From cehoyos at ag.or.at  Sat Dec  1 00:30:02 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 30 Nov 2012 23:30:02 +0000 (UTC)
Subject: [Libav-user] Uploading large video files to bugtracker/ftp?
References: <CALCauYcF2mLXwFSuiBeSSTtGTEsQv6zbKYSkMDXRT+K=05ozMA@mail.gmail.com>
Message-ID: <loom.20121201T002438-368@post.gmane.org>

Matthew Einhorn <moiein2000 at ...> writes:

> There's an option to attach a file to the bug report, 
> but I imagine 300MB is too much.

It is not, there is no hard filesize limit.

To elaborate a bit what is written on 
http://ffmpeg.org/bugreports.html :
There were many bugreports in the past (esp. 
crashes) were 10MB is far too much, 100k were 
often enough. I remember a bug about a buggy 
first frame in a an asf stream, the user 
uploaded 30MB, one would have been more than 
enough. Otoh, if you want to report A/V sync 
issues with a HD stream, 100MB can be appropriate 
(or even not enough). So you should upload enough 
to (very easily) reproduce your problem, not ten 
times as much.

Use tools/qt-faststart to produce a mov file 
that you can cut with dd.

Carl Eugen


From secu_tech at yahoo.com  Sat Dec  1 08:43:29 2012
From: secu_tech at yahoo.com (amir amir)
Date: Fri, 30 Nov 2012 23:43:29 -0800 (PST)
Subject: [Libav-user] Request
Message-ID: <1354347809.92641.YahooMailNeo@web164506.mail.gq1.yahoo.com>

Dear Sir/Madam,

Hello,
Is there any plugin or API Ref or any vendor that enables the FFMPEG run on GPU in order to make use of the GPU acceleration?

Best Regards,

Reza Hosseini
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121130/9c0cb582/attachment.html>

From cehoyos at ag.or.at  Sat Dec  1 11:59:29 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sat, 1 Dec 2012 10:59:29 +0000 (UTC)
Subject: [Libav-user] Request
References: <1354347809.92641.YahooMailNeo@web164506.mail.gq1.yahoo.com>
Message-ID: <loom.20121201T115637-994@post.gmane.org>

amir amir <secu_tech at ...> writes:

> Is there any plugin or API Ref or any vendor that enables 
> the FFMPEG run on GPU in order to make use of the GPU 
> acceleration?

FFmpeg contains code to support VDPAU, VA-API, DXVA2 and vda, 
see MPlayer and vlc for examples on how to use them.

libstagefright is supported for h264.

Carl Eugen


From secu_tech at yahoo.com  Sat Dec  1 12:15:13 2012
From: secu_tech at yahoo.com (amir amir)
Date: Sat, 1 Dec 2012 03:15:13 -0800 (PST)
Subject: [Libav-user] Request
In-Reply-To: <loom.20121201T115637-994@post.gmane.org>
References: <1354347809.92641.YahooMailNeo@web164506.mail.gq1.yahoo.com>
	<loom.20121201T115637-994@post.gmane.org>
Message-ID: <1354360513.28198.YahooMailNeo@web164505.mail.gq1.yahoo.com>

Dear Carl,

I am really new to FFMPEG. it is a very comprehensive project.
could you please give me links of Mplayer and VLC and examples how to use the GPU-acceleration?

Many thanks


________________________________
 From: Carl Eugen Hoyos <cehoyos at ag.or.at>
To: libav-user at ffmpeg.org 
Sent: Saturday, December 1, 2012 2:29 PM
Subject: Re: [Libav-user] Request
 
amir amir <secu_tech at ...> writes:

> Is there any plugin or API Ref or any vendor that enables 
> the FFMPEG run on GPU in order to make use of the GPU 
> acceleration?

FFmpeg contains code to support VDPAU, VA-API, DXVA2 and vda, 
see MPlayer and vlc for examples on how to use them.

libstagefright is supported for h264.

Carl Eugen

_______________________________________________
Libav-user mailing list
Libav-user at ffmpeg.org
http://ffmpeg.org/mailman/listinfo/libav-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121201/c4e5e49d/attachment.html>

From cehoyos at ag.or.at  Sat Dec  1 15:52:51 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sat, 1 Dec 2012 14:52:51 +0000 (UTC)
Subject: [Libav-user] Request
References: <1354347809.92641.YahooMailNeo@web164506.mail.gq1.yahoo.com>
	<loom.20121201T115637-994@post.gmane.org>
	<1354360513.28198.YahooMailNeo@web164505.mail.gq1.yahoo.com>
Message-ID: <loom.20121201T155119-618@post.gmane.org>

amir amir <secu_tech at ...> writes:

> I am really new to FFMPEG. it is a very comprehensive project.
> could you please give me links of Mplayer and VLC

www.mplayerhq.hu
www.videolan.org

> and examples how to use the GPU-acceleration?

For vdpau, look at libvo/vo_vdpau.c in MPlayer, 
I don't know where you find the relevant code 
in vlc.

Carl Eugen


From igorm at stats.com  Sat Dec  1 00:41:15 2012
From: igorm at stats.com (Morduhaev, Igor (igorm@stats.com))
Date: Fri, 30 Nov 2012 23:41:15 +0000
Subject: [Libav-user] How to create video using avcodec from OpenCV::Mat?
Message-ID: <F43D3702A40CBB40A1902FC1AD899909320328@SEAVER.stats.local>

I get colored jpeg images of OpenCV::Mat type and I create from them video using avcodec. The video that I get is upside-down, black & white and each row of each frame is shifted and I got diagonal line. What could be the reason for such output?
Follow this link (https://picasaweb.google.com/lh/photo/ZNQeoPw8YFBa-iFpzuYi4dMTjNZETYmyPJy0liipFm0?feat=directlink) to watch the video I get using  avcodec.
I'm using acpicture_fill to create avFrame from cv::Mat frame!

P.S.
Each cv::Mat cvFrame has  width=810, height=610, step=2432
I noticed that avFrame (that is filled by acpicture_fill) has linesize[0]=2430
I tried manually setting avFrame->linesizep0]=2432 and not 2430 but it still didn't helped.


======== CODE =========================================================

AVCodec *encoder = avcodec_find_encoder(AV_CODEC_ID_H264);
AVStream *outStream = avformat_new_stream(outContainer, encoder);
avcodec_get_context_defaults3(outStream->codec, encoder);

outStream->codec->pix_fmt = AV_PIX_FMT_YUV420P;
outStream->codec->width = 810;
outStream->codec->height = 610;
//...

SwsContext *swsCtx = sws_getContext(outStream->codec->width, outStream->codec->height, PIX_FMT_RGB24,
       outStream->codec->width, outStream->codec->height,  outStream->codec->pix_fmt, SWS_BICUBIC, NULL, NULL, NULL);

for (uint i=0; i < frameNums; i++)
{
// get frame at location I using OpenCV
       cv::Mat cvFrame;
       myReader.getFrame(cvFrame, i);
       cv::Size frameSize = cvFrame.size();
//Each cv::Mat cvFrame has  width=810, height=610, step=2432


       // create AVPicture from cv::Mat frame
       avpicture_fill((AVPicture*)avFrame, cvFrame.data, PIX_FMT_BGR24, outStream->codec->width, outStream->codec->height);
avFrame->width = frameSize.width;
       avFrame->height = frameSize.height;

// rescale to outStream format
       sws_scale(swsCtx, avFrame->data, avFrame->linesize, 0, outStream->codec->height, avFrameRescaledFrame->data, avFrameRescaledFrame ->linesize);
encoderRescaledFrame->pts=i;
avFrameRescaledFrame->width = frameSize.width;
       avFrameRescaledFrame->height = frameSize.height;

av_init_packet(&avEncodedPacket);
       avEncodedPacket.data = NULL;
       avEncodedPacket.size = 0;

       // encode rescaled frame
       if(avcodec_encode_video2(outStream->codec, &avEncodedPacket, avFrameRescaledFrame, &got_frame) < 0) exit(1);
       if(got_frame)
       {
              if (avEncodedPacket.pts != AV_NOPTS_VALUE)
                     avEncodedPacket.pts =  av_rescale_q(avEncodedPacket.pts, outStream->codec->time_base, outStream->time_base);
              if (avEncodedPacket.dts != AV_NOPTS_VALUE)
                     avEncodedPacket.dts = av_rescale_q(avEncodedPacket.dts, outStream->codec->time_base, outStream->time_base);

              // outContainer is "mp4"
              av_write_frame(outContainer, & avEncodedPacket);

              av_free_packet(&encodedPacket);
       }
}


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121130/4c306ce7/attachment.html>

From rjvbertin at gmail.com  Sun Dec  2 17:16:09 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Sun, 2 Dec 2012 17:16:09 +0100
Subject: [Libav-user] ffmpeg default for -b:v ?
Message-ID: <0C6D0D09-7F61-419B-AF5C-6F140A28E646@gmail.com>

Hello list,

A couple of questions about the way ffmpeg handles bit rates while converting:
I'm using ffmpeg to convert a raw m4v file into a .mov container. The input file does not have any information about its frequency, size or bitrate in the header, so everything comes from the individual packets (frames). When I repackage (-vcodec copy), I obtain a .mov at 12Hz, with a bitrate of approximately 4000kb/s.

Questions:
- when encoding, say to mjpeg, what is the bitrate chosen by ffmpeg when one doesn't specify a rate explicitely?
- to obtain a .mov with approximately 4000kb/s, I need to add -b:v 1000k under Mac OS X, but -b:v 2000kb/s on win32 ... why this difference (and why don't I get a rate closer to the requested rate)?

Thanks in advance,
Ren?

From cehoyos at ag.or.at  Sun Dec  2 17:27:40 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sun, 2 Dec 2012 16:27:40 +0000 (UTC)
Subject: [Libav-user] ffmpeg default for -b:v ?
References: <0C6D0D09-7F61-419B-AF5C-6F140A28E646@gmail.com>
Message-ID: <loom.20121202T172546-293@post.gmane.org>

Ren? J.V. Bertin <rjvbertin at ...> writes:

> - when encoding, say to mjpeg, what is the bitrate chosen 
> by ffmpeg when one doesn't specify a rate explicitely?

200kb

> - to obtain a .mov with approximately 4000kb/s, I need to 
> add -b:v 1000k under Mac OS X, but -b:v 2000kb/s on win32

Sounds unlikely / command line(s) and complete, uncut 
console output missing.

> why this difference 

> (and why don't I get a rate closer to the requested rate)?

Are you using 1-pass or 2-pass encoding?

Carl Eugen


From adrozdoff at gmail.com  Mon Dec  3 08:59:58 2012
From: adrozdoff at gmail.com (hatred)
Date: Mon, 3 Dec 2012 18:59:58 +1100
Subject: [Libav-user] AVFilterGraph,
	AVFilterContext and AVFilterInOut together
Message-ID: <CAP9Vioe=JAUNZgZGsfYTSifftykfdNx+ExLPa3278GSW+jCiZg@mail.gmail.com>

Hi List!

AVFilterInOut is linked list that we can pass to
avfilter_graph_parse/avfilter_graph_parse2. More interesting use case where
we pass this list to avfilter_graph_parse(): as I understand I must fill
fields with valid values,
so I have a question:
Does filter context that stored in 'filter_ctx' field should be also
registered in FilterGraph (like with avfilter_graph_create_filter(),
avfilter_graph_add_filter())?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121203/4dba90c9/attachment.html>

From aw.wisch at googlemail.com  Mon Dec  3 15:55:16 2012
From: aw.wisch at googlemail.com (Alexander Wischnewski)
Date: Mon, 3 Dec 2012 22:55:16 +0800
Subject: [Libav-user] How to connect to an IP camera in libav
In-Reply-To: <005301cdce32$7705fc20$6511f460$@caelumspace.com>
References: <005301cdce32$7705fc20$6511f460$@caelumspace.com>
Message-ID: <CAJg_av9=D7gjzYL8Z2eMSxDyZxWHVv_5jbrpONiP1pB=TnztaQ@mail.gmail.com>

Hello Brian,

not long ago i ran into a similar problem so i have a clue what it could
be. In some cases, when it comes to receiving a live stream, you have to
tell ffmpeg which format the input should have. Attached you find some code
how to do that, i am using it for several mjpeg cameras.
AVInputFormat* avInformat = NULL;

avInformat = av_find_input_format("mjpeg");

if(avformat_open_input(&pFormatCtx, path, avInformat, NULL)!=0)
{ ...

Best regards,
Alex
Am 29.11.2012 21:07 schrieb "Brian Martin" <brian at caelumspace.com>:

> Hello,****
>
> ** **
>
> I?m trying to connect to a Foscam MJpeg IP camera (http) using libav on
> Windows and am getting an error code of -5 after ****
>
> ** **
>
> AVFormatContext *pFormatCtx=NULL;****
>
>                 int             i, videoStream;****
>
>                 AVCodecContext  *pCodecCtx;****
>
>                 AVCodec         *pCodec;****
>
>                 AVFrame         *pFrame; ****
>
>                 AVFrame         *pFrameRGB;****
>
>                 AVPacket        packet;****
>
>                 int             frameFinished;****
>
>                 int             numBytes;****
>
>                 uint8_t         *buffer;****
>
>                 struct SwsContext *img_convert_ctx;****
>
>                 int frameCount=0;****
>
> ** **
>
>                 if(!videoStreamAddress) {****
>
>                                 printf("Please provide a movie file\n");**
> **
>
>                                 return -1;****
>
>                 }****
>
>                 // Register all formats and codecs****
>
>                 av_register_all();****
>
>                 avformat_network_init();****
>
> ** **
>
>                 //Open video file****
>
>                 if(avformat_open_input(&pFormatCtx, videoStreamAddress,
> NULL, NULL)!=0)****
>
>                 {****
>
>                                 printf("Couldn't open file %s\n",
> videoStreamAddress);****
>
>                                 return -1; // Couldn't open file****
>
>                 }****
>
> ** **
>
> I tried to get the string value of this error but am getting a write
> protected memory error using the built-in av_strerror command.****
>
> Any ideas would be appreciated. Thanks!****
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121203/a80f0493/attachment.html>

From rafi.fisher at wywy.com  Mon Dec  3 16:52:57 2012
From: rafi.fisher at wywy.com (Rafi Fisher)
Date: Mon, 3 Dec 2012 17:52:57 +0200
Subject: [Libav-user] problem while using swresample library
Message-ID: <CAOr2bae_ZMtLEE3QVEj0Hk2PeC5WBWGN7v+GdS+jAZjJfHQ3aQ@mail.gmail.com>

hello

  i'm using *swresample lib*. my code is used to convert mp2 or mp4 into
wav file.
  the code fragment :
    .......

    SwrContext *swr = swr_alloc();
    ret = av_opt_set_int(swr, "in_channel_layout",  AV_CH_LAYOUT_STEREO, 0);
    ret = av_opt_set_int(swr, "out_channel_layout",  AV_CH_LAYOUT_MONO,
 0);
    ret = av_opt_set_int(swr, "in_sample_rate",     IN_SAMPLE_RATE,      0);
    ret = av_opt_set_int(swr, "out_sample_rate",   OUT_SAMPLE_RATE,     0);
    ret = av_opt_set_int(swr, "in_sample_fmt",      AV_SAMPLE_FMT_S16,   0);
    ret = av_opt_set_int(swr, "out_sample_fmt",     AV_SAMPLE_FMT_S16,   0);
    ret = swr_init(swr);

    .......

    out_samples = swr_convert(swr, output, out_samples,(const uint8_t **)
(inframe->data), in_samples);
    if out_channel_layout = AV_CH_LAYOUT_ STEREO resampling works fine

    if out_channel_layout = AV_CH_LAYOUT_MONO  & other set options remain
same it fails that is:

    the swr_convert fails that is somwhere inside the code of this function
there is a call to* exit*
    which is  problematic by itself (exit a program from av. library)
instead of returning some error code.

        help will be appreciated thanks rafi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121203/dac2672f/attachment.html>

From moiein2000 at gmail.com  Mon Dec  3 22:26:11 2012
From: moiein2000 at gmail.com (Matthew Einhorn)
Date: Mon, 3 Dec 2012 16:26:11 -0500
Subject: [Libav-user] Uploading large video files to bugtracker/ftp?
In-Reply-To: <loom.20121201T002438-368@post.gmane.org>
References: <CALCauYcF2mLXwFSuiBeSSTtGTEsQv6zbKYSkMDXRT+K=05ozMA@mail.gmail.com>
	<loom.20121201T002438-368@post.gmane.org>
Message-ID: <CALCauYdjkdMFMSxA0hkaaT8xNvRY+HHz4jB3uTB6uUxjbkPGpA@mail.gmail.com>

On Fri, Nov 30, 2012 at 6:30 PM, Carl Eugen Hoyos <cehoyos at ag.or.at> wrote:

> Matthew Einhorn <moiein2000 at ...> writes:
>
> > There's an option to attach a file to the bug report,
> > but I imagine 300MB is too much.
>
> It is not, there is no hard filesize limit.
>
> To elaborate a bit what is written on
> http://ffmpeg.org/bugreports.html :
> There were many bugreports in the past (esp.
> crashes) were 10MB is far too much, 100k were
> often enough. I remember a bug about a buggy
> first frame in a an asf stream, the user
> uploaded 30MB, one would have been more than
> enough. Otoh, if you want to report A/V sync
> issues with a HD stream, 100MB can be appropriate
> (or even not enough). So you should upload enough
> to (very easily) reproduce your problem, not ten
> times as much.
>
> Use tools/qt-faststart to produce a mov file
> that you can cut with dd.
>

qt-faststart with dd did the trick.

On a side note, I initially tried to make qt-faststart on Windows in
cygwin, however, I kept getting this error "make: *** No rule to make
target `tools/qt-faststart'.  Stop." But when I did the identical think on
ubuntu it worked perfectly. I'm not that good with scripts so I'm not
exactly sure how to debug this. But I did run configure successfully in
cygwin. But it's not that important.

Matt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121203/bc22c799/attachment.html>

From cehoyos at ag.or.at  Mon Dec  3 23:30:41 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Mon, 3 Dec 2012 22:30:41 +0000 (UTC)
Subject: [Libav-user] Uploading large video files to bugtracker/ftp?
References: <CALCauYcF2mLXwFSuiBeSSTtGTEsQv6zbKYSkMDXRT+K=05ozMA@mail.gmail.com>
	<loom.20121201T002438-368@post.gmane.org>
	<CALCauYdjkdMFMSxA0hkaaT8xNvRY+HHz4jB3uTB6uUxjbkPGpA@mail.gmail.com>
Message-ID: <loom.20121203T232949-510@post.gmane.org>

Matthew Einhorn <moiein2000 at ...> writes:

> On a side note, I initially tried to make 
> qt-faststart on Windows in cygwin, however, 
> I kept getting this error "make: *** No rule 
> to make target `tools/qt-faststart'.? Stop."

My guess is that you have to do:
$ make tools/qt-faststart.exe

Carl Eugen


From moiein2000 at gmail.com  Tue Dec  4 00:37:38 2012
From: moiein2000 at gmail.com (Matthew Einhorn)
Date: Mon, 3 Dec 2012 18:37:38 -0500
Subject: [Libav-user] Uploading large video files to bugtracker/ftp?
In-Reply-To: <loom.20121203T232949-510@post.gmane.org>
References: <CALCauYcF2mLXwFSuiBeSSTtGTEsQv6zbKYSkMDXRT+K=05ozMA@mail.gmail.com>
	<loom.20121201T002438-368@post.gmane.org>
	<CALCauYdjkdMFMSxA0hkaaT8xNvRY+HHz4jB3uTB6uUxjbkPGpA@mail.gmail.com>
	<loom.20121203T232949-510@post.gmane.org>
Message-ID: <CALCauYe3A=Uf5EWx30c-VkxpGJv1V2WJ4qjRUNm8GXWMHFcdpg@mail.gmail.com>

On Mon, Dec 3, 2012 at 5:30 PM, Carl Eugen Hoyos <cehoyos at ag.or.at> wrote:
>
> Matthew Einhorn <moiein2000 at ...> writes:
>
> > On a side note, I initially tried to make
> > qt-faststart on Windows in cygwin, however,
> > I kept getting this error "make: *** No rule
> > to make target `tools/qt-faststart'.  Stop."
>
> My guess is that you have to do:
> $ make tools/qt-faststart.exe


You're right, that did the trick.

Thanks,
Matt

From stefasab at gmail.com  Tue Dec  4 11:14:59 2012
From: stefasab at gmail.com (Stefano Sabatini)
Date: Tue, 4 Dec 2012 11:14:59 +0100
Subject: [Libav-user] problem while using swresample library
In-Reply-To: <CAOr2bae_ZMtLEE3QVEj0Hk2PeC5WBWGN7v+GdS+jAZjJfHQ3aQ@mail.gmail.com>
References: <CAOr2bae_ZMtLEE3QVEj0Hk2PeC5WBWGN7v+GdS+jAZjJfHQ3aQ@mail.gmail.com>
Message-ID: <20121204101458.GG2959@arborea>

On date Monday 2012-12-03 17:52:57 +0200, Rafi Fisher encoded:
> hello
> 
>   i'm using *swresample lib*. my code is used to convert mp2 or mp4 into
> wav file.
>   the code fragment :
>     .......
> 
>     SwrContext *swr = swr_alloc();
>     ret = av_opt_set_int(swr, "in_channel_layout",  AV_CH_LAYOUT_STEREO, 0);
>     ret = av_opt_set_int(swr, "out_channel_layout",  AV_CH_LAYOUT_MONO,
>  0);
>     ret = av_opt_set_int(swr, "in_sample_rate",     IN_SAMPLE_RATE,      0);
>     ret = av_opt_set_int(swr, "out_sample_rate",   OUT_SAMPLE_RATE,     0);
>     ret = av_opt_set_int(swr, "in_sample_fmt",      AV_SAMPLE_FMT_S16,   0);
>     ret = av_opt_set_int(swr, "out_sample_fmt",     AV_SAMPLE_FMT_S16,   0);
>     ret = swr_init(swr);
> 
>     .......
> 
>     out_samples = swr_convert(swr, output, out_samples,(const uint8_t **)
> (inframe->data), in_samples);
>     if out_channel_layout = AV_CH_LAYOUT_ STEREO resampling works fine
> 
>     if out_channel_layout = AV_CH_LAYOUT_MONO  & other set options remain
> same it fails that is:
> 

>     the swr_convert fails that is somwhere inside the code of this function
> there is a call to* exit* which is  problematic by itself (exit a program from av. library)
> instead of returning some error code.
> 

There is no call to exit() in libswresample, it seems more likely like
a crash. Make sure that you correctly allocated the samples buffers.

Also check my recently posted resampling example:
http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/155151

>         help will be appreciated thanks rafi
-- 
FFmpeg = Forgiving and Fostering Mega Puristic Explosive Genius

From stefasab at gmail.com  Tue Dec  4 11:29:50 2012
From: stefasab at gmail.com (Stefano Sabatini)
Date: Tue, 4 Dec 2012 11:29:50 +0100
Subject: [Libav-user] How to create video using avcodec from OpenCV::Mat?
In-Reply-To: <F43D3702A40CBB40A1902FC1AD899909320328@SEAVER.stats.local>
References: <F43D3702A40CBB40A1902FC1AD899909320328@SEAVER.stats.local>
Message-ID: <20121204102950.GH2959@arborea>

On date Friday 2012-11-30 23:41:15 +0000, Morduhaev, Igor (igorm at stats.com) encoded:
> I get colored jpeg images of OpenCV::Mat type and I create from them video using avcodec. The video that I get is upside-down, black & white and each row of each frame is shifted and I got diagonal line. What could be the reason for such output?
> Follow this link (https://picasaweb.google.com/lh/photo/ZNQeoPw8YFBa-iFpzuYi4dMTjNZETYmyPJy0liipFm0?feat=directlink) to watch the video I get using  avcodec.
> I'm using acpicture_fill to create avFrame from cv::Mat frame!
> 
> P.S.
> Each cv::Mat cvFrame has  width=810, height=610, step=2432
> I noticed that avFrame (that is filled by acpicture_fill) has linesize[0]=2430
> I tried manually setting avFrame->linesizep0]=2432 and not 2430 but it still didn't helped.
> 
> 
> ======== CODE =========================================================
> 
> AVCodec *encoder = avcodec_find_encoder(AV_CODEC_ID_H264);
> AVStream *outStream = avformat_new_stream(outContainer, encoder);
> avcodec_get_context_defaults3(outStream->codec, encoder);
> 
> outStream->codec->pix_fmt = AV_PIX_FMT_YUV420P;
> outStream->codec->width = 810;
> outStream->codec->height = 610;
> //...
> 
> SwsContext *swsCtx = sws_getContext(outStream->codec->width, outStream->codec->height, PIX_FMT_RGB24,
>        outStream->codec->width, outStream->codec->height,  outStream->codec->pix_fmt, SWS_BICUBIC, NULL, NULL, NULL);
> 
> for (uint i=0; i < frameNums; i++)
> {
> // get frame at location I using OpenCV
>        cv::Mat cvFrame;
>        myReader.getFrame(cvFrame, i);
>        cv::Size frameSize = cvFrame.size();
> //Each cv::Mat cvFrame has  width=810, height=610, step=2432
> 
> 

>        // create AVPicture from cv::Mat frame
>        avpicture_fill((AVPicture*)avFrame, cvFrame.data, PIX_FMT_BGR24, outStream->codec->width, outStream->codec->height);

avpicture_fill() assumes aligned data, which is not your case. Check
av_image_fill_arrays(), or simply set the linesize to the correct
value (which is 2432 if I understand correctly). Also this code is
potentially inconsistent, since you read the size from outStream
(which may be totally unrelated).

