- gpl --enable-version3 --enable-nonfree --enable-libx264 --enable-libspeex --enable-libtheora --enable-libvorbis --enable-libfaac --e nable-libxvid --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libmp3lame --enable-libvpx --disable-decoder=libvpx libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 6. 0 / 53. 6. 0 libavformat 53. 1. 0 / 53. 1. 0 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 5. 0 / 2. 5. 0 libswscale 0. 14. 0 / 0. 14. 0 libpostproc 51. 2. 0 / 51. 2. 0 Input #0, avi, from 'test.avi': Duration: 00:06:00.12, start: 0.000000, bitrate: 938 kb/s Stream #0.0: Video: mpeg4, yuv420p, 624x352 [PAR 1:1 DAR 39:22], 29.97 tbr, 29.97 tbn, 29.97 tbc Stream #0.1: Audio: mp3, 48000 Hz, stereo, s16, 128 kb/s [buffer @ 0x31386c0] w:624 h:352 pixfmt:yuv420p tb:1/1000000 sar:1/1 [libx264 @ 0x3137960] using SAR=1/1 [libx264 @ 0x3137960] using cpu capabilities: MMX2 SSE2Fast FastShuffle SSEMisalign LZCNT [libx264 @ 0x3137960] profile High, level 3.0 [libx264 @ 0x3137960] 264 - core 115 r1995 c1e60b9 - H.264/MPEG-4 AVC codec - Copyleft 2003-2011 - http://www.videolan.org/x264.html - options: cabac=1 ref=5 deblock=1:0:0 analyse=0x3:0x113 me=umh subme=8 psy=1 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=1 sliced_threads=0 nr=0 decimate=1 interlaced=0 blu ray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=2 b_bias=0 direct=3 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_ min=25 scenecut=40 intra_refresh=0 rc_lookahead=50 rc=crf mbtree=1 crf=26.0 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1: 1.00 Output #0, mp4, to 'test.mp4': Metadata: encoder : Lavf53.1.0 Stream #0.0: Video: libx264, yuv420p, 624x352 [PAR 1:1 DAR 39:22], q=2-31, 200 kb/s, 30k tbn, 29.97 tbc Stream mapping: Stream #0.0 -> #0.0 Press [q] to stop, [?] for help frame= 308 fps= 24 q=32.0 size= 182kB time=8.51 bitrate= 175.3kbits/s bits/s From blutearuby at yahoo.com Tue May 10 05:36:39 2011 From: blutearuby at yahoo.com (ruby2010) Date: Mon, 9 May 2011 20:36:39 -0700 (PDT) Subject: [FFmpeg-user] How to convert MPEG, VOB, MP4, M4V, WMV, MKV, TS, 3GP to WMV Message-ID: <1304998599321-3510916.post@n4.nabble.com> WMV is streaming media format released by Microsoft's, which is an extension of upgrades with the "fellow" of the ASF (Advanced Stream Format) format. Under the same video quality, the size of WMV format is very small, so it is suitable for online playback and transmission. AVI files will be packaged in a video and audio files, and allow the audio and video formats to play at the same time. As the DVD video format , AVI files support multiple video and audio streams. iCoolsoft [url=http://www.icoolsoft.com/wmv-converter/]WMV Converter[/url] is a professional WMV video conversion tool, which supports the WMV format, having high speed and lossless quality media file conversion to AVI, DivX, XviD, MPG, MPEG, WMV, MOV, ASF, QT, RM, RMVB files. iCoolsoft WMV Converters is specially designed to convert files to WMV format. You can use it to convert files for Zune, Creative Zen, Archos, iRiver, Xbox 360, Dell Player, HP iPAQ, General Pocket PC, Gphone, BlackBerry, and Palm Pre. Powerful as it is, it can assure you with the fastest conversion speed and the highest output quality. iCoolsoft [url=http://www.icoolsoft.com/dvd-to-wmv-converter/]DVD to WMV Converter[/url] can help you convert DVD movie to WMV, WMA, MP3 and other video/audio files for storage on computer, playback on portable players and so on. Supported devices of this DVD to WMV Converter include Wii, PS3, Xbox 360, Zune, Archos, BlackBerry, Pocket PC, mobile phones, etc. [b]Steps of converting DVD to WMV format[/b] Step 1: Click button [b]Load DVD [/b]to load DVD movie. [b]Tip:[/b] You can preview the DVD file in the preview pane and take snapshot of the scenes you like the most. Click button and the picture will be saved as default format in default folder. Click button then you can open the folder and find the pictures. Step 2: Specify the output format as WMV. Step 3: You can click "Effect", "Trim", or "Crop" button on the tool bar to edit the DVD as you like. [img]http://www.icoolsoft.com/guides/images/dvd-to-wmv-converter/edit.jpg[/img] Step 4: Set the destination and click[img]http://www.icoolsoft.com/guides/images/dvd-ripper/start.jpg[/img] button to start converting DVD to WMV format. In the "Conversion" window, check "Open output folder when conversion completed" or you can click "Open Folder" button to find the converted files. [img]http://www.icoolsoft.com/guides/images/dvd-to-wmv-converter/converting.jpg[/img] -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/How-to-convert-MPEG-VOB-MP4-M4V-WMV-MKV-TS-3GP-to-WMV-tp3510916p3510916.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From kumar.pavan463 at gmail.com Wed May 11 06:57:17 2011 From: kumar.pavan463 at gmail.com (pavankuk) Date: Tue, 10 May 2011 21:57:17 -0700 (PDT) Subject: [FFmpeg-user] HI Message-ID: <1305089837708-3513868.post@n4.nabble.com> Hi AM NEW TO THE FFMPEG, WHEN AM USING THE FUNCTION AVCODEC_FIND_DECODER() IT IS NOT ABLE TO FIND OUT THE CORRESPONDING DECODER, IS THERE ANY PROCEDURE TO FOLLOW FOR IDENTIFYING THE REGISTERED DECODERS BY FFMPEG... THANKS IN ADVANCE PAVAKUK :) -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/HI-tp3513868p3513868.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From Niranjana at gloriatech.com Wed May 11 10:53:16 2011 From: Niranjana at gloriatech.com (Dev_com) Date: Wed, 11 May 2011 01:53:16 -0700 (PDT) Subject: [FFmpeg-user] ffmpeg with rtmp user authentication Message-ID: <1305103996143-3514174.post@n4.nabble.com> hi friends.. i am using ffmpeg encoding with wowza.. i want to stream a video file to live stream.. For Example: ffmpeg -i myfile.flv -re -sameq -acodec copy -vcodec copy -f flv rtmp://localhost/live/myStream here my live application publish was protected by username and password.. how can i pass this username and password in ffmpeg to wowza? can anyone help me? -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/ffmpeg-with-rtmp-user-authentication-tp3514174p3514174.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From kumar.pavan463 at gmail.com Thu May 12 06:54:04 2011 From: kumar.pavan463 at gmail.com (pavankuk) Date: Wed, 11 May 2011 21:54:04 -0700 (PDT) Subject: [FFmpeg-user] HI In-Reply-To: <1305089837708-3513868.post@n4.nabble.com> References: <1305089837708-3513868.post@n4.nabble.com> Message-ID: <1305176044995-3516494.post@n4.nabble.com> hi frnds the problem wat i posted is cleard -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/HI-tp3513868p3516494.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From kumar.pavan463 at gmail.com Thu May 12 06:57:23 2011 From: kumar.pavan463 at gmail.com (pavankuk) Date: Wed, 11 May 2011 21:57:23 -0700 (PDT) Subject: [FFmpeg-user] Problem facing while decoding Message-ID: <1305176243564-3516501.post@n4.nabble.com> hi frnds while decoding the video data using the avcodec_decode_video() function am gettng the error message like [h263 @ 0x10002bc0] Bad picture start code [h263 @ 0x10002bc0] header damaged i googled the problem but i didnt get any proepr information why we will get the above error and how to clear it please help out on it thanks in advance pavan kumar.k -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/Problem-facing-while-decoding-tp3516501p3516501.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From denis.muraviev at me.com Thu May 12 16:06:16 2011 From: denis.muraviev at me.com (Denis Muraviev) Date: Thu, 12 May 2011 18:06:16 +0400 Subject: [FFmpeg-user] ac3 6ch -> 6 mono wav files In-Reply-To: <4DCBCEA7.5060706@googlemail.com> References: <7C2FDC84-1149-40B0-9EA4-86EC431A6B10@mac.com> <4DCBCEA7.5060706@googlemail.com> Message-ID: Thanks, guys I'll check both of them. On May 12, 2011, at 4:12 PM, betonpfeiler wrote: > Am 11.05.2011 23:19, schrieb Denis Muraviev: >> Hi, >> >> I have .ac3 file with 6 channels, and I want to decode it to 6 different mono .wav files (every channel need to be in separate wav file). >> I found that there is `-ac channels' to set the number of audio channels, but how can I select the channel I need (not the first one). >> >> Thanks, >> Denis. > > Hey! > If you don't want to use SOX, you can also achieve this with the -map_audio_channel xi:yi:zi:xo:yo:zo command (x=file, y=stream, z=channel; i=in, o=out) you can find a more detailed explanation under ffmbc.wordpress.com > > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From andrey.vul at gmail.com Fri May 13 09:30:00 2011 From: andrey.vul at gmail.com (Andrey Moshbear) Date: Fri, 13 May 2011 03:30:00 -0400 Subject: [FFmpeg-user] xvid blockiness Message-ID: Using the default options for -vcodec xvid, I get results about as blocky as -vcodec libx264 -b 200, regardless of bitrate. Unfortunately, I can't -vpre slower / -profile slower for xvid and mpeg4 can't do 2pass as it dies whining about too low bitrate, for -b 800k. Which cmdline options reduce blockiness? Solutions? Xbox 360 can only do x264+aac-lc _or_ xvid+(mp3|ac3). I can't do x264+ac3, and my box isn't jtagged, so I can't run mplayer or equivalent. 001100 Andrey "m05hbear" Vul 010010 011110 andrey at moshbear dot net 100001 andrey dot vul at gmail 101101 110011 From kksreddy at umich.edu Fri May 13 17:23:51 2011 From: kksreddy at umich.edu (Kumar Sricharan) Date: Fri, 13 May 2011 11:23:51 -0400 Subject: [FFmpeg-user] error when compiling ffmpeg Message-ID: <4DCD4D07.4000107@umich.edu> Hello, When trying to install ffmpeg, I get the error "ERROR: libx264 not found". I am attaching the config.log file. Please let me know how to resolve this problem. Thanks, Kumar -------------- next part -------------- A non-text attachment was scrubbed... Name: config.log Type: text/x-log Size: 120182 bytes Desc: not available URL: From Marc-Olivier.Bergeron at marketel.com Mon May 16 21:59:01 2011 From: Marc-Olivier.Bergeron at marketel.com (Marc-Olivier Bergeron) Date: Mon, 16 May 2011 15:59:01 -0400 Subject: [FFmpeg-user] Configuration is empty?? Message-ID: Hay! I have a weird problem, or I'm just being stupid. I installed ffmpeg with a load of codecs (mp3lame, vpx, ogg, vorbis, libfaac, libfaad, theora, etc) and they are not in the "ffmpeg -codecs" list or in the "ffmpeg -formats" list. Upon installation, I entered this: "./configure --prefix=/usr --libdir=/usr/lib64 --mandir=/usr/share/man --disable-avisynth --enable-avfilter --enable-libfaac --enable-libgsm --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libx264 --enable-gpl --enable-nonfree --enable-postproc --enable-pthreads --enable-shared --enable-swscale --enable-vdpau --enable-version3 --enable-x11grab --enable-libvorbis --enable-libvpx --enable-libtheora" And everything went well after "make && make install", but then, if I just write "ffmpeg", I do not get a single library that was installed, I get this: " ffmpeg version 0.7_beta2, Copyright (c) 2000-2011 the Libav developers built on May 13 2011 12:34:43 with gcc 4.1.2 20080704 (Red Hat 4.1.2-50) configuration: libavutil 51. 1. 0 / 51. 1. 0 libavcodec 53. 3. 0 / 53. 3. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 4. 0 / 2. 4. 0 libswscale 1. 1. 0 / 1. 1. 0 Hyper fast Audio and Video encoder usage: ffmpeg [options] [[infile options] -i infile]... {[outfile options] outfile}... Use -h to get full help or, even better, run 'man ffmpeg'" Why is "configuration:" empty? Is it because 0.7 version does not give the configuration anymore? ANYWAY, here's all the info I know and what I tried: When I try to encode a video like this: "ffmpeg -i sample.mp4 -vcodec h263 -acodec libfaac -y sample.3gp" It says: "Unknown encoder 'libfaac'" We had 0.6.1 on our server before, installed 0.7 on top of it Friday and upon entering "ffmpeg", it said there were multiple config files. I told myself I would correct this Monday but now there is no configuration, perhaps because of a server reboot. I checked on the internet how to unsinstall, it said "make clean" and/or "make uninstall". Tried both but ffmpeg is still there afterwards (locate with grep) In the "./configuration" command, I tried putting incorrect values, like "--enable-libfaaaaaaac" and it gives me an error message, so it means that it works with the correct synthax "--enable-libfaac" I have set all the paths after installation: ln -s /usr/local/lib/libfaac.so.0 /usr/lib/libfaac.so.0 ln -s /usr/local/lib/libfaad.so.2 /usr/lib/libfaad.so.2 ln -s /usr/local/lib/libmp3lame.so.0 /usr/lib/libmp3lame.so.0 ln -s /usr/local/lib/libogg.so.0 /usr/lib/libogg.so.0 ln -s /usr/local/lib/libvorbisenc.so.2 /usr/lib/libvorbisenc.so.2 ln -s /usr/local/lib/libvorbis.so.0 /usr/lib/libvorbis.so.0 ln -s /usr/local/lib/libvorbis.so.0 /usr/lib/libvorbis.so.0 ln -s /usr/local/lib/libtheoradec.so.1 /usr/lib/libtheoradec.so.1 ln -s /usr/local/lib/libtheoraenc.so.1 /usr/lib/libtheoraenc.so.1 ln -s /usr/local/lib/libavcodec.so.52 /usr/lib/libavcodec.so.52 ln -s /usr/local/lib/libavformat.so.52 /usr/lib/libavformat.so.52 ln -s /usr/local/lib/libavutil.so.50 /usr/lib/libavutil.so.50 ln -s /usr/local/bin/ffmpeg2theora /usr/bin/ffmpeg2theora I have tried that line that was supposed to fix it: "sudo ldconfig" Our server is 64bits, but I still managed to compile everything with "-fPIC" when was needed so that is not the problem. My lib directories has all the libs needed (ls listing): codecs libfaac.so.0.0.0 libogg.la libtheoraenc.so libvorbisfile.a libavcodec.a libfaad.a libogg.so libtheoraenc.so.1 libvorbisfile.la libavcodec.so.52 libfaad.la libogg.so.0 libtheoraenc.so.1.1.2 libvorbisfile.so libavdevice.a libfaad.so libogg.so.0.7.1 libtheora.la libvorbisfile.so.3 libavfilter.a libfaad.so.2 libswscale.a libtheora.so libvorbisfile.so.3.3.4 libavformat.a libfaad.so.2.0.0 libtheora.a libtheora.so.0 libvorbis.la libavformat.so.52 libmp3lame.a libtheoradec.a libtheora.so.0.3.10 libvorbis.so libavutil.a libmp3lame.la libtheoradec.la libvorbis.a libvorbis.so.0 libavutil.so.50 libmp3lame.so libtheoradec.so libvorbisenc.a libvorbis.so.0.4.5 libfaac.a libmp3lame.so.0 libtheoradec.so.1 libvorbisenc.la libvpx.a libfaac.la libmp3lame.so.0.0.0 libtheoradec.so.1.1.4 libvorbisenc.so perl5 libfaac.so libmp4ff.a libtheoraenc.a libvorbisenc.so.2 pkgconfig libfaac.so.0 libogg.a libtheoraenc.la libvorbisenc.so.2.0.8 I'm about to throw myself out the window so if anyone has any idea, it would be appreciated. Thanks! Marco From kern at clavain.co.uk Tue May 17 20:23:26 2011 From: kern at clavain.co.uk (Enrico Kern) Date: Tue, 17 May 2011 20:23:26 +0200 Subject: [FFmpeg-user] converting a previous cuted file, Invalid pixel format string '-1' Message-ID: <4DD2BD1E.5030708@clavain.co.uk> Hi all, maybe someone can help me with this issue. I use the latest ffmpeg from trunk from today but i guess older ones are affected too. - I have a FLV file which contains H264 video and mp3 audio, the file is recorded by WowzaMediaServer - I cut this file with ffmpeg in multiple parts with ffmpeg -i $input -ss $time -t $time $output.flv That works fine, the file also plays in flash and players. Now however i want to transcode this cuted file to another format, that doesnt work. ffmpeg returns with this: ffmpeg -i cuterror.flv -y test.webm FFmpeg version SVN-r26402, Copyright (c) 2000-2011 the FFmpeg developers built on May 17 2011 07:24:46 with gcc 4.1.2 20080704 (Red Hat 4.1.2-50) configuration: --enable-version3 --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libvpx --enable-libfaac --enable-libmp3lame --enable-libtheora --enable-libvorbis --enable-libx264 --enable-libxvid --disable-ffplay --enable-shared --enable-gpl --enable-postproc --enable-nonfree --enable-avfilter --enable-pthreads --extra-cflags=-fPIC libavutil 50.36. 