> avFrame->width = frameSize.width;
>        avFrame->height = frameSize.height;
> 
> // rescale to outStream format
>        sws_scale(swsCtx, avFrame->data, avFrame->linesize, 0, outStream->codec->height, avFrameRescaledFrame->data, avFrameRescaledFrame ->linesize);

From rafi.fisher at wywy.com  Tue Dec  4 17:29:42 2012
From: rafi.fisher at wywy.com (Rafi Fisher)
Date: Tue, 4 Dec 2012 18:29:42 +0200
Subject: [Libav-user] problem while using swresample library
In-Reply-To: <20121204101458.GG2959@arborea>
References: <CAOr2bae_ZMtLEE3QVEj0Hk2PeC5WBWGN7v+GdS+jAZjJfHQ3aQ@mail.gmail.com>
	<20121204101458.GG2959@arborea>
Message-ID: <CAOr2bafUe1mK1kjfFNL6+SBx9TK4URCbRPksKQaMDVg-yz=p+Q@mail.gmail.com>

hello

  the above  problem was solved. since i did not have the latest ffmpeg
libs version downloaded.

  still i have some problem when for example transcoding :
     from: mp4 file  48khz 16bit stereo  downsampled  to
     to :     wav file   11khz 16bit stereo
 i dont get the same wav file compared to the wav file i got by ruuning the
ffmpeg.exe utility
 my file size is smaller & the quality of the audio less good compared to
the ffmpeg.exe wav file.

 do you have an idea why i get different results ?
 because when i transcode without downsampling i get excatlly the same wav
file.

    thanks a lot rafi

On Tue, Dec 4, 2012 at 12:14 PM, Stefano Sabatini <stefasab at gmail.com>wrote:

> On date Monday 2012-12-03 17:52:57 +0200, Rafi Fisher encoded:
> > hello
> >
> >   i'm using *swresample lib*. my code is used to convert mp2 or mp4 into
> > wav file.
> >   the code fragment :
> >     .......
> >
> >     SwrContext *swr = swr_alloc();
> >     ret = av_opt_set_int(swr, "in_channel_layout",  AV_CH_LAYOUT_STEREO,
> 0);
> >     ret = av_opt_set_int(swr, "out_channel_layout",  AV_CH_LAYOUT_MONO,
> >  0);
> >     ret = av_opt_set_int(swr, "in_sample_rate",     IN_SAMPLE_RATE,
>  0);
> >     ret = av_opt_set_int(swr, "out_sample_rate",   OUT_SAMPLE_RATE,
> 0);
> >     ret = av_opt_set_int(swr, "in_sample_fmt",      AV_SAMPLE_FMT_S16,
> 0);
> >     ret = av_opt_set_int(swr, "out_sample_fmt",     AV_SAMPLE_FMT_S16,
> 0);
> >     ret = swr_init(swr);
> >
> >     .......
> >
> >     out_samples = swr_convert(swr, output, out_samples,(const uint8_t **)
> > (inframe->data), in_samples);
> >     if out_channel_layout = AV_CH_LAYOUT_ STEREO resampling works fine
> >
> >     if out_channel_layout = AV_CH_LAYOUT_MONO  & other set options remain
> > same it fails that is:
> >
>
> >     the swr_convert fails that is somwhere inside the code of this
> function
> > there is a call to* exit* which is  problematic by itself (exit a
> program from av. library)
> > instead of returning some error code.
> >
>
> There is no call to exit() in libswresample, it seems more likely like
> a crash. Make sure that you correctly allocated the samples buffers.
>
> Also check my recently posted resampling example:
> http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/155151
>
> >         help will be appreciated thanks rafi
> --
> FFmpeg = Forgiving and Fostering Mega Puristic Explosive Genius
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121204/b989ab1d/attachment.html>

From rwoods at vaytek.com  Tue Dec  4 22:05:39 2012
From: rwoods at vaytek.com (Ron Woods)
Date: Tue, 4 Dec 2012 15:05:39 -0600
Subject: [Libav-user] warbly audio playback when syncing to video
Message-ID: <27FB82F7EB0DA04C92361A01E7BA9A9301A71DF829A5@GODFATHERII.vaytek2003.local>

In a video previewer that I am writing, I am trying to synchronize the audio to the video as in ffplay.c with av_sync_type = AV_SYNC_VIDEO_MASTER.   The resulting audio seems to be in sync with the video; however, it is warbly, like it's playing through water or something.  Obviously, I am missing something in mimicking the technique from ffplay.c

In synchronize_audio() of ffplay.c, there is this line:
diff = get_audio_clock(is) - get_master_clock(is);
I am using the get_audio_clock() as in ffplay, but for the get_master_clock(), I am using a custom get_video_clock() directly, as I will always be syncing to the video.  However, I am not using the same function as in ffplay.c, because I am not handling the video stream through a PacketQueue.

My app performs video analysis on the video frames as they are decoded and then passes them to a Windows C# UI for display and optional saving to disk.   The throughput of the video analysis drives the video playback.  That is, when it is done processing a frame, then it requests the next one.   If the processing takes less time than a frame interval, then I insert a delay to ensure that the playback is close to real time for the user.  (Maybe this is not the best design?)

Therefore, I would like the audio to follow the video. The problem seems to be in getting the proper value for the video "clock".  To this end, I have been using a function as in Dranger's tutorial06.c:
double get_video_clock(VideoState *is) {
                double delta;
                delta = (av_gettime() - video_current_pts_time) / 1000000.0;
                return video_current_pts + delta;
}
where delta is amount of time since the video_current_pts was obtained.  Seems to me that this should work in my context, yet the result is not yet acceptable.  If I change the function to simply
return video_current_pts;
Then the warble is more pronounced.

How do I get the audio to properly synchronize with the video?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121204/09598a88/attachment.html>

From barikondunar at hotmail.com  Wed Dec  5 11:51:12 2012
From: barikondunar at hotmail.com (Dolevo Jay)
Date: Wed, 5 Dec 2012 10:51:12 +0000
Subject: [Libav-user] avcodec_decode_video2 problem
Message-ID: <DUB114-W1352D69A1767CCD89D796B2D1460@phx.gbl>


Hi experts,


I am having a problem when I am decoding my encoded data which 
is done by x264_encoder_encode. Basically, I get a frame and encode it 
with h264_encoder_encode. Then I send the NAL to the decoder and there I
 try to decode it with avcodec_decode_video2.


The problem is,  


lenDecoded = avcodec_decode_video2(c1, av_pic, &got_picture, &pkt);



lenDecoded returns always -1 and got_picture returns always 0
 while "no frame!" error appears on the screen. I waited some time, I 
thought that it needs some time to decode but no joy.


I will provide you all the code here and I am expecting your priceless guidance and help. 


Thank you.



//ENCODING PARAMETERS, It is currently set for zero latency and ultra fast encoding.

void SetParameters(EncoderType et, int aWidth, int aHeight, int quantizationFactor, int gopSize )

{

    ////////

    convertCtx = sws_getContext(aWidth, aHeight, PIX_FMT_RGB24, 
aWidth, aHeight, PIX_FMT_YUV420P, SWS_FAST_BILINEAR, NULL, NULL, NULL);

    x264_picture_alloc(&pic_in, X264_CSP_I420, aWidth, aHeight);    //memory allocation for being encoded data.

    srcstride = aWidth*3;                               //RGB stride is just 3*width

    ////////


    Framebuffersize=0;


    x264_param_default( ?m );


    x264_param_default_preset(?m, "veryfast", "zerolatency");       //Default preset.


    Framebuffer = (uint8_t*) malloc(MAX_FRAME_SIZE);


    mux_buffer_size = 0;

    mux_buffer = NULL;


    //Set video resolution

    param.i_width = aWidth;

    param.i_height = aHeight;


    //We don't need bframe

    param.i_bframe = 0;

    //param.i_bframe_adaptive = 0;


    //Maximum dimension of the gop

    param.i_keyint_min = 2;

    param.i_keyint_max = 5;


    param.rc.i_lookahead = 0;

    param.i_threads = 1;

    //param.i_sync_lookahead = 0;


    //More high, more compressed but loss quality

    quantization_factor = 30;


    //Setting CRF

    param.rc.i_rc_method = X264_RC_CRF;

    //Specify quality of encoding

    param.rc.f_rf_constant = quantization_factor;


    //only pass1 mode for low latency

    param.rc.b_stat_read = 0;

    param.rc.b_stat_write = 1;


    param.i_threads = 1;                                    //The number of threads that x264 will use

    param.i_width = aWidth;

    param.i_height = aHeight;

    param.i_fps_num = 5;                                    //Apparently GOP size

    param.i_fps_den = 1;

    // Intra refres:

    param.b_intra_refresh = 1;

    //Rate control:

    param.rc.i_rc_method = X264_RC_CRF;

    param.rc.f_rf_constant = 25;

    param.rc.f_rf_constant_max = 35;

    //For streaming:

    param.b_annexb = 1;

    param.b_repeat_headers = 1;

    //my params:

    param.i_log_level = X264_LOG_DEBUG;

    //AA added

    param.i_bframe_adaptive = 0;

    param.rc.i_lookahead = 0;

    param.i_bframe = 0;

    param->i_frame_reference = 1;

    param->i_scenecut_threshold = 0;

    param->b_deblocking_filter = 0;

    param->b_cabac = 0;

    param->i_bframe = 0;

    param->analyse.intra = 0;

    param->analyse.inter = 0;

    param->analyse.b_transform_8x8 = 0;

    param->analyse.i_me_method = X264_ME_DIA;

    param->analyse.i_subpel_refine = 0;

    param->rc.i_aq_mode = 0;

    param->analyse.b_mixed_references = 0;

    param->analyse.i_trellis = 0;


    x264_param_apply_profile(?m, "baseline");           //baseline profile is selected.


    //If ultrafast flag is enabled the transcoding latency is reduced,

    //the bandwith of the encoder is also duplicated, but the dimension

    //of the encoded frame increase, also by a factor of 2

    if(et == X264_ULTRAFAST)

    {

        setParamForUltrafastEncoding( ?m );

    }


    //Open the encoder

    if( ( h = x264_encoder_open( ?m ) ) == NULL )

    {

        fprintf( stderr, "x264 [error]: x264_encoder_open failed\n" );

        exit(1);

    }


    this->quantization_factor = quantizationFactor;

    //param.i_keyint_max = gopSize;

}




//ENCODING STARTS HERE

void mpStartEncodingFrame(const char ** bits, int aWidth, int aHeight, int mvGOPsize, qint32 tstamp)

{

    sws_scale(convertCtx, (uint8_t**)bits, &srcstride, 0, aHeight, pic_in.img.plane, pic_in.img.i_stride);

    pic_in.i_dts=tstamp;

    pic_in.i_type = X264_TYPE_I;

    pic_in.i_qpplus1 = 0;


    int a = Encode_frame( h , &pic_in );


    //The encoded data (Framebuffer) is here sent to the decoder side,


    free(pBits);

}



//Actual encoding,

int  Encode_frame( x264_t *h, x264_picture_t *pic )

{


    int i_nal, i_nalu_size;

    i_nalu_size = 0;


    int frame_size=x264_encoder_encode( h, &nal, &i_nal, pic, &pic_out );

    if(i_nal>1)

        int yy=1;


    if( frame_size < 0 )

    {

        fprintf( stderr, "x264 [error]: x264_encoder_encode failed\n" );

        return -1;

    }


    memcpy(Framebuffer, nal[0].p_payload, frame_size);

    Framebuffersize = frame_size;


    return Framebuffersize;

}



//DECODING

int decoderDecodeFrame(uint8_t* buffer, int buffersize)

{

    AVPacket pkt;

    int got_picture = 0, lenDecoded;

    av_init_packet(&pkt);

    AVFormatContext *a;

    AVCodecContext *c1;

    avcodec_register_all();


    av_init_packet(&pkt);


    codec = avcodec_find_decoder(CODEC_ID_H264);

    if ( !codec ) {

        fprintf(stderr, "[avcodec][Err] codec not found!\n");

         exit(1);

    }

    c1 = avcodec_alloc_context3(codec);

    c1->width = 1600;

    c1->height = 900;

    c1->codec_id = CODEC_ID_H264;


    c1->codec_type = AVMEDIA_TYPE_VIDEO;


    av_pic = avcodec_alloc_frame();


    if(codec->capabilities&CODEC_CAP_TRUNCATED)

               c1->flags|= CODEC_FLAG_TRUNCATED; /* we do not send complete

    frames */


        /* open it */

        if (avcodec_open2(c1, codec, NULL) < 0) {

            fprintf(stderr, "[avcodec][Err] could not open codec\n");

            exit(1);

        }


    pkt.data=buffer;

    pkt.size=buffersize;


    lenDecoded = avcodec_decode_video2(c1, av_pic, &got_picture, &pkt);


    if (lenDecoded < 0)

    {

           fprintf(stderr, "Error while decoding frame %d\n %d", av_pic, lenDecoded);

           exit(1);

    }


    if (got_picture)

        printf("OK");

        return 1;


    return 0;

}
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121205/4fbf690b/attachment.html>

From rafi.fisher at wywy.com  Wed Dec  5 14:51:17 2012
From: rafi.fisher at wywy.com (Rafi Fisher)
Date: Wed, 5 Dec 2012 15:51:17 +0200
Subject: [Libav-user] some problem in using swresample library
Message-ID: <CAOr2badpEPRaF7k39QGcfcDoU+bsoVhEa1W2UyxD8h8pCT9-mg@mail.gmail.com>

hello

 hello



  i have some problem when for example transcoding :
     from: mp4 file  48khz 16bit stereo  downsampled  to
     to :     wav file   11khz 16bit stereo
using swresample functions.

 i dont get the same wav file compared to the wav file i got by running the
ffmpeg.exe utility
 i use ffmpeg.exe as the standart.

 my file size is smaller & the quality of the audio less good compared to
the ffmpeg.exe wav file.

 do you have an idea why i get different results ?

 but when i transcode from mp2/mp4 to wav file without downsampling i get
excatlly the same wav file
 as i get from the ffmpeg utility

    thanks a lot rafi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121205/f2b4052a/attachment.html>

From panzhaotai at gmail.com  Wed Dec  5 07:02:41 2012
From: panzhaotai at gmail.com (zt p)
Date: Wed, 5 Dec 2012 14:02:41 +0800
Subject: [Libav-user] example code of H264 decoder using ffmpeg 1.0
Message-ID: <CADiK3X9v2Bm9Hjxkg7s2hfrj2D4EBx5fkt_gNBrq=Ft=t3RRmw@mail.gmail.com>

Dear all,

Is there any example code to decode H264 video using ffmpeg 1.0? It seems
the interfaces of ffmpeg 1.0 are different from ffmpeg 0.7.

Best regards,
Zhaotai
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121205/0a282f2b/attachment.html>

From panzhaotai at gmail.com  Wed Dec  5 14:51:05 2012
From: panzhaotai at gmail.com (zt p)
Date: Wed, 5 Dec 2012 21:51:05 +0800
Subject: [Libav-user] avcodec_decode_video2 gets wrong frame dimensions
Message-ID: <CADiK3X-VLwQ2S81KV9B_Du=LD6Z_4Z2fuoJQN2=DcXzfxBXo0Q@mail.gmail.com>

Hi,

I'm using ffmpeg1.0 (with code name of "Angle") as H264 decoder, following
the sample code in doc/examples/decoding_encoding.c. I use it to decode a
compressed H264 bit stream. The resolution is 1280x800. However, after
calling avcodec_decode_video2, coded_width = 800 and codec_height = 1280,
while width = 0 and height = 1280. It seems that the values of width and
height are exchanged.
Is it a bug?

Regards,
Zhaotai
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121205/cc0d43a4/attachment.html>

From billconan at gmail.com  Wed Dec  5 23:52:56 2012
From: billconan at gmail.com (Shi Yan)
Date: Wed, 5 Dec 2012 14:52:56 -0800
Subject: [Libav-user] unable to decode h264 video within mp4
Message-ID: <CAKO7S=JEGh9SN8Z7RgpomUANORTqxOTr+MyxEeve4XaN5=kB7Q@mail.gmail.com>

May I ask you a question regarding h264 decoding with libavcodec?

 I?m working on a launch animation for our application. The animation is a
h264 mp4 file. I?m not very familiar with video decoding, I started by
following this tutorial:

 http://blog.tomaka17.com/2012/03/libavcodeclibavformat-tutorial/



I verified the tutorial can work fine on other video formats, for example
mpeg, but can?t work on h264.

 The reason is that a call to this function:

        const auto processedLength = avcodec_decode_video2(avCodecContext,
avFrame, &isFrameAvailable, &packetToSend);

        if (processedLength < 0) {

            av_free_packet(&packet);

            qDebug() << "Error while processing the data";

            exit(0);

        }

 Always returns -1. libav prints an error "no frame!" or "nal is a large
number".

I know that there could be missing nal header that I need to tell the
decoder. But I couldn't find any articles about how to do it.

I?m stuck here.

Thank you,



This is my code:



    avcodec_init();

    av_register_all();

    av_log_set_callback(my_log_callback);

    av_log_set_level(AV_LOG_DEBUG);

    AVFormatContext *formatContext = avformat_alloc_context();



    if (av_open_input_file(&formatContext, "C:\\ani.mp4", NULL, NULL, NULL)
!= 0)

    {

        qDebug() << "can't open animation";

    }



    if (av_find_stream_info(formatContext) < 0)

    {

        qDebug() << "can't find stream info";

    }



    AVStream* videoStream;

    for (unsigned int i = 0; i < formatContext->nb_streams; ++i)

    {

        AVStream *stream = formatContext->streams[i];        // pointer to
a structure about the stream

        AVMediaType codecType = stream->codec->codec_type;   // the type of
data in this stream, notable values are AVMEDIA_TYPE_VIDEO and
AVMEDIA_TYPE_AUDIO

        CodecID codecID = stream->codec->codec_id;           // identifier
for the codec



        qDebug() <<"media type"<< codecType << "codec_id:" << codecID;

        if (codecType == AVMEDIA_TYPE_VIDEO)

        {

            videoStream = stream;

        }

    }



    AVCodec *codec = avcodec_find_decoder(videoStream->codec->codec_id);

    if (codec == NULL)

    {

        qDebug() << "can't create codec";

    }



    AVCodecContext *avCodecContext =
avcodec_alloc_context2(videoStream->codec->codec_type);



    avCodecContext->width = 1920;

    avCodecContext->height = 1080;

    //avCodecContext->flags2 |= CODEC_FLAG2_FAST;

    avCodecContext->pix_fmt =  (enum PixelFormat)29;



    if (avcodec_open(avCodecContext,codec ) < 0)

    {

        qDebug() << "can't open animation file";

    }





    qDebug() << "codec width:" << avCodecContext->width << "height:" <<
avCodecContext->height <<avCodecContext->pix_fmt ;



    AVFrame * avFrame =avcodec_alloc_frame();



    if (avFrame == NULL)

    {

        qDebug() << "can't allocation avFrame";



    }



    avcodec_get_frame_defaults(avFrame);



    // the current packet of data

    AVPacket packet;

    av_init_packet(&packet);

    packet.size = 0;

    // data in the packet of data already processed

    size_t offsetInData = 0;



    // the decoding loop, running until EOF

    while (true)

    {

        // reading a packet using libavformat

        if (offsetInData >= packet.size)

        {

            av_free_packet(&packet);

            while (true)

            {

                if (av_read_frame(formatContext, &packet) < 0)

                {

                    qDebug() << "EOF";

                    goto finished;

                }

                if (packet.stream_index == videoStream->index)

                    break;

                av_free_packet(&packet);

            }

        }



        qDebug() << "packet" << packet.size;



        // preparing the packet that we will send to libavcodec

        AVPacket packetToSend;

        packetToSend.data = packet.data + offsetInData;

        packetToSend.size = packet.size - offsetInData;



        // sending data to libavcodec

        int isFrameAvailable = 0;

        const auto processedLength = avcodec_decode_video2(avCodecContext,
avFrame, &isFrameAvailable, &packetToSend);

        if (processedLength < 0) {

            av_free_packet(&packet);

            qDebug() << "Error while processing the data";

            exit(0);

        }

        qDebug() << "processed Length:" << processedLength;



        offsetInData += processedLength;



        // processing the image if available

        if (isFrameAvailable) {

            qDebug() << "one frame is ready!";

        }



    }



finished:
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121205/63780863/attachment.html>

From drrobison at openroadsconsulting.com  Fri Dec  7 00:35:21 2012
From: drrobison at openroadsconsulting.com (David R Robison)
Date: Thu, 06 Dec 2012 18:35:21 -0500
Subject: [Libav-user] Duplicate frames?
Message-ID: <50C12BB9.6010907@openroadsconsulting.com>

I am having a problem with poor quality of H.264 video. I've turned on 
some debugging and I'm getting the following output from within 
decode_slice_header in h264.c

[8288] slice:1 T mb:0 P fix pps:0 frame:23 poc:65582/2147483647/65582 
ref:1/1 qp:24 loop:1:0:0 weight:0
[8288] slice:1 B mb:0 P fix pps:0 frame:23 poc:65582/65583/65582 ref:1/1 
qp:24 loop:1:0:0 weight:0
[8288] slice:1 T mb:0 P fix pps:0 frame:24 poc:65584/2147483647/65584 
ref:1/1 qp:24 loop:1:0:0 weight:0
[8288] slice:1 T mb:0 P fix pps:0 frame:24 poc:65584/2147483647/65584 
ref:1/1 qp:24 loop:1:0:0 weight:0
[8288] mmco: unref short failure
[8288] mmco: unref short failure

I've added the current POC at the end of POC[0] and POC[1]. Am I correct 
in assuming that the last two slices are the same slice being processed 
twice?
David

-- 

David R Robison
Open Roads Consulting, Inc.
103 Watson Road, Chesapeake, VA 23320
phone: (757) 546-3401
e-mail: drrobison at openroadsconsulting.com
web: http://openroadsconsulting.com
blog: http://therobe.blogspot.com
book: http://www.xulonpress.com/bookstore/bookdetail.php?PB_ISBN=9781597816526



This email communication (including any attachments) may contain confidential and/or privileged material intended solely for the individual or entity to which it is addressed.
If you are not the intended recipient, please delete this email immediately.


From cehoyos at ag.or.at  Fri Dec  7 00:37:16 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Thu, 6 Dec 2012 23:37:16 +0000 (UTC)
Subject: [Libav-user] Duplicate frames?
References: <50C12BB9.6010907@openroadsconsulting.com>
Message-ID: <loom.20121207T003636-826@post.gmane.org>

David R Robison <drrobison at ...> writes:

> I am having a problem with poor quality of H.264 video.

Can you reproduce the poor quality with ffmpeg (the application)?
If yes, please provide command line together with complete, uncut 
console output.

Carl Eugen


From amir.rouhi at rmit.edu.au  Fri Dec  7 00:37:50 2012
From: amir.rouhi at rmit.edu.au (Rouhi Amirhossein2)
Date: Fri, 7 Dec 2012 10:37:50 +1100
Subject: [Libav-user] Mapping functions of intra-prediction
Message-ID: <CAJAZ5paMDbRdcTuJSWQQXHuK-QNyzoWAkiwYb4UR5FU9g-NfvQ@mail.gmail.com>

Hi Guys
I need just those functions of ffmpeg, that produces intra-prediction
modes. Do you know which functions are they? i want to use them separately
in another program.
Cheers
Amir
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121207/96062c1f/attachment.html>

From phamsyquybk at gmail.com  Fri Dec  7 01:50:18 2012
From: phamsyquybk at gmail.com (Quy Pham Sy)
Date: Fri, 7 Dec 2012 09:50:18 +0900
Subject: [Libav-user] How can i suspend ffmpeg's file reading?
Message-ID: <CAEEaHNJVg=7G8d9K6ccO5nZLcYe7oCWf+SOpsi+Y7D2o6a8Kwg@mail.gmail.com>

Hi,

My ffmpeg-based player plays a mpeg-ts file that downloaded from a server.
I want my player be able to play the video file
even when it is being downloaded.

There is no problem if the download speed is faster than reading speed,
otherwise, it just stop playing.
My question is that is there anyway to suspend ffmpeg file reading to wait
for data available to proceed?

Thanks
Jack
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121207/bf8df479/attachment.html>

From jettoblack at gmail.com  Fri Dec  7 04:05:39 2012
From: jettoblack at gmail.com (Jason Livingston)
Date: Thu, 6 Dec 2012 22:05:39 -0500
Subject: [Libav-user] Options to minimize decoder CPU use
Message-ID: <CAOJT_J3yCNw2=ys1bTXxtyaYSOfVvBDvO07jcb6DihezVvcyQQ@mail.gmail.com>

Hello,
When decoding MPEG-2 or H.264 video with lavc, what decoder flags or
options can I use to decode as quickly as possible?  I have zero
concern for the quality of the decoded picture (my application does
not even need the image), just as long as no frames are dropped or
skipped.

I see AVCodecContext has some options such as skip_idct and
skip_loop_filter, which I set to AVDISCARD_ALL, but at least for
MPEG-2 the decode is not significantly faster.  Are there any other
good options to use?

Thanks!

From cehoyos at ag.or.at  Fri Dec  7 09:33:25 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 7 Dec 2012 08:33:25 +0000 (UTC)
Subject: [Libav-user] Options to minimize decoder CPU use
References: <CAOJT_J3yCNw2=ys1bTXxtyaYSOfVvBDvO07jcb6DihezVvcyQQ@mail.gmail.com>
Message-ID: <loom.20121207T091205-230@post.gmane.org>

Jason Livingston <jettoblack at ...> writes:

> When decoding MPEG-2 or H.264 video with lavc, what decoder 
> flags or options can I use to decode as quickly as possible?

Do I understand you correctly that you want to maximize CPU use?

[...]

> I see AVCodecContext has some options such as skip_idct and
> skip_loop_filter, which I set to AVDISCARD_ALL, but at least 
> for MPEG-2 the decode is not significantly faster.

Could you elaborate?
Skipping idct makes a large difference here for mpeg-2 
decoding.
Perhaps test with the cli (ffmpeg -benchmark -skip_idct) 
and post some output.

> Are there any other good options to use?

You can use "fast" if you know that the input is valid 
(not corrupt) but it does not add much additional 
speedup.

Carl Eugen


From rjvbertin at gmail.com  Fri Dec  7 11:32:15 2012
From: rjvbertin at gmail.com (=?ISO-8859-1?Q?Ren=E9_J=2EV=2E_Bertin?=)
Date: Fri, 07 Dec 2012 11:32:15 +0100
Subject: [Libav-user] How can i suspend ffmpeg's file reading?
In-Reply-To: <CAEEaHNJVg=7G8d9K6ccO5nZLcYe7oCWf+SOpsi+Y7D2o6a8Kwg@mail.gmail.com>
References: <CAEEaHNJVg=7G8d9K6ccO5nZLcYe7oCWf+SOpsi+Y7D2o6a8Kwg@mail.gmail.com>
Message-ID: <f60e6958-1ca7-47bd-b761-bbb1b7407b92@email.android.com>

Hello Jack

How do you launch ffmpeg and on what os? If you have a process id for it, you could sigstop/suspend it while you wait for data, but that too would of course halt the playing. I've never tried, but I suspect that it would however resume playing when you sigcont/release the process.
Alternatively, buffer your stream, measuring download speed and expected bitrate during playback, so that you can launch ffmpeg only when you have downloaded enough content to be reasonably sure that you won't catch up with the download head. This is more or less how QuickTime handles streaming playback, but it isn't failure proof against net congestion or other unforeseen events.

Ren?

Quy Pham Sy <phamsyquybk at gmail.com> wrote:

>Hi,
>
>My ffmpeg-based player plays a mpeg-ts file that downloaded from a
>server.
>I want my player be able to play the video file
>even when it is being downloaded.
>
>There is no problem if the download speed is faster than reading speed,
>otherwise, it just stop playing.
>My question is that is there anyway to suspend ffmpeg file reading to
>wait
>for data available to proceed?
>
>Thanks
>Jack
>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>Libav-user mailing list
>Libav-user at ffmpeg.org
>http://ffmpeg.org/mailman/listinfo/libav-user


From rjvbertin at gmail.com  Fri Dec  7 12:35:12 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Fri, 7 Dec 2012 12:35:12 +0100
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
Message-ID: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>

I discovered yesterday that it's possible to split an input movie (a raw m4v stream) into 4 separate output movies (.mov) each containing a quarter of the original image (obtained with -vf crop=w/2:h/2:x:y), using a single invocation of ffmpeg.
Is it possible to generate a single output movie with 4 tracks instead, either in that same invocation (would make most sense, I presume) or else by piping the output into a separate ffmpeg command?

Thanks in advance!
Ren?

From cehoyos at ag.or.at  Fri Dec  7 13:19:02 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 7 Dec 2012 12:19:02 +0000 (UTC)
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
Message-ID: <loom.20121207T131805-592@post.gmane.org>

Ren? J.V. Bertin <rjvbertin at ...> writes:

> I discovered yesterday that it's possible to split an input 
> movie (a raw m4v stream) into 4 separate output movies (.mov) 
> each containing a quarter of the original image (obtained with 
> -vf crop=w/2:h/2:x:y), using a single invocation of ffmpeg.
> Is it possible to generate a single output movie with 4 
> tracks instead

The complex filter allows this, please RTFM for examples.

Carl Eugen


From rjvbertin at gmail.com  Fri Dec  7 16:33:28 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Fri, 7 Dec 2012 16:33:28 +0100
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
In-Reply-To: <loom.20121207T131805-592@post.gmane.org>
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
Message-ID: <E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>

Are we really speaking about the same thing, generating multiple tracks in a QuickTime movie, and not a track with 4 overlays or a multi-channel (audio) track? May I ask what the name of the filter is that merges the crop filter outputs in 4 tracks in a single output (not overlay, not amerge ... I must have overlooked the relevant one)?

For reference, what I'm doing now is create a pipe to

ffmpeg -y -v 1 -i - -vf crop=360:288:0:0 -b:v 500k -vcodec mjpeg cam1.mov -vcodec mjpeg -vf crop=360:288:360:0 -b:v 500k cam2.mov -vf crop=360:288:0:288 -vcodec mjpeg -b:v 500k cam3.mov -vf crop=360:288:360:288 -vcodec mjpeg -b:v 500k cam4.mov

to generate single-track 4 quicktime movies from an m4v 4000kbit/s stream that contains the recorded output of 4 cameras, in 'quad' arrangement. I then combine those 4 movies into a single QuickTime movie with 4 tracks using QuickTime SDK functions.
NB: at 12Hz there isn't much gain of mpeg4 encoding over mjpeg encoding, which however plays back much smoother on a wider range of hardware.


Ren?

On Dec 07, 2012, at 13:19, Carl Eugen Hoyos wrote:

> Ren? J.V. Bertin <rjvbertin at ...> writes:
> 
>> I discovered yesterday that it's possible to split an input 
>> movie (a raw m4v stream) into 4 separate output movies (.mov) 
>> each containing a quarter of the original image (obtained with 
>> -vf crop=w/2:h/2:x:y), using a single invocation of ffmpeg.
>> Is it possible to generate a single output movie with 4 
>> tracks instead
> 
> The complex filter allows this, please RTFM for examples.
> 
> Carl Eugen


From drrobison at openroadsconsulting.com  Fri Dec  7 16:41:48 2012
From: drrobison at openroadsconsulting.com (David R Robison)
Date: Fri, 07 Dec 2012 10:41:48 -0500
Subject: [Libav-user] Duplicate frames?
In-Reply-To: <loom.20121207T003636-826@post.gmane.org>
References: <50C12BB9.6010907@openroadsconsulting.com>
	<loom.20121207T003636-826@post.gmane.org>
Message-ID: <50C20E3C.6010300@openroadsconsulting.com>

Hmm... So I cannot reproduce it with ffplay. I am using ffmpeg though 
the ffdshow DirectShow filters. It does appear to be some kind of 
buffering issue as if I buffer 500ms of video it plays fine, otherwise I 
get video quality issues. Maybe I need to discuss this with the ffdshow 
folks. David

David R Robison
Open Roads Consulting, Inc.
103 Watson Road, Chesapeake, VA 23320
phone: (757) 546-3401
e-mail: drrobison at openroadsconsulting.com
web: http://openroadsconsulting.com
blog: http://therobe.blogspot.com
book: http://www.xulonpress.com/bookstore/bookdetail.php?PB_ISBN=9781597816526

On 12/6/2012 6:37 PM, Carl Eugen Hoyos wrote:
> David R Robison <drrobison at ...> writes:
>
>> I am having a problem with poor quality of H.264 video.
> Can you reproduce the poor quality with ffmpeg (the application)?
> If yes, please provide command line together with complete, uncut
> console output.
>
> Carl Eugen
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user



This email communication (including any attachments) may contain confidential and/or privileged material intended solely for the individual or entity to which it is addressed.
If you are not the intended recipient, please delete this email immediately.


From cehoyos at ag.or.at  Fri Dec  7 17:10:45 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 7 Dec 2012 16:10:45 +0000 (UTC)
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
Message-ID: <loom.20121207T170627-929@post.gmane.org>

Ren? J.V. Bertin <rjvbertin at ...> writes:

> Are we really speaking about the same thing, generating 
> multiple tracks in a QuickTime movie, and not a
> track with 4 overlays or a multi-channel (audio) track? 
> May I ask what the name of the filter is that merges
> the crop filter outputs in 4 tracks in a single output 
> (not overlay, not amerge ... I must have overlooked
> the relevant one)?

$ ffmpeg -i - -filter_complex "split=4 [a][b][c][d]; \
[a]crop=360:288:0:0;[b]crop=360:288:360:0; \
[c]crop=360:288:0:288;[d]crop=360:288:360:288" \
-vcodec mjpeg cam.mov

Please do not top-post here, it is considered rude, Carl Eugen


From cehoyos at ag.or.at  Fri Dec  7 17:14:48 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 7 Dec 2012 16:14:48 +0000 (UTC)
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
Message-ID: <loom.20121207T171133-538@post.gmane.org>

Ren? J.V. Bertin <rjvbertin at ...> writes:

> NB: at 12Hz there isn't much gain of mpeg4 encoding over 
> mjpeg encoding

Depending on what you mean with "gain", this is not correct.
(Your sentence is equivalent to "intra-only codecs have 
no real disadvantage over using P and B-frames for 
sufficiently small fps." And this does not even take 
into account that I suspect ASP compresses better than 
jpeg even for images.)

Carl Eugen


From drrobison at openroadsconsulting.com  Fri Dec  7 17:36:09 2012
From: drrobison at openroadsconsulting.com (David R Robison)
Date: Fri, 07 Dec 2012 11:36:09 -0500
Subject: [Libav-user] Duplicate frames?
In-Reply-To: <50C20E3C.6010300@openroadsconsulting.com>
References: <50C12BB9.6010907@openroadsconsulting.com>
	<loom.20121207T003636-826@post.gmane.org>
	<50C20E3C.6010300@openroadsconsulting.com>
Message-ID: <50C21AF9.8060608@openroadsconsulting.com>

After further review, it appears that the problem ceneters around 
interlaced video. From the logging, the Top field is processed twice 
(lines 3 & 4) and this seems to cause problems when the Bottom field 
finally arrives (line 8). Does this make since? Should/could this cause 
video quality problems?

[4064] slice:1 T mb:0 P fix pps:0 frame:8 poc:65552/2147483647/65552 
ref:1/1 qp:24 loop:1:0:0 weight:0
[4064] slice:1 B mb:0 P fix pps:0 frame:8 poc:65552/65553/65552 ref:1/1 
qp:24 loop:1:0:0 weight:0
[4064] slice:1 T mb:0 P fix pps:0 frame:9 poc:65554/2147483647/65554 
ref:1/1 qp:24 loop:1:0:0 weight:0
[4064] slice:1 T mb:0 P fix pps:0 frame:9 poc:65554/2147483647/65554 
ref:1/1 qp:24 loop:1:0:0 weight:0
[4064] mmco: unref short failure
[4064] mmco: unref short failure
[4064] illegal short term buffer state detected
[4064] slice:1 B mb:0 P fix pps:0 frame:9 poc:65554/65555/65554 ref:1/1 
qp:24 loop:1:0:0 weight:0

David

David R Robison
Open Roads Consulting, Inc.
103 Watson Road, Chesapeake, VA 23320
phone: (757) 546-3401
e-mail: drrobison at openroadsconsulting.com
web: http://openroadsconsulting.com
blog: http://therobe.blogspot.com
book: http://www.xulonpress.com/bookstore/bookdetail.php?PB_ISBN=9781597816526

On 12/7/2012 10:41 AM, David R Robison wrote:
> Hmm... So I cannot reproduce it with ffplay. I am using ffmpeg though
> the ffdshow DirectShow filters. It does appear to be some kind of
> buffering issue as if I buffer 500ms of video it plays fine, otherwise I
> get video quality issues. Maybe I need to discuss this with the ffdshow
> folks. David
>
> David R Robison
> Open Roads Consulting, Inc.
> 103 Watson Road, Chesapeake, VA 23320
> phone: (757) 546-3401
> e-mail: drrobison at openroadsconsulting.com
> web: http://openroadsconsulting.com
> blog: http://therobe.blogspot.com
> book: http://www.xulonpress.com/bookstore/bookdetail.php?PB_ISBN=9781597816526
>
> On 12/6/2012 6:37 PM, Carl Eugen Hoyos wrote:
>> David R Robison <drrobison at ...> writes:
>>
>>> I am having a problem with poor quality of H.264 video.
>> Can you reproduce the poor quality with ffmpeg (the application)?
>> If yes, please provide command line together with complete, uncut
>> console output.
>>
>> Carl Eugen
>>
>> _______________________________________________
>> Libav-user mailing list
>> Libav-user at ffmpeg.org
>> http://ffmpeg.org/mailman/listinfo/libav-user
>
>
> This email communication (including any attachments) may contain confidential and/or privileged material intended solely for the individual or entity to which it is addressed.
> If you are not the intended recipient, please delete this email immediately.
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user



This email communication (including any attachments) may contain confidential and/or privileged material intended solely for the individual or entity to which it is addressed.
If you are not the intended recipient, please delete this email immediately.


From rjvbertin at gmail.com  Fri Dec  7 19:00:38 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Fri, 7 Dec 2012 19:00:38 +0100
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
In-Reply-To: <alpine.DEB.2.00.1212071840120.20512@rainbow.studorg.tuwien.ac.at>
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
	<loom.20121207T171133-538@post.gmane.org>
	<9925CF46-57D5-45D2-AA98-E9B7E336C6B7@gmail.com>
	<alpine.DEB.2.00.1212071840120.20512@rainbow.studorg.tuwien.ac.at>
Message-ID: <A16E0B7E-999C-4AED-8EC0-F18638B5C16E@gmail.com>

Top-hi :P

On Dec 07, 2012, at 18:42, Carl Eugen Hoyos wrote:
>> What I mean is the size of the files I get on my sample videos. Guess I should have made that clearer. And of course I never meant to say it depends on the framerate only, evidently it depends on how much the images vary between frames.
> 
> Allow me to repeat that what you believe is simply not true.
> If you want me to elaborate, please post on the mailing list.

Ok, please do elaborate - but also please accept the fact that I know what I observe ... if you have an alternate explanation I'd love to hear it.

> Allow me to add that I had neither used complex_filter nor split filter before, I only used the documentation.
> (And it was your luck that I had not read your private mail before searching for an answer.)

May I presume that you're both familiar with the general filter syntax and the documentation? I'm not.

>> Am I right that there's no filter to control the offsets of the 4 tracks
> 
> There are filters that allow to manipulate the timestamps, yes.

That I saw. I'm not talking about temporal offset, but spatial ones (that would allow to layout the tracks such that the original quad image is restored).

R.

From mpupilli at gmail.com  Fri Dec  7 19:51:08 2012
From: mpupilli at gmail.com (Mark Pupilli)
Date: Fri, 7 Dec 2012 18:51:08 +0000
Subject: [Libav-user] Double free/corruption problem.
Message-ID: <CAF93UxJTVvvm_L+0rOd=q5p=1894fwWT6S6DTfFOkTy-w3r1xg@mail.gmail.com>

I get a double free or corruption problem, should really do a bug
report but don't have time to decipher the FFmpeg bug reporting page.
I'll post it here in case anyone is interested, feel free to ignore.

When using custom IO for reading involves  something like the
following when creating a context:

buffer = (uint8_t*)av_malloc( BUFFER_SIZE );
ctxt = avio_alloc_context( buffer, BUFFER_SIZE, 0, this,
read_function, NULL, NULL );

and then to free it:

av_free( ctxt );
av_free( buffer );

However, if BUFFER_SIZE > 32768 bytes then av_free( buffer ) results
in a crash due to invalid free/delete (which can also be detected by
valgrind).
Not calling av_free( buffer ) leads to a memory leak (reported by valgrind).

In the case of calling av_free() Valgrind reports the buffer as
already being free'd in:

==5471==  Address 0x9181680 is 0 bytes inside a block of size 32,784 free'd
==5471==    at 0x4C2A82E: free (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==5471==    by 0x712EC6F: av_free (mem.c:186)
==5471==    by 0x6DBEFC4: ffio_set_buf_size (aviobuf.c:707)
==5471==    by 0x6DBE1E5: fill_buffer (aviobuf.c:408)
==5471==    by 0x6DBE570: avio_read (aviobuf.c:495)
==5471==    by 0x6DE8971: ff_id3v2_read (id3v2.c:748)
==5471==    by 0x6EA4F42: avformat_open_input (utils.c:638)

which if you follow the call in a debugger leads to the culprit at
aviobuf.c:408:

    /* make buffer smaller in case it ended up large after probing */
    if (s->read_packet && s->buffer_size > max_buffer_size) {
        ffio_set_buf_size(s, max_buffer_size);

        s->checksum_ptr = dst = s->buffer;
        len = s->buffer_size;
    }

This only gets called when we are reading. So in the case of using
custom input libav takes it upon itself to tinker with the user
supplied buffer. The crash only occurs if the buffer size is larger
than 32KB
which in my case it was because I was trying to follow some examples
which show allocating (32768 + FF_INPUT_BUFFER_PADDING_SIZE) bytes.

I don't know if the bug is that libav is supposed to give an error if
the buffer is larger than 32KB, or not tinker with user supplied
buffers, or avio_alloc_context should take a ** to the buffer so it
can reallocate properly and clients don't have
to choose between a double free or a memory leak, but its definitely
wrong one way or another. I actually do not understand why the user
needs to pass in the buffer at all as it seems avio_alloc_context
could do it internally if it is going to
reallocate it on a whim anyway.

regards,
Mark

From cehoyos at ag.or.at  Sat Dec  8 00:06:12 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 7 Dec 2012 23:06:12 +0000 (UTC)
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
	<loom.20121207T171133-538@post.gmane.org>
	<9925CF46-57D5-45D2-AA98-E9B7E336C6B7@gmail.com>
	<alpine.DEB.2.00.1212071840120.20512@rainbow.studorg.tuwien.ac.at>
	<A16E0B7E-999C-4AED-8EC0-F18638B5C16E@gmail.com>
Message-ID: <loom.20121207T235742-798@post.gmane.org>

Ren? J.V. Bertin <rjvbertin at ...> writes:

> >> What I mean is the size of the files I get on my sample videos.
> >> Guess I should have made that clearer. And of course I never 
> >> meant to say it depends on the framerate only, evidently it 
> >> depends on how much the images vary between frames.

("Guess" is probably the right word since it is not easy for me 
to understand above.)
Allow me to repeat:
mpeg-4 asp is (at least) one generation ahead of jpeg encoding, 
I strongly suspect it has advantages when encoding still images 
(although I don't know of any tests), the two encoders certainly 
cannot be compared for encoding of a real-world video stream (no 
matter the framerate) because mpeg-4 will always win a quality / 
size comparison by a large margin.
My guess now is that you argue that for two given codecs (no 
matter if mjpeg, mpeg-4 asp, mpeg2video, flv1 or snow), the 
following command line always produces files of nearly identical 
size:
$ ffmpeg -i input -vcodec xxx out.avi
This is certainly correct, the reason is that ffmpeg defaults 
to -vb 200k, so all codecs try to reach the same filesize 
(that may or may not be reached for given input etc.), the 
quality will vary significantly because the codecs do not 
have the same quality.
(The order is probably mjpeg->mpeg2video->flv1->mpeg4, with 
snow having great advantage for small bitrates.)

Carl Eugen


From rjvbertin at gmail.com  Sat Dec  8 02:07:33 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Sat, 8 Dec 2012 02:07:33 +0100
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
In-Reply-To: <loom.20121207T235742-798@post.gmane.org>
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
	<loom.20121207T171133-538@post.gmane.org>
	<9925CF46-57D5-45D2-AA98-E9B7E336C6B7@gmail.com>
	<alpine.DEB.2.00.1212071840120.20512@rainbow.studorg.tuwien.ac.at>
	<A16E0B7E-999C-4AED-8EC0-F18638B5C16E@gmail.com>
	<loom.20121207T235742-798@post.gmane.org>
Message-ID: <072932B6-EF4B-46C7-B5F9-B8CF0AF7F8B1@gmail.com>

On Dec 08, 2012, at 00:06, Carl Eugen Hoyos wrote:
> My guess now is that you argue that for two given codecs (no 
> matter if mjpeg, mpeg-4 asp, mpeg2video, flv1 or snow), the 
> following command line always produces files of nearly identical 
> size:
> $ ffmpeg -i input -vcodec xxx out.avi

No, not always, but for the files I'm working with. When I compare

> ffmpeg -i 4000kbaud.m4v -vcodec copy out1.mov
> ffmpeg -i 4000kbaud.m4v -vcodec mpeg4 -b:v 4000k out2.mov
> ffmpeg -i 4000kbaud.m4v -vcodec mjpeg -b:v 1150k out3.mov

I get 3 output movies that are all at (approximately) 4000kbits/s, and all approximately the same size as the input.
You may not agree, but it my experience that material that has already been compressed with a lossy codec is often hard to compress further - and in this case the input already shows visible blocking artefacts. In fact, 720x576 @ 12Hz, that gives almost 5000k pixels/s, so if I'm not missing something the content is already compressed to less than a bit per pixel ... and I have no idea if that's stretching it or not, in fact.

'night!
R.

From cehoyos at ag.or.at  Sat Dec  8 19:53:35 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sat, 8 Dec 2012 18:53:35 +0000 (UTC)
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
	<loom.20121207T171133-538@post.gmane.org>
	<9925CF46-57D5-45D2-AA98-E9B7E336C6B7@gmail.com>
	<alpine.DEB.2.00.1212071840120.20512@rainbow.studorg.tuwien.ac.at>
	<A16E0B7E-999C-4AED-8EC0-F18638B5C16E@gmail.com>
	<loom.20121207T235742-798@post.gmane.org>
	<072932B6-EF4B-46C7-B5F9-B8CF0AF7F8B1@gmail.com>
Message-ID: <loom.20121208T194544-794@post.gmane.org>

Ren? J.V. Bertin <rjvbertin at ...> writes:

> On Dec 08, 2012, at 00:06, Carl Eugen Hoyos wrote:
> > My guess now is that you argue that for two given codecs (no 
> > matter if mjpeg, mpeg-4 asp, mpeg2video, flv1 or snow), the 
> > following command line always produces files of nearly identical 
> > size:
> > $ ffmpeg -i input -vcodec xxx out.avi
> 
> No, not always, but for the files I'm working with. When I compare
> 
> > ffmpeg -i 4000kbaud.m4v -vcodec copy out1.mov
> > ffmpeg -i 4000kbaud.m4v -vcodec mpeg4 -b:v 4000k out2.mov
> > ffmpeg -i 4000kbaud.m4v -vcodec mjpeg -b:v 1150k out3.mov

(I don't think -vcodec mjpeg supports setting a bitrate, sorry 
I did not really consider this.)

> I get 3 output movies that are all at (approximately) 
> 4000kbits/s, and all approximately the same size as
> the input.
> You may not agree, but it my experience that material 
> that has already been compressed with a lossy codec is
> often hard to compress further

This is true for material that has been compressed with 
maximum quantiser but this does not seem like a typical 
use-case to me.

Carl Eugen


From cehoyos at ag.or.at  Sat Dec  8 19:55:20 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sat, 8 Dec 2012 18:55:20 +0000 (UTC)
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
	<loom.20121207T171133-538@post.gmane.org>
	<9925CF46-57D5-45D2-AA98-E9B7E336C6B7@gmail.com>
	<alpine.DEB.2.00.1212071840120.20512@rainbow.studorg.tuwien.ac.at>
	<A16E0B7E-999C-4AED-8EC0-F18638B5C16E@gmail.com>
	<loom.20121207T235742-798@post.gmane.org>
	<072932B6-EF4B-46C7-B5F9-B8CF0AF7F8B1@gmail.com>
	<loom.20121208T194544-794@post.gmane.org>
Message-ID: <loom.20121208T195411-910@post.gmane.org>

Carl Eugen Hoyos <cehoyos at ...> writes:

> > You may not agree, but it my experience that material 
> > that has already been compressed with a lossy codec is
> > often hard to compress further
> 
> This is true for material that has been compressed with 
> maximum quantiser but this does not seem like a typical 
> use-case to me.