0 / 50.36. 0 libavcore 0.16. 1 / 0.16. 1 libavcodec 52.108. 0 / 52.108. 0 libavformat 52.93. 0 / 52.93. 0 libavdevice 52. 2. 3 / 52. 2. 3 libavfilter 1.74. 0 / 1.74. 0 libswscale 0.12. 0 / 0.12. 0 libpostproc 51. 2. 0 / 51. 2. 0 [flv @ 0x1e698510] max_analyze_duration reached [flv @ 0x1e698510] Estimating duration from bitrate, this may be inaccurate Input #0, flv, from 'cuterror.flv': Metadata: duration : 573 width : 1152 height : 720 videodatarate : 200 framerate : 1000 videocodecid : 7 audiodatarate : 100 audiosamplerate : 44100 audiosamplesize : 16 stereo : false audiocodecid : 2 encoder : Lavf52.84.0 filesize : 37793482 Duration: 00:09:32.92, start: 0.002000, bitrate: 307 kb/s Stream #0.0: Video: h264, 204 kb/s, 1k tbr, 1k tbn, 1k tbc Stream #0.1: Audio: mp3, 44100 Hz, 1 channels, s16, 102 kb/s [buffer @ 0x1e6abc30] Invalid pixel format string '-1' Error opening filters! I can of course write the file to this format when i cut them, but thats not possible in our production environment, as we need the flash files for our website players. I need the ability to convert them later to webm for html5 integration. But it doesnt matter if i use webm or any other codec/container, the error stays the same even if i just want todo a copy of the streams. Anyone got a idea of whats wrong here? A cutted demonstration file can be downloaded here: http://master.inacdn.com/cuterror.flv (36MB) Regards, Enrico -- [Clavain Technologies Ltd.] Enrico Kern, Managing Director kern at clavain.co.uk www.clavain.co.uk Herdernstr. 16 8004 Zurich SWITZERLAND TEL: +41 (44) 586 55 84 FAX: +49 (180) 5233633-54820 Registered in 1A Pope Street, SE1 3PR London UK, Company No. 07245410 From tom_a_sparks at yahoo.com.au Wed May 18 03:23:06 2011 From: tom_a_sparks at yahoo.com.au (Tom Sparks) Date: Tue, 17 May 2011 18:23:06 -0700 (PDT) Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: <4DD24B66.3010800@zonnet.nl> Message-ID: <415964.3708.qm@web36408.mail.mud.yahoo.com> --- On Tue, 17/5/11, belcampo wrote: > From: belcampo > Subject: Re: [FFmpeg-user] Convert some hundreds files to a H264 > To: "FFmpeg user questions and RTFMs" > Received: Tuesday, 17 May, 2011, 8:18 PM > On 05/17/11 11:16, marcos rebelo > wrote: > > My problem is to do the [insert your commands here], > > > > In my opinion, it should be easy to give some commands > defining the > > final format of the video and ignore the original > format, a smart > > script should analyse the video and it should create > the [insert your > > commands here]. > > > > Note that my files are correctly visible on mplayer. > If mplayer detects/analyses your files correct, so will > ffmpeg, compiled > with the same codecs, do the same. Nothing special to do > AFAIK. > > > > Thanks for all help > > > > this code may give you a starting place, it was brrowed from http://www.mythtv.org/wiki/Stream_mythtv_recordings_from_mythweb_using_flash_video #!/bin/bash directory=$1; file=$2; # Determine aspect ratio original_name=$1/$2; #Grab input file information videoheight=`mplayer -vo null -ao null -frames 0 -identify "$original_name" 2>/dev/null | grep "ID_VIDEO_HEIGHT" | sed -e 's/ID_VIDEO_HEIGHT=//'` height=240; if [ "$videoheight" -gt 480 ]; then # "HDTV" width=426 else # "SDTV" width=320 fi dimensions=$width"x"$height; # Create the flash video (flv) file with a frame rate of 20fps # deinterlace the video and set an apropriate audio sample rate ffmpeg -y -i $directory/$file -r 20 -s $dimensions -deinterlace -ar 22050 $directory/$file.flv 1>/dev/null 2>/dev/null # Add metadata to file file (optional) cat $directory/$file.flv | flvtool2 -U stdin $directory/$file.flv tom From blutearuby at yahoo.com Wed May 18 04:09:04 2011 From: blutearuby at yahoo.com (ruby2010) Date: Tue, 17 May 2011 19:09:04 -0700 (PDT) Subject: [FFmpeg-user] How to rip DVD movies and convert DVD to AVI, DivX, MKV, MP4, WMV Message-ID: <1305684544873-3531363.post@n4.nabble.com> Let me recommend a software program--DVD ripper.It can copy the content of a DVD to a hard disk drive and transfer video on DVDs to different formats,such as convert DVD video for playback on media players and mobile devices.iCoolsoft DVD Ripper is our new product to you. iCoolsoft [b][url=http://www.icoolsoft.com/dvd-ripper/]DVD Ripper[/url][/b] is a professional DVD movie ripping and converting software, that can rip DVD movies and convert DVD to all popular video formats like ripping DVD to AVI, DivX, MKV, MP4, WMV, ASF, 3GP, etc. even output HD videos such as HD AVI, HD H.264, HD WMV, and so on. This DVD Ripper also helps extract DVD audio tracks to popular audio formats, like MP3, WMA, M4A, AAC, WAV, AC3, etc. We also have iCoolsoft [b][url=http://www.icoolsoft.com/dvd-copy//]DVD Copy[/url] [/b]is a powerful DVD copier and DVD duplicate tool, that can help you copy DVD and clone DVD in several ways, such as copy DVD to a new disc (DVD-5 to DVD-5, DVD-9 to DVD-9, DVD-9 to DVD-5), backup DVD to DVD folder or ISO image file on local disc. This DVD Copy also enables you to burn DVD folder and ISO image to DVD disc. [b]How to Rip DVD Movie?[/b] [b]Step 1[/b]: Download iCoolsoft DVD Ripper for free, then install and launch this tool. [b] Step 2[/b]: Click button and choose "[b]Load DVD[/b]", then select your DVD driver and click "OK". The following window will show up. [b] Step 3[/b]: Select Audio Track and Subtitle. Choose the output format from "Profile" drop-down list. Click "Apply to all" to set the format for all source videos. Click [b]"Browse"[/b] button to set output folder, use "Open Folder" to quickly open that folder. [b]Step 4[/b]: Click [b]start [/b] button to begin ripping and converting. When finished, you will get DVD movies stored in specified format in output folder. [img]http://www.icoolsoft.com/guides/images/dvd-ripper/edit.jpg[/img] -- View this message in context: http://ffmpeg-users.933282.n4.nabble.com/How-to-rip-DVD-movies-and-convert-DVD-to-AVI-DivX-MKV-MP4-WMV-tp3531363p3531363.html Sent from the FFmpeg-users mailing list archive at Nabble.com. From rodney.baker at iinet.net.au Wed May 18 05:17:21 2011 From: rodney.baker at iinet.net.au (Rodney Baker) Date: Wed, 18 May 2011 12:47:21 +0930 Subject: [FFmpeg-user] error when compiling ffmpeg In-Reply-To: <4DCD4D07.4000107@umich.edu> References: <4DCD4D07.4000107@umich.edu> Message-ID: <201105181247.21689.rodney.baker@iinet.net.au> On Sat, 14 May 2011 00:53:51 Kumar Sricharan wrote: > Hello, > > When trying to install ffmpeg, I get the error "ERROR: libx264 not > found". I am attaching the config.log file. > > Please let me know how to resolve this problem. > > Thanks, > Kumar Install the libx264-devel package (whatever that happends to be for your distro) or build x264 from source before building ffmpeg. Either way will install the required headers. If you do choose to build libx264, I'm pretty sure it needs --enable-shared during configure to build shared libraries. -- =================================================== Rodney Baker VK5ZTV rodney.baker at iinet.net.au =================================================== From rodney.baker at iinet.net.au Wed May 18 05:20:29 2011 From: rodney.baker at iinet.net.au (Rodney Baker) Date: Wed, 18 May 2011 12:50:29 +0930 Subject: [FFmpeg-user] How to rip DVD movies and convert DVD to AVI, DivX, MKV, MP4, WMV In-Reply-To: <1305684544873-3531363.post@n4.nabble.com> References: <1305684544873-3531363.post@n4.nabble.com> Message-ID: <201105181250.29536.rodney.baker@iinet.net.au> On Wed, 18 May 2011 11:39:04 ruby2010 wrote: > Let me recommend a software program--DVD ripper.It can copy the content of >[...] If you're not posting about using ffmpeg, please go away. -- =================================================== Rodney Baker VK5ZTV rodney.baker at iinet.net.au =================================================== From stefano.sabatini-lala at poste.it Wed May 18 11:13:00 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Wed, 18 May 2011 11:13:00 +0200 Subject: [FFmpeg-user] converting a previous cuted file, Invalid pixel format string '-1' In-Reply-To: <4DD2BD1E.5030708@clavain.co.uk> References: <4DD2BD1E.5030708@clavain.co.uk> Message-ID: <20110518091300.GC5814@geppetto> On date Tuesday 2011-05-17 20:23:26 +0200, Enrico Kern encoded: > Hi all, > > maybe someone can help me with this issue. I use the latest ffmpeg > from trunk from today but i guess older ones are affected too. > > - I have a FLV file which contains H264 video and mp3 audio, the > file is recorded by WowzaMediaServer > - I cut this file with ffmpeg in multiple parts with ffmpeg -i > $input -ss $time -t $time $output.flv > > That works fine, the file also plays in flash and players. Now > however i want to transcode this cuted file to another format, > that doesnt work. ffmpeg returns with this: > > ffmpeg -i cuterror.flv -y test.webm > FFmpeg version SVN-r26402, Copyright (c) 2000-2011 the FFmpeg developers > built on May 17 2011 07:24:46 with gcc 4.1.2 20080704 (Red Hat 4.1.2-50) > configuration: --enable-version3 --enable-libopencore-amrnb > --enable-libopencore-amrwb --enable-libvpx --enable-libfaac > --enable-libmp3lame --enable-libtheora --enable-libvorbis > --enable-libx264 --enable-libxvid --disable-ffplay --enable-shared > --enable-gpl --enable-postproc --enable-nonfree --enable-avfilter > --enable-pthreads --extra-cflags=-fPIC > libavutil 50.36. 0 / 50.36. 0 > libavcore 0.16. 1 / 0.16. 1 > libavcodec 52.108. 0 / 52.108. 0 > libavformat 52.93. 0 / 52.93. 0 > libavdevice 52. 2. 3 / 52. 2. 3 > libavfilter 1.74. 0 / 1.74. 0 > libswscale 0.12. 0 / 0.12. 0 > libpostproc 51. 2. 0 / 51. 2. 0 > [flv @ 0x1e698510] max_analyze_duration reached > [flv @ 0x1e698510] Estimating duration from bitrate, this may be inaccurate > Input #0, flv, from 'cuterror.flv': > Metadata: > duration : 573 > width : 1152 > height : 720 > videodatarate : 200 > framerate : 1000 > videocodecid : 7 > audiodatarate : 100 > audiosamplerate : 44100 > audiosamplesize : 16 > stereo : false > audiocodecid : 2 > encoder : Lavf52.84.0 > filesize : 37793482 > Duration: 00:09:32.92, start: 0.002000, bitrate: 307 kb/s > Stream #0.0: Video: h264, 204 kb/s, 1k tbr, 1k tbn, 1k tbc > Stream #0.1: Audio: mp3, 44100 Hz, 1 channels, s16, 102 kb/s > [buffer @ 0x1e6abc30] Invalid pixel format string '-1' > Error opening filters! > > > I can of course write the file to this format when i cut them, but > thats not possible in our production environment, as we need > the flash files for our website players. I need the ability to > convert them later to webm for html5 integration. But it doesnt > matter > if i use webm or any other codec/container, the error stays the same > even if i just want todo a copy of the streams. > > Anyone got a idea of whats wrong here? A cutted demonstration file > can be downloaded here: The problem is that the input file is missing the pixel format information, the decoder doesn't guess it and ffmpeg fails because of that. File a bug report, and we'll try to fix it. > http://master.inacdn.com/cuterror.flv (36MB) > > Regards, -- ffmpeg-user random tip #6 Please follow netiquette rules while posting to ffmpeg-user: http://linux.sgms-centre.com/misc/netiquette.php From joolzg at btinternet.com Wed May 18 12:05:23 2011 From: joolzg at btinternet.com (JULIAN GARDNER) Date: Wed, 18 May 2011 11:05:23 +0100 (BST) Subject: [FFmpeg-user] -g command Message-ID: <816669.14707.qm@web86401.mail.ird.yahoo.com> Has anything changed in relation to this command and the new svn code. I ask this as i have a an encoder now running latest svn code and it takes 10 seconds to get the first frame, on my other encoders its within 3 seconds. all encoders use "-g 75" joolz From stefano.sabatini-lala at poste.it Wed May 18 12:07:21 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Wed, 18 May 2011 12:07:21 +0200 Subject: [FFmpeg-user] Convert some hundreds files to a H264 In-Reply-To: References: Message-ID: <20110518100721.GI5814@geppetto> On date Tuesday 2011-05-17 12:24:08 -0400, Mailing List encoded: > marcos rebelo wrote in > news:BANLkTi=43-RGdG9h4LPx+mi2AcxdZyv-zQ at mail.gmail.com: > > > I was giving some hundreds of files (more will come) with > > different formats (div, avi, mov, ...) and different sizes to be > > converted to H264 with: > > a maximum width 720, > > -b 500k > > -bt 500k > > -acodec libfaac > > As others have suggested, using a simple loop to process the files is not > particularly difficult. The only issue is your "maximum width 720". Using > -vf "scale=720:-1" is unreliable because it throws errors for odd-numbered > heights, instead of rounding to a usable value as mencoder can. You can rely on expression evaluation (recently added): -vf "scale=720:trunc(ow/a/vsub)*vsub" ow/a is the scaled height, in this case you get the final height an integer multiple of the chroma vertical subsampling value. Many variations are possible. -- ffmpeg-user random tip #14 Smart questions (un)official document: http://www.catb.org/~esr/faqs/smart-questions.html From nicolas.george at normalesup.org Wed May 18 15:11:03 2011 From: nicolas.george at normalesup.org (Nicolas George) Date: Wed, 18 May 2011 15:11:03 +0200 Subject: [FFmpeg-user] What does "-vol" parameter mean in ffmpeg? In-Reply-To: <20110517110530.c7c0023d.dashing.meng@gmail.com> References: <20110517110530.c7c0023d.dashing.meng@gmail.com> Message-ID: <20110518131103.GA1576@phare.normalesup.org> L'octidi 28 flor?al, an CCXIX, littlebat a ?crit?: > I can't understand how to use "-vol" parameter to increase volume in > ffmpeg. I know "-vol 256" mean keep volume unchanged. But I don't know > how to choose a proper number for a video file to increase volume. For > example, does "-vol 512" mean volume intensity is twice as original > file? Just try it for yourself: perl -e 'print pack("s*", 3000, 2000, 1000, 0, -1000, -2000)' | ffmpeg -f s16le -i - -vol 512 -f s16le - 2> /dev/null | perl -e '$/=\2; while(<>) { print unpack("s", $_), "\n" }' Of course, you can do exactly the same with any GUI PCM editor. By looking at the result, you can see that obviously each sample is multiplied by vol/256. Regards, -- Nicolas George -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 198 bytes Desc: Digital signature URL: From nicolas.george at normalesup.org Wed May 18 15:15:33 2011 From: nicolas.george at normalesup.org (Nicolas George) Date: Wed, 18 May 2011 15:15:33 +0200 Subject: [FFmpeg-user] how to maximize the volume without distorting the sound using ffmpeg? In-Reply-To: <20110517103427.e0b2cacc.dashing.meng@gmail.com> References: <20110517103427.e0b2cacc.dashing.meng@gmail.com> Message-ID: <20110518131533.GB1576@phare.normalesup.org> L'octidi 28 flor?al, an CCXIX, littlebat a ?crit?: > Is there a way to maximize the volume without distorting the sound > using Ffmpeg? just like the "-af volnorm" filter in Mencoder? Your message is self-contradictory: -af volnorm does not do what you think, it adjusts the gain to keep the volume mostly uniform, this is very much a distortion. Maximizing the volume without distortion requires a constant gain (because a variable gain is a distortion) and finding the greatest gain that will avoid clipping. The latest operation can only be done by decoding the whole stream once first, obviously. I think ffmpeg does not have a tool to compute the maximum value of a sample; mplayer does, it's -af stats. Regards, -- Nicolas George -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 198 bytes Desc: Digital signature URL: From mblainml at earthlink.net Wed May 18 15:36:38 2011 From: mblainml at earthlink.net (Mark Blain) Date: Wed, 18 May 2011 09:36:38 -0400 Subject: [FFmpeg-user] Convert some hundreds files to a H264 References: <20110518100721.GI5814@geppetto> Message-ID: Stefano Sabatini wrote: > >You can rely on expression evaluation (recently added): > >-vf "scale=720:trunc(ow/a/vsub)*vsub" > >ow/a is the scaled height, in this case you get the final height an >integer multiple of the chroma vertical subsampling value. Many >variations are possible. Thanks! I read some development discussion about that, but I didn't know it was in the trunk yet. The details are in http://www.ffmpeg.org/libavfilter.html#SEC37 From alexandre.ferrieux at orange-ftgroup.com Wed May 18 18:24:20 2011 From: alexandre.ferrieux at orange-ftgroup.com (Alexandre Ferrieux) Date: Wed, 18 May 2011 18:24:20 +0200 Subject: [FFmpeg-user] Asynchronous overlays Message-ID: <4DD3F2B4.1070201@orange-ftgroup.com> Hello, I'd like to put side-by-side two video sources (with the overlay filter), with a completely asynchronous/decoupled scheme. Indeed my two video sources have varied and unstable frame rates, and in a naive setup of the overlay filter, ffmpeg insists on fetching a frame from each one for every output frame. The idea is do decode both streams independently (threads) into the same overlay buffer, and have a third thread sample this at a regular rate (the wanted output frame rate), and feed that into the output chain (encoder + container). So, if at any given time one of the sources lags, the same frozen frame from it will be reused in several output frames (ie overlay buffer not updated in that area), but the overall output will not be stalled (as it is today). Q1: is this doable with command-line flags to the ffmpeg executable ? Q2: if not, I'd appreciate a sketch of where to look in the sources to do this in C. -Alex From baptiste.coudurier at gmail.com Wed May 18 22:29:26 2011 From: baptiste.coudurier at gmail.com (Baptiste Coudurier) Date: Wed, 18 May 2011 13:29:26 -0700 Subject: [FFmpeg-user] Asynchronous overlays In-Reply-To: <4DD3F2B4.1070201@orange-ftgroup.com> References: <4DD3F2B4.1070201@orange-ftgroup.com> Message-ID: <4DD42C26.9000904@gmail.com> Hi, On 05/18/2011 09:24 AM, Alexandre Ferrieux wrote: > Hello, > > I'd like to put side-by-side two video sources (with the overlay > filter), with a completely asynchronous/decoupled scheme. Indeed my two > video sources have varied and unstable frame rates, and in a naive setup > of the overlay filter, ffmpeg insists on fetching a frame from each one > for every output frame. > > The idea is do decode both streams independently (threads) into the same > overlay buffer, and have a third thread sample this at a regular rate > (the wanted output frame rate), and feed that into the output chain > (encoder + container). So, if at any given time one of the sources lags, > the same frozen frame from it will be reused in several output frames > (ie overlay buffer not updated in that area), but the overall output > will not be stalled (as it is today). > > Q1: is this doable with command-line flags to the ffmpeg executable ? No. > Q2: if not, I'd appreciate a sketch of where to look in the sources to > do this in C. Doing this is way more simple I think, it's just a matter of fetching a new overlay frame at the right time depending on pts of the main video, I don't think you need any threads here. -- Baptiste COUDURIER Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org From jsd at cluttered.com Thu May 19 00:48:05 2011 From: jsd at cluttered.com (Jon Drukman) Date: Wed, 18 May 2011 22:48:05 +0000 (UTC) Subject: [FFmpeg-user] =?utf-8?b?YXZfaW50ZXJsZWF2ZWRfd3JpdGVfZnJhbWUoKTog?= =?utf-8?q?Invalid_argument?= Message-ID: I've noticed this happening a lot with wmv files. Is there anything I can do to fix it? /usr/local/bin/ffmpeg -loglevel quiet -v 0 -i /tmp/videoproc-24481/video_00006.wmv -s 480x352 -an -pass 1 -vcodec libx264 -vpre fast_firstpass -b 300k -bt 300k -threads 0 /tmp/videoproc-24481/video_00006.wmv.mp4 ffmpeg version git-N-29818-gb7c7f89, Copyright (c) 2000-2011 the FFmpeg developers built on May 11 2011 15:23:54 with gcc 4.1.2 20080704 (Red Hat 4.1.2-48) configuration: --enable-gpl --enable-libfaac --enable-libmp3lame --enable-libx264 --enable-pthreads --enable-static --disable-shared --disable-network --enable-nonfree --extra-ldflags=-static libavutil 51. 2. 1 / 51. 2. 1 libavcodec 53. 5. 0 / 53. 5. 0 libavformat 53. 0. 3 / 53. 0. 3 libavdevice 53. 0. 0 / 53. 0. 0 libavfilter 2. 5. 0 / 2. 5. 0 libswscale 0. 14. 0 / 0. 14. 0 Stream mapping: Stream #0.1 -> #0.0 Press [q] to stop encoding av_interleaved_write_frame(): Invalid argument From axu_69 at yahoo.com Thu May 19 01:48:13 2011 From: axu_69 at yahoo.com (Aoxiang Xu) Date: Wed, 18 May 2011 16:48:13 -0700 (PDT) Subject: [FFmpeg-user] libavcodec.so crashed under arm v7 Message-ID: <554378.86671.qm@web36901.mail.mud.yahoo.com> Hi: We have an application that uses libavcodec to play mpeg4 video. It works on arm v5. Now, I recompiled it under arm v7. The application crashed. I traced the problem to the line of MPV_common_init() in function ff_h263_decode_frame() of h263dec.c. After some investigations, i realized that MPV_common_init_arm() that is called in ff_dct_common_init() is arm version dependent. I would like to know if libavcodec is ready to work on arm v7. Thank you very much, Cheers, From dashing.meng at gmail.com Thu May 19 05:29:35 2011 From: dashing.meng at gmail.com (littlebat) Date: Thu, 19 May 2011 11:29:35 +0800 Subject: [FFmpeg-user] how to maximize the volume without distorting the sound using ffmpeg? In-Reply-To: <20110518131533.GB1576@phare.normalesup.org> References: <20110517103427.e0b2cacc.dashing.meng@gmail.com> <20110518131533.GB1576@phare.normalesup.org> Message-ID: <20110519112935.d7ca8c34.dashing.meng@gmail.com> On Wed, 18 May 2011 15:15:33 +0200 Nicolas George wrote: > L'octidi 28 flor?al, an CCXIX, littlebat a ?crit?: > > Is there a way to maximize the volume without distorting the sound > > using Ffmpeg? just like the "-af volnorm" filter in Mencoder? > > Your message is self-contradictory: -af volnorm does not do what you > think, it adjusts the gain to keep the volume mostly uniform, this is > very much a distortion. > > Maximizing the volume without distortion requires a constant gain > (because a variable gain is a distortion) and finding the greatest > gain that will avoid clipping. > > The latest operation can only be done by decoding the whole stream > once first, obviously. I think ffmpeg does not have a tool to compute > the maximum value of a sample; mplayer does, it's -af stats. Thanks for your reply. This is documented in mplayer(mencoder) man page: ( http://www.mplayerhq.hu/DOCS/man/en/mplayer.1.html#AUDIO%20FILTERS ) **************************************** volnorm[=method:target] Maximizes the volume without distorting the sound. Sets the used method. 1: Use a single sample to smooth the variations via the standard weighted mean over past samples (default). 2: Use several samples to smooth the variations via the standard weighted mean over past samples. Sets the target amplitude as a fraction of the maximum for the sample type (default: 0.25). **************************** And I found "mencoder -af volnorm" is really useful after some tests. I attach two Audaity(audio editor) audio track screenshots, one is orginal (orig.jpg) and another is "-af volnorm" by mencoder(volnorm.jpg). But, there are still some places is clipped in volnormed file(red part). -------------- next part -------------- A non-text attachment was scrubbed... Name: orig.jpg Type: image/jpeg Size: 9997 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: volnorm.jpg Type: image/jpeg Size: 10232 bytes Desc: not available URL: From alexandre.ferrieux at orange-ftgroup.com Thu May 19 08:52:23 2011 From: alexandre.ferrieux at orange-ftgroup.com (Alexandre Ferrieux) Date: Thu, 19 May 2011 08:52:23 +0200 Subject: [FFmpeg-user] Asynchronous overlays In-Reply-To: <4DD42C26.9000904@gmail.com> References: <4DD3F2B4.1070201@orange-ftgroup.com> <4DD42C26.9000904@gmail.com> Message-ID: <4DD4BE27.9030000@orange-ftgroup.com> On 18/05/2011 22:29, Baptiste Coudurier wrote: > Hi, > > On 05/18/2011 09:24 AM, Alexandre Ferrieux wrote: >> Hello, >> >> I'd like to put side-by-side two video sources (with the overlay >> filter), with a completely asynchronous/decoupled scheme. Indeed my two >> video sources have varied and unstable frame rates, and in a naive setup >> of the overlay filter, ffmpeg insists on fetching a frame from each one >> for every output frame. >> >> The idea is do decode both streams independently (threads) into the same >> overlay buffer, and have a third thread sample this at a regular rate >> (the wanted output frame rate), and feed that into the output chain >> (encoder + container). So, if at any given time one of the sources lags, >> the same frozen frame from it will be reused in several output frames >> (ie overlay buffer not updated in that area), but the overall output >> will not be stalled (as it is today). >> >> Q1: is this doable with command-line flags to the ffmpeg executable ? > > No. OK, that's clear at least :) >> Q2: if not, I'd appreciate a sketch of where to look in the sources to >> do this in C. > > Doing this is way more simple I think, it's just a matter of fetching a > new overlay frame at the right time depending on pts of the main video, > I don't think you need any threads here. Oh yes, everything can always be done single-threadedly with a big select(), but I don't know the overall architecture of ffmpeg regarding this. Links appreciated. In the limit case where there's a single source that is stalled, if we want the output to continue emitting frames at OUTPUT-FRAME-PERIOD, we need: - either a nonblocking "fetch a new input frame" - or a select() monitoring the input with a timeout of OUTPUT-FRAME-PERIOD Which one is supported by (or in the spirit of) ffmpeg's overall design ? -Alex From nicolas.george at normalesup.org Thu May 19 10:42:30 2011 From: nicolas.george at normalesup.org (Nicolas George) Date: Thu, 19 May 2011 10:42:30 +0200 Subject: [FFmpeg-user] how to maximize the volume without distorting the sound using ffmpeg? In-Reply-To: <20110519112935.d7ca8c34.dashing.meng@gmail.com> References: <20110517103427.e0b2cacc.dashing.meng@gmail.com> <20110518131533.GB1576@phare.normalesup.org> <20110519112935.d7ca8c34.dashing.meng@gmail.com> Message-ID: <20110519084229.GA16356@phare.normalesup.org> Le decadi 30 flor?al, an CCXIX, littlebat a ?crit?: > This is documented in mplayer(mencoder) man page: > Maximizes the volume without distorting the sound. Well, in that case, all I can say is that the documentation is misleading at best. Regards, -- Nicolas George -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 198 bytes Desc: Digital signature URL: From d.maddau at tiscali.it Thu May 19 17:05:09 2011 From: d.maddau at tiscali.it (Gavino L. Maddau) Date: Thu, 19 May 2011 17:05:09 +0200 Subject: [FFmpeg-user] Number of frames decreased Message-ID: Hi everyone! When i try to encode a yv12 file into a compressed one (i'm using mpeg1video, mpeg2video and x264), the number of frames of the encoded file is smaller than that of the original one. In fact, the number of frames of the yv12 file is 29, the other one is 27, whichever coder i use. Usually i write, in the mpeg1 case: ffmpeg -s qcif -pix_fmt yuv420p -i in.yuv -vcodec mpeg1video -f mpeg out.mpg I also tried to set the input and output framerate, the number of frames to code (-vframes 29), the output bit rate (-b), the buffer size and the maximum bit rate, but nothing changed. I have to calculate the psnr between the original video and the encoded one, but if the number of frames is different, i can't know which frame in the original file originate a coded one, because there's not a one-to-one correspondence. How can i solve this problem? Thank you. Sorry for my bad english. From diarrhio at gmail.com Thu May 19 23:46:17 2011 From: diarrhio at gmail.com (Dario) Date: Thu, 19 May 2011 14:46:17 -0700 Subject: [FFmpeg-user] Error extracting images from .mov with timecode Message-ID: Hi, We have a .mov file which has a timecode track in it, which seems to be causing ffmpeg lots of fits when extracting jpegs out of it. I'm using the latest windows binary found on http://ffmpeg.zeranoe.com/builds/ I first tried extracting the jpegs out like this: ffmpeg.exe -i inputmovie.mov -y Output\Image%06d.jpg When running the above, I get the first frame of video duplicated about 2700 times, then I get *948* frames of the actual video. So, I then ran ffprobe.exe to extract the timecode information out of the video: Here is the output from ffprobe.exe: Unsupported codec with id 0 for input stream 2 [STREAM] index=0 codec_name=pcm_s16be codec_long_name=PCM signed 16-bit big-endian codec_type=audio codec_time_base=0/1 codec_tag_string=twos codec_tag=0x736f7774 sample_rate=48000.000000 channels=2 bits_per_sample=16 r_frame_rate=0/0 avg_frame_rate=7575/158 time_base=1/48000 start_time=0.000000 duration=39.528333 nb_frames=1897360 TAG:creation_time=2011-05-03 20:16:26 TAG:language=eng [/STREAM] [STREAM] index=1 codec_name=mjpeg codec_long_name=MJPEG (Motion JPEG) codec_type=video codec_time_base=1/23976 codec_tag_string=jpeg codec_tag=0x6765706a width=960 height=540 has_b_frames=0 sample_aspect_ratio=72:72 display_aspect_ratio=16:9 pix_fmt=yuvj420p r_frame_rate=2997/125 avg_frame_rate=2997/125 time_base=1/23976 start_time=-0.003003 duration=39.539540 nb_frames=948 TAG:creation_time=2011-05-03 20:16:26 TAG:language=eng [/STREAM] [STREAM] index=2 codec_name=unknown codec_type=data codec_time_base=0/1 codec_tag_string=tmcd codec_tag=0x64636d74 r_frame_rate=0/0 avg_frame_rate=0/0 time_base=1/600 start_time=-108.695000 duration=162.286667 nb_frames=1 TAG:creation_time=2011-05-03 20:16:26 TAG:language=eng [/STREAM] As you can see there are *948* frames in the video (as is confirmed by Quicktime). In addition, the start_time in the timecode stream is -108.695000. So, I then tried to pass the timecode start_time value as the -ss value like so: ffmpeg.exe -i inputmovie.mov -ss 108.695 -y Output\Image%06d.jpg However, this only extracts *947* jpegs. I also tried passing the duration as the -t value, but it didn't help. I am stuck either getting tons of duplicate images or getting one frame less than the number of frames in the video. I also tried the -vcodec copy option instead of specifying the -ss option, which worked wonderfully and extracted 948 frames, except that it only seems to work on motion jpegs and no other codecs... my utility needs to work on any type of codec ffmpeg supports. So, I'm stuck. I'm wondering if this is a bug or if I'm just using ffmpeg improperly. Any help would be appreciated. Thanks, Dario From oceantear at gmail.com Fri May 20 09:15:14 2011 From: oceantear at gmail.com (=?UTF-8?B?5bCP576K?=) Date: Fri, 20 May 2011 15:15:14 +0800 Subject: [FFmpeg-user] Meet problem when cross compiler(using powerpc-linux-gnu-gcc ) FFMPEG source code Message-ID: Hi All, I am a newbie for using FFMPEG lib to develop a tool. Firstly, i downloaded the source code from git server and tried to cross compiler(powerpc-linux-gnu-gcc ) the source code on my LINUX server. Unfortunately, it is not smooth. My compiler environment: OS: Ubuntu 11.04 VMWare: 7.0 My configuration: ./configure --enable-cross-compile --arch=powerpc --host-cc=powerpc-linux-gnu-gcc --target-os=linux Error Message: You need a compiler that supports {} in AltiVec vector declarations Can someone give me some suggestion? Thanks in advance. From andrey.vul at gmail.com Fri May 20 09:50:03 2011 From: andrey.vul at gmail.com (Andrey m0shbear Vul) Date: Fri, 20 May 2011 03:50:03 -0400 Subject: [FFmpeg-user] Meet problem when cross compiler(using powerpc-linux-gnu-gcc ) FFMPEG source code Message-ID: From: ?? Hi All, I am a newbie for using FFMPEG lib to develop a tool. Firstly, i downloaded the source code from git server and tried to cross compiler(powerpc-linux-gnu-gcc ) the source code on my LINUX server. Unfortunately, it is not smooth. My compiler environment: OS: Ubuntu 11.04 VMWare: 7.0 My configuration: ./configure --enable-cross-compile --arch=powerpc --host-cc=powerpc-linux-gnu-gcc --target-os=linux Error Message: You need a compiler that supports {} in AltiVec vector declarations Can someone give me some suggestion? Thanks in advance. _______________________________________________ gcc version? From oceantear at gmail.com Fri May 20 10:09:50 2011 From: oceantear at gmail.com (Shu Chaio Yang) Date: Fri, 20 May 2011 01:09:50 -0700 Subject: [FFmpeg-user] Meet problem when cross compiler(using powerpc-linux-gnu-gcc ) FFMPEG source code In-Reply-To: References: Message-ID: thks for your reply My gcc version is (Ubuntu/Linaro 4.5.2-8ubuntu4) 4.5.2 And,My target machine is powerpc On Fri, May 20, 2011 at 12:50 AM, Andrey m0shbear Vul wrote: > > From: ?? > > Hi All, > I am a newbie for using FFMPEG lib to develop a tool. > Firstly, i downloaded the source code from git server and tried to cross > compiler(powerpc-linux-gnu-gcc ) the source code on my LINUX server. > Unfortunately, it is not smooth. > > My compiler environment: > OS: Ubuntu 11.04 > VMWare: 7.0 > > My configuration: > ./configure --enable-cross-compile --arch=powerpc > --host-cc=powerpc-linux-gnu-gcc --target-os=linux > > Error Message: > You need a compiler that supports {} in AltiVec vector declarations > > > Can someone give me some suggestion? > Thanks in advance. > _______________________________________________ > gcc version? > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From anders at flauntkit.com Fri May 20 10:26:20 2011 From: anders at flauntkit.com (Anders Gunnarsson) Date: Fri, 20 May 2011 10:26:20 +0200 Subject: [FFmpeg-user] libavfilter movie overlay Message-ID: <4DD625AC.9090709@flauntkit.com> Hi I'm trying to overlay a movie with another movie using the -vf movie and overlay filter. I'm trying the examples at http://www.ffmpeg.org/libavfilter.html#SEC51 But getting the error [movie @ 0x2ed6190] init() expected 3 arguments:'full.mp4' Error initializing filter 'movie' with args 'full.mp4' I've tried adding all [:options] to the movie command, but I'm getting the same error. This is using the libavfilter-SoC version. I've tried with the original ffmpeg too, were it also fails, only with no error messages. Have anybody gotten this to work? regards Anders From alexandre.ferrieux at orange-ftgroup.com Fri May 20 10:45:23 2011 From: alexandre.ferrieux at orange-ftgroup.com (Alexandre Ferrieux) Date: Fri, 20 May 2011 10:45:23 +0200 Subject: [FFmpeg-user] libavfilter movie overlay In-Reply-To: <4DD625AC.9090709@flauntkit.com> References: <4DD625AC.9090709@flauntkit.com> Message-ID: <4DD62A23.1080606@orange-ftgroup.com> On 20/05/2011 10:26, Anders Gunnarsson wrote: > Hi > > I'm trying to overlay a movie with another movie using the -vf movie and > overlay filter. > > I'm trying the examples at > http://www.ffmpeg.org/libavfilter.html#SEC51 > > But getting the error > [movie @ 0x2ed6190] init() expected 3 arguments:'full.mp4' > Error initializing filter 'movie' with args 'full.mp4' > > I've tried adding all [:options] to the movie command, but I'm getting > the same error. > This is using the libavfilter-SoC version. I've tried with the original > ffmpeg too, were it also fails, only with no error messages. > > Have anybody gotten this to work? My experience is with the ffmpeg trunk, where it works like a charm (especially since Stefano's recent fix of a memory leak). Please elaborate on "also fails, only with no error messages", by giving: - the exact command line (including the complete vfilter) - the full ffmpeg output. -Alex From stefano.sabatini-lala at poste.it Fri May 20 12:17:34 2011 From: stefano.sabatini-lala at poste.it (Stefano Sabatini) Date: Fri, 20 May 2011 12:17:34 +0200 Subject: [FFmpeg-user] libavfilter movie overlay In-Reply-To: <4DD625AC.9090709@flauntkit.com> References: <4DD625AC.9090709@flauntkit.com> Message-ID: <20110520101734.GA8494@geppetto> On date Friday 2011-05-20 10:26:20 +0200, Anders Gunnarsson encoded: > Hi > > I'm trying to overlay a movie with another movie using the -vf movie > and overlay filter. > > I'm trying the examples at > http://www.ffmpeg.org/libavfilter.html#SEC51 > > But getting the error > [movie @ 0x2ed6190] init() expected 3 arguments:'full.mp4' > Error initializing filter 'movie' with args 'full.mp4' > > I've tried adding all [:options] to the movie command, but I'm > getting the same error. > This is using the libavfilter-SoC version. I've tried with the > original ffmpeg too, were it also fails, only with no error > messages. > > Have anybody gotten this to work? movie source is now integrated in the main repository, and its syntax is different from that of the soc-repo movie. You're recommended to use the main repository version, the soc-repo is no more maintained/updated (which reminds me that we should drop the link). From mike.scheutzow at alcatel-lucent.com Fri May 20 14:34:55 2011 From: mike.scheutzow at alcatel-lucent.com (Mike Scheutzow) Date: Fri, 20 May 2011 08:34:55 -0400 Subject: [FFmpeg-user] Error extracting images from .mov with timecode In-Reply-To: References: Message-ID: <4DD65FEF.9060003@alcatel-lucent.com> Dario wrote: > Hi, > > We have a .mov file which has a timecode track in it, which seems to be > causing ffmpeg lots of fits when extracting jpegs out of it. I'm using the > latest windows binary found on http://ffmpeg.zeranoe.com/builds/ > > I first tried extracting the jpegs out like this: > > ffmpeg.exe -i inputmovie.mov -y Output\Image%06d.jpg > > When running the above, I get the first frame of video duplicated about 2700 > times, then I get *948* frames of the actual video. > > ... > > As you can see there are *948* frames in the video (as is confirmed by > Quicktime). In addition, the start_time in the timecode stream is > -108.695000. > > So, I then tried to pass the timecode start_time value as the -ss value like > so: > > ffmpeg.exe -i inputmovie.mov -ss 108.695 -y Output\Image%06d.jpg > > However, this only extracts *947* jpegs. I also tried passing the duration > as the -t value, but it didn't help. I am stuck either getting tons of > duplicate images or getting one frame less than the number of frames in the > video. > > I also tried the -vcodec copy option instead of specifying the -ss option, > which worked wonderfully and extracted 948 frames, except that it only seems > to work on motion jpegs and no other codecs... my utility needs to work on > any type of codec ffmpeg supports. > > So, I'm stuck. I'm wondering if this is a bug or if I'm just using ffmpeg > improperly. Any help would be appreciated. > > Thanks, > Dario Try '-vsync 0', placed before the -i option. This tells ffmpeg not to rate match (i.e do not duplicate or drop any video frames.) From what I can tell, -ss is not reliable for most of ffmpeg's codecs. Mike Scheutzow From richard at richsim900.plus.com Fri May 20 15:43:41 2011 From: richard at richsim900.plus.com (richard) Date: Fri, 20 May 2011 14:43:41 +0100 Subject: [FFmpeg-user] Problem editing m4a files with ffmpeg Message-ID: <1305899021.8150.2.camel@base-desktop> When I edit an m4a file using ffmpeg using command like this: ffmpeg -i filename.m4a -ss 00:02:34 -t 00:28:35 -acodec copy outputfile.m4a the edited file will play OK on software media players, but large m4a files won't play on my Marantz CD6003. The files timeout after 30 seconds. If I use mp4creator to optimize the edited m4a file using command like this: mp4creator -optimize filename.m4a the optimized file plays OK on the Marantz. I found out using AtomicParsley that editing an m4a file with ffmpeg moves the mdat atoms from the end of the file to near the front. This seems to be an issue for the Marantz. mp4creator moves the mdat atoms back to the end of the file again. Is this a bug in ffmpeg? You kindly fixed another issue with the Marantz here:- http://ffmpeg.org/pipermail/ffmpeg-user/2011-April/000374.html From diarrhio at gmail.com Fri May 20 18:37:25 2011 From: diarrhio at gmail.com (Dario) Date: Fri, 20 May 2011 09:37:25 -0700 Subject: [FFmpeg-user] Error extracting images from .mov with timecode In-Reply-To: <4DD65FEF.9060003@alcatel-lucent.com> References: <4DD65FEF.9060003@alcatel-lucent.com> Message-ID: That did the trick. Thanks! On Fri, May 20, 2011 at 5:34 AM, Mike Scheutzow < mike.scheutzow at alcatel-lucent.com> wrote: > > Try '-vsync 0', placed before the -i option. This tells ffmpeg not to rate > match (i.e do not duplicate or drop any video frames.) > > From what I can tell, -ss is not reliable for most of ffmpeg's codecs. > > Mike Scheutzow > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > From pbasista at hotmail.com Sat May 21 02:52:06 2011 From: pbasista at hotmail.com (Peter Basista) Date: Sat, 21 May 2011 00:52:06 +0000 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy Message-ID: Hello everyone! I would like ask for an advice. I have some mpegts files captured from DVB card. They contain some video streams and some audio streams. I know how to extract the streams from the mpegts file and how to use them to create a mkv file, for example. I do it like this:ffmpeg -i DVBchannel.mpegts -vcodec copy -acodec copy movie.mkv But I have a problem with the resulting file. The original video streams (in file DVBchannel.mpegts) are interlaced at 50fps. I want the final file to play as simply as possible (e.g. without the deinterlace filters), so I would like to deinterlace the video in advance. But I assume it is not possible with -vcodec copy. Or am I mistaken? If it would be achievable, it would probably only work with those video codecs, for which it is possible to somehow easily combine the two adjacent video frames into one. Anyway, is there such a codec? My point here is that I do not want to reencode the video "just" to deinterlace it. But as far as I know it is necessary in order to use the deinterlacing filters. Am I correct? To my knowledge, the deinterlace filters just somehow combine the frames together, but they do not change the framerate. And here, I _want_ the framerate to be changed from 50fps to 25fps. So, basically, I _want_ half of the frames to be deleted. But, I want the resulting frames to be complete, not just odd or even lines. So, is it possible with -vcodec copy? And my second question is about the framerate itself. Let's say I have managed to get a 25fps noninterlaced file movie.mkv. Is there any "easy" way to change the framerate to 24fps? By "easy" I mean without reencoding (or: with -vcodec copy). This, in my opinion, should be possible, because it seems like a rather minor change. Here I would like to emphasize that what I want to change is the _framerate_ not the number of frames. So, I expect the video to be longer (take more time to play) after such a framerate change. Here: http://www.hdslr-cinema.com/news/workflow/convert-between-framerates/the author mentions that such a framerate change is referred to as "conforming". I tried the intermediate rawvideo format method described on the mentioned website, but it is not a solution for me, because it requires reencoding. I would really like to do it with -vcodec copy. Is is possible? It it is, which codecs allow for framerate changes like this? Thank you for your time. Peter Basista From mark at mdsh.com Sat May 21 11:42:34 2011 From: mark at mdsh.com (Mark Himsley) Date: Sat, 21 May 2011 10:42:34 +0100 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: References: Message-ID: <4DD7890A.8050102@mdsh.com> On 21/05/2011 01:52, Peter Basista wrote: > > Hello everyone! > I would like ask for an advice. > I have some mpegts files captured from DVB card. They contain some video streams and some audio streams. > I know how to extract the streams from the mpegts file and how to use them to create a mkv file, for example. I do it like this:ffmpeg -i DVBchannel.mpegts -vcodec copy -acodec copy movie.mkv > But I have a problem with the resulting file. The original video streams (in file DVBchannel.mpegts) are interlaced at 50fps. No, I very much doubt that is true. There are no DVB broadcasters that I know of which transmit at 100 fields per second, which is what you are implying. Please supply the output of "ffmpeg -i DVBchannel.mpegts" > I want the final file to play as simply as possible (e.g. without the deinterlace filters), so I would like to deinterlace the video in advance. But I assume it is not possible with -vcodec copy. Or am I mistaken? > If it would be achievable, it would probably only work with those video codecs, for which it is possible to somehow easily combine the two adjacent video frames into one. Anyway, is there such a codec? > My point here is that I do not want to reencode the video "just" to deinterlace it. But as far as I know it is necessary in order to use the deinterlacing filters. Am I correct? > To my knowledge, the deinterlace filters just somehow combine the frames together, but they do not change the framerate. And here, I _want_ the framerate to be changed from 50fps to 25fps. So, basically, I _want_ half of the frames to be deleted. But, I want the resulting frames to be complete, not just odd or even lines. So, is it possible with -vcodec copy? > And my second question is about the framerate itself. > Let's say I have managed to get a 25fps noninterlaced file movie.mkv. Is there any "easy" way to change the framerate to 24fps? By "easy" I mean without reencoding (or: with -vcodec copy). > This, in my opinion, should be possible, because it seems like a rather minor change. Here I would like to emphasize that what I want to change is the _framerate_ not the number of frames. So, I expect the video to be longer (take more time to play) after such a framerate change. > Here: http://www.hdslr-cinema.com/news/workflow/convert-between-framerates/the author mentions that such a framerate change is referred to as "conforming". > I tried the intermediate rawvideo format method described on the mentioned website, but it is not a solution for me, because it requires reencoding. I would really like to do it with -vcodec copy. Is is possible? It it is, which codecs allow for framerate changes like this? > Thank you for your time. > Peter Basista > > > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user From bourgon.a at gmail.com Sat May 21 13:27:24 2011 From: bourgon.a at gmail.com (Armel Bourgon-Drouot) Date: Sat, 21 May 2011 13:27:24 +0200 Subject: [FFmpeg-user] mpeg-ts to mkv without sound delay Message-ID: Hi everyone, I wood like to change the container of some videos I record with my TV, and compress it for archiving purpose. The container is mpeg-ts, the video codec is h264 and the audio one is mp2. I wood like mkv, h264, aac. I also want the video stream to be more compress. Here is the result of ffpeg -i test.ts : Input #0, mpegts, from 'test.ts': Duration: 01:55:45.60, start: 1.400000, bitrate: 5050 kb/s Program 1 Service01 Metadata: name : Service01 provider_name : FFmpeg Stream #0.0[0x100]: Video: h264, yuv420p, 1440x1080 [PAR 4:3 DAR 16:9], 25 fps, 25 tbr, 90k tbn, 50 tbc Stream #0.1[0x101](fre): Audio: mp2, 48000 Hz, 2 channels, s16, 64 kb/s At least one output file must be specified File size : 4.1 GB So the first step is to change the container, I try : ffmpeg -i test.ts -vcodec copy -acodec copy movie.mkv The process run normally except this message : [NULL @ 0x645bb0]missing picture in access unit The output file is fine but there is a delay between sound and video. So my questions are : 1. How to avoid this delay ? 