And you can of course further compress such material by 
using some of the picture processing abilities that 
FFmpeg offers...

Carl Eugen


From rjvbertin at gmail.com  Sat Dec  8 20:14:13 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Sat, 8 Dec 2012 20:14:13 +0100
Subject: [Libav-user] ffmpeg ;
	splitting an input movie into multiple tracks in a single output
	movie, possible?
In-Reply-To: <loom.20121208T195411-910@post.gmane.org>
References: <593F4504-0126-424F-AB65-A38EE4DA0790@gmail.com>
	<loom.20121207T131805-592@post.gmane.org>
	<E6924971-A4D3-4F71-9310-FC75DA06ABEE@gmail.com>
	<loom.20121207T171133-538@post.gmane.org>
	<9925CF46-57D5-45D2-AA98-E9B7E336C6B7@gmail.com>
	<alpine.DEB.2.00.1212071840120.20512@rainbow.studorg.tuwien.ac.at>
	<A16E0B7E-999C-4AED-8EC0-F18638B5C16E@gmail.com>
	<loom.20121207T235742-798@post.gmane.org>
	<072932B6-EF4B-46C7-B5F9-B8CF0AF7F8B1@gmail.com>
	<loom.20121208T194544-794@post.gmane.org>
	<loom.20121208T195411-910@post.gmane.org>
Message-ID: <3734DCA9-F0BB-4F6A-91DD-2745B67D73E0@gmail.com>

On Dec 08, 2012, at 19:55, Carl Eugen Hoyos wrote:

> Carl Eugen Hoyos <cehoyos at ...> writes:
> 
>>> You may not agree, but it my experience that material 
>>> that has already been compressed with a lossy codec is
>>> often hard to compress further
>> 
>> This is true for material that has been compressed with 
>> maximum quantiser but this does not seem like a typical 
>> use-case to me.
> 
> And you can of course further compress such material by 
> using some of the picture processing abilities that 
> FFmpeg offers...
> 
Possibly, but that's not the point here, nor necessary. Maybe for those recordings that approach the maximum size (4Gb), but in that case there'd also be the issue of the transcoding duration. Given the (mediocre) image quality I'd be more interested in an approach without transcoding, but that turns out not to be feasible for playback using the existing QuickTime-based player on win32. It really surprised me that some of the code path could be that badly optimised, but that isn't the point here either :)

> (I don't think -vcodec mjpeg supports setting a bitrate, sorry 
> I did not really consider this.)

It clearly doesn't support it as expected, but the option does have an effect. Fortunately ...

I don't know if you're the one to suggest this too, but I really feel something like a -b:v auto or -b:v copy is missing (but maybe you'll tell me that it doesn't make sense to transcode while maintaining the original bit rate?)

Ren?


From bilthon at gmail.com  Sun Dec  9 01:19:02 2012
From: bilthon at gmail.com (Nelson Perez)
Date: Sat, 8 Dec 2012 22:19:02 -0200
Subject: [Libav-user] Problem with rtmp stream
Message-ID: <CAFOd1atNCd6BQc5r5m+FxQHdRaf1QXY03ZAovPUcd=e602vebQ@mail.gmail.com>

Hi there, I've been experiencing some trouble with a few rtmp streams
lately. The context is that I'm trying to capture some streams from street
cameras avaiable in the interent. Some cameras are not working though but
their streams have not been shut down.

What happens then is that since these streams have no data, the play
command doesnt do anything. I suppose some socket calls a blocking read
somewhere and the program just hangs there. What I would like to have is
some sort of timeout of some kind for these "faulty streams".

To give you guys an example, here's the command to grab video from a camera
that works.

ffmpeg -i "rtmp://200.141.78.68:1935/cet-rio/1.stream pageUrl=
http://transito.rio.rj.gov.br --live" -vcodec copy -an out.flv

And here is an example of the problematic one:

ffmpeg -i "rtmp://200.141.78.68:1935/cet-rio/74.stream pageUrl=
http://transito.rio.rj.gov.br --live" -vcodec copy -an out.flv

Does somebody have a suggestion to fix this issue?

Nelson
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121208/43147cf3/attachment.html>

From cehoyos at ag.or.at  Sun Dec  9 10:49:45 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sun, 9 Dec 2012 09:49:45 +0000 (UTC)
Subject: [Libav-user] Problem with rtmp stream
References: <CAFOd1atNCd6BQc5r5m+FxQHdRaf1QXY03ZAovPUcd=e602vebQ@mail.gmail.com>
Message-ID: <loom.20121209T104908-579@post.gmane.org>

Nelson Perez <bilthon at ...> writes:

> ffmpeg -i "rtmp://200.141.78.68:1935/cet-rio/74.stream 
> pageUrl=http://transito.rio.rj.gov.br --live" -vcodec copy -an out.flv
> 
> 
> Does somebody have a suggestion to fix this issue?

Since I cannot reproduce your issue:
Please provide complete, uncut console output.

Carl Eugen


From donmoir at comcast.net  Sun Dec  9 18:04:00 2012
From: donmoir at comcast.net (Don Moir)
Date: Sun, 9 Dec 2012 12:04:00 -0500
Subject: [Libav-user] How to connect to an IP camera in libav
References: <005301cdce32$7705fc20$6511f460$@caelumspace.com>
	<CAJg_av9=D7gjzYL8Z2eMSxDyZxWHVv_5jbrpONiP1pB=TnztaQ@mail.gmail.com>
Message-ID: <892934367069402A98161DBD557A4CEF@MANLAP>

> I?m trying to connect to a Foscam MJpeg IP camera (http) using libav on Windows and am getting an error code of -5 after 

For the foscam F18918W the default http string is: (dont forget &pwd=)

http://x.x.x.x/videostream.asf?user=admin&pwd=

For a list of possible strings for foscam cameras goto:

http://www.ispyconnect.com/man.aspx?n=foscam

I dont have any problem connecting but the time is way off like 6 seconds plus so large delay in animation.

I am assuming there is a way to control this but after looking around still don't know.

If I just ignore the timing its perfect. ffplay has the same problem.

I dont care about ffplay and this for internal code.

Alex, you got any info on this ?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121209/f7808a85/attachment.html>

From aw.wisch at googlemail.com  Mon Dec 10 03:43:52 2012
From: aw.wisch at googlemail.com (Alexander Wischnewski)
Date: Mon, 10 Dec 2012 10:43:52 +0800
Subject: [Libav-user] How to connect to an IP camera in libav
In-Reply-To: <892934367069402A98161DBD557A4CEF@MANLAP>
References: <005301cdce32$7705fc20$6511f460$@caelumspace.com>
	<CAJg_av9=D7gjzYL8Z2eMSxDyZxWHVv_5jbrpONiP1pB=TnztaQ@mail.gmail.com>
	<892934367069402A98161DBD557A4CEF@MANLAP>
Message-ID: <CAJg_av_jxV_AcsRF_UK8DOEmQJfaCv-3OfTX3-FiGE5y66=mmg@mail.gmail.com>

Hello Don,

is the *avformat_open_input *function* *taking long or maybe the *
avformat_find_stream_info* ?

In the second case you have to reduce the maximal analyze duration of
the AVFormatContext, you can do it like this:

    pFormatCtx->max_analyze_duration = 1.0 * AV_TIME_BASE;

    // Retrieve stream information
    avformat_find_stream_info(pFormatCtx, NULL)

Seems that the default value is set to a very high value to make sure there
are all important informations gathered, but this is not necessary for a
simple mjpeg stream. I even use this setting for decoding h264 streams and
had no problems so far.


Best regards,
Alex


2012/12/10 Don Moir <donmoir at comcast.net>

> **
> > I?m trying to connect to a Foscam MJpeg IP camera (http) using libav on
> Windows and am getting an error code of -5 after
>
> For the foscam F18918W the default http string is: (dont forget &pwd=)
>
> http://x.x.x.x/videostream.asf?user=admin&pwd=
>
> For a list of possible strings for foscam cameras goto:
>
> http://www.ispyconnect.com/man.aspx?n=foscam
>
> I dont have any problem connecting but the time is way off like 6 seconds
> plus so large delay in animation.
>
> I am assuming there is a way to control this but after looking around
> still don't know.
>
> If I just ignore the timing its perfect. ffplay has the same problem.
>
> I dont care about ffplay and this for internal code.
>
> Alex, you got any info on this ?
>
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121210/bf94e5d4/attachment.html>

From bilthon at gmail.com  Mon Dec 10 05:03:07 2012
From: bilthon at gmail.com (Nelson Perez)
Date: Mon, 10 Dec 2012 02:03:07 -0200
Subject: [Libav-user] Problem with rtmp stream
In-Reply-To: <loom.20121209T104908-579@post.gmane.org>
References: <CAFOd1atNCd6BQc5r5m+FxQHdRaf1QXY03ZAovPUcd=e602vebQ@mail.gmail.com>
	<loom.20121209T104908-579@post.gmane.org>
Message-ID: <CAFOd1aui+oB6xzuFSwwMaVAi48v7_ZtYRkF20Ej64MFWeXsC5Q@mail.gmail.com>

My bad, I used the wrong address to illustrate the problem. It actually
happens with camera 71. So you should type:

ffmpeg -i "rtmp://200.141.78.68:1935/cet-rio/71.stream<http://200.141.78.68:1935/cet-rio/74.stream>
 pageUrl=http://transito.rio.rj.gov.br --live" -vcodec copy -an out.flv

I just did it again, this time with a more verbose output, and this is what
I get:

----------------------------------------
nelson at rigel:/opt/ffmpeg/ffmpeg$ sudo ./ffmpeg -i "rtmp://
200.141.78.68:1935/cet-rio/71.stream
pageUrl=http://transito.rio.rj.gov.br--live" -vcodec copy -an
-loglevel verbose out.flv
ffmpeg version N-47645-gf494647 Copyright (c) 2000-2012 the FFmpeg
developers
  built on Dec  8 2012 21:54:25 with gcc 4.7 (Ubuntu/Linaro 4.7.2-2ubuntu1)
  configuration: --enable-gpl --enable-libfaac --enable-libfdk-aac
--enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb
--enable-librtmp --enable-libtheora --enable-libvorbis --enable-libvpx
--enable-x11grab --enable-libx264 --enable-nonfree --enable-version3
  libavutil      52. 12.100 / 52. 12.100
  libavcodec     54. 79.100 / 54. 79.100
  libavformat    54. 48.100 / 54. 48.100
  libavdevice    54.  3.102 / 54.  3.102
  libavfilter     3. 26.101 /  3. 26.101
  libswscale      2.  1.103 /  2.  1.103
  libswresample   0. 17.102 /  0. 17.102
  libpostproc    52.  2.100 / 52.  2.100
Parsing...
Parsed protocol: 0
Parsed host    : 200.141.78.68
Parsed app     : cet-rio
RTMP_Connect1, ... connected, handshaking
HandShake: Type Answer   : 03
HandShake: Server Uptime : 96850052
HandShake: FMS Version   : 3.0.1.1
HandShake: Handshaking finished....
RTMP_Connect1, handshaked
Invoking connect
HandleServerBW: server BW = 2500000
HandleClientBW: client BW = 2500000 2
HandleCtrl, received ctrl. type: 0, len: 6
HandleCtrl, Stream Begin 0
HandleChangeChunkSize, received: chunk size change to 512
RTMP_ClientPacket, received: invoke 259 bytes
(object begin)
Property: <Name:           no-name., STRING: _result>
Property: <Name:           no-name., NUMBER: 1.00>
Property: <Name:           no-name., OBJECT>
(object begin)
Property: <Name:             fmsVer, STRING: FMS/3,5,4,210>
Property: <Name:       capabilities, NUMBER: 31.00>
Property: <Name:               mode, NUMBER: 1.00>
(object end)
Property: <Name:           no-name., OBJECT>
(object begin)
Property: <Name:              level, STRING: status>
Property: <Name:               code, STRING: NetConnection.Connect.Success>
Property: <Name:        description, STRING: Connection succeeded.>
Property: <Name:               data, OBJECT>
(object begin)
Property: <Name:            version, STRING: 3,5,4,210>
(object end)
Property: <Name:           clientid, NUMBER: 1304507915.00>
Property: <Name:     objectEncoding, NUMBER: 0.00>
(object end)
(object end)
HandleInvoke, server invoking <_result>
HandleInvoke, received result for method call <connect>
sending ctrl. type: 0x0003
Invoking createStream
RTMP_ClientPacket, received: invoke 29 bytes
(object begin)
Property: <Name:           no-name., STRING: _result>
Property: <Name:           no-name., NUMBER: 2.00>
Property: NULL
Property: <Name:           no-name., NUMBER: 1.00>
(object end)
HandleInvoke, server invoking <_result>
HandleInvoke, received result for method call <createStream>
SendPlay, seekTime=0, stopTime=0, sending play: 71.stream
Invoking play
sending ctrl. type: 0x0003
HandleCtrl, received ctrl. type: 6, len: 6
HandleCtrl, Ping 21175
sending ctrl. type: 0x0007
HandleCtrl, received ctrl. type: 6, len: 6
HandleCtrl, Ping 41580
sending ctrl. type: 0x0007
HandleCtrl, received ctrl. type: 6, len: 6
HandleCtrl, Ping 61860
sending ctrl. type: 0x0007
HandleCtrl, received ctrl. type: 6, len: 6
HandleCtrl, Ping 82265
sending ctrl. type: 0x0007
HandleCtrl, received ctrl. type: 6, len: 6
HandleCtrl, Ping 102857
--------------------------------------

And this goes on forever. I also found this post
http://lists.mplayerhq.hu/pipermail/rtmpdump/2010-March/000622.html, which
describes the exact problem.

But it doesn't look like the problem has been fixed, as the same thing
happens even when I try it with rtmpdump v2.4.

Any ideas?

Nelson



On Sun, Dec 9, 2012 at 7:49 AM, Carl Eugen Hoyos <cehoyos at ag.or.at> wrote:

> Nelson Perez <bilthon at ...> writes:
>
> > ffmpeg -i "rtmp://200.141.78.68:1935/cet-rio/74.stream
> > pageUrl=http://transito.rio.rj.gov.br --live" -vcodec copy -an out.flv
> >
> >
> > Does somebody have a suggestion to fix this issue?
>
> Since I cannot reproduce your issue:
> Please provide complete, uncut console output.
>
> Carl Eugen
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121210/acab2e31/attachment.html>

From cehoyos at ag.or.at  Mon Dec 10 12:14:12 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Mon, 10 Dec 2012 11:14:12 +0000 (UTC)
Subject: [Libav-user] Problem with rtmp stream
References: <CAFOd1atNCd6BQc5r5m+FxQHdRaf1QXY03ZAovPUcd=e602vebQ@mail.gmail.com>
	<loom.20121209T104908-579@post.gmane.org>
	<CAFOd1aui+oB6xzuFSwwMaVAi48v7_ZtYRkF20Ej64MFWeXsC5Q@mail.gmail.com>
Message-ID: <loom.20121210T120457-876@post.gmane.org>

Nelson Perez <bilthon at ...> writes:

> My bad, I used the wrong address to illustrate the problem. 
> It actually happens with camera 71. So you should type:
> 
> 
> ffmpeg -i
"rtmp://200.141.78.68:1935/cet-rio/71.stream?pageUrl=http://transito.rio.rj.gov.br?--live"
-vcodec copy -an out.flv

I originally could not reproduce the problem (ffmpeg immediately 
exits for me with an appropriate error message), I recompiled 
with --enable-librtmp and it appears that the bug is in librtmp 
(which does not return).
rtmpdump also hangs:
$ rtmpdump -i "rtmp://200.141.78.68:1935/cet-rio/71.stream
pageUrl=http://transito.rio.rj.gov.br --live"

Carl Eugen


From donmoir at comcast.net  Mon Dec 10 19:29:48 2012
From: donmoir at comcast.net (Don Moir)
Date: Mon, 10 Dec 2012 13:29:48 -0500
Subject: [Libav-user] How to connect to an IP camera in libav
References: <005301cdce32$7705fc20$6511f460$@caelumspace.com><CAJg_av9=D7gjzYL8Z2eMSxDyZxWHVv_5jbrpONiP1pB=TnztaQ@mail.gmail.com><892934367069402A98161DBD557A4CEF@MANLAP>
	<CAJg_av_jxV_AcsRF_UK8DOEmQJfaCv-3OfTX3-FiGE5y66=mmg@mail.gmail.com>
Message-ID: <05728DAB83454BD6BC1565011C7E05CD@MANLAP>

> From: Alexander Wischnewski 
> To: This list is about using libavcodec, libavformat, libavutil,libavdevice and libavfilter. 
> Sent: Sunday, December 09, 2012 9:43 PM
> Subject: Re: [Libav-user] How to connect to an IP camera in libav
> 
> Hello Don,
> 
> is the avformat_open_input function taking long or maybe the avformat_find_stream_info ? 
> 
> In the second case you have to reduce the maximal analyze duration of the AVFormatContext, you can do it like this:
> 
>     pFormatCtx->max_analyze_duration = 1.0 * AV_TIME_BASE;
> 
>     // Retrieve stream information
>     avformat_find_stream_info(pFormatCtx, NULL)
> 
> Seems that the default value is set to a very high value to make sure there are all important informations gathered,
> but this is not necessary for a simple mjpeg stream. I even use this setting for decoding h264 streams and had no problems so far.
> 
> Best regards,
> Alex

Hi Alex,

That helps but doesn't completely resolve it. Even if I ignore timing completely it's about 1/4 second off. The software that came with it is right on. Arrgh more work :)

Also the framerate guessed by ffmpeg is always 25 fps in this case which is wrong so need to do something about that.

I would be interested in a database of strings to control a variety of IP cameras. Note that I just care about the various strings for acquisition and controlling pan/tilt/zoom for various IP cameras and not interested in controlling software so much.

I know it exist but mostly packaged up with a program. I would be willing to purchase such a thing. 

Would you happen to know where I could my hands on that ?'

I have the strings for my foscam but thats just one camera. I would need them for many.

Thanks,

Don
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121210/366713bd/attachment.html>

From aw.wisch at googlemail.com  Tue Dec 11 14:21:23 2012
From: aw.wisch at googlemail.com (Alexander Wischnewski)
Date: Tue, 11 Dec 2012 21:21:23 +0800
Subject: [Libav-user] How to connect to an IP camera in libav
In-Reply-To: <05728DAB83454BD6BC1565011C7E05CD@MANLAP>
References: <005301cdce32$7705fc20$6511f460$@caelumspace.com>
	<CAJg_av9=D7gjzYL8Z2eMSxDyZxWHVv_5jbrpONiP1pB=TnztaQ@mail.gmail.com>
	<892934367069402A98161DBD557A4CEF@MANLAP>
	<CAJg_av_jxV_AcsRF_UK8DOEmQJfaCv-3OfTX3-FiGE5y66=mmg@mail.gmail.com>
	<05728DAB83454BD6BC1565011C7E05CD@MANLAP>
Message-ID: <CAJg_av_JRsWW6fJV0=eYLrJzzceWg7_+=GRcwOsip1aXMEBhJw@mail.gmail.com>

Hi Don,

take a look at the open source project iSpy (Connect), it is a free
software for network cameras and recorders supporting all kind of
manufactors. If you are lucky you will find the database of .cgi commands
they are using in the source code.

Best regards,
Alex


2012/12/11 Don Moir <donmoir at comcast.net>

> **
> > From: Alexander Wischnewski
> > To: This list is about using libavcodec, libavformat,
> libavutil,libavdevice and libavfilter.
> > Sent: Sunday, December 09, 2012 9:43 PM
> > Subject: Re: [Libav-user] How to connect to an IP camera in libav
> >
> > Hello Don,
> >
> > is the avformat_open_input function taking long or maybe the
> avformat_find_stream_info ?
> >
> > In the second case you have to reduce the maximal analyze duration of
> the AVFormatContext, you can do it like this:
> >
> >     pFormatCtx->max_analyze_duration = 1.0 * AV_TIME_BASE;
> >
> >     // Retrieve stream information
> >     avformat_find_stream_info(pFormatCtx, NULL)
> >
> > Seems that the default value is set to a very high value to make sure
> there are all important informations gathered,
> > but this is not necessary for a simple mjpeg stream. I even use this
> setting for decoding h264 streams and had no problems so far.
> >
> > Best regards,
> > Alex
>
> Hi Alex,
>
> That helps but doesn't completely resolve it. Even if I ignore timing
> completely it's about 1/4 second off. The software that came with it is
> right on. Arrgh more work :)
>
> Also the framerate guessed by ffmpeg is always 25 fps in this case which
> is wrong so need to do something about that.
>
> I would be interested in a database of strings to control a variety of IP
> cameras. Note that I just care about the various strings for acquisition
> and controlling pan/tilt/zoom for various IP cameras and not interested in
> controlling software so much.
>
> I know it exist but mostly packaged up with a program. I would be willing
> to purchase such a thing.
>
> Would you happen to know where I could my hands on that ?'
>
> I have the strings for my foscam but thats just one camera. I would need
> them for many.
>
> Thanks,
>
> Don
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121211/0ff7f40d/attachment.html>

From mike at mikeversteeg.com  Tue Dec 11 17:40:44 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Tue, 11 Dec 2012 17:40:44 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2 recorder
In-Reply-To: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
Message-ID: <CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>

Using the famous (and only) output example I developed an MPEG-2
recorder for audio and video. Both come from different sources with
their own clock (audio as PCM and video as bitmaps) and as a result
over time audio will get out of sync due to small differences in the
two clocks. To compensate I figured to tweak the audio PTS, but it
appears not to be used (it is marked no pts value).

What would be the proper way to keep A & V in sync?

Thanks,

Mike

From phamsyquybk at gmail.com  Wed Dec 12 05:35:52 2012
From: phamsyquybk at gmail.com (Quy Pham Sy)
Date: Wed, 12 Dec 2012 13:35:52 +0900
Subject: [Libav-user] [libav-users] How to detect max packet size for audio
	stream
Message-ID: <CAEEaHNJhTWALqG7f3=k4VhgHu9FZDKQ=mLFNaMFdJjOUXxgntw@mail.gmail.com>

Hi,

How to get the maximum audio packet size for an audio stream to  properly
allocate audio buffer?

(i my audio stream encode in AAC)

Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121212/77b32a67/attachment.html>

From hwcy at hotmail.com  Wed Dec 12 17:03:11 2012
From: hwcy at hotmail.com (Yoon Chris)
Date: Wed, 12 Dec 2012 11:03:11 -0500
Subject: [Libav-user] ffmpeg - Where can I find rtpvideo.cpp?
Message-ID: <SNT106-W5130D3FAC233CC32D2D572CA4F0@phx.gbl>



Hi,

I'm looking for a file named "rtpvideo.cpp" that is supposedly a part of ffmpeg.  I couldn't find the file in the latest version of ffmpeg code I downloaded.  Do you know where I can get this file?  Thanks.

 
Chris
  		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121212/9dbe548e/attachment.html>

From cehoyos at ag.or.at  Wed Dec 12 19:34:49 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Wed, 12 Dec 2012 18:34:49 +0000 (UTC)
Subject: [Libav-user] ffmpeg - Where can I find rtpvideo.cpp?
References: <SNT106-W5130D3FAC233CC32D2D572CA4F0@phx.gbl>
Message-ID: <loom.20121212T193207-439@post.gmane.org>

Yoon Chris <hwcy at ...> writes:

> Hi,I'm looking for a file named "rtpvideo.cpp" that is 
> supposedly a part of ffmpeg.

The file is definitely not part of FFmpeg and I would be 
very surprised if it ever has been.

(I suspect it is part of an inactive (spanish) open-souce 
project called teartp.)

Carl Eugen


From mike at mikeversteeg.com  Thu Dec 13 11:36:35 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 13 Dec 2012 11:36:35 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
Message-ID: <CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>

Anyone?

From mark.kenna at sureviewsystems.com  Thu Dec 13 11:48:31 2012
From: mark.kenna at sureviewsystems.com (Mark Kenna)
Date: Thu, 13 Dec 2012 10:48:31 +0000
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
Message-ID: <CAAxE61gx9wUoJZ-ChsR1szbg-p=EZK4itgXYmsFTnumBxJsjvA@mail.gmail.com>

On 13 December 2012 10:36, Mike Versteeg <mike at mikeversteeg.com> wrote:

> Anyone?
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>


Try this: http://dranger.com/ffmpeg/tutorial05.html

Slightly out of date but still a great tutorial
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121213/703c717d/attachment.html>

From mike at mikeversteeg.com  Thu Dec 13 12:26:03 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 13 Dec 2012 12:26:03 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAAxE61gx9wUoJZ-ChsR1szbg-p=EZK4itgXYmsFTnumBxJsjvA@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<CAAxE61gx9wUoJZ-ChsR1szbg-p=EZK4itgXYmsFTnumBxJsjvA@mail.gmail.com>
Message-ID: <CAE=zCuomSmTnWoUP2cG01Pvc4Fd2Q7PGZTGtr-R5+vX16WPi4w@mail.gmail.com>

Thanks, but that's for decoding, not encoding.