2. How to compress more the video stream ? Maybe try with a different bitrate ? Is there a way to automatically choose a suitable bitrate? 3. To change the audio codec -acodec aac will be ok ? Thank you for your time -- Armel From pbasista at hotmail.com Sat May 21 14:54:09 2011 From: pbasista at hotmail.com (Peter Basista) Date: Sat, 21 May 2011 12:54:09 +0000 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: References: Message-ID: >> Hello everyone!>>>> I ?would like ask for an advice.>>>> I have some mpegts files captured from DVB card.>> They contain some video streams and some audio streams.>> I know how to extract the streams from the mpegts file>> and how to use them to create a mkv file, for example.>> I do it like this:>> ffmpeg -i DVBchannel.mpegts -vcodec copy -acodec copy movie.mkv>>>> But I have a problem with the resulting file. The original>> video streams (in file DVBchannel.mpegts) are interlaced at 50fps. > No, I very much doubt that is true. There are no DVB broadcasters that I> know of which transmit at 100 fields per second, which is what you are> implying. Please supply the output of "ffmpeg -i DVBchannel.mpegts" Okay, I am sorry, you are right. I have posted somethingthat is definitely confusing. I should have written 50 fields per second, not frames per second.By "interlaced at 50fps" I was trying to tell that the video contains50 fields (or half frames) per second. But I did not know aboutthe difference between frames and fields before you replied.So, thank you for pointing that out. You wanted me to post output from ffmpeg -i DVBchannel.mpegts, so here it is: Seems stream 0 codec frame rate differs from container frame rate: 50.00 (50/1) -> 50.00 (50/1)Input #0, mpegts, from 'DVBchannel.mpegts':? Duration: 01:23:55.58, start: 54068.412678, bitrate: 5916 kb/s? Program 1?? ? Stream #0.0[0x44d]: Video: h264 (Main), yuv420p, 1440x1080 [PAR 4:3 DAR 16:9], 55.22 fps, 50 tbr, 90k tbn, 50 tbc? ? Stream #0.1[0x44e](eng): Audio: mp2, 48000 Hz, stereo, s16, 192 kb/s I have reformulated the remaining part of my original email, so I repost it. I want the final file to play as simply as possible(e.g. without the deinterlace filters), so I would like to deinterlacethe video in advance. But I assume it is not possible with -vcodec copy.Or am I mistaken? My point here is that I do not want to reencode the video "just"to deinterlace it. I want to do it as simply as possible, utilizingas much previous encoding work as possible. But as far as I know,it is necessary to completely reencode the video in order touse the deinterlacing filters. Am I correct? To my knowledge, the deinterlace filters just somehowcombine the fields (half frames) together, but they do not changethe framerate. So, my first question is: Is there a video codecfor which it is possible to deinterlace video with -vcodec copy?I would really appreciate if it would be possible for h264. And my second question is about the framerate itself. Let's say I have managed to get a 25fps noninterlaced file movie.mkv.Is there any "easy" way to change the framerate to 24fps?By "easy" I mean without reencoding (or: with -vcodec copy). This, in my opinion, should be possible, because it seems likea rather minor change. Here I would like to emphasize thatwhat I want to change is the framerate, not the number of frames. So, I expect the video to be longer (take more time to play)after such a framerate change. Here:http://www.hdslr-cinema.com/news/workflow/convert-between-framerates/the author mentions that such a framerate changeis referred to as "conforming". I tried the intermediate rawvideo format method describedon the mentioned website, but it is of no use for me,because it requires reencoding. I would really like to do itwith -vcodec copy. Is is possible? If it is, which codecsallow for framerate changes like this? Does h264 support it? Thank you for your time. Peter Basista From pbasista at hotmail.com Sat May 21 15:10:09 2011 From: pbasista at hotmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Sat, 21 May 2011 15:10:09 +0200 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy Message-ID: I am sorry for the previous post. I did not know why, but Hotmail just mercilessly discards all my newlines. I will try to repost correctly. Peter Basista From pbasista at gmail.com Sat May 21 15:24:13 2011 From: pbasista at gmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Sat, 21 May 2011 15:24:13 +0200 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: References: Message-ID: >> Hello everyone! >> >> I ?would like ask for an advice. >> >> I have some mpegts files captured from DVB card. >> They contain some video streams and some audio streams. >> I know how to extract the streams from the mpegts file >> and how to use them to create a mkv file, for example. >> I do it like this: >> ffmpeg -i DVBchannel.mpegts -vcodec copy -acodec copy movie.mkv >> >> But I have a problem with the resulting file. The original >> video streams (in file DVBchannel.mpegts) are interlaced at 50fps. > No, I very much doubt that is true. There are no DVB broadcasters that I > know of which transmit at 100 fields per second, which is what you are > implying. Please supply the output of "ffmpeg -i DVBchannel.mpegts" Okay, I am sorry, you are right. I have posted something that is definitely confusing. I should have written 50 fields per second, not frames per second. By "interlaced at 50fps" I was trying to tell that the video contains 50 fields (or half frames) per second. But I did not know about the difference between frames and fields before you replied. So, thank you for pointing that out. You wanted me to post output from ffmpeg -i DVBchannel.mpegts, so here it is: Seems stream 0 codec frame rate differs from container frame rate: 50.00 (50/1) -> 50.00 (50/1) Input #0, mpegts, from 'DVBchannel.mpegts': ?Duration: 01:23:55.58, start: 54068.412678, bitrate: 5916 kb/s ?Program 1 ? ?Stream #0.0[0x44d]: Video: h264 (Main), yuv420p, 1440x1080 [PAR 4:3 DAR 16:9], 55.22 fps, 50 tbr, 90k tbn, 50 tbc ? ?Stream #0.1[0x44e](eng): Audio: mp2, 48000 Hz, stereo, s16, 192 kb/s I have reformulated the remaining part of my original email, so I repost it. I want the final file to play as simply as possible (e.g. without the deinterlace filters), so I would like to deinterlace the video in advance. But I assume it is not possible with -vcodec copy. Or am I mistaken? My point here is that I do not want to reencode the video "just" to deinterlace it. I want to do it as simply as possible, utilizing as much previous encoding work as possible. But as far as I know, it is necessary to completely reencode the video in order to use the deinterlace filters. Am I correct? To my knowledge, the deinterlace filters just somehow combine the fields (half frames) together, but they do not change the framerate. So, my first question is: Is there a video codec for which it is possible to deinterlace video with -vcodec copy? I would really appreciate if it would be possible for h264. And my second question is about the framerate itself. Let's say I have managed to get a 25fps noninterlaced file movie.mkv. Is there any "easy" way to change the framerate to 24fps? By "easy" I mean without reencoding (or: with -vcodec copy). This, in my opinion, should be possible, because it seems like a rather minor change. Here I would like to emphasize that what I want to change is the framerate, not the number of frames. So, I expect the video to be longer (take more time to play) after such a framerate change. Here: http://www.hdslr-cinema.com/news/workflow/convert-between-framerates/ the author mentions that such a framerate change is referred to as "conforming". I tried the intermediate rawvideo format method described on the mentioned website, but it is of no use for me, because it requires reencoding. I would really like to do it with -vcodec copy. Is is possible? If it is, which codecs allow for framerate changes like this? Does h264 support it? Thank you for your time. Peter Basista From pbasista at gmail.com Sat May 21 17:07:11 2011 From: pbasista at gmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Sat, 21 May 2011 17:07:11 +0200 Subject: [FFmpeg-user] mpeg-ts to mkv without sound delay Message-ID: > Hi everyone, > > I wood like to change the container of some videos I record with my TV, and > compress it for archiving purpose. > The container is mpeg-ts, the video codec is h264 and the audio one is mp2. > I wood like mkv, h264, aac. I also want the video stream to be more > compress. We seem to have similar goals :) > So the first step is to change the container, I try : > > ffmpeg -i test.ts -vcodec copy -acodec copy movie.mkv > > The process run normally except this message : > > [NULL @ 0x645bb0]missing picture in access unit > > The output file is fine but there is a delay between sound and video. Can you be more specific? If I do the same thing with my captured mpeg-ts file, I have no delay at all. Only for the first a few seconds after the playback starts, there is an audio delay, but it is continuously fading away and after about 10 seconds, everything is in sync. A piece from mplayer's status line: A: 1.9 V: 3.4 A-V: -1.534 That is, I suppose, mplayer's issue. It is just not able to sync audio and video from this file immediately after the start of the playback. But then it all plays in sync. I don't think the reason is that the audio and video are not synchronized in the resulting file. > So my questions are : > > 1. How to avoid this delay ? Explain in more detail what kind of delay you experience. > 2. How to compress more the video stream ? Maybe try with a > different bitrate ? Is there a way to automatically choose a suitable > bitrate? I do not have any bright ideas about that. However, since the source is h264, I would try very hard to avoid reencoding, unless it is absolutely necessary. I don't know how much you care about the resulting file size, but I doubt that you can achieve significant file size reduction (say, 50%) while maintaining the video quality. On the other hand, I may be wrong. I am no expert in this field. Another question is whether it is worth to reencode the video if you know that it will reduce its size only by, say 20%. Would you sitll want to do it? Most of the time, I would not. > 3. To change the audio codec -acodec aac will be ok ? This will only work if you add -strict experimental to your command line. I do not have any experience with that, but you can also use the non experimental codec libfaac. Peter Basista From armel.bourgon-drouot at esial.net Sat May 21 17:35:16 2011 From: armel.bourgon-drouot at esial.net (Armel Bourgon-Drouot) Date: Sat, 21 May 2011 17:35:16 +0200 Subject: [FFmpeg-user] mpeg-ts to mkv without sound delay In-Reply-To: References: Message-ID: Hi, thanks a lot for your reply Peter. >Explain in more detail what kind of delay you experience. I think video and audio streams are not synchronized at all. The delay is huge (about 2s) and it's present during the whole movie. Maybe "delay" is not the correct word, sorry for my poor English. I read somewhere this append because some errors in the input file (due to transmission) lead to frame deletion while audio stream is not modified. >I do not have any bright ideas about that. However, since the source >is h264, I would try very hard to avoid reencoding, unless it is >absolutely necessary. I don't know how much you care about the >resulting file size, but I doubt that you can achieve significant file >size reduction (say, 50%) while maintaining the video quality. On the >other hand, I may be wrong. I am no expert in this field. >Another question is whether it is worth to reencode the video if you >know that it will reduce its size only by, say 20%. Would you sitll >want to do it? Most of the time, I would not. I expect a significant file size reduction, at least 50%. I know it will be probably time consuming but it's not a problem. On Sat, May 21, 2011 at 5:07 PM, Peter Ba?ista wrote: > > Hi everyone, > > > > I wood like to change the container of some videos I record with my TV, > and > > compress it for archiving purpose. > > The container is mpeg-ts, the video codec is h264 and the audio one is > mp2. > > I wood like mkv, h264, aac. I also want the video stream to be more > > compress. > > We seem to have similar goals :) > > > So the first step is to change the container, I try : > > > > ffmpeg -i test.ts -vcodec copy -acodec copy movie.mkv > > > > The process run normally except this message : > > > > [NULL @ 0x645bb0]missing picture in access unit > > > > The output file is fine but there is a delay between sound and video. > > Can you be more specific? If I do the same thing with my captured > mpeg-ts file, I have no delay at all. Only for the first a few seconds > after the playback starts, there is an audio delay, but it is > continuously fading away and after about 10 seconds, everything is in > sync. > > A piece from mplayer's status line: > A: 1.9 V: 3.4 A-V: -1.534 > > That is, I suppose, mplayer's issue. It is just not able to sync audio > and video from this file immediately after the start of the playback. > But then it all plays in sync. I don't think the reason is that the > audio and video are not synchronized in the resulting file. > > > So my questions are : > > > > 1. How to avoid this delay ? > > Explain in more detail what kind of delay you experience. > > > 2. How to compress more the video stream ? Maybe try with a > > different bitrate ? Is there a way to automatically choose a suitable > > bitrate? > > I do not have any bright ideas about that. However, since the source > is h264, I would try very hard to avoid reencoding, unless it is > absolutely necessary. I don't know how much you care about the > resulting file size, but I doubt that you can achieve significant file > size reduction (say, 50%) while maintaining the video quality. On the > other hand, I may be wrong. I am no expert in this field. > > Another question is whether it is worth to reencode the video if you > know that it will reduce its size only by, say 20%. Would you sitll > want to do it? Most of the time, I would not. > > > 3. To change the audio codec -acodec aac will be ok ? > > This will only work if you add -strict experimental to your command > line. I do not have any experience with that, but you can also use the > non experimental codec libfaac. > > Peter Basista > _______________________________________________ > ffmpeg-user mailing list > ffmpeg-user at ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > -- Armel Bourgon-Drouot ?l?ve-ing?nieur Esial 2A IL From mark at mdsh.com Sat May 21 18:15:03 2011 From: mark at mdsh.com (Mark Himsley) Date: Sat, 21 May 2011 17:15:03 +0100 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: References: Message-ID: <4DD7E507.2090806@mdsh.com> On 21/05/2011 14:24, Peter Ba?ista wrote: >>> Hello everyone! >>> >>> I would like ask for an advice. >>> >>> I have some mpegts files captured from DVB card. >>> They contain some video streams and some audio streams. >>> I know how to extract the streams from the mpegts file >>> and how to use them to create a mkv file, for example. >>> I do it like this: >>> ffmpeg -i DVBchannel.mpegts -vcodec copy -acodec copy movie.mkv >>> >>> But I have a problem with the resulting file. The original >>> video streams (in file DVBchannel.mpegts) are interlaced at 50fps. > >> No, I very much doubt that is true. There are no DVB broadcasters that I >> know of which transmit at 100 fields per second, which is what you are >> implying. Please supply the output of "ffmpeg -i DVBchannel.mpegts" > > Okay, I am sorry, you are right. I have posted something > that is definitely confusing. > > I should have written 50 fields per second, not frames per second. > By "interlaced at 50fps" I was trying to tell that the video contains > 50 fields (or half frames) per second. But I did not know about > the difference between frames and fields before you replied. > So, thank you for pointing that out. Ok. So you say you have a video that is encoded at 25i (50 fields per second)(and if anyone wants the 25i verses 50i argument then please start a new thread, I'll be happy to join in ;-). I hope you understand that each frame will contain 2 fields, that is, two picture each taken 20 milliseconds apart that are encoded onto alternating lines of the frame. And what you want to do is output 25 frames per second without throwing away any data and without re-encoding? Before I go much further, I think I should point out that that will look horrible. Turning an interlaced frames into progressive frames without doing some de-interlacing will leave you with comb artefacts on all movement. It is not something I would want to do, but I do have a vague recollection of some bit of software being able to twiddle with the flags on an mpeg2 video stream to alter the interlaced/progressive flags. But since, in general, doing so would be horrible, I've forgotten what it is. > You wanted me to post output from ffmpeg -i DVBchannel.mpegts, so here it is: > > Seems stream 0 codec frame rate differs from container frame rate: > 50.00 (50/1) -> 50.00 (50/1) > Input #0, mpegts, from 'DVBchannel.mpegts': > Duration: 01:23:55.58, start: 54068.412678, bitrate: 5916 kb/s > Program 1 > Stream #0.0[0x44d]: Video: h264 (Main), yuv420p, 1440x1080 [PAR > 4:3 DAR 16:9], 55.22 fps, 50 tbr, 90k tbn, 50 tbc > Stream #0.1[0x44e](eng): Audio: mp2, 48000 Hz, stereo, s16, 192 kb/s Interesting that FFmpeg says the file is 55.22 fps... FFmpeg generally tells you the field rate for interlaced files - but that is a strange rate. > I have reformulated the remaining part of my original email, so I repost it. > > I want the final file to play as simply as possible > (e.g. without the deinterlace filters), so I would like to deinterlace > the video in advance. But I assume it is not possible with -vcodec copy. > Or am I mistaken? > > My point here is that I do not want to reencode the video "just" > to deinterlace it. I want to do it as simply as possible, utilizing > as much previous encoding work as possible. But as far as I know, > it is necessary to completely reencode the video in order to > use the deinterlace filters. Am I correct? > > To my knowledge, the deinterlace filters just somehow > combine the fields (half frames) together, but they do not change > the framerate. So, my first question is: Is there a video codec > for which it is possible to deinterlace video with -vcodec copy? > I would really appreciate if it would be possible for h264. The yadif filter can take the 50 fields and convert them into 50 frames, which doubles the frame rate. It can also drop every other field - preserving the 25 fps. > And my second question is about the framerate itself. > > Let's say I have managed to get a 25fps noninterlaced file movie.mkv. > Is there any "easy" way to change the framerate to 24fps? > By "easy" I mean without reencoding (or: with -vcodec copy). That's an easy answer. No. > This, in my opinion, should be possible, because it seems like > a rather minor change. Here I would like to emphasize that > what I want to change is the framerate, not the number of frames. Ok. Think about how h.264 is encoded. It contains few 'I' frames (effectively a full frame - only compressed) but it also contains 'P' frames and 'B' frames that both just encode the differences between other frames and this frame. If you just simply throw away frames, as your opinion suggests should be possible, then you are just throwing away frames that OTHER frames REQUIRE in order to be able to create them. So, no, you cannot just throw away frames in a h.264 stream (unless your h.264 stream is 'I' frame only, which yours will not be). > So, I expect the video to be longer (take more time to play) > after such a framerate change. Here: > http://www.hdslr-cinema.com/news/workflow/convert-between-framerates/ > the author mentions that such a framerate change > is referred to as "conforming". > > I tried the intermediate rawvideo format method described > on the mentioned website, but it is of no use for me, > because it requires reencoding. I would really like to do it > with -vcodec copy. Is is possible? If it is, which codecs > allow for framerate changes like this? Does h264 support it? No. To change frame rate you have to decode, throw away 'baseband' 'I' frames, and re-encode. > Thank you for your time. > > Peter Basista From pbasista at gmail.com Sat May 21 20:05:12 2011 From: pbasista at gmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Sat, 21 May 2011 20:05:12 +0200 Subject: [FFmpeg-user] mpeg-ts to mkv without sound delay In-Reply-To: References: Message-ID: > Hi, > > thanks a lot for your reply Peter. > >>Explain in more detail what kind of delay you experience. > > I think video and audio streams are not synchronized at all. The delay is > huge (about 2s) and it's present during the whole movie. Maybe "delay" is > not the correct word, sorry for my poor English. You can use that word. But you should also say what is delayed (audio or video). > I read somewhere this append because some errors in the input file (due to > transmission) lead to frame deletion while audio stream is not modified. That sounds reasonable, but I don' have any experience with fixing this kind of problems. However, if your video and audio streams differ by a constant time throughout the whole file, I believe you can use simple -itsoffset option to shift the audio stream appropriately. >>I do not have any bright ideas about that. However, since the source >>is h264, I would try very hard to avoid reencoding, unless it is >>absolutely necessary. I don't know how much you care about the >>resulting file size, but I doubt that you can achieve significant file >>size reduction (say, 50%) while maintaining the video quality. On the >>other hand, I may be wrong. I am no expert in this field. > >>Another question is whether it is worth to reencode the video if you >>know that it will reduce its size only by, say 20%. Would you sitll >>want to do it? Most of the time, I would not. > > I expect a significant file size reduction, at least 50%. I know it will be > probably time consuming but it's not a problem. Okay, then I can suggest you look up some advanced libx264 presets and options. I too wish I knew how to set up libx264 to produce small and quality videos. It would be nice to have something like a competitive libx264 presets database. Maybe it already exists, but I am not aware of it. Peter Basista From batguano999 at hotmail.com Sat May 21 23:22:26 2011 From: batguano999 at hotmail.com (bat guano) Date: Sat, 21 May 2011 21:22:26 +0000 Subject: [FFmpeg-user] Differences downloading RA files with FFMpeg and mplayer Message-ID: Dear all, I have noted some differences downloading ra files (via rtsp) using FFMpeg and mplayer, for example: fmpeg -i rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra output1.ra mplayer rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra -dumpstream if I listen the resulting files with FFPlay, they seems to be the same audio stream, but, if I open them with media player cassic (Real Alternative codecs), the file downloaded with FFMpeg have a bad quality. Someone knows the reason? Maybe, seeking problems, I think, but I am not sure. I use Windows XP, but I don't think it is the reason. Thanks Sandro ********************************************** Hi Sandro When you download with FFmpeg with command:- ffmpeg -i rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra output1.ra you are re-encoding the file. That's why the quality is poor. When you download with mPlayer using command:- mplayer rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra -dumpstream you are dumping the file without re-encoding. The quality is the same as original. So FFmpeg is not a good choice for this job. mPlayer is best in this case. Especially if you use '-bandwidth 999999'. Like this:- mplayer -bandwidth 999999 rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra -dumpstream -dumpfile cuoreditenebra2011_03_26.ra By using '-bandwidth 999999' the 50 minute show is downloaded in less than 2 minutes! Here is the result:- http://www.mediafire.com/?8rqr1y8bkjj5e59 I don't understand Italian, but in this show they have been talking about Led Zeppelin! From fcassia at gmail.com Sat May 21 23:39:10 2011 From: fcassia at gmail.com (Fernando Cassia) Date: Sat, 21 May 2011 18:39:10 -0300 Subject: [FFmpeg-user] Differences downloading RA files with FFMpeg and mplayer In-Reply-To: References: Message-ID: On Sat, May 21, 2011 at 18:22, bat guano wrote: > > Like this:- > mplayer -bandwidth 999999 rtsp://mm6.rai.it/radiofonia/radio3/napoli/cuoreditenebra/2011/cuoreditenebra2011_03_26.ra -dumpstream -dumpfile cuoreditenebra2011_03_26.ra Amazing that RAI still offers something as RealAudio. I thought they completely sold their souls to Microsoft?s big wallet... their video site is completely unwatcheable, thanks to its reliance on mandatory Silverlight plug-in and player. FC From pbasista at gmail.com Sun May 22 00:37:58 2011 From: pbasista at gmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Sun, 22 May 2011 00:37:58 +0200 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: <4DD7E507.2090806@mdsh.com> References: <4DD7E507.2090806@mdsh.com> Message-ID: >>> No, I very much doubt that is true. There are no DVB broadcasters that I >>> know of which transmit at 100 fields per second, which is what you are >>> implying. Please supply the output of "ffmpeg -i DVBchannel.mpegts" >> >> Okay, I am sorry, you are right. I have posted something >> that is definitely confusing. >> >> I should have written 50 fields per second, not frames per second. >> By "interlaced at 50fps" I was trying to tell that the video contains >> 50 fields (or half frames) per second. But I did not know about >> the difference between frames and fields before you replied. >> So, thank you for pointing that out. > > Ok. So you say you have a video that is encoded at 25i (50 fields per > second)(and if anyone wants the 25i verses 50i argument then please > start a new thread, I'll be happy to join in ;-). > > I hope you understand that each frame will contain 2 fields, that is, > two picture each taken 20 milliseconds apart that are encoded onto > alternating lines of the frame. Yes, that I understand. > And what you want to do is output 25 frames per second without throwing > away any data and without re-encoding? Precisely. That is what I originally wanted to do. > Before I go much further, I think I should point out that that will look > horrible. Turning an interlaced frames into progressive frames without > doing some de-interlacing will leave you with comb artefacts on all > movement. Ok, that's not good. I am sorry, now I see it would be "a little" weird to do that. I recall that my original intention to deinterlace like that came from (obviously incorrect) idea, that the two adjacent fields of an interlaced video are actually formed from a single original frame. That would mean that in 25 fps interlaced video some of the adjacent fields would be 40 ms apart and some would be 0 ms apart. I just wanted to merge the ones that are 0 ms apart into single frames. And I thought it might be posiible without (much) reencoding. Now I know it can not be done like this. There are not any adjacent fields which are 0 ms apart. That's why it would be inappropriate to simply merge them. > It is not something I would want to do, but I do have a vague > recollection of some bit of software being able to twiddle with the > flags on an mpeg2 video stream to alter the interlaced/progressive > flags. But since, in general, doing so would be horrible, I've forgotten > what it is. Now I am a little bit confused. Just to make it clear, here you are talking about video format or video codec? Mpeg-2 transport stream (mpegts) or mpeg-2 video codec (mpeg2video)? But either way, I thought that you have just been trying to point out that what makes a video interlaced or deinterlaced is the whole nature of encoding the frames, fileds, etc. And now you say something about a flag that could change a video from interlaced to deinterlaced and vice versa? Just like that? By altering flag? How is that possible? >> You wanted me to post output from ffmpeg -i DVBchannel.mpegts, so here it is: >> >> Seems stream 0 codec frame rate differs from container frame rate: >> 50.00 (50/1) -> 50.00 (50/1) >> Input #0, mpegts, from 'DVBchannel.mpegts': >> ?Duration: 01:23:55.58, start: 54068.412678, bitrate: 5916 kb/s >> ?Program 1 >> ? ?Stream #0.0[0x44d]: Video: h264 (Main), yuv420p, 1440x1080 [PAR >> 4:3 DAR 16:9], 55.22 fps, 50 tbr, 90k tbn, 50 tbc >> ? ?Stream #0.1[0x44e](eng): Audio: mp2, 48000 Hz, stereo, s16, 192 kb/s > > Interesting that FFmpeg says the file is 55.22 fps... FFmpeg generally > tells you the field rate for interlaced files - but that is a strange rate. I agree, it is strange, but that's what ffmpeg says. I have another video from the same source and the reported fps is 59.81. So, the reported value is "unstable" :) and I think it is just estimated incorrectly. As far as I know, ffmpeg estimates fps from the first a few milliseconds (or frames, Bytes, I am not sure) of the video. In mpegts recorded from DVB, I think it might be possible that the leading part of the file is recorded incorrectly due to the delay caused by overhead of starting the recording. I have also forgotten to mention that when I run ffmpeg -i DVBstream.mpegts, there is a lot of warnings like this: [h264 @ 0xce3c20] non-existing PPS referenced [h264 @ 0xce3c20] non-existing PPS 0 referenced [h264 @ 0xce3c20] decode_slice_header error [h264 @ 0xce3c20] no frame! ... [mpegts @ 0xcde6c0] max_analyze_duration 5000000 reached at 5016000 So, something is definitely wrong and that might, in my opinion, result in wrong frame rate detection. >> I have reformulated the remaining part of my original email, so I repost it. >> >> I want the final file to play as simply as possible >> (e.g. without the deinterlace filters), so I would like to deinterlace >> the video in advance. But I assume it is not possible with -vcodec copy. >> Or am I mistaken? >> >> My point here is that I do not want to reencode the video "just" >> to deinterlace it. I want to do it as simply as possible, utilizing >> as much previous encoding work as possible. But as far as I know, >> it is necessary to completely reencode the video in order to >> use the deinterlace filters. Am I correct? >> >> To my knowledge, the deinterlace filters just somehow >> combine the fields (half frames) together, but they do not change >> the framerate. So, my first question is: Is there a video codec >> for which it is possible to deinterlace video with -vcodec copy? >> I would really appreciate if it