From info at denisgottardello.it  Thu Dec 13 15:53:08 2012
From: info at denisgottardello.it (Denis)
Date: Thu, 13 Dec 2012 15:53:08 +0100
Subject: [Libav-user] av_opt_set, x264
Message-ID: <201212131553.09147.info@denisgottardello.it>

In order to reduce the latency of a live x264 video stream I have tested these rows:

                    if (pAVStreamVideo->codec->codec_id== CODEC_ID_H264) {
                        qDebug() << "CODEC_ID_H264";
                        av_opt_set(pAVFormatContext->priv_data, "preset", "veryfast", 0);
                        av_opt_set(pAVFormatContext->priv_data, "tune", "zerolatency", 0);
                        av_opt_set(pAVFormatContext->priv_data, "x264opts", "no-mbtree:sliced-threads:sync-lookahead=0", 0);
                    }

Without any effect. Is there another way for pass a parameter or must I use another api?
Regards.


-- 
www.denisgottardello.it
Skype: mrdebug
Videosurveillance and home automation! 
http://www.denisgottardello.it/DomusBoss/DomusBossIndice.php

From mike at mikeversteeg.com  Thu Dec 13 16:42:57 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 13 Dec 2012 16:42:57 +0100
Subject: [Libav-user] Commercial support
Message-ID: <CAE=zCuot1jeigoBEPDEqLTRZN7xhmE_twYi7p71SdcYBAbPrgA@mail.gmail.com>

Anyone willing & able to offer me (paid) support?

Thanks,

Mike

From nitinkumgoyal at gmail.com  Thu Dec 13 17:27:36 2012
From: nitinkumgoyal at gmail.com (NITIN GOYAL)
Date: Thu, 13 Dec 2012 21:57:36 +0530
Subject: [Libav-user] Commercial support
In-Reply-To: <CAE=zCuot1jeigoBEPDEqLTRZN7xhmE_twYi7p71SdcYBAbPrgA@mail.gmail.com>
References: <CAE=zCuot1jeigoBEPDEqLTRZN7xhmE_twYi7p71SdcYBAbPrgA@mail.gmail.com>
Message-ID: <CADig5u-nz74hm2oyAs2RO5qytSN6ng+hoaSc0y9LEZZs1WQSgg@mail.gmail.com>

I think if you can explain your requirement it will be much better and what
kind of support you are actually looking for?



On Thu, Dec 13, 2012 at 9:12 PM, Mike Versteeg <mike at mikeversteeg.com>wrote:

> Anyone willing & able to offer me (paid) support?
>
> Thanks,
>
> Mike
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121213/78299f4f/attachment.html>

From mike at mikeversteeg.com  Thu Dec 13 17:36:04 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 13 Dec 2012 17:36:04 +0100
Subject: [Libav-user] Commercial support
In-Reply-To: <CADig5u-nz74hm2oyAs2RO5qytSN6ng+hoaSc0y9LEZZs1WQSgg@mail.gmail.com>
References: <CAE=zCuot1jeigoBEPDEqLTRZN7xhmE_twYi7p71SdcYBAbPrgA@mail.gmail.com>
	<CADig5u-nz74hm2oyAs2RO5qytSN6ng+hoaSc0y9LEZZs1WQSgg@mail.gmail.com>
Message-ID: <CAE=zCuq14_om72kjcc2NPA_ySTqkgUFkW3uD6bmMtXPaTZ0=6w@mail.gmail.com>

> I think if you can explain your requirement it will be much better and what
> kind of support you are actually looking for?
>
>
>
> On Thu, Dec 13, 2012 at 9:12 PM, Mike Versteeg <mike at mikeversteeg.com>
> wrote:
>>
>> Anyone willing & able to offer me (paid) support?

These would be questions related to the use of the libav API,
particularly questions that relate to my lack of knowledge of the API
(due to lack of documentation and examples). A good example would be
this question: http://libav-users.943685.n4.nabble.com/Libav-user-How-to-perform-drift-compensation-in-MPEG-2-recorder-td4656246.html.

From cehoyos at ag.or.at  Thu Dec 13 18:07:53 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Thu, 13 Dec 2012 17:07:53 +0000 (UTC)
Subject: [Libav-user] Commercial support
References: <CAE=zCuot1jeigoBEPDEqLTRZN7xhmE_twYi7p71SdcYBAbPrgA@mail.gmail.com>
	<CADig5u-nz74hm2oyAs2RO5qytSN6ng+hoaSc0y9LEZZs1WQSgg@mail.gmail.com>
	<CAE=zCuq14_om72kjcc2NPA_ySTqkgUFkW3uD6bmMtXPaTZ0=6w@mail.gmail.com>
Message-ID: <loom.20121213T180532-693@post.gmane.org>

Mike Versteeg <mike at ...> writes:

> These would be questions related to the use of the libav API,
> particularly questions that relate to my lack of knowledge of the API

> (due to lack of documentation and examples).

Did you look into the doc/examples directory?
It also contains an encoding example which iirc 
is what you have asked for.

Was your original question about two independent inputs 
(like alsa and v4l) that do not use the same time-source 
or do you have a file that is converted fine with ffmpeg 
(the command line application) but not with your app?

Carl Eugen


From cehoyos at ag.or.at  Thu Dec 13 18:10:38 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Thu, 13 Dec 2012 17:10:38 +0000 (UTC)
Subject: [Libav-user] =?utf-8?q?av=5Fopt=5Fset=2C_x264?=
References: <201212131553.09147.info@denisgottardello.it>
Message-ID: <loom.20121213T180823-434@post.gmane.org>

Denis <info at ...> writes:

>     if (pAVStreamVideo->codec->codec_id== CODEC_ID_H264) {
>         qDebug() << "CODEC_ID_H264";
>         av_opt_set(pAVFormatContext->priv_data, "preset", "veryfast", 0);
>         av_opt_set(pAVFormatContext->priv_data, "tune", "zerolatency", 0);
>         av_opt_set(pAVFormatContext->priv_data, "x264opts",
> "no-mbtree:sliced-threads:sync-lookahead=0", 0);
>         }
> 
> Without any effect.

How do you mean "Without any effect"?
I wonder if "priv_data" is the right parameter though, 
shouldn't this be something like "st->codec" ?

Carl Eugen


From effemm at mykmk.com  Thu Dec 13 19:03:17 2012
From: effemm at mykmk.com (mattes)
Date: Thu, 13 Dec 2012 10:03:17 -0800
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>,
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
Message-ID: <ed1230c622e3e0ee5efa0db4387764b5@mykmk.com>

I assume that you notice the drift during playback.
Many player are using different approaches in how to sync
the two streams (AV). You might encounter different results
with different players.

One way is to stuff the continuous audio stream in the audio channel
and then sync in the video frames. Playing back the audio, you have to
take two clocks into account and compensate for it. The clock of the
recorder and the playback host will slightly different.

A simple way on the recorder site is to stuff an additional frame 
or toss one. You could soft 'calibrate' it against the real time clock.
This might ease the AV sync on the playback site. But nevertheless 
playback needs to compensate on its own in similar fashion. 


On Tue, 11 Dec 2012 17:40:44 +0100 Mike Versteeg <mike at mikeversteeg.com> wrote

> Using the famous (and only) output example I developed an MPEG-2
> recorder for audio and video. Both come from different sources with
> their own clock (audio as PCM and video as bitmaps) and as a result
> over time audio will get out of sync due to small differences in the
> two clocks. To compensate I figured to tweak the audio PTS, but it
> appears not to be used (it is marked no pts value).
> 
> What would be the proper way to keep A & V in sync?
> 
> Thanks,
> 
> Mike
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user



From info at denisgottardello.it  Thu Dec 13 19:03:36 2012
From: info at denisgottardello.it (Denis)
Date: Thu, 13 Dec 2012 19:03:36 +0100
Subject: [Libav-user] av_opt_set, x264
In-Reply-To: <loom.20121213T180823-434@post.gmane.org>
References: <201212131553.09147.info@denisgottardello.it>
	<loom.20121213T180823-434@post.gmane.org>
Message-ID: <201212131903.36859.info@denisgottardello.it>

In data gioved? 13 dicembre 2012 18:10:38, Carl Eugen Hoyos ha scritto:
> How do you mean "Without any effect"?
> I wonder if "priv_data" is the right parameter though, 
> shouldn't this be something like "st->codec" ?


Which is the right way for change the profile? Have you got an example?
I need a no latency solution.


See the ffmpeg dump:

[libx264 @ 0x92497c0] using cpu capabilities: none!
[libx264 @ 0x92497c0] profile High, level 1.2
[libx264 @ 0x92497c0] frame I:4     Avg QP:21,97  size:  8554
[libx264 @ 0x92497c0] frame P:59    Avg QP:22,14  size:  1394

-- 
www.denisgottardello.it
Skype: mrdebug
Videosurveillance and home automation! 
http://www.denisgottardello.it/DomusBoss/DomusBossIndice.php
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121213/7537637c/attachment.html>

From mike at mikeversteeg.com  Thu Dec 13 19:41:17 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 13 Dec 2012 19:41:17 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <ed1230c622e3e0ee5efa0db4387764b5@mykmk.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<ed1230c622e3e0ee5efa0db4387764b5@mykmk.com>
Message-ID: <CAE=zCup-j0eqUWMDbL1dxxKi=+A-XwVK89CcTzHe404tY-64Qw@mail.gmail.com>

> I assume that you notice the drift during playback.
> Many player are using different approaches in how to sync
> the two streams (AV). You might encounter different results
> with different players.
>
> One way is to stuff the continuous audio stream in the audio channel
> and then sync in the video frames. Playing back the audio, you have to
> take two clocks into account and compensate for it. The clock of the
> recorder and the playback host will slightly different.
>
> A simple way on the recorder site is to stuff an additional frame
> or toss one. You could soft 'calibrate' it against the real time clock.
> This might ease the AV sync on the playback site. But nevertheless
> playback needs to compensate on its own in similar fashion.

Thanks. But as I mentioned both sources have their own clock and
doubling or omitting frames is unacceptable in the world of
broadcasting. The problem is also not at playback but at recording, as
one can easily detect in the changing size of the audio buffer. I can
overcome this by changing the sample rate of the audio slightly so the
buffer's size will remain constant, but figured there would be an
easier way by modulating the PTS appropriately. At least I figured
that's what it's for. The question here is how to change it in this
particular case.

From mike at mikeversteeg.com  Thu Dec 13 19:50:45 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 13 Dec 2012 19:50:45 +0100
Subject: [Libav-user] Commercial support
In-Reply-To: <loom.20121213T180532-693@post.gmane.org>
References: <CAE=zCuot1jeigoBEPDEqLTRZN7xhmE_twYi7p71SdcYBAbPrgA@mail.gmail.com>
	<CADig5u-nz74hm2oyAs2RO5qytSN6ng+hoaSc0y9LEZZs1WQSgg@mail.gmail.com>
	<CAE=zCuq14_om72kjcc2NPA_ySTqkgUFkW3uD6bmMtXPaTZ0=6w@mail.gmail.com>
	<loom.20121213T180532-693@post.gmane.org>
Message-ID: <CAE=zCupybOMAE70OhYZ8VGHWfXdajqgu=wcz2wfeGUROM674MQ@mail.gmail.com>

>> These would be questions related to the use of the libav API,
>> particularly questions that relate to my lack of knowledge of the API
>
>> (due to lack of documentation and examples).
>
> Did you look into the doc/examples directory?

Yes.

> It also contains an encoding example which iirc
> is what you have asked for.

No, it does not do any drift compensation if you feed it with two
independently clocked streams.

> Was your original question about two independent inputs
> (like alsa and v4l) that do not use the same time-source
> or do you have a file that is converted fine with ffmpeg
> (the command line application) but not with your app?

Totally independent. Problem with the output example is that it
assumes the audio is in complete sync with the video (for each frame a
fixed number of audio samples is taken in). I need to modulate that
number, and I assume that can be done by changing the PTS. But I do
not know how, and in fact when I examine it is seems not to be used at
all.

Anyway, this it rather OT and should go to the other thread. This one
is about the availability of someone who knows all about stuff like
this and is willing to help me for an hourly fee. Considering libav*
is (almost) free I gladly support this financially.

From cmst at live.com  Wed Dec 12 20:32:54 2012
From: cmst at live.com (Dolevo Jay)
Date: Wed, 12 Dec 2012 19:32:54 +0000
Subject: [Libav-user] ffmpeg - Where can I find rtpvideo.cpp?
In-Reply-To: <SNT106-W5130D3FAC233CC32D2D572CA4F0@phx.gbl>
References: <SNT106-W5130D3FAC233CC32D2D572CA4F0@phx.gbl>
Message-ID: <DUB114-W24F83A19321C55FF7B3A71AA4F0@phx.gbl>



Can it be this one? 
http://teartp.googlecode.com/svn-history/r9/trunk/Comun/Video/RTPVideo.cpp

From: hwcy at hotmail.com
To: libav-user at ffmpeg.org
Date: Wed, 12 Dec 2012 11:03:11 -0500
Subject: [Libav-user] ffmpeg - Where can I find rtpvideo.cpp?






Hi,


I'm looking for a file named "rtpvideo.cpp" that is supposedly a part of ffmpeg.  I couldn't find the file in the latest version of ffmpeg code I downloaded.  Do you know where I can get this file?  Thanks.


 
Chris
  		 	   		  

_______________________________________________
Libav-user mailing list
Libav-user at ffmpeg.org
http://ffmpeg.org/mailman/listinfo/libav-user 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121212/d6cff6ff/attachment.html>

From gohar_h at instigatemobile.com  Thu Dec 13 13:57:32 2012
From: gohar_h at instigatemobile.com (Gohar Hovhannisyan)
Date: Thu, 13 Dec 2012 16:57:32 +0400
Subject: [Libav-user] Decoding HD videos on iPad devices
Message-ID: <D03BBEB0-1A2F-4EF3-BB13-28FD05D41075@instigatemobile.com>

Dear FFMpeg team,

We are creating player for iPad and using FFMPeg with libavcodec to decode video frames.

FFMpeg helped to process a lot of formats and really reduced the development time.

Unfortunately with HD videos we experience a poor performance and the video playback is breaking up.
According to profile stats the decode time is taking longer then the playback

Profile shows ~100% CPU usage on the call to avcodec_decode_video2 function. In the same time the GPU is almost not used.
 
We wonder whether it is possible to perform decoding on iPad GPU to speed up decoding process (by compiling it with special flags, or with some patches) or run it parallel on dual core (iPad2) / quad-core (new iPad) or are there any other methods to speed up decoding process?

Thanks in advance.

Regards,
Gohar

From info at denisgottardello.it  Fri Dec 14 14:11:05 2012
From: info at denisgottardello.it (Denis)
Date: Fri, 14 Dec 2012 14:11:05 +0100
Subject: [Libav-user] av_opt_set, x264
In-Reply-To: <201212131903.36859.info@denisgottardello.it>
References: <201212131553.09147.info@denisgottardello.it>
	<loom.20121213T180823-434@post.gmane.org>
	<201212131903.36859.info@denisgottardello.it>
Message-ID: <201212141411.05879.info@denisgottardello.it>

In data gioved? 13 dicembre 2012 19:03:36, Denis ha scritto:
> Which is the right way for change the profile? Have you got an example?
> I need a no latency solution.
> 
> 
> 
> 
> See the ffmpeg dump:
> 
> 
> [libx264 @ 0x92497c0] using cpu capabilities: none!
> [libx264 @ 0x92497c0] profile High, level 1.2
> [libx264 @ 0x92497c0] frame I:4     Avg QP:21,97  size:  8554
> [libx264 @ 0x92497c0] frame P:59    Avg QP:22,14  size:  1394

Nobody?


-- 
www.denisgottardello.it
Skype: mrdebug
Videosurveillance and home automation! 
http://www.denisgottardello.it/DomusBoss/DomusBossIndice.php

From flamz3d at gmail.com  Fri Dec 14 17:38:53 2012
From: flamz3d at gmail.com (Dominic Laflamme)
Date: Fri, 14 Dec 2012 11:38:53 -0500
Subject: [Libav-user] what's wrong with this loop?
Message-ID: <CALkdtKbhOvQ4rORAhp+t0FkpCPQue74Ur8WaYWwUnzHonHwxVQ@mail.gmail.com>

Hello,
I am attempting to sequentially fetch frames from a movie encoded with
libx264. I am using a typical av_read_frame / av_decode_video2 loop (found
in most demos out there) to fetch each frames. Like this:

int nbFrames = 0;
while ( av_read_frame () >= 0 )
{
av_decode_video2(&frameFinished);
if ( frameFinished )
{
nbFrames++;
}
}

Now, my video file has 1000 frames, but after the loop runs, nbFrames only
got incremented to 994.

So, my question is this: why would I only get 994 complete frames out of
1000? Clearly this has to do with some sort of DTS / PTS issue, but I was
under the impression that this type of loop ensured that frames would be
reordered properly (hence if(frameFinished)).

(btw, if I dump all frames using cmdline ffmpeg, I get 1000 files..., so my
video file is not corrupted or anything like that);

Any insight would be greatly appreciated!

regards,
?Dominic
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121214/1735eda6/attachment.html>

From jettoblack at gmail.com  Fri Dec 14 18:30:17 2012
From: jettoblack at gmail.com (Jason Livingston)
Date: Fri, 14 Dec 2012 12:30:17 -0500
Subject: [Libav-user] what's wrong with this loop?
In-Reply-To: <CALkdtKbhOvQ4rORAhp+t0FkpCPQue74Ur8WaYWwUnzHonHwxVQ@mail.gmail.com>
References: <CALkdtKbhOvQ4rORAhp+t0FkpCPQue74Ur8WaYWwUnzHonHwxVQ@mail.gmail.com>
Message-ID: <CAOJT_J1VQhQ1PLRVAzHKHkOMPbyWScB=auQGvwEJkeSfqZYpWA@mail.gmail.com>

On Fri, Dec 14, 2012 at 11:38 AM, Dominic Laflamme <flamz3d at gmail.com>wrote:

> Hello,
> I am attempting to sequentially fetch frames from a movie encoded with
> libx264. I am using a typical av_read_frame / av_decode_video2 loop (found
> in most demos out there) to fetch each frames. Like this:
>
> int nbFrames = 0;
> while ( av_read_frame () >= 0 )
> {
>  av_decode_video2(&frameFinished);
> if ( frameFinished )
>  {
> nbFrames++;
> }
> }
>
> Now, my video file has 1000 frames, but after the loop runs, nbFrames only
> got incremented to 994.
>

Hi Dominic,
Your loop stops when av_read_frame() returns an error such as EOF.  There
may still be decoded frames in memory.  To get those frames out, pass an
empty AVPacket (size = 0, data = NULL) into av_decode_video2() until
frameFinished returns 0.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121214/5126d3bb/attachment.html>

From flamz3d at gmail.com  Fri Dec 14 18:57:13 2012
From: flamz3d at gmail.com (Dominic Laflamme)
Date: Fri, 14 Dec 2012 12:57:13 -0500
Subject: [Libav-user] what's wrong with this loop?
In-Reply-To: <CAOJT_J1VQhQ1PLRVAzHKHkOMPbyWScB=auQGvwEJkeSfqZYpWA@mail.gmail.com>
References: <CALkdtKbhOvQ4rORAhp+t0FkpCPQue74Ur8WaYWwUnzHonHwxVQ@mail.gmail.com>
	<CAOJT_J1VQhQ1PLRVAzHKHkOMPbyWScB=auQGvwEJkeSfqZYpWA@mail.gmail.com>
Message-ID: <CALkdtKZ6JTbYVG5Z_OPruDr6FO5_1pEXDwx-U38CsRKWPS7peg@mail.gmail.com>

that did it.
Thanks Jason!

?Dominic
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121214/b2a181da/attachment.html>

From nkipe at tatapowersed.com  Mon Dec 17 11:19:56 2012
From: nkipe at tatapowersed.com (Navin)
Date: Mon, 17 Dec 2012 15:49:56 +0530
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
	array using Purify
Message-ID: <50CEF1CC.50308@tatapowersed.com>

I've got a piece of code which extracts every rgb pixel from a video and 
displays it without any problem. But when I ran it with Purify, I got 
these errors:
*
[E] ABR: Array bounds read in GetFrame {197 occurrences}
         Reading 1 byte from 0x04ae0040 (1 byte at 0x04ae0040 illegal)
         Address 0x04ae0040 is 16 bytes before the beginning of a 
1769472 byte block at 0x04ae0050
         Address 0x04ae0040 points to a malloc'd block in heap 0x003a0000
         Thread ID: 0x17c4*

So I reduced the code to just this
*pAreaInMemory[0] = g_pFrameRGB->data[0][0];*
and I'm still getting the error (which I've shown above).
The code goes something like this (I've shown many of the 
initializations in GetFrame itself although they're supposed to be outside):

bool GetFrame()
{
    ...blah...
    AVFrame*  g_pFrameRGB = NULL;
    avpicture_fill_proc((AVPicture *) g_pFrameRGB, pbuffer, 
PIX_FMT_RGB24, g_metadata->WIDTH, g_metadata->HEIGHT);
    ...blah...
    sws_scale_proc(g_sws_ctx, (uint8_t const * const *)g_pFrame->data, 
g_pFrame->linesize, 0, g_metadata->HEIGHT, g_pFrameRGB->data, 
g_pFrameRGB->linesize );
    ...blah...
    unsigned char* pAreaInMemory = (unsigned char*) malloc( 
sizeof(unsigned char) * 1024 * 768 );
    ...get the video data here...
    pAreaInMemory[0] = g_pFrameRGB->data[0][0];
    ...blah...
    free(pAreaInMemory);
}

Why is there a bounds problem when reading from data[0][0] ? I noticed 
the way data is initialized in ffmpeg. Is this purify error something I 
can ignore (there's no memory leak. Just the bounds error) or can it 
lead to serious problems?

-- 
Navin

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121217/88e1fcfd/attachment.html>

From rjvbertin at gmail.com  Mon Dec 17 11:43:37 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Mon, 17 Dec 2012 11:43:37 +0100
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
	array using Purify
In-Reply-To: <50CEF1CC.50308@tatapowersed.com>
References: <50CEF1CC.50308@tatapowersed.com>
Message-ID: <C0B8ACDC-DDA6-4D4F-AE15-EEE0085A6F6D@gmail.com>

Just some general thoughts:

On Dec 17, 2012, at 11:19, Navin wrote:

> I've got a piece of code which extracts every rgb pixel from a video and displays it without any problem. But when I ran it with Purify, I got these errors:
> 
> [E] ABR: Array bounds read in GetFrame {197 occurrences}
>         Reading 1 byte from 0x04ae0040 (1 byte at 0x04ae0040 illegal)
>         Address 0x04ae0040 is 16 bytes before the beginning of a 1769472 byte block at 0x04ae0050
>         Address 0x04ae0040 points to a malloc'd block in heap 0x003a0000
>         Thread ID: 0x17c4
> 
> So I reduced the code to just this
> pAreaInMemory[0] = g_pFrameRGB->data[0][0];
> and I'm still getting the error (which I've shown above).  
> The code goes something like this (I've shown many of the initializations in GetFrame itself although they're supposed to be outside):


Are you compiling in 32 or 64 bits, i.e. are you reading 1 or 2 addresses ahead of the allocated block? 2 address locations ahead could mean that the data matrix indexing is 1-based and not 0-based, for instance.

Checking avcodec/avcodec.h, there's the following doxygen comment associated with the AVFrame->data member:
    /**
     * pointer to the picture/channel planes.
     * This might be different from the first allocated byte
     * - encoding: Set by user
     * - decoding: set by AVCodecContext.get_buffer()
     */

Maybe you're seeing exactly what's stated, that data[0][0] is different from the first allocated byte?

> Why is there a bounds problem when reading from data[0][0] ? I noticed the way data is initialized in ffmpeg. Is this purify error something I can ignore (there's no memory leak. Just the bounds error) or can it lead to serious problems? 

A leak error you could chose to ignore as they're often false alarms (memory deallocated elsewhere). A bounds error is another issue because in this case you're not just checking the location for a specific value but you're assuming it contains a valid pixel. Which it most likely will not if the location is indeed not within the allocated block. Whether that can lead to serious problems only you can assess, but on some platforms you'd get an exception even for reading an illegal address.

Ren?

From cmst at live.com  Mon Dec 17 09:59:01 2012
From: cmst at live.com (Dolevo Jay)
Date: Mon, 17 Dec 2012 08:59:01 +0000
Subject: [Libav-user] How to perform drift compensation in MPEG-2
 recorder
In-Reply-To: <CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>,
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>,
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
Message-ID: <DUB114-W498929C4AE9E7318463455AA320@phx.gbl>


Hello all,

I am having a very similar problem in my project. I capture the video frame by frame from the desktop screen and send them over network after encoding. This happens on the PC processor and I have no problem with this. It works perfect for now. However, I capture the audio separately with the help of an HID device which has its own processor. I can send the audio even without encoding but still the problem is the video and audio get out of sync and drift occurs although I sync the processors at the very beginning. Could anyone shed any light on this issue?

Thanks.
Jay

> Date: Thu, 13 Dec 2012 11:36:35 +0100
> From: mike at mikeversteeg.com
> To: libav-user at ffmpeg.org
> Subject: Re: [Libav-user] How to perform drift compensation in MPEG-2	recorder
> 
> Anyone?
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121217/a2d14f12/attachment.html>

From perez.didac at gmail.com  Mon Dec 17 09:31:06 2012
From: perez.didac at gmail.com (=?ISO-8859-1?Q?D=EDdac_P=E9rez?=)
Date: Mon, 17 Dec 2012 09:31:06 +0100
Subject: [Libav-user] Syncronize RTSP with computer time
Message-ID: <CAM0v-hJ2ZHF922=wbrUu1sbE66dw4DiOS7JZBunb0088KKm-SQ@mail.gmail.com>

I am successfully using libav to receive the video stream from an RTSP
network source. The point is that I need to syncronize my computer's time
with the video capturing, meaning that I need to know which datetime of my
computer corresponds to the first frame (pts = 0). My API calls are the
following ones:

    av_register_all()
    avcodec_register_all()
    avformat_network_init()
    avformat_open_input()
    avformat_find_stream_info()
    av_read_play()
    loop
      av_init_packet()
      av_read_frame()
      [...]
      av_free_packet
    end loop

With the calls above, I successfully read frames, but I do need to know how
can I know the exact absolute datetime that corresponds to the first frame,
since it has a pts of 0. Maybe I can use a time() function or GetSystemTime
(I am using Windows) between two calls of the above, but do not really know
how libav works. I will appreciate your help,

Kind regards,

--
D?dac P?rez
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121217/1ce70702/attachment.html>

From alexcohn at netvision.net.il  Tue Dec 18 10:27:38 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Tue, 18 Dec 2012 11:27:38 +0200
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
 array using Purify
In-Reply-To: <50CEF1CC.50308@tatapowersed.com>
References: <50CEF1CC.50308@tatapowersed.com>
Message-ID: <CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>

On Mon, Dec 17, 2012 at 12:19 PM, Navin <nkipe at tatapowersed.com> wrote:
> So I reduced the code to just this
> pAreaInMemory[0] = g_pFrameRGB->data[0][0];
> and I'm still getting the error (which I've shown above).
> The code goes something like this (I've shown many of the initializations in
> GetFrame itself although they're supposed to be outside):
>
> bool GetFrame()
> {
>    ...blah...
>    AVFrame*  g_pFrameRGB = NULL;
>    avpicture_fill_proc((AVPicture *) g_pFrameRGB, pbuffer, PIX_FMT_RGB24,
> g_metadata->WIDTH, g_metadata->HEIGHT);

avipicture_fill does not expect NULL pointer to AVPicture as first
parameter. The pointer must be initialized (see doc/examples/muxing.c
for the correct usage).

BR,
Alex

From mtaha.ansari at gmail.com  Tue Dec 18 12:06:27 2012
From: mtaha.ansari at gmail.com (Taha Ansari)
Date: Tue, 18 Dec 2012 16:06:27 +0500
Subject: [Libav-user] Webcam streaming - network transfer rate does not
 change irrespective of decoder resolution
Message-ID: <CAK2cdgDVKyQsByRcEnJfbKJ8-fA_6HomnxmLuYhJiQWsjOa2Zg@mail.gmail.com>

Hi,

I am streaming webcam over udp using h264 raw format (Windows). When I try
this:

ffmpeg -f dshow -r 25 -s 160x120 -i video="A4 tech USB2.0 Camera"
udp://<ipaddr>:8888/a.h264

Changing from 160x120 to 640x480 (different allowable resolutions), I see
there is significant change in data transmitted over network per minute
that I have monitored.

But, when I try to do the same from my code, either by:

i- opening up decoder at native resolution (640x480 default), then scaling
it to required resolution, then stream this scaled frame to network, or
ii- opening up decoder at different resolution, using AVDictionary, like:

av_dict_set( &dict, "video_size", "320x240",0);
if(avformat_open_input(&this->
pFormatCtx, this->finalInputName.c_str(), pInputFormat, &dict)!=0)
      return -1; // Couldn't open file

so I am capturing at 320x240 resolution from example above, monitoring for
above both cases, I get approximately the same data transmission rate, i.e.
around 2.7 MB per minute.

I use baseline profile and ultrafast preset.

I think using -s 320x240 is the same as setting up dictionary options to
open up a decoder stream with a smaller resolution, but it doesn't seem to
work.

Logically, capturing at any resolution, then scaling it, and sending the
scaled frame over network should be taking up less space as well, but
that's not the case observed here.

Can anyone provide some guidance what I may be doing wrong?

Thanks for your time!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121218/c345ba50/attachment.html>

From rjvbertin at gmail.com  Tue Dec 18 13:02:31 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Tue, 18 Dec 2012 13:02:31 +0100
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
	array using Purify
In-Reply-To: <CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
Message-ID: <A87BF8EC-1DE3-4A76-BDA8-B8EEDBADFC1C@gmail.com>


On Dec 18, 2012, at 10:27, Alex Cohn wrote:

> On Mon, Dec 17, 2012 at 12:19 PM, Navin <nkipe at tatapowersed.com> wrote:
>> So I reduced the code to just this
>> pAreaInMemory[0] = g_pFrameRGB->data[0][0];
>> and I'm still getting the error (which I've shown above).
>> The code goes something like this (I've shown many of the initializations in
>> GetFrame itself although they're supposed to be outside):
>> 
>> bool GetFrame()
>> {
>>   ...blah...
>>   AVFrame*  g_pFrameRGB = NULL;
>>   avpicture_fill_proc((AVPicture *) g_pFrameRGB, pbuffer, PIX_FMT_RGB24,
>> g_metadata->WIDTH, g_metadata->HEIGHT);
> 
> avipicture_fill does not expect NULL pointer to AVPicture as first
> parameter. The pointer must be initialized (see doc/examples/muxing.c
> for the correct usage).
> 

Doh, everyone could have seen that g_pFrameRGB can't possibly be initialised in the snippet above, so much for my general suggestions :)

R.


From alexcohn at netvision.net.il  Tue Dec 18 13:32:05 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Tue, 18 Dec 2012 14:32:05 +0200
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
Message-ID: <CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>

On Mon, Dec 17, 2012 at 10:59 AM, Dolevo Jay <cmst at live.com> wrote:
> I capture the video frame by frame from the desktop screen and send them
> over network after encoding. This happens on the PC processor and I have no
> problem with this. It works perfect for now. However, I capture the audio
> separately with the help of an HID device which has its own processor. I can
> send the audio even without encoding but still the problem is the video and
> audio get out of sync and drift occurs although I sync the processors at the
> very beginning. Could anyone shed any light on this issue?

You can set the video PTS in sync with the current audio "clock". Most
likely, your audio packets arrive at fixed nominal frequency which is
not precise (from the point of view of your PC clock). In this case
you should calculate PTS by multiplying the expected audio packet
duration by number of actual audio packets received.

BR,
Alex

From rjvbertin at gmail.com  Tue Dec 18 13:44:49 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Tue, 18 Dec 2012 13:44:49 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
Message-ID: <BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>


On Dec 18, 2012, at 13:32, Alex Cohn wrote:

> On Mon, Dec 17, 2012 at 10:59 AM, Dolevo Jay <cmst at live.com> wrote:
>> I capture the video frame by frame from the desktop screen and send them
>> over network after encoding. This happens on the PC processor and I have no
>> problem with this. It works perfect for now. However, I capture the audio
>> separately with the help of an HID device which has its own processor. I can
>> send the audio even without encoding but still the problem is the video and
>> audio get out of sync and drift occurs although I sync the processors at the
>> very beginning. Could anyone shed any light on this issue?
> 
> You can set the video PTS in sync with the current audio "clock". Most
> likely, your audio packets arrive at fixed nominal frequency which is
> not precise (from the point of view of your PC clock). In this case
> you should calculate PTS by multiplying the expected audio packet
> duration by number of actual audio packets received.


Even if a nominal fixed frequency were precise enough with respect to the video frequency it seems you're probably looking at transport delays that are neither constant nor identical (enough) for the two streams. It's not clear if you send video and audio over the same network stream if you maybe even capture the audio on a separate host (possibly the remote)? How big is the drift, I suppose we're not talking about a 1 or 2 frames jitter?

R.

From david.longest at apx-labs.com  Tue Dec 18 23:14:33 2012
From: david.longest at apx-labs.com (David Longest)
Date: Tue, 18 Dec 2012 22:14:33 +0000
Subject: [Libav-user] H263 dts from avformat_find_stream_info
Message-ID: <D806546EE181BF499186ECAB48AA8D205692C6E0@CH1PRD0610MB392.namprd06.prod.outlook.com>

Hello All,

I am trying to use libav to decode an h263 stream. While I can successfully decode the stream and display decoded frames, I have found that I get about a 5 second stall in avformat_find_stream_info. After further investigation, it seems that the AVStream's first_dts is always set to AV_NOPTS_VALUE and the for loop in avformat_find_stream_info attempts to wait for that data to be populated.

The only conclusion I have come to is that the h263 parser does not ever populate the "dts" portion of the incoming packets. What is the proper way to handle setting the decompression timestamp on h263 data?

Thanks,
David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121218/52e91eee/attachment.html>

From mtaha.ansari at gmail.com  Tue Dec 18 12:07:29 2012
From: mtaha.ansari at gmail.com (Taha Ansari)
Date: Tue, 18 Dec 2012 16:07:29 +0500
Subject: [Libav-user] Scaled down h264 video has almost same size as normal
	sized video; why?
Message-ID: <CAK2cdgC8PeUvf-evgNmYJSJ2-SgpCKVhPhubSt_eQu-vN__TBQ@mail.gmail.com>

Hi,

I have modified muxer program that saves h264 video to disk of 01 minute
duration, as follows:

******************************
************************************************************
/*
 * Copyright (c) 2001 Fabrice Bellard
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
copy
 * of this software and associated documentation files (the "Software"), to
deal
 * in the Software without restriction, including without limitation the
rights
 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 * copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included
in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 * THE SOFTWARE.
 */

/**
 * @file
 * libavcodec API use example.
 *
 * Note that libavcodec only handles codecs (mpeg, mpeg4, etc...),
 * not file formats (avi, vob, mp4, mov, mkv, mxf, flv, mpegts, mpegps,
etc...). See library 'libavformat' for the
 * format handling
 */

#include <Windows.h>
#include <math.h>
extern "C"
{
#include "libavcodec/avcodec.h"
#include "libavformat/avformat.h"
#include "libavdevice/avdevice.h"
#include "libswscale/swscale.h"
#include "libavutil/dict.h"
#include "libavutil/error.h"
#include "libavutil/opt.h"
}

#define DECODER_WIDTH    640
#define DECODER_HEIGHT    480

#define ENCODER_WIDTH    640
#define ENCODER_HEIGHT    480

/* 5 seconds stream duration */
#define STREAM_DURATION   60.0
#define STREAM_FRAME_RATE 25 /* 25 images/s */
#define STREAM_NB_FRAMES  ((int)(STREAM_DURATION * STREAM_FRAME_RATE))
#define STREAM_PIX_FMT    AV_PIX_FMT_YUV420P /* default pix_fmt */

static int sws_flags = SWS_BICUBIC;

/**************************************************************/
/* audio output */

static float t, tincr, tincr2;
static int16_t *samples;
static int audio_input_frame_size;

/* Add an output stream. */
static AVStream *add_stream(AVFormatContext *oc, AVCodec **codec,
                            enum AVCodecID codec_id)
{
    AVCodecContext *c;
    AVStream *st;

    /* find the encoder */
    *codec = avcodec_find_encoder(codec_id);
    if (!(*codec)) {
        fprintf(stderr, "Could not find codec\n");
        exit(1);
    }

    st = avformat_new_stream(oc, *codec);
    if (!st) {
        fprintf(stderr, "Could not allocate stream\n");
        exit(1);
    }
    st->id = oc->nb_streams-1;
    c = st->codec;

    switch ((*codec)->type) {
    case AVMEDIA_TYPE_AUDIO:
        st->id = 1;
        c->sample_fmt  = AV_SAMPLE_FMT_S16;
        c->bit_rate    = 64000;
        c->sample_rate = 44100;
        c->channels    = 2;
        break;

    case AVMEDIA_TYPE_VIDEO:
        avcodec_get_context_defaults3(c, *codec);
        c->codec_id = codec_id;

        c->bit_rate = 400000;
        /* Resolution must be a multiple of two. */
        c->width    = ENCODER_WIDTH;
        c->height   = ENCODER_HEIGHT;
        /* timebase: This is the fundamental unit of time (in seconds) in
terms
         * of which frame timestamps are represented. For fixed-fps content,
         * timebase should be 1/framerate and timestamp increments should be
         * identical to 1. */
        c->time_base.den = STREAM_FRAME_RATE;
        c->time_base.num = 1;
        c->gop_size      = 12; /* emit one intra frame every twelve frames
at most */
        c->pix_fmt       = STREAM_PIX_FMT;
        if (c->codec_id == AV_CODEC_ID_MPEG2VIDEO) {
            /* just for testing, we also add B frames */
            c->max_b_frames = 2;
        }
        if (c->codec_id == AV_CODEC_ID_MPEG1VIDEO) {
            /* Needed to avoid using macroblocks in which some coeffs
overflow.
             * This does not happen with normal video, it just happens here
as
             * the motion of the chroma plane does not match the luma
plane. */
            c->mb_decision = 2;
        }
    break;

    default:
        break;
    }

    /* Some formats want stream headers to be separate. */
    if (oc->oformat->flags & AVFMT_GLOBALHEADER)
        c->flags |= CODEC_FLAG_GLOBAL_HEADER;

    return st;
}

/**************************************************************/
/* audio output */


static void open_audio(AVFormatContext *oc, AVCodec *codec, AVStream *st)
{
    AVCodecContext *c;

    c = st->codec;

    /* open it */
    if (avcodec_open2(c, codec, NULL) < 0) {
        fprintf(stderr, "Could not open audio codec\n");
        exit(1);
    }

    /* init signal generator */
    t     = 0;
    tincr = 2 * M_PI * 110.0 / c->sample_rate;
    /* increment frequency by 110 Hz per second */
    tincr2 = 2 * M_PI * 110.0 / c->sample_rate / c->sample_rate;

    if (c->codec->capabilities & CODEC_CAP_VARIABLE_FRAME_SIZE)
        audio_input_frame_size = 10000;
    else
        audio_input_frame_size = c->frame_size;
    samples = (int16_t *) av_malloc(audio_input_frame_size *
                        av_get_bytes_per_sample(c->sample_fmt) *
                        c->channels);
    if (!samples) {
        fprintf(stderr, "Could not allocate audio samples buffer\n");
        exit(1);
    }
}

/* Prepare a 16 bit dummy audio frame of 'frame_size' samples and
 * 'nb_channels' channels. */
static void get_audio_frame(int16_t *samples, int frame_size, int
nb_channels)
{
    int j, i, v;
    int16_t *q;

    q = samples;
    for (j = 0; j < frame_size; j++) {
        v = (int)(sin(t) * 10000);
        for (i = 0; i < nb_channels; i++)
            *q++ = v;
        t     += tincr;
        tincr += tincr2;
    }
}

static void write_audio_frame(AVFormatContext *oc, AVStream *st)
{
    AVCodecContext *c;
    AVPacket pkt = { 0 }; // data and size must be 0;
    AVFrame *frame = avcodec_alloc_frame();
    int got_packet, ret;

    av_init_packet(&pkt);
    c = st->codec;

    get_audio_frame(samples, audio_input_frame_size, c->channels);
    frame->nb_samples = audio_input_frame_size;
    avcodec_fill_audio_frame(frame, c->channels, c->sample_fmt,
                             (uint8_t *)samples,
                             audio_input_frame_size *
                             av_get_bytes_per_sample(c->sample_fmt) *
                             c->channels, 1);

    ret = avcodec_encode_audio2(c, &pkt, frame, &got_packet);
    if (ret < 0) {
        fprintf(stderr, "Error encoding audio frame\n");
        exit(1);
    }

    if (!got_packet)
        return;

    pkt.stream_index = st->index;

    /* Write the compressed frame to the media file. */
    if (av_interleaved_write_frame(oc, &pkt) != 0) {
        fprintf(stderr, "Error while writing audio frame\n");
        exit(1);
    }
    avcodec_free_frame(&frame);
}

static void close_audio(AVFormatContext *oc, AVStream *st)
{
    avcodec_close(st->codec);

    av_free(samples);
}

/**************************************************************/
/* video output */

static AVFrame *frame;
static AVPicture src_picture, dst_picture;
static int frame_count;

static void open_video(AVFormatContext *oc, AVCodec *codec, AVStream *st)
{
    int ret;
    AVCodecContext *c = st->codec;

    /* open the codec */
    if (avcodec_open2(c, codec, NULL) < 0) {
        fprintf(stderr, "Could not open video codec\n");
        exit(1);
    }

    /* allocate and init a re-usable frame */
    frame = avcodec_alloc_frame();
    if (!frame) {
        fprintf(stderr, "Could not allocate video frame\n");
        exit(1);
    }

    /* Allocate the encoded raw picture. */
    ret = avpicture_alloc(&dst_picture, c->pix_fmt, c->width, c->height);
    if (ret < 0) {
        fprintf(stderr, "Could not allocate picture\n");
        exit(1);
    }

    /* If the output format is not YUV420P, then a temporary YUV420P
     * picture is needed too. It is then converted to the required
     * output format. */
    if (1) {
        ret = avpicture_alloc(&src_picture, PIX_FMT_YUV420P, DECODER_WIDTH,
DECODER_HEIGHT);
        if (ret < 0) {
            fprintf(stderr, "Could not allocate temporary picture\n");
            exit(1);
        }
    }
    //if (c->pix_fmt != AV_PIX_FMT_YUV420P) {
    //    ret = avpicture_alloc(&src_picture, AV_PIX_FMT_YUV420P, c->width,
c->height);
    //    if (ret < 0) {
    //        fprintf(stderr, "Could not allocate temporary picture\n");
    //        exit(1);
    //    }
    //}

    /* copy data and linesize picture pointers to frame */
    *((AVPicture *)frame) = dst_picture;
}

/* Prepare a dummy image. */
static void fill_yuv_image(AVPicture *pict, int frame_index,
                           int width, int height)
{
    int x, y, i;

    i = frame_index;

    /* Y */
    for (y = 0; y < height; y++)
        for (x = 0; x < width; x++)
            pict->data[0][y * pict->linesize[0] + x] = x + y + i * 3;

    /* Cb and Cr */
    for (y = 0; y < height / 2; y++) {
        for (x = 0; x < width / 2; x++) {
            pict->data[1][y * pict->linesize[1] + x] = 128 + y + i * 2;
            pict->data[2][y * pict->linesize[2] + x] = 64 + x + i * 5;
        }
    }
}

static void write_video_frame(AVFormatContext *oc, AVStream *st)
{
    int ret;
    static struct SwsContext *sws_ctx;
    AVCodecContext *c = st->codec;

    if (frame_count >= STREAM_NB_FRAMES) {
        /* No more frames to compress. The codec has a latency of a few
         * frames if using B-frames, so we get the last frames by
         * passing the same picture again. */
    } else {
        if (c->pix_fmt != AV_PIX_FMT_YUV420P) {
            /* as we only generate a YUV420P picture, we must convert it
             * to the codec pixel format if needed */
            if (!sws_ctx) {
                sws_ctx = sws_getContext(c->width, c->height,
AV_PIX_FMT_YUV420P,
                                         c->width, c->height, c->pix_fmt,
                                         sws_flags, NULL, NULL, NULL);
                if (!sws_ctx) {
                    fprintf(stderr,
                            "Could not initialize the conversion
context\n");
                    exit(1);
                }
            }
            fill_yuv_image(&src_picture, frame_count, c->width, c->height);
            sws_scale(sws_ctx,
                      (const uint8_t * const *)src_picture.data,
src_picture.linesize,
                      0, c->height, dst_picture.data, dst_picture.linesize);
        } else
        {
             /* as we only generate a YUV420P picture, we must convert it
            * to the codec pixel format if needed */
            if (!sws_ctx)
            {
                sws_ctx = sws_getContext( DECODER_WIDTH, DECODER_HEIGHT,
PIX_FMT_YUV420P,
                    c->width, c->height, c->pix_fmt,
                    sws_flags, NULL, NULL, NULL);
                if (!sws_ctx)
                {
                    fprintf(stderr,
                        "Could not initialize the conversion context\n");
                    exit(1);
                }
            }
            fill_yuv_image(&src_picture, frame_count, DECODER_WIDTH,
DECODER_HEIGHT);
            sws_scale(sws_ctx,
                (const uint8_t * const *)src_picture.data,
src_picture.linesize,
                0, DECODER_WIDTH, dst_picture.data, dst_picture.linesize);
            //fill_yuv_image(&dst_picture, frame_count, c->width,
c->height);
        }
    }

    if (oc->oformat->flags & AVFMT_RAWPICTURE) {
        /* Raw video case - directly store the picture in the packet */
        AVPacket pkt;
        av_init_packet(&pkt);

        pkt.flags        |= AV_PKT_FLAG_KEY;
        pkt.stream_index  = st->index;
        pkt.data          = dst_picture.data[0];
        pkt.size          = sizeof(AVPicture);

        ret = av_interleaved_write_frame(oc, &pkt);
    } else {
        /* encode the image */
        AVPacket pkt;
        int got_output;

        av_init_packet(&pkt);
        pkt.data = NULL;    // packet data will be allocated by the encoder
        pkt.size = 0;

        ret = avcodec_encode_video2(c, &pkt, frame, &got_output);
        if (ret < 0) {
            fprintf(stderr, "Error encoding video frame\n");
            exit(1);
        }

        /* If size is zero, it means the image was buffered. */
        if (got_output) {
            if (c->coded_frame->key_frame)
                pkt.flags |= AV_PKT_FLAG_KEY;

            pkt.stream_index = st->index;

            /* Write the compressed frame to the media file. */
            ret = av_interleaved_write_frame(oc, &pkt);
        } else {
            ret = 0;
        }
    }
    if (ret != 0) {
        fprintf(stderr, "Error while writing video frame\n");
        exit(1);
    }
    frame_count++;
}

static void close_video(AVFormatContext *oc, AVStream *st)
{
    avcodec_close(st->codec);
    av_free(src_picture.data[0]);
    av_free(dst_picture.data[0]);
    av_free(frame);
}

/**************************************************************/
/* media file output */

int main(int argc, char **argv)
{
    char filename[99]="";
    AVOutputFormat *fmt;
    AVFormatContext *oc;
    AVStream *audio_st, *video_st;
    AVCodec *audio_codec, *video_codec;
    double audio_pts, video_pts;
    int i;

    /* Initialize libavcodec, and register all codecs and formats. */
    av_register_all();

    if (argc != 2) {
        printf("usage: %s output_file\n"
               "API example program to output a media file with
libavformat.\n"
               "This program generates a synthetic audio and video stream,
encodes and\n"
               "muxes them into a file named output_file.\n"
               "The output format is automatically guessed according to the
file extension.\n"
               "Raw images can also be output by using '%%d' in the
filename.\n"
               "\n", argv[0]);
        return 1;
    }

    strcpy(filename,"a.h264");
    /* allocate the output media context */
    avformat_alloc_output_context2(&oc, NULL, NULL, filename);
    if (!oc) {
        printf("Could not deduce output format from file extension: using
MPEG.\n");
        avformat_alloc_output_context2(&oc, NULL, "mpeg", filename);
    }
    if (!oc) {
        return 1;
    }
    fmt = oc->oformat;

    /* Add the audio and video streams using the default format codecs
     * and initialize the codecs. */
    video_st = NULL;
    audio_st = NULL;

    if (fmt->video_codec != AV_CODEC_ID_NONE) {
        video_st = add_stream(oc, &video_codec, fmt->video_codec);
    }
    if (fmt->audio_codec != AV_CODEC_ID_NONE) {
        audio_st = add_stream(oc, &audio_codec, fmt->audio_codec);
    }

    /* Now that all the parameters are set, we can open the audio and
     * video codecs and allocate the necessary encode buffers. */
    if (video_st)
        open_video(oc, video_codec, video_st);
    if (audio_st)
        open_audio(oc, audio_codec, audio_st);

    av_dump_format(oc, 0, filename, 1);

    /* open the output file, if needed */
    if (!(fmt->flags & AVFMT_NOFILE)) {
        if (avio_open(&oc->pb, filename, AVIO_FLAG_WRITE) < 0) {
            fprintf(stderr, "Could not open '%s'\n", filename);
            return 1;
        }
    }

    /* Write the stream header, if any. */
    if (avformat_write_header(oc, NULL) < 0) {
        fprintf(stderr, "Error occurred when opening output file\n");
        return 1;
    }

    if (frame)
        frame->pts = 0;
    for (;;) {
        /* Compute current audio and video time. */
        if (audio_st)
            audio_pts = (double)audio_st->pts.val * audio_st->time_base.num
/ audio_st->time_base.den;
        else
            audio_pts = 0.0;

        if (video_st)
            video_pts = (double)video_st->pts.val * video_st->time_base.num
/
                        video_st->time_base.den;
        else
            video_pts = 0.0;

        if ((!audio_st || audio_pts >= STREAM_DURATION) &&
            (!video_st || video_pts >= STREAM_DURATION))
            break;

        /* write interleaved audio and video frames */
        if (!video_st || (video_st && audio_st && audio_pts < video_pts)) {
            write_audio_frame(oc, audio_st);
        } else {
            write_video_frame(oc, video_st);
            frame->pts += av_rescale_q(1, video_st->codec->time_base,
video_st->time_base);
        }
        Sleep(40);
    }

    /* Write the trailer, if any. The trailer must be written before you
     * close the CodecContexts open when you wrote the header; otherwise
     * av_write_trailer() may try to use memory that was freed on
     * av_codec_close(). */
    av_write_trailer(oc);

    /* Close each codec. */
    if (video_st)
        close_video(oc, video_st);
    if (audio_st)
        close_audio(oc, audio_st);

    /* Free the streams. */
    for (i = 0; i < oc->nb_streams; i++) {
        av_freep(&oc->streams[i]->codec);
        av_freep(&oc->streams[i]);
    }

    if (!(fmt->flags & AVFMT_NOFILE))
        /* Close the output file. */
        avio_close(oc->pb);

    /* free the stream */
    av_free(oc);

    return 0;
}
*******************************************************************************

After first run, if we change

#define ENCODER_WIDTH 640
#define ENCODER_HEIGHT 480

to

#define ENCODER_WIDTH 160
#define ENCODER_HEIGHT 120

and run again, file size is almost the same. Question now is why? shouldn't
smaller sized video be taking up much lesser space than the original sized
video?