would be possible for h264. > > The yadif filter can take the 50 fields and convert them into 50 frames, > which doubles the frame rate. It can also drop every other field - > preserving the 25 fps. So, I have two options: either double the frame rate to 50 fps (double the number of frames) and dramatically increase the video size or leave the fps at 25 but lose the smoothness of 50 fields per second. Hmm ... I think I will stick to the interlaced video and, despite my original intention, keep using the real time deinterlace filter. If I use the one which doubles the video framerate, I should be happy enough. I will not lose smoothness and my video size remains the same. The only drawback would be the deinterlacing overhead. >> And my second question is about the framerate itself. >> >> Let's say I have managed to get a 25fps noninterlaced file movie.mkv. >> Is there any "easy" way to change the framerate to 24fps? >> By "easy" I mean without reencoding (or: with -vcodec copy). > > That's an easy answer. No. Oh, that's too bad! >> This, in my opinion, should be possible, because it seems like >> a rather minor change. Here I would like to emphasize that >> what I want to change is the framerate, not the number of frames. > > Ok. Think about how h.264 is encoded. It contains few 'I' frames > (effectively a full frame - only compressed) but it also contains 'P' > frames and 'B' frames that both just encode the differences between > other frames and this frame. If you just simply throw away frames, as > your opinion suggests should be possible, then you are just throwing > away frames that OTHER frames REQUIRE in order to be able to create them. All right, here we completely don't understand each other. I do not suggest that it is possible to throw away frames. By far not! I just want the frames to "remain longer on the screen" when playing :) ... For example, a frame will not to be displayed for 1/25th of a second but for 1/24th of a second. And that, in my opinion, should be pretty simple to achieve, ... I would expect that changing some video codec flag would do the trick. > So, no, you cannot just throw away frames in a h.264 stream (unless your > h.264 stream is 'I' frame only, which yours will not be). All right, we made that clear and I know I can not. >> So, I expect the video to be longer (take more time to play) >> after such a framerate change. Here: >> http://www.hdslr-cinema.com/news/workflow/convert-between-framerates/ >> the author mentions that such a framerate change >> is referred to as "conforming". >> >> I tried the intermediate rawvideo format method described >> on the mentioned website, but it is of no use for me, >> because it requires reencoding. I would really like to do it >> with -vcodec copy. Is is possible? If it is, which codecs >> allow for framerate changes like this? Does h264 support it? > > No. To change frame rate you have to decode, throw away 'baseband' 'I' > frames, and re-encode. I don't understand why. I mean, why can't I just flag the video to be played at the slower speed "by default"? It would then be very simple to change frame rate. If it is not possible in every video codec, does at least h264 support it? If not, why? Thank you for your former response and for taking part in this discussion. Peter Basista From mark at mdsh.com Sun May 22 12:41:13 2011 From: mark at mdsh.com (Mark Himsley) Date: Sun, 22 May 2011 11:41:13 +0100 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: References: <4DD7E507.2090806@mdsh.com> Message-ID: <4DD8E849.5080309@mdsh.com> On 21/05/2011 23:37, Peter Ba?ista wrote: >>>> No, I very much doubt that is true. There are no DVB broadcasters that I >>>> know of which transmit at 100 fields per second, which is what you are >>>> implying. Please supply the output of "ffmpeg -i DVBchannel.mpegts" >>> >>> Okay, I am sorry, you are right. I have posted something >>> that is definitely confusing. >>> >>> I should have written 50 fields per second, not frames per second. >>> By "interlaced at 50fps" I was trying to tell that the video contains >>> 50 fields (or half frames) per second. But I did not know about >>> the difference between frames and fields before you replied. >>> So, thank you for pointing that out. >> >> Ok. So you say you have a video that is encoded at 25i (50 fields per >> second)(and if anyone wants the 25i verses 50i argument then please >> start a new thread, I'll be happy to join in ;-). >> >> I hope you understand that each frame will contain 2 fields, that is, >> two picture each taken 20 milliseconds apart that are encoded onto >> alternating lines of the frame. > > Yes, that I understand. > >> And what you want to do is output 25 frames per second without throwing >> away any data and without re-encoding? > > Precisely. That is what I originally wanted to do. > >> Before I go much further, I think I should point out that that will look >> horrible. Turning an interlaced frames into progressive frames without >> doing some de-interlacing will leave you with comb artefacts on all >> movement. > > Ok, that's not good. I am sorry, now I see it would be "a little" > weird to do that. > > I recall that my original intention to deinterlace like that came from > (obviously incorrect) idea, that the two adjacent fields of an > interlaced video are actually formed from a single original frame. > > That would mean that in 25 fps interlaced video some of the adjacent > fields would be 40 ms apart and some would be 0 ms apart. I just > wanted to merge the ones that are 0 ms apart into single frames. And I > thought it might be posiible without (much) reencoding. That would be called Segmented Field and has a shorthand of 25PsF. Most films are broadcast as 25PsF in 50Hz territories although the MPEG frames will indicate that they are interlaced. http://en.wikipedia.org/wiki/Interlaced http://en.wikipedia.org/wiki/Progressive_segmented_frame > Now I know it can not be done like this. There are not any adjacent > fields which are 0 ms apart. That's why it would be inappropriate to > simply merge them. > >> It is not something I would want to do, but I do have a vague >> recollection of some bit of software being able to twiddle with the >> flags on an mpeg2 video stream to alter the interlaced/progressive >> flags. But since, in general, doing so would be horrible, I've forgotten >> what it is. > > Now I am a little bit confused. Just to make it clear, here you are > talking about video format or video codec? Mpeg-2 transport stream > (mpegts) or mpeg-2 video codec (mpeg2video)? Each MPEG2/4 frame includes a flag to indicate if that frame is interlaced and whether its top-field-first or bottom field first. > But either way, I thought that you have just been trying to point out > that what makes a video interlaced or deinterlaced is the whole nature > of encoding the frames, fileds, etc. And now you say something about a > flag that could change a video from interlaced to deinterlaced and > vice versa? Just like that? By altering flag? How is that possible? It is possible to have a piece of software that reads the transport stream and alters the interlaced/progressive/TFF/BFF flags on each frame. I do not know of software that can do that. >>> You wanted me to post output from ffmpeg -i DVBchannel.mpegts, so here it is: >>> >>> Seems stream 0 codec frame rate differs from container frame rate: >>> 50.00 (50/1) -> 50.00 (50/1) >>> Input #0, mpegts, from 'DVBchannel.mpegts': >>> Duration: 01:23:55.58, start: 54068.412678, bitrate: 5916 kb/s >>> Program 1 >>> Stream #0.0[0x44d]: Video: h264 (Main), yuv420p, 1440x1080 [PAR >>> 4:3 DAR 16:9], 55.22 fps, 50 tbr, 90k tbn, 50 tbc >>> Stream #0.1[0x44e](eng): Audio: mp2, 48000 Hz, stereo, s16, 192 kb/s >> >> Interesting that FFmpeg says the file is 55.22 fps... FFmpeg generally >> tells you the field rate for interlaced files - but that is a strange rate. > > I agree, it is strange, but that's what ffmpeg says. I have another > video from the same source and the reported fps is 59.81. So, the > reported value is "unstable" :) and I think it is just estimated > incorrectly. > > As far as I know, ffmpeg estimates fps from the first a few > milliseconds (or frames, Bytes, I am not sure) of the video. In mpegts > recorded from DVB, I think it might be possible that the leading part > of the file is recorded incorrectly due to the delay caused by > overhead of starting the recording. I have also forgotten to mention > that when I run ffmpeg -i DVBstream.mpegts, there is a lot of warnings > like this: > > [h264 @ 0xce3c20] non-existing PPS referenced > [h264 @ 0xce3c20] non-existing PPS 0 referenced > [h264 @ 0xce3c20] decode_slice_header error > [h264 @ 0xce3c20] no frame! > ... > [mpegts @ 0xcde6c0] max_analyze_duration 5000000 reached at 5016000 > > So, something is definitely wrong and that might, in my opinion, > result in wrong frame rate detection. > >>> I have reformulated the remaining part of my original email, so I repost it. >>> >>> I want the final file to play as simply as possible >>> (e.g. without the deinterlace filters), so I would like to deinterlace >>> the video in advance. But I assume it is not possible with -vcodec copy. >>> Or am I mistaken? >>> >>> My point here is that I do not want to reencode the video "just" >>> to deinterlace it. I want to do it as simply as possible, utilizing >>> as much previous encoding work as possible. But as far as I know, >>> it is necessary to completely reencode the video in order to >>> use the deinterlace filters. Am I correct? >>> >>> To my knowledge, the deinterlace filters just somehow >>> combine the fields (half frames) together, but they do not change >>> the framerate. So, my first question is: Is there a video codec >>> for which it is possible to deinterlace video with -vcodec copy? >>> I would really appreciate if it would be possible for h264. >> >> The yadif filter can take the 50 fields and convert them into 50 frames, >> which doubles the frame rate. It can also drop every other field - >> preserving the 25 fps. > > So, I have two options: either double the frame rate to 50 fps (double > the number of frames) and dramatically increase the video size or > leave the fps at 25 but lose the smoothness of 50 fields per second. > Hmm ... I think I will stick to the interlaced video and, despite my > original intention, keep using the real time deinterlace filter. > > If I use the one which doubles the video framerate, I should be happy > enough. I will not lose smoothness and my video size remains the same. > The only drawback would be the deinterlacing overhead. > >>> And my second question is about the framerate itself. >>> >>> Let's say I have managed to get a 25fps noninterlaced file movie.mkv. >>> Is there any "easy" way to change the framerate to 24fps? >>> By "easy" I mean without reencoding (or: with -vcodec copy). >> >> That's an easy answer. No. > > Oh, that's too bad! > >>> This, in my opinion, should be possible, because it seems like >>> a rather minor change. Here I would like to emphasize that >>> what I want to change is the framerate, not the number of frames. >> >> Ok. Think about how h.264 is encoded. It contains few 'I' frames >> (effectively a full frame - only compressed) but it also contains 'P' >> frames and 'B' frames that both just encode the differences between >> other frames and this frame. If you just simply throw away frames, as >> your opinion suggests should be possible, then you are just throwing >> away frames that OTHER frames REQUIRE in order to be able to create them. > > All right, here we completely don't understand each other. I do not > suggest that it is possible to throw away frames. By far not! > > I just want the frames to "remain longer on the screen" when playing > :) ... For example, a frame will not to be displayed for 1/25th of a > second but for 1/24th of a second. And that, in my opinion, should be > pretty simple to achieve, ... I would expect that changing some video > codec flag would do the trick. > >> So, no, you cannot just throw away frames in a h.264 stream (unless your >> h.264 stream is 'I' frame only, which yours will not be). > > All right, we made that clear and I know I can not. Each frame of video and audio in your transport stream includes a Presentation Time Stamp. Audio frames are a different size to video frames. The PTS is used to (a) present the video/audio in sync and (b) to present the video/audio at the correct speed. It is not inconceivable for a piece of software to read the transport stream and alter the PTS of every frame to make the video run slow. Now, what happens to the audio. The audio stream is encoded at (say) 48,000 samples per second. If you have altered the PTS of the video and audio frames so that they are presented at 96% of full speed then the audio sample rate needs to be altered to 46,080 samples per second - which is not (as far as I know) something that can be done. I think your idea of messing with the file is doomed. >>> So, I expect the video to be longer (take more time to play) >>> after such a framerate change. Here: >>> http://www.hdslr-cinema.com/news/workflow/convert-between-framerates/ >>> the author mentions that such a framerate change >>> is referred to as "conforming". >>> >>> I tried the intermediate rawvideo format method described >>> on the mentioned website, but it is of no use for me, >>> because it requires reencoding. I would really like to do it >>> with -vcodec copy. Is is possible? If it is, which codecs >>> allow for framerate changes like this? Does h264 support it? >> >> No. To change frame rate you have to decode, throw away 'baseband' 'I' >> frames, and re-encode. > > I don't understand why. I mean, why can't I just flag the video to be > played at the slower speed "by default"? It would then be very simple > to change frame rate. If it is not possible in every video codec, does > at least h264 support it? If not, why? > > Thank you for your former response and for taking part in this discussion. > > Peter Basista From pbasista at gmail.com Sun May 22 14:38:49 2011 From: pbasista at gmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Sun, 22 May 2011 14:38:49 +0200 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: <4DD8E849.5080309@mdsh.com> References: <4DD7E507.2090806@mdsh.com> <4DD8E849.5080309@mdsh.com> Message-ID: Hello once again! >> That would mean that in 25 fps interlaced video some of the adjacent >> fields would be 40 ms apart and some would be 0 ms apart. I just >> wanted to merge the ones that are 0 ms apart into single frames. And I >> thought it might be posiible without (much) reencoding. > > That would be called Segmented Field and has a shorthand of 25PsF. Most > films are broadcast as 25PsF in 50Hz territories although the MPEG > frames will indicate that they are interlaced. > > http://en.wikipedia.org/wiki/Interlaced > http://en.wikipedia.org/wiki/Progressive_segmented_frame Thank you for explaining this to me. >> Now I know it can not be done like this. There are not any adjacent >> fields which are 0 ms apart. That's why it would be inappropriate to >> simply merge them. >> >>> It is not something I would want to do, but I do have a vague >>> recollection of some bit of software being able to twiddle with the >>> flags on an mpeg2 video stream to alter the interlaced/progressive >>> flags. But since, in general, doing so would be horrible, I've forgotten >>> what it is. >> >> Now I am a little bit confused. Just to make it clear, here you are >> talking about video format or video codec? Mpeg-2 transport stream >> (mpegts) or mpeg-2 video