Same thing happens if instead of saving to hard-disk I transmit over udp;
it is taking up similar bandwidth consumption.

Can anyone provide guidance for above?

Thanks for your time!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121218/cab0a8c3/attachment.html>

From mtaha.ansari at gmail.com  Wed Dec 19 08:58:50 2012
From: mtaha.ansari at gmail.com (Taha Ansari)
Date: Wed, 19 Dec 2012 12:58:50 +0500
Subject: [Libav-user] Changing decoder resolution does not change file
	output size for h264 format
Message-ID: <CAK2cdgA+YR8nGrbepYvv9AbD9MrZ=H10GjT932w432T0Bdkefg@mail.gmail.com>

Hi all!

I have made a test application that captures from webcam (using dshow
filter), and saves it as a file on hard disk. My camera supports natively
640x480 resolution, so I had to forcefully open it in smaller resolution to
conduct this test:

    pFormatCtx = avformat_alloc_context();
    AVDictionary *dict = NULL;
    av_dict_set( &dict, "video_size", "160x120" ,0);
    if(avformat_open_input(&this->pFormatCtx, this->finalInputName.c_str(),
pInputFormat, &dict)!=0)
        return -1; // Couldn't open file

finalInputName is the correctly formatted video:audio string ffmpeg
recognizes for dshow.

Problem now is, no matter what the decoder's resolution is (160x120,
320x240, 640x480 etc.), output file size for approximately one minute
duration is always the same - around 3 MB.

I tried encoding to other formats like 'mpg', 'ogg', their video sizes
seems to be appropriately scaled, then what am I doing wrong with h264
format, fail to understand?

Can anyone kindly provide some guidance?

Thanks for your time!

Best regards...
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121219/77d873c9/attachment.html>

From alexcohn at netvision.net.il  Wed Dec 19 09:02:35 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Wed, 19 Dec 2012 10:02:35 +0200
Subject: [Libav-user] Scaled down h264 video has almost same size as
 normal sized video; why?
In-Reply-To: <CAK2cdgC8PeUvf-evgNmYJSJ2-SgpCKVhPhubSt_eQu-vN__TBQ@mail.gmail.com>
References: <CAK2cdgC8PeUvf-evgNmYJSJ2-SgpCKVhPhubSt_eQu-vN__TBQ@mail.gmail.com>
Message-ID: <CADiKgog6H6ftFA2+x3=Gazk8jfsuN==eB=vqnOSPRQyJNqpEuw@mail.gmail.com>

On Tue, Dec 18, 2012 at 1:07 PM, Taha Ansari <mtaha.ansari at gmail.com> wrote:
> After first run, if we change
>
> #define ENCODER_WIDTH 640
> #define ENCODER_HEIGHT 480
>
> to
>
> #define ENCODER_WIDTH 160
> #define ENCODER_HEIGHT 120
>
> and run again, file size is almost the same. Question now is why? shouldn't
> smaller sized video be taking up much lesser space than the original sized
> video?

Note that you have the encoder bit rate hard coded:

> c->bit_rate = 400000;

... so, one would expect a well-tempered encoder to produce same
encoded frame sizes. You could try to compare the resulting video
quality in two cases. E.g., decode the two streams and display them at
160x120. You are expected to see very similar images. If you increase
the requested bit rate, the VGA stream will still be improving, but
the 160x120 stream is probably saturated. You can consult the "video
bitrate calculator" at http://www.dr-lex.be/info-stuff/videocalc.html
to find the optimal bit rate - it would be more than 1Mbps for VGA at
25 FPS.

BR,
Alex

From mtaha.ansari at gmail.com  Wed Dec 19 10:29:15 2012
From: mtaha.ansari at gmail.com (Taha Ansari)
Date: Wed, 19 Dec 2012 14:29:15 +0500
Subject: [Libav-user] Scaled down h264 video has almost same size as
 normal sized video; why?
In-Reply-To: <CADiKgog6H6ftFA2+x3=Gazk8jfsuN==eB=vqnOSPRQyJNqpEuw@mail.gmail.com>
References: <CAK2cdgC8PeUvf-evgNmYJSJ2-SgpCKVhPhubSt_eQu-vN__TBQ@mail.gmail.com>
	<CADiKgog6H6ftFA2+x3=Gazk8jfsuN==eB=vqnOSPRQyJNqpEuw@mail.gmail.com>
Message-ID: <CAK2cdgDa=umFLsSLrsYPRWJfPVymunLbVz0yVmr6VbUyQfX3XA@mail.gmail.com>

Note that you have the encoder bit rate hard coded:

>
> > c->bit_rate = 400000;
>
> ... so, one would expect a well-tempered encoder to produce same
> encoded frame sizes. You could try to compare the resulting video
> quality in two cases. E.g., decode the two streams and display them at
> 160x120. You are expected to see very similar images. If you increase
> the requested bit rate, the VGA stream will still be improving, but
> the 160x120 stream is probably saturated. You can consult the "video
> bitrate calculator" at http://www.dr-lex.be/info-stuff/videocalc.html
> to find the optimal bit rate - it would be more than 1Mbps for VGA at
> 25 FPS.
>
> BR,
> Alex
>

Hi Alex!

Thanks for the link - you simply nailed down the problem for me; it was
bit_rate factor after all!

Thanks again for your guidance!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121219/34655486/attachment.html>

From mike at mikeversteeg.com  Wed Dec 19 11:31:05 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Wed, 19 Dec 2012 11:31:05 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
Message-ID: <CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>

>> On Mon, Dec 17, 2012 at 10:59 AM, Dolevo Jay <cmst at live.com> wrote:
>>> I capture the video frame by frame from the desktop screen and send them
>>> over network after encoding. This happens on the PC processor and I have no
>>> problem with this. It works perfect for now. However, I capture the audio
>>> separately with the help of an HID device which has its own processor. I can
>>> send the audio even without encoding but still the problem is the video and
>>> audio get out of sync and drift occurs although I sync the processors at the
>>> very beginning. Could anyone shed any light on this issue?
>>
>> You can set the video PTS in sync with the current audio "clock". Most
>> likely, your audio packets arrive at fixed nominal frequency which is
>> not precise (from the point of view of your PC clock). In this case
>> you should calculate PTS by multiplying the expected audio packet
>> duration by number of actual audio packets received.
>
>
> Even if a nominal fixed frequency were precise enough with respect to the video frequency it seems you're probably looking at transport delays that are neither constant nor identical (enough) for the two streams. It's not clear if you send video and audio over the same network stream if you maybe even capture the audio on a separate host (possibly the remote)? How big is the drift, I suppose we're not talking about a 1 or 2 frames jitter?

I have a minor jitter, there is no network involved.

I aforementioned example all PTS's are calculated automatically. Where
do I insert my own video PTS? I assume it is in the write_video_frame
function, but is it before the call to avcodec_encode_video() or
after? Do I set the codec PTS or the packet, or both?

Here's (I think) the relevant piece of code:

// encode the image
int out_size = avcodec_encode_video(c, pVideoOut, iVideoOutSize, stPicture);
// if zero size, it means the image was buffered
if (out_size > 0) {
  AVPacket pkt;
  av_init_packet(&pkt);
  if (c->coded_frame->pts != AV_NOPTS_VALUE)
    pkt.pts = av_rescale_q(c->coded_frame->pts, c->time_base, st->time_base);
}

Thanks,

Mike

From cehoyos at ag.or.at  Wed Dec 19 11:59:23 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Wed, 19 Dec 2012 10:59:23 +0000 (UTC)
Subject: [Libav-user] Decoding HD videos on iPad devices
References: <D03BBEB0-1A2F-4EF3-BB13-28FD05D41075@instigatemobile.com>
Message-ID: <loom.20121219T115830-976@post.gmane.org>

Gohar Hovhannisyan <gohar_h at ...> writes:

> We wonder whether it is possible to perform decoding on 
> iPad GPU to speed up decoding process (by compiling
> it with special flags, or with some patches) 

Only if you implement it.

> or run it parallel on dual core (iPad2) / quad-core (new iPad)
> or are there any other methods to speed up decoding process?

You can set the number of threads that decoding should use.

Carl Eugen


From mike at mikeversteeg.com  Wed Dec 19 13:46:12 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Wed, 19 Dec 2012 13:46:12 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
Message-ID: <CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>

PS: This (http://lists.libav.org/pipermail/libav-api/2012-April/000473.html)
thread is about the exact same question and again no one seems to know
the answer.

From mike at mikeversteeg.com  Wed Dec 19 14:07:01 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Wed, 19 Dec 2012 14:07:01 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
Message-ID: <CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>

Found another thread with similar question:
http://libav-users.943685.n4.nabble.com/Libav-user-PTS-values-and-writing-encoded-audio-and-video-frames-td4551001.html

Again, not answered.

I do not think no one has ever considered video and audio clocks can
slightly drift away from each other, so the lack of feedback is likely
because the questions are wrong. I am therefore trying to rephrase my,
and all these similar questions, as follows..

Using the sample code referred to, there is a direct connection
between a video frame (lasting 1/fps) and an audio frame (fixed length
lasting length/samplerate). This works fine as long as the audio and
video clock are in perfect sync. However when has only the smallest
difference from the other, after a while (could be hours) an error
accumulates resulting in a slight AV sync problem. Obviously this
problem can be fixed by monitoring this difference and resampling the
audio data, but it uses resources. Changing the audio packet size
dynamically seems a solution, but it does not seem possible. Is there
another way, like "modulating" some time stamp and if so, how?

Mike

From rjvbertin at gmail.com  Wed Dec 19 15:26:25 2012
From: rjvbertin at gmail.com (RJV Bertin)
Date: Wed, 19 Dec 2012 15:26:25 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
Message-ID: <496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>



On 19 Dec 2012, at 14:07, Mike Versteeg <mike at mikeversteeg.com> wrote:

> Found another thread with similar question:
> http://libav-users.943685.n4.nabble.com/Libav-user-PTS-values-and-writing-encoded-audio-and-video-frames-td4551001.html
> 
> Again, not answered.
> 
> I do not think no one has ever considered video and audio clocks can
> slightly drift away from each other, so the lack of feedback is likely
> because the questions are wrong. I am therefore trying to rephrase my,
> and all these similar questions, as follows..
> 
> Using the sample code referred to, there is a direct connection
> between a video frame (lasting 1/fps) and an audio frame (fixed length
> lasting length/samplerate). This works fine as long as the audio and
> video clock are in perfect sync. However when has only the smallest
> difference from the other, after a while (could be hours) an error
> accumulates resulting in a slight AV sync problem. Obviously this
> problem can be fixed by monitoring this difference and resampling the
> audio data, but it uses resources. Changing the audio packet size
> dynamically seems a solution, but it does not seem possible. Is there
> another way, like "modulating" some time stamp and if so, how?
> 
> Mike
> _______________________________________________
> 

Maybe you should ask which clock is the most precise, and use that one as your master clock? If that's not possible you won't have many other options if everything has to be done on realtime; adapting frame display times as soon as you detect the need will be much easier than modifying sound chunks, I think.

From mike at mikeversteeg.com  Wed Dec 19 15:34:51 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Wed, 19 Dec 2012 15:34:51 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
Message-ID: <CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>

> Maybe you should ask which clock is the most precise, and use that one as your master clock? If that's not possible you won't have many other options if everything has to be done on realtime; adapting frame display times as soon as you detect the need will be much easier than modifying sound chunks, I think.

Either one can be used as master clock for the recorder. But I cannot
sync both clocks with each other if that's what you mean. So are you
saying this cannot be handled by libav, i.e. it always requires a
fixed ratio between frames and audio samples? Earlier comments let to
believe "something" can be done with timestamps, but I fail to see
what.

From alexcohn at netvision.net.il  Wed Dec 19 16:20:43 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Wed, 19 Dec 2012 17:20:43 +0200
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
Message-ID: <CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>

On Wed, Dec 19, 2012 at 4:34 PM, Mike Versteeg <mike at mikeversteeg.com> wrote:
>> Maybe you should ask which clock is the most precise, and use that one as your master clock? If that's not possible you won't have many other options if everything has to be done on realtime; adapting frame display times as soon as you detect the need will be much easier than modifying sound chunks, I think.

It's not a question of clock precision; video should follow audio
because it's FPS is much lower, and it is more resilient to slight
adjustments.

> Either one can be used as master clock for the recorder. But I cannot sync both clocks with each other if that's what you mean.

Yes you can! You should set the PTS for the video stream according to
the audio clock, that's it. I believe that you can know the audio
clock by counting the audio packets.

BR,
Alex

From mike at mikeversteeg.com  Wed Dec 19 16:27:15 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Wed, 19 Dec 2012 16:27:15 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
Message-ID: <CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>

>> Either one can be used as master clock for the recorder. But I cannot sync both clocks with each other if that's what you mean.
>
> Yes you can! You should set the PTS for the video stream according to
> the audio clock, that's it. I believe that you can know the audio
> clock by counting the audio packets.

I can't as this would introduce double/missing frames, which is
unacceptable for my app.

I guess what I want is not possible, so I have to resort to audio
resampling. Thanks everyone.

Mike

From alexcohn at netvision.net.il  Wed Dec 19 16:38:01 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Wed, 19 Dec 2012 17:38:01 +0200
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
Message-ID: <CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>

On Wed, Dec 19, 2012 at 5:27 PM, Mike Versteeg <mike at mikeversteeg.com> wrote:
> I can't as this would introduce double/missing frames, which is
> unacceptable for my app.

Why? The nominal frame rate of a video stream is only a guideline; the
actual PTS can handle reasonable jitter.

Alex

From mike at mikeversteeg.com  Wed Dec 19 16:46:15 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Wed, 19 Dec 2012 16:46:15 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
Message-ID: <CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>

>> I can't as this would introduce double/missing frames, which is
>> unacceptable for my app.
>
> Why? The nominal frame rate of a video stream is only a guideline; the
> actual PTS can handle reasonable jitter.

My initial thought was just that: adjust the PTS to effectively
slightly change the frame rate in the file, matching my video clock to
the audio clock. But I have since learned (offline) that this is not
possible as indeed the video PTS, at least for MPEG-2, increments as
integer (0, 1, 2, ...). There's no fiddling with it.  So short of
doubling/dropping the occasional video frame, I can't see any other
way to "sync" my video to audio.

Mike

From alexcohn at netvision.net.il  Wed Dec 19 17:01:39 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Wed, 19 Dec 2012 18:01:39 +0200
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
Message-ID: <CADiKgojN2KSvNhNUe1xQZf=JJVDKZNB8GTs-aMRzdo5UNtZpbA@mail.gmail.com>

On Wed, Dec 19, 2012 at 5:46 PM, Mike Versteeg <mike at mikeversteeg.com> wrote:
> video PTS, at least for MPEG-2, increments as integer (0, 1, 2, ...)

Oops, my bad... I lost your original requirement to be MPEG-2
compliant. But maybe it would be enough to set REPEAT_FIRST_FIELD once
in a while.

Alex

From mike at mikeversteeg.com  Wed Dec 19 17:14:31 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Wed, 19 Dec 2012 17:14:31 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CADiKgojN2KSvNhNUe1xQZf=JJVDKZNB8GTs-aMRzdo5UNtZpbA@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
	<CADiKgojN2KSvNhNUe1xQZf=JJVDKZNB8GTs-aMRzdo5UNtZpbA@mail.gmail.com>
Message-ID: <CAE=zCur0cn-bvkZvs1HPFLe7fJZxZJO3egKAodVyT1JQg2J49w@mail.gmail.com>

>> video PTS, at least for MPEG-2, increments as integer (0, 1, 2, ...)
>
> Oops, my bad... I lost your original requirement to be MPEG-2
> compliant. But maybe it would be enough to set REPEAT_FIRST_FIELD once
> in a while.

Thanks. I am not familiar with that. Won't that introduce a double
frame? And if that's a trick that works I assume it can only be used
to delay my video, what if I need to speed it up?

So other formats do support "true" time stamping? I have chosen MPEG-2
as my master for video recording as it is light on the CPU, compresses
well and is accepted by most editing software. If there is a similar
format that would solve my problem, I can look into using that. I was
not even aware MPEG-2 was stricter than other codecs.

Mike

From alexcohn at netvision.net.il  Wed Dec 19 17:28:19 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Wed, 19 Dec 2012 18:28:19 +0200
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCur0cn-bvkZvs1HPFLe7fJZxZJO3egKAodVyT1JQg2J49w@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
	<CADiKgojN2KSvNhNUe1xQZf=JJVDKZNB8GTs-aMRzdo5UNtZpbA@mail.gmail.com>
	<CAE=zCur0cn-bvkZvs1HPFLe7fJZxZJO3egKAodVyT1JQg2J49w@mail.gmail.com>
Message-ID: <CADiKgohBbebpDgKZv724jTo9g5q=VTF1W-ugT4O4AjPWSOzXWA@mail.gmail.com>

On Wed, Dec 19, 2012 at 6:14 PM, Mike Versteeg <mike at mikeversteeg.com> wrote:
>>> video PTS, at least for MPEG-2, increments as integer (0, 1, 2, ...)
>>
>> Oops, my bad... I lost your original requirement to be MPEG-2
>> compliant. But maybe it would be enough to set REPEAT_FIRST_FIELD once
>> in a while.
>
> Thanks. I am not familiar with that. Won't that introduce a double frame?

No, it is sued to fight interlaced video (a.k.a telecine). The effect
of this flag is that the frame is slowed down by 50%. It should be
good enough to keep a ~30 fps video seamlessly synchronized with the
audio stream that is a bit slower.

> And if that's a trick that works I assume it can only be used to delay my video, what if I need to speed it up?

That's where you can have problems.

> So other formats do support "true" time stamping?

Absolutely. MPEG-4 h264 stream has full support of flexible PTS per
frame, and it will deliver much better quality. But it really needs
more CPU, and you may face questions of licenses, patents, and
royalties for use of different h264 encoders.

Anyway, if you are looking at player support, MPEG-2 is the king. I
hope that other people on this list will have more to share on this
topic.

Alex

From rjvbertin at gmail.com  Wed Dec 19 19:50:49 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Wed, 19 Dec 2012 19:50:49 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
Message-ID: <98CDC337-B28B-485A-8CE6-C08053BEB4F3@gmail.com>


On Dec 19, 2012, at 16:46, Mike Versteeg wrote:

> the audio clock. But I have since learned (offline) that this is not
> possible as indeed the video PTS, at least for MPEG-2, increments as
> integer (0, 1, 2, ...). There's no fiddling with it.  So short of

Can't you set the video's time scale such that an integer increment corresponds to an amount of real time that's relevant to your situation?
(This is more or less what I did when I had to sync 12.5Hz video with 1kHz dynamic data; I reconstructed new QuickTime movies with a time scale of 12500)

Of course I know next to nothing about libav O:^)

R


From gohar_h at instigatemobile.com  Thu Dec 20 06:47:37 2012
From: gohar_h at instigatemobile.com (Gohar Hovhannisyan)
Date: Thu, 20 Dec 2012 09:47:37 +0400
Subject: [Libav-user] Decoding HD videos on iPad devices
In-Reply-To: <loom.20121219T115830-976@post.gmane.org>
References: <D03BBEB0-1A2F-4EF3-BB13-28FD05D41075@instigatemobile.com>
	<loom.20121219T115830-976@post.gmane.org>
Message-ID: <50D2A679.4090206@instigatemobile.com>

Hi Carl,

thank you for your answer.

I've set the number of threads to 10, but every time my app crashes on 
avcodec_decode_video2. Could you please assist me what I'm doing wrong?

Thanks,
Gohar

On 12/19/12 2:59 PM, Carl Eugen Hoyos wrote:
> Gohar Hovhannisyan <gohar_h at ...> writes:
>
>> We wonder whether it is possible to perform decoding on
>> iPad GPU to speed up decoding process (by compiling
>> it with special flags, or with some patches)
> Only if you implement it.
>
>> or run it parallel on dual core (iPad2) / quad-core (new iPad)
>> or are there any other methods to speed up decoding process?
> You can set the number of threads that decoding should use.
>
> Carl Eugen
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user


From mike at mikeversteeg.com  Thu Dec 20 11:12:50 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 20 Dec 2012 11:12:50 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CADiKgohBbebpDgKZv724jTo9g5q=VTF1W-ugT4O4AjPWSOzXWA@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
	<CADiKgojN2KSvNhNUe1xQZf=JJVDKZNB8GTs-aMRzdo5UNtZpbA@mail.gmail.com>
	<CAE=zCur0cn-bvkZvs1HPFLe7fJZxZJO3egKAodVyT1JQg2J49w@mail.gmail.com>
	<CADiKgohBbebpDgKZv724jTo9g5q=VTF1W-ugT4O4AjPWSOzXWA@mail.gmail.com>
Message-ID: <CAE=zCuqB0dhc7C6QccDwSDGQ1AyQsMGUK_aVeNWJqm3Tar9a_A@mail.gmail.com>

>>>> video PTS, at least for MPEG-2, increments as integer (0, 1, 2, ...)
>>>
>>> Oops, my bad... I lost your original requirement to be MPEG-2
>>> compliant. But maybe it would be enough to set REPEAT_FIRST_FIELD once
>>> in a while.
>>
>> Thanks. I am not familiar with that. Won't that introduce a double frame?
>
> No, it is sued to fight interlaced video (a.k.a telecine). The effect
> of this flag is that the frame is slowed down by 50%. It should be
> good enough to keep a ~30 fps video seamlessly synchronized with the
> audio stream that is a bit slower.

If this does not introduce any other problems, like in editing
software, then yes it should work. Unfortunately I'd still need a
trick to speed up.

>> So other formats do support "true" time stamping?
>
> Absolutely. MPEG-4 h264 stream has full support of flexible PTS per
> frame, and it will deliver much better quality. But it really needs
> more CPU, and you may face questions of licenses, patents, and
> royalties for use of different h264 encoders.

So far I've not been able to buy a licence for x264, but also it is
too heavy on CPU.

> Anyway, if you are looking at player support, MPEG-2 is the king. I
> hope that other people on this list will have more to share on this
> topic.

Thanks Alex, you've been very helpful!

Mike

From mike at mikeversteeg.com  Thu Dec 20 11:18:56 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Thu, 20 Dec 2012 11:18:56 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <98CDC337-B28B-485A-8CE6-C08053BEB4F3@gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
	<98CDC337-B28B-485A-8CE6-C08053BEB4F3@gmail.com>
Message-ID: <CAE=zCupNsE8X_RESqHMM-72_yzEz_GU_+Cyrx5Wrb1Oidz7tag@mail.gmail.com>

>> the audio clock. But I have since learned (offline) that this is not
>> possible as indeed the video PTS, at least for MPEG-2, increments as
>> integer (0, 1, 2, ...). There's no fiddling with it.  So short of
>
> Can't you set the video's time scale such that an integer increment corresponds to an amount of real time that's relevant to your situation?