codec (mpeg2video)? > > Each MPEG2/4 frame includes a flag to indicate if that frame is > interlaced and whether its top-field-first or bottom field first. I did not know that either. Thank you again! >> But either way, I thought that you have just been trying to point out >> that what makes a video interlaced or deinterlaced is the whole nature >> of encoding the frames, fileds, etc. And now you say something about a >> flag that could change a video from interlaced to deinterlaced and >> vice versa? Just like that? By altering flag? How is that possible? > > It is possible to have a piece of software that reads the transport > stream and alters the interlaced/progressive/TFF/BFF flags on each > frame. I do not know of software that can do that. All right. But let's say that there is a software which alters these flags. What exactly would this kind of software do? Take, for example, a frame with flags: interlaced (I), TFF. How could this kind of software make a progressive (P) frame out of this interlaced frame just by altering the flags? Because that's my point. I don't see a way to merge two half fields of a frame into one full field without reencoding. I mean: the half fields are interlaced, right? The first one contains the odd lines of a frame, the second one the even lines of a frame. So, if you just put the data they contain in a single field one after another, you get a pretty messed up frame, which is not the same as the original full frame, am I correct? >>> Ok. Think about how h.264 is encoded. It contains few 'I' frames >>> (effectively a full frame - only compressed) but it also contains 'P' >>> frames and 'B' frames that both just encode the differences between >>> other frames and this frame. If you just simply throw away frames, as >>> your opinion suggests should be possible, then you are just throwing >>> away frames that OTHER frames REQUIRE in order to be able to create them. >> >> All right, here we completely don't understand each other. I do not >> suggest that it is possible to throw away frames. By far not! >> >> I just want the frames to "remain longer on the screen" when playing >> :) ... For example, a frame will not to be displayed for 1/25th of a >> second but for 1/24th of a second. And that, in my opinion, should be >> pretty simple to achieve, ... I would expect that changing some video >> codec flag would do the trick. >> >>> So, no, you cannot just throw away frames in a h.264 stream (unless your >>> h.264 stream is 'I' frame only, which yours will not be). >> >> All right, we made that clear and I know I can not. > > Each frame of video and audio in your transport stream includes a > Presentation Time Stamp. Audio frames are a different size to video > frames. The PTS is used to (a) present the video/audio in sync and (b) > to present the video/audio at the correct speed. That's another thing I did not know. Thank you! Do I understand correctly that the PTS are present in a mpeg transport stream? What happens to them if I change the video format to mkv or avi, for example? Are they still present and provide video/audio sync or these formats provide different ways of video/audio sync? > It is not inconceivable for a piece of software to read the transport > stream and alter the PTS of every frame to make the video run slow. Now that would be really awesome! I think it is exactly what I am looking for. > Now, what happens to the audio. > The audio stream is encoded at (say) 48,000 > samples per second. If you have altered the PTS of the video and audio > frames so that they are presented at 96% of full speed then the audio > sample rate needs to be altered to 46,080 samples per second - which is > not (as far as I know) something that can be done. Well, as far as I know, it is no problem to resample an audio file to arbitrary sample rate. For example, using sox: sox input.wav output.wav rate 46080 The resulting file has the same audio length but its sample rate is 46,080. And it can be played just fine using mplayer. You can also just alter the header of an input file and force the different sample rate like this: sox -r 46080 input.wav output.wav In this case, provided that the input file has the sample rate of 48,000, the resulting file will take more time to play, but its size will be exactly the same. And it still can be played just fine using mplayer. > I think your idea of messing with the file is doomed. :) I still think there is a way to do it correctly. Okay. Let's not worry about the audio for now. Let's suppose it can have arbitrary sample rate. Now, if I would like to change the PTS of the video and audio frames, how can I do that? Is it possible to do it with ffmpeg? Even if I had to do some audio reencoding afterwards, it is still much less time consuming than video reencoding. That's my point after all. I want to avoid video reencoding whenever possible. Peter Basista From bahamutzero8825 at gmail.com Sun May 22 16:25:46 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Sun, 22 May 2011 09:25:46 -0500 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: References: Message-ID: <4DD91CEA.9080504@gmail.com> On 2011.05.20 07:52 PM, Peter Basista wrote: > But I assume it is not possible with -vcodec copy. Or am I mistaken? You are not mistaken. Deinterlacing only works on uncompressed video and changes the video data (or more accurately, it creates new video data based on the input), so it is impossible to deinterlace without decoding. > To my knowledge, the deinterlace filters just somehow combine the frames together, but they do not change the framerate. It depends on the strategy used by the filter. Yadif, for example, can produce one frame for each field, doubling the framerate, or it can produce a frame for each pair of fields, leaving the framerate the same. > And here, I _want_ the framerate to be changed from 50fps to 25fps. So, basically, I _want_ half of the frames to be deleted. But, I want the resulting frames to be complete, not just odd or even lines. 2 fields = 1 frame. Each field is half of the picture. Uncompressed, 50 fields contain the same amount of data as 25 frames. > Let's say I have managed to get a 25fps noninterlaced file movie.mkv. Is there any "easy" way to change the framerate to 24fps? By "easy" I mean without reencoding (or: with -vcodec copy). You could simply drop every 25th frame or simply change the headers (I recommend changing container headers) to tell the player to play 24fps (which will make it slower, and you'll lose sync if you don't adjust the audio). If you're trying to get the original 23.976fps of a movie from a 50i source, I can't really give much advice because I don't know all the details involved in handling film content in PAL broadcast systems. > This, in my opinion, should be possible, because it seems like a rather minor change. Here I would like to emphasize that what I want to change is the _framerate_ not the number of frames. So, I expect the video to be longer (take more time to play) after such a framerate change. In this case, you should be able to change the framerate in the container headers. mkvtoolnix would be suitable for manipulating Matroska containers. From pbasista at gmail.com Sun May 22 20:43:54 2011 From: pbasista at gmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Sun, 22 May 2011 20:43:54 +0200 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: <4DD91CEA.9080504@gmail.com> References: <4DD91CEA.9080504@gmail.com> Message-ID: >> But I assume it is not possible with -vcodec copy. Or am I mistaken? > You are not mistaken. Deinterlacing only works on uncompressed video and > changes the video data (or more accurately, it creates new video data > based on the input), so it is impossible to deinterlace without decoding. Thank you for pointing that out and explaining it in more detail. Mark has already suggested the same view, so I think it is clear now. >> To my knowledge, the deinterlace filters just somehow combine the frames together, but they do not change the framerate. > It depends on the strategy used by the filter. Yadif, for example, can > produce one frame for each field, doubling the framerate, or it can > produce a frame for each pair of fields, leaving the framerate the same. We have been discussing this as well. But thank you anyway for more information. >> ?And here, I _want_ the framerate to be changed from 50fps to 25fps. So, basically, I _want_ half of the frames to be deleted. But, I want the resulting frames to be complete, not just odd or even lines. > 2 fields = 1 frame. Each field is half of the picture. Uncompressed, 50 > fields contain the same amount of data as 25 frames. Yes, now I know that. Mark has pointed that out and explained some details. It should be clear now. >> Let's say I have managed to get a 25fps noninterlaced file movie.mkv. Is there any "easy" way to change the framerate to 24fps? By "easy" I mean without reencoding (or: with -vcodec copy). > You could simply drop every 25th frame Actually, according to information from Mark, this would require reencoding. And I agree with that, because it sounds reasonable. You have to somehow fill in the "blank spot" created by a dropped frame. Let's say you delete frame number 100. Then you have to completely change the way frame 101 is computed, because it was originally computed from frame 100, which is now gone. So, you will have to encode at least the transition from the frame 99 to 100 and from 100 to 101 into one new transition (from frame 99 to frame 101)... So, I believe encoding is required, at least for the frames just after the dropped frames. But, as I mentioned, this is not what I would like to do. > or simply change the headers (I recommend changing container > headers) to tell the player to play 24fps > (which will make it slower, and you'll lose sync if you don't adjust the > audio). If you're trying to get the original 23.976fps of a movie from a > 50i source, I can't really give much advice because I don't know all the > details involved in handling film content in PAL broadcast systems. All right, that's okay. Actually you made the point. This is exactly what I am trying to achieve. Although I am not sure at the moment what should the target frame rate be (24 fps or 23.976 fps), it is not important right now. The challenge is to decrease the frame rate in general without changing the number of frames and without reencoding. >> This, in my opinion, should be possible, because it seems like a rather minor change. Here I would like to emphasize that what I want to change is the _framerate_ not the number of frames. So, I expect the video to be longer (take more time to play) after such a framerate change. > In this case, you should be able to change the framerate in the > container headers. mkvtoolnix would be suitable for manipulating > Matroska containers. Thank you for a suggestion. I will try that. But even if I succeed in changing the container's headers in a way that the video will play at 96% speed, I am not sure if I want to do it. The reason is simple. If someone later extracts the video out of this container, what would its frame rate be? I suppose it will remain unchanged at the "old" value, which means 25 fps. And this is what I want to avoid. I want the fps change to be more robust than that. I don't want the careless container changes to blow the fps change away. That's why I consider changing PTS (presentation time stamps) of the video frames as suggested by Mark as better, more stable solution. The only problem is I don't know how to do it. (yet) :) Peter Basista From bahamutzero8825 at gmail.com Mon May 23 16:30:32 2011 From: bahamutzero8825 at gmail.com (Andrew Berg) Date: Mon, 23 May 2011 09:30:32 -0500 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: References: <4DD91CEA.9080504@gmail.com> Message-ID: <4DDA6F88.5070304@gmail.com> On 2011.05.22 01:43 PM, Peter Ba?ista wrote: > >> Let's say I have managed to get a 25fps noninterlaced file movie.mkv. Is there any "easy" way to change the framerate to 24fps? By "easy" I mean without reencoding (or: with -vcodec copy). > > You could simply drop every 25th frame > > Actually, according to information from Mark, this would require > reencoding. And I agree with that, because it sounds reasonable. You > have to somehow fill in the "blank spot" created by a dropped frame. I was a little simplistic there. If nothing depends on a frame for information, that frame can be dropped (IIRC, B-frames can be discarded; otherwise all frames must be I-frames). However, especially with the short GOPs in broadcast streams, you are still likely to drop a lot of P-frames and even a few I-frames, which would result in a lot of GOPs being corrupted. As a general rule, Mark is right. > But even if I succeed in changing the container's headers in a way > that the video will play at 96% speed, I am not sure if I want to do > it. The reason is simple. If someone later extracts the video out of > this container, what would its frame rate be? Depends on the video stream. x264 writes the frame rate into the video streams it creates (this is overridden by the container, however). I don't know if this is common with other encoders. > I suppose it will remain unchanged at the "old" value, which means 25 > fps. And this is what I want to avoid. I want the fps change to be > more robust than that. I don't want the careless container changes to > blow the fps change away. > > That's why I consider changing PTS (presentation time stamps) of the > video frames as suggested by Mark as better, more stable solution. The > only problem is I don't know how to do it. (yet) :) It is possible to change such information in a video stream, but I'm not familiar with the tools. There are ways to go from 50i to 24p, but IIRC, it's not usually pretty and it will require decoding. I can't help much with the process, though, since I've never had to do it (almost all the material I work with just need to be deinterlaced, and the little that needs to be converted to 24p is originally 60i). From pbasista at gmail.com Mon May 23 19:36:13 2011 From: pbasista at gmail.com (=?UTF-8?Q?Peter_Ba=C5=A1ista?=) Date: Mon, 23 May 2011 19:36:13 +0200 Subject: [FFmpeg-user] Deinterlace and change framerate with -vcodec copy In-Reply-To: <4DDA6F88.5070304@gmail.com> References: <4DD91CEA.9080504@gmail.com> <4DDA6F88.5070304@gmail.com> Message-ID: On Mon, May 23, 2011 at 4:30 PM, Andrew Berg wrote: >> That's why I consider changing PTS (presentation time stamps) of the >> video frames as suggested by Mark as better, more stable solution. The >> only problem is I don't know how to do it. (yet) :) > It is possible to change such information in a video stream, but I'm not > familiar with the tools. There are ways to go from 50i to 24p, It would be pretty sufficient for me to change the video attributes from "25 fps, 50 fields per second, interlaced" to "24 fps, 48 fields per second, still interlaced". I have given up deinterlacing, because I have spotted this sentence in the manual page of ffmpeg: "... but deinterlacing introduces losses." which was enough for me to give it up. > but IIRC, > it's not usually pretty and it will require decoding. I can't help much > with the process, though, since I've never had to do it (almost all the > material I work with just need to be deinterlaced, and the little that > needs to be converted to 24p is originally 60i). So, I will try to summarize and I will welcome any corrections: Deinterlacing: 1) is only possible with reencoding 2) always introduces losses The way to do it with ffmpeg: ffmpeg -i input.file -vcodec