No. The MPEG-2 encoder seems very picky (possibly part of the
protocol) when it comes to that. In fact a time base other than 1/25
or 1/30 and it either returns an error or smacks me with an AV. Won't
even do 29.97 fps or 50 or 60 fps.

> Of course I know next to nothing about libav O:^)

I've been using it for about a year and still feel I know nothing
about it. It requires a lot of dedication, surfing and reading to get
comfortable with it. Great code though, and (mostly) free.

Mike

From cehoyos at ag.or.at  Thu Dec 20 12:13:56 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Thu, 20 Dec 2012 11:13:56 +0000 (UTC)
Subject: [Libav-user] Decoding HD videos on iPad devices
References: <D03BBEB0-1A2F-4EF3-BB13-28FD05D41075@instigatemobile.com>
	<loom.20121219T115830-976@post.gmane.org>
Message-ID: <loom.20121220T121310-402@post.gmane.org>

Carl Eugen Hoyos <cehoyos at ...> writes:

> > We wonder whether it is possible to perform decoding on 
> > iPad GPU to speed up decoding process (by compiling
> > it with special flags, or with some patches) 
> 
> Only if you implement it.

Reconsidering this, it is possible that -vcodec h264_vda 
is what you are searching for (I don't know for sure).

Carl Eugen


From cehoyos at ag.or.at  Thu Dec 20 12:12:39 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Thu, 20 Dec 2012 11:12:39 +0000 (UTC)
Subject: [Libav-user] Decoding HD videos on iPad devices
References: <D03BBEB0-1A2F-4EF3-BB13-28FD05D41075@instigatemobile.com>
	<loom.20121219T115830-976@post.gmane.org>
	<50D2A679.4090206@instigatemobile.com>
Message-ID: <loom.20121220T121148-431@post.gmane.org>

Gohar Hovhannisyan <gohar_h at ...> writes:

> I've set the number of threads to 10, but every time my app 
> crashes on avcodec_decode_video2.

(Backtrace etc. missing, see https://ffmpeg.org/bugreports.html )
Is the crash reproducible with ffmpeg (the application)?

Please do not top-post here, Carl Eugen


From rjvbertin at gmail.com  Thu Dec 20 18:36:36 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Thu, 20 Dec 2012 18:36:36 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <CAE=zCuqB0dhc7C6QccDwSDGQ1AyQsMGUK_aVeNWJqm3Tar9a_A@mail.gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
	<CADiKgojN2KSvNhNUe1xQZf=JJVDKZNB8GTs-aMRzdo5UNtZpbA@mail.gmail.com>
	<CAE=zCur0cn-bvkZvs1HPFLe7fJZxZJO3egKAodVyT1JQg2J49w@mail.gmail.com>
	<CADiKgohBbebpDgKZv724jTo9g5q=VTF1W-ugT4O4AjPWSOzXWA@mail.gmail.com>
	<CAE=zCuqB0dhc7C6QccDwSDGQ1AyQsMGUK_aVeNWJqm3Tar9a_A@mail.gmail.com>
Message-ID: <1205BAD9-76B2-4925-839A-6BB83E2674C6@gmail.com>

On Dec 20, 2012, at 11:12, Mike Versteeg wrote:

> If this does not introduce any other problems, like in editing
> software, then yes it should work. Unfortunately I'd still need a
> trick to speed up.

Speeding up (probably) means reducing the duration of the preceding frame - which seems easy enough to accomplish with a simple caching scheme.

> No. The MPEG-2 encoder seems very picky (possibly part of the
> protocol) when it comes to that. In fact a time base other than 1/25
> or 1/30 and it either returns an error or smacks me with an AV. Won't
> even do 29.97 fps or 50 or 60 fps.

I'm not talking about frame rate or time base. I'm talking about the unit in which presentation durations are encoded, a factor that makes it possible to encode a continuous variable (time) on an arbitrary fixed point scale that's most suited to the required frame rate. A 1/25 time base (25Hz, 0.04s) would be amount to a frame duration of 24 'ticks' with the standard (QuickTime) timescale of 600/s, or 100 with a timescale of 2500/s . You don't see any difference between the two when watching a video, but with the latter timescale one has more precise control over individual frame timing. I doubt things work differently in libav.

R.

From mike at mikeversteeg.com  Fri Dec 21 09:00:59 2012
From: mike at mikeversteeg.com (Mike Versteeg)
Date: Fri, 21 Dec 2012 09:00:59 +0100
Subject: [Libav-user] How to perform drift compensation in MPEG-2
	recorder
In-Reply-To: <1205BAD9-76B2-4925-839A-6BB83E2674C6@gmail.com>
References: <CAE=zCuo4Q4ma3xcqn86ZWn4v8yZQnF1vC+UG2t5yK3xpf7syfw@mail.gmail.com>
	<CAE=zCuoJzhs_RiXORvOep0J9Vr=pLMGe29aYE+enpqF5ossJww@mail.gmail.com>
	<CAE=zCuoRD5AiUds_CRJNeD2v3rNSsQCobieNM0hHYwh3NPb=xQ@mail.gmail.com>
	<DUB114-W498929C4AE9E7318463455AA320@phx.gbl>
	<CADiKgojij57Afn-b4FAdCYM_bLJxK=sHL1mK1sDNNHxCzUqRMQ@mail.gmail.com>
	<BE28CE4C-B4BD-4B94-87C2-379E00347F81@gmail.com>
	<CAE=zCuqS53+9pXfbR0e5EyX=8aENVYW5PhCuwBLz+5idDtNXag@mail.gmail.com>
	<CAE=zCupYBJJ90T6rD8nAGqL9MbnG5F8QFC4Q7CzKFNbWEHxEMQ@mail.gmail.com>
	<CAE=zCuqBtf1-c6O1p4hP+vW1dL8zxL+2ukxrpzU5j0=bmOSvpw@mail.gmail.com>
	<496155AF-A561-4036-BDCA-B9E1047871D3@gmail.com>
	<CAE=zCur0bCivy+=dGZTAZQK_2_CTnT3t4Q5-L7ESPto6e73uRQ@mail.gmail.com>
	<CADiKgoi9UiO3DDmQe-AMchpMTF8dRwbGWEiA-HWUUpoBP6gu-w@mail.gmail.com>
	<CAE=zCupMJWpDFDA8XOJCopJRg+dKktYcg+yo9DRYVDJOUGbpwA@mail.gmail.com>
	<CADiKgogQ4Nutf69O-EmHu_9n4DOoH2fQqaH=3fCArPQo+3Xh-g@mail.gmail.com>
	<CAE=zCuo2Oqr_Qyrx3o1sGzMb1HwA8GHiry1bse_34Z+3WZj=vg@mail.gmail.com>
	<CADiKgojN2KSvNhNUe1xQZf=JJVDKZNB8GTs-aMRzdo5UNtZpbA@mail.gmail.com>
	<CAE=zCur0cn-bvkZvs1HPFLe7fJZxZJO3egKAodVyT1JQg2J49w@mail.gmail.com>
	<CADiKgohBbebpDgKZv724jTo9g5q=VTF1W-ugT4O4AjPWSOzXWA@mail.gmail.com>
	<CAE=zCuqB0dhc7C6QccDwSDGQ1AyQsMGUK_aVeNWJqm3Tar9a_A@mail.gmail.com>
	<1205BAD9-76B2-4925-839A-6BB83E2674C6@gmail.com>
Message-ID: <CAE=zCup62TsKOTVXNCTAXq39KzUR=qzn08m26V5104ZfVhqdDA@mail.gmail.com>

>> No. The MPEG-2 encoder seems very picky (possibly part of the
>> protocol) when it comes to that. In fact a time base other than 1/25
>> or 1/30 and it either returns an error or smacks me with an AV. Won't
>> even do 29.97 fps or 50 or 60 fps.
>
> I'm not talking about frame rate or time base. I'm talking about the unit in which presentation durations are encoded, a factor that makes it possible to encode a continuous variable (time) on an arbitrary fixed point scale that's most suited to the required frame rate. A 1/25 time base (25Hz, 0.04s) would be amount to a frame duration of 24 'ticks' with the standard (QuickTime) timescale of 600/s, or 100 with a timescale of 2500/s . You don't see any difference between the two when watching a video, but with the latter timescale one has more precise control over individual frame timing. I doubt things work differently in libav.

That's my original thought, unfortunately in ffmpeg the PTS is
expressed in units of time_base, and (at least in MPEG-2) that equals
the frame rate. Try anything other than 1/25 or 1/30 and the encoder
refuses to do its job.

Mike

From trn200190 at gmail.com  Fri Dec 21 10:55:22 2012
From: trn200190 at gmail.com (i m what i m ~~~~)
Date: Fri, 21 Dec 2012 15:25:22 +0530
Subject: [Libav-user] Please help me will ffmpeg be able to decode the h264
	data
Message-ID: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>

Hello All

i am new to ffmpeg i have worked with VC++ before in VS2010.
want to make an application in which i will stream h264 encrypted and
encoded video using live media libraries and then decrypt and decode it and
display it using direct show the video.
For now i am able to encrypt and encode the video and decrypt the video but
i need some free source decoder to which i can give NAL units and then give
the decoded data to direct show libraries to show the video.....
Please help me......

Will ffmpeg libraries be able to decode h 264 data and then will be able to
give data to direct show and play it.......
Please give me the source link for ffmpeg source code which can be built in
VS2010 and steps to built it.

Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/6de60054/attachment.html>

From mybrokenbeat at gmail.com  Fri Dec 21 11:18:00 2012
From: mybrokenbeat at gmail.com (Oleg)
Date: Fri, 21 Dec 2012 12:18:00 +0200
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
	h264 data
In-Reply-To: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
Message-ID: <51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>

FFmpeg CAN'T be built in VS2010 by default as VS doesn't support C89 standard. You can build ffmpeg with MINGW compiler and use output libraries in VS. 

Btw, Windows Vista and above have built-in support of decoding h264 in Media Foundation framework as well as GPU acceleration support.

21.12.2012, ? 11:55, i m what i m ~~~~ ???????(?):

> Hello All
> 
> i am new to ffmpeg i have worked with VC++ before in VS2010.
> want to make an application in which i will stream h264 encrypted and encoded video using live media libraries and then decrypt and decode it and display it using direct show the video.
> For now i am able to encrypt and encode the video and decrypt the video but i need some free source decoder to which i can give NAL units and then give the decoded data to direct show libraries to show the video.....
> Please help me......
> 
> Will ffmpeg libraries be able to decode h 264 data and then will be able to give data to direct show and play it.......
> Please give me the source link for ffmpeg source code which can be built in VS2010 and steps to built it.
> 
> Thanks
> 
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user


From trn200190 at gmail.com  Fri Dec 21 11:21:22 2012
From: trn200190 at gmail.com (i m what i m ~~~~)
Date: Fri, 21 Dec 2012 15:51:22 +0530
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
 h264 data
In-Reply-To: <51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
Message-ID: <CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>

Sir

Is  Media Foundation framework  free i mean open source ?
How can we use  Media Foundation framework  to decode the h264 video?

On Fri, Dec 21, 2012 at 3:48 PM, Oleg <mybrokenbeat at gmail.com> wrote:

> Media Foundation framework
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/bde50b39/attachment.html>

From ryltsov at gmail.com  Fri Dec 21 11:42:24 2012
From: ryltsov at gmail.com (Roman Ryltsov)
Date: Fri, 21 Dec 2012 12:42:24 +0200
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
 h264 data
In-Reply-To: <CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
	<CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
Message-ID: <CAE9ZOGFNdR08SK6T7J7pXQj0q5rOA9BV=VK6auLRrxs2ANH8YQ@mail.gmail.com>

Media Foundation is core Windows API. To correct Oleg on his previous post,
its H.264 decoder is available starting Windows 7 and is a part of baseline
Windows (no additional installations necessary).

The decoder is available through both video processing APIs, DirectShow and
Media Foundation, which you can leverage in your application. There is a
multitude of community samples available for DirectShow and there are Media
Foundation samples in Windows SDK.

You can also use FFmpeg's H.264 video decoder, the quick start for which is
pre-built FFmpeg for Windows available from
http://ffmpeg.zeranoe.com/builds/ where you can grab redistributable DLLs
and LIBs to link your Visual Studio project against.

Hope this helps,
Roman



On Fri, Dec 21, 2012 at 12:21 PM, i m what i m ~~~~ <trn200190 at gmail.com>wrote:

> Sir
>
> Is  Media Foundation framework  free i mean open source ?
> How can we use  Media Foundation framework  to decode the h264 video?
>
> On Fri, Dec 21, 2012 at 3:48 PM, Oleg <mybrokenbeat at gmail.com> wrote:
>
>> Media Foundation framework
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/df6a817a/attachment.html>

From trn200190 at gmail.com  Fri Dec 21 11:48:50 2012
From: trn200190 at gmail.com (i m what i m ~~~~)
Date: Fri, 21 Dec 2012 16:18:50 +0530
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
 h264 data
In-Reply-To: <CAE9ZOGFNdR08SK6T7J7pXQj0q5rOA9BV=VK6auLRrxs2ANH8YQ@mail.gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
	<CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
	<CAE9ZOGFNdR08SK6T7J7pXQj0q5rOA9BV=VK6auLRrxs2ANH8YQ@mail.gmail.com>
Message-ID: <CAJmxhGU=ad8USdh0Tnoq7SQ4MqgELxtDFGus9LBQVXDot_CxPg@mail.gmail.com>

Thanks for the useful info Roman sir i downloaded the "FFmpeg git-8b6aeb1
64-bit Static<http://ffmpeg.zeranoe.com/builds/win64/static/ffmpeg-20121120-git-8b6aeb1-win64-static.7z>
"
build form the link you have given but i am not able to decode the h264
video how to decode the video i have what i have to type in cmd prompt

On Fri, Dec 21, 2012 at 4:12 PM, Roman Ryltsov <ryltsov at gmail.com> wrote:

> Media Foundation is core Windows API. To correct Oleg on his previous
> post, its H.264 decoder is available starting Windows 7 and is a part of
> baseline Windows (no additional installations necessary).
>
> The decoder is available through both video processing APIs, DirectShow
> and Media Foundation, which you can leverage in your application. There is
> a multitude of community samples available for DirectShow and there are
> Media Foundation samples in Windows SDK.
>
> You can also use FFmpeg's H.264 video decoder, the quick start for which
> is pre-built FFmpeg for Windows available from
> http://ffmpeg.zeranoe.com/builds/ where you can grab redistributable DLLs
> and LIBs to link your Visual Studio project against.
>
> Hope this helps,
> Roman
>
>
>
>
> On Fri, Dec 21, 2012 at 12:21 PM, i m what i m ~~~~ <trn200190 at gmail.com>wrote:
>
>> Sir
>>
>> Is  Media Foundation framework  free i mean open source ?
>> How can we use  Media Foundation framework  to decode the h264 video?
>>
>> On Fri, Dec 21, 2012 at 3:48 PM, Oleg <mybrokenbeat at gmail.com> wrote:
>>
>>> Media Foundation framework
>>
>>
>>
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/1cbbbcf2/attachment.html>

From ryltsov at gmail.com  Fri Dec 21 11:52:32 2012
From: ryltsov at gmail.com (Roman Ryltsov)
Date: Fri, 21 Dec 2012 12:52:32 +0200
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
 h264 data
In-Reply-To: <CAJmxhGU=ad8USdh0Tnoq7SQ4MqgELxtDFGus9LBQVXDot_CxPg@mail.gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
	<CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
	<CAE9ZOGFNdR08SK6T7J7pXQj0q5rOA9BV=VK6auLRrxs2ANH8YQ@mail.gmail.com>
	<CAJmxhGU=ad8USdh0Tnoq7SQ4MqgELxtDFGus9LBQVXDot_CxPg@mail.gmail.com>
Message-ID: <CAE9ZOGGm_8iaHVs3P5-Gm-c4qYdxau5pwB3tWJfRPx9mZw4BaA@mail.gmail.com>

You don't need "static" unless you are going to use command line pre-built
executable.

Have a look here:
http://stackoverflow.com/questions/11701635/use-ffmpeg-in-visual-studio/11701737#11701737if
you are going to link to FFmpeg (libav*) as "to API/library". There is
a
link to minimal app there as well which you can quick start from.

Roman


On Fri, Dec 21, 2012 at 12:48 PM, i m what i m ~~~~ <trn200190 at gmail.com>wrote:

> Thanks for the useful info Roman sir i downloaded the "FFmpeg git-8b6aeb1
> 64-bit Static<http://ffmpeg.zeranoe.com/builds/win64/static/ffmpeg-20121120-git-8b6aeb1-win64-static.7z> "
> build form the link you have given but i am not able to decode the h264
> video how to decode the video i have what i have to type in cmd prompt
>
> On Fri, Dec 21, 2012 at 4:12 PM, Roman Ryltsov <ryltsov at gmail.com> wrote:
>
>> Media Foundation is core Windows API. To correct Oleg on his previous
>> post, its H.264 decoder is available starting Windows 7 and is a part of
>> baseline Windows (no additional installations necessary).
>>
>> The decoder is available through both video processing APIs, DirectShow
>> and Media Foundation, which you can leverage in your application. There is
>> a multitude of community samples available for DirectShow and there are
>> Media Foundation samples in Windows SDK.
>>
>> You can also use FFmpeg's H.264 video decoder, the quick start for which
>> is pre-built FFmpeg for Windows available from
>> http://ffmpeg.zeranoe.com/builds/ where you can grab redistributable
>> DLLs and LIBs to link your Visual Studio project against.
>>
>> Hope this helps,
>> Roman
>>
>>
>>
>>
>> On Fri, Dec 21, 2012 at 12:21 PM, i m what i m ~~~~ <trn200190 at gmail.com>wrote:
>>
>>> Sir
>>>
>>> Is  Media Foundation framework  free i mean open source ?
>>> How can we use  Media Foundation framework  to decode the h264 video?
>>>
>>> On Fri, Dec 21, 2012 at 3:48 PM, Oleg <mybrokenbeat at gmail.com> wrote:
>>>
>>>> Media Foundation framework
>>>
>>>
>>>
>>
>> _______________________________________________
>> Libav-user mailing list
>> Libav-user at ffmpeg.org
>> http://ffmpeg.org/mailman/listinfo/libav-user
>>
>>
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/cc6e0424/attachment.html>

From trn200190 at gmail.com  Fri Dec 21 12:21:27 2012
From: trn200190 at gmail.com (i m what i m ~~~~)
Date: Fri, 21 Dec 2012 16:51:27 +0530
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
 h264 data
In-Reply-To: <CAE9ZOGGm_8iaHVs3P5-Gm-c4qYdxau5pwB3tWJfRPx9mZw4BaA@mail.gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
	<CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
	<CAE9ZOGFNdR08SK6T7J7pXQj0q5rOA9BV=VK6auLRrxs2ANH8YQ@mail.gmail.com>
	<CAJmxhGU=ad8USdh0Tnoq7SQ4MqgELxtDFGus9LBQVXDot_CxPg@mail.gmail.com>
	<CAE9ZOGGm_8iaHVs3P5-Gm-c4qYdxau5pwB3tWJfRPx9mZw4BaA@mail.gmail.com>
Message-ID: <CAJmxhGVOfsAiUQ3q-unShm5=zi8ph+drWNA__je0SHYk1Y92Wg@mail.gmail.com>

Sir

i have directly downloaded the exe for ffmpeg now how do i pass the file
name and command line parameter to decode h264 file?

On Fri, Dec 21, 2012 at 4:22 PM, Roman Ryltsov <ryltsov at gmail.com> wrote:

> You don't need "static" unless you are going to use command line pre-built
> executable.
>
> Have a look here:
> http://stackoverflow.com/questions/11701635/use-ffmpeg-in-visual-studio/11701737#11701737if you are going to link to FFmpeg (libav*) as "to API/library". There is a
> link to minimal app there as well which you can quick start from.
>
> Roman
>
>
>
> On Fri, Dec 21, 2012 at 12:48 PM, i m what i m ~~~~ <trn200190 at gmail.com>wrote:
>
>> Thanks for the useful info Roman sir i downloaded the "FFmpeg
>> git-8b6aeb1 64-bit Static<http://ffmpeg.zeranoe.com/builds/win64/static/ffmpeg-20121120-git-8b6aeb1-win64-static.7z> "
>> build form the link you have given but i am not able to decode the h264
>> video how to decode the video i have what i have to type in cmd prompt
>>
>> On Fri, Dec 21, 2012 at 4:12 PM, Roman Ryltsov <ryltsov at gmail.com> wrote:
>>
>>> Media Foundation is core Windows API. To correct Oleg on his previous
>>> post, its H.264 decoder is available starting Windows 7 and is a part of
>>> baseline Windows (no additional installations necessary).
>>>
>>> The decoder is available through both video processing APIs, DirectShow
>>> and Media Foundation, which you can leverage in your application. There is
>>> a multitude of community samples available for DirectShow and there are
>>> Media Foundation samples in Windows SDK.
>>>
>>> You can also use FFmpeg's H.264 video decoder, the quick start for which
>>> is pre-built FFmpeg for Windows available from
>>> http://ffmpeg.zeranoe.com/builds/ where you can grab redistributable
>>> DLLs and LIBs to link your Visual Studio project against.
>>>
>>> Hope this helps,
>>> Roman
>>>
>>>
>>>
>>>
>>> On Fri, Dec 21, 2012 at 12:21 PM, i m what i m ~~~~ <trn200190 at gmail.com
>>> > wrote:
>>>
>>>> Sir
>>>>
>>>> Is  Media Foundation framework  free i mean open source ?
>>>> How can we use  Media Foundation framework  to decode the h264 video?
>>>>
>>>> On Fri, Dec 21, 2012 at 3:48 PM, Oleg <mybrokenbeat at gmail.com> wrote:
>>>>
>>>>> Media Foundation framework
>>>>
>>>>
>>>>
>>>
>>> _______________________________________________
>>> Libav-user mailing list
>>> Libav-user at ffmpeg.org
>>> http://ffmpeg.org/mailman/listinfo/libav-user
>>>
>>>
>>
>> _______________________________________________
>> Libav-user mailing list
>> Libav-user at ffmpeg.org
>> http://ffmpeg.org/mailman/listinfo/libav-user
>>
>>
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/611be7f7/attachment.html>

From mybrokenbeat at gmail.com  Fri Dec 21 12:22:08 2012
From: mybrokenbeat at gmail.com (Oleg)
Date: Fri, 21 Dec 2012 13:22:08 +0200
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
	h264 data
In-Reply-To: <CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
	<CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
Message-ID: <14287177-D4A9-4E9F-A3A9-14E151BC01D0@gmail.com>

It's closed source, but it's a part of Windows OS. So you can use it free in your commercial application and even doesn't have h264 patenting problems as Microsoft has already payed for that. 

This framework has documentation where you can find all answers on your questions:
http://msdn.microsoft.com/ru-ru/library/windows/desktop/dd797815(v=vs.85).aspx

Btw, not it's available not from Vista(as I mentioned before), but from Windows 7.

21.12.2012, ? 12:21, i m what i m ~~~~ ???????(?):

> Sir
> 
> Is  Media Foundation framework  free i mean open source ?
> How can we use  Media Foundation framework  to decode the h264 video?
> 
> On Fri, Dec 21, 2012 at 3:48 PM, Oleg <mybrokenbeat at gmail.com> wrote:
> Media Foundation framework
> 
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/efb856ad/attachment.html>

From rjvbertin at gmail.com  Fri Dec 21 12:30:02 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Fri, 21 Dec 2012 12:30:02 +0100
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
	h264 data
In-Reply-To: <14287177-D4A9-4E9F-A3A9-14E151BC01D0@gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
	<CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
	<14287177-D4A9-4E9F-A3A9-14E151BC01D0@gmail.com>
Message-ID: <957E0E08-F3B2-45FD-848B-CA6DA9F2F90F@gmail.com>

On Dec 21, 2012, at 12:22, Oleg wrote:

> It's closed source, but it's a part of Windows OS. So you can use it free in your commercial application and even doesn't have h264 patenting problems as Microsoft has already payed for that. 

Very interesting, though I have some hopes to use this to improve QuickTime's handling of mpeg4 content on MS Windows O:-)

> This framework has documentation where you can find all answers on your questions:
> http://msdn.microsoft.com/ru-ru/library/windows/desktop/dd797815(v=vs.85).aspx

http://msdn.microsoft.com/library/windows/desktop/dd797815(v=vs.85).aspx will probably work for more readers ;)


R


From trn200190 at gmail.com  Fri Dec 21 12:37:29 2012
From: trn200190 at gmail.com (i m what i m ~~~~)
Date: Fri, 21 Dec 2012 17:07:29 +0530
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
 h264 data
In-Reply-To: <957E0E08-F3B2-45FD-848B-CA6DA9F2F90F@gmail.com>
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
	<CAJmxhGUK5exzdpLf+VS8NNK-3bd9UZh-5_sc35bh96Z5OfpamQ@mail.gmail.com>
	<14287177-D4A9-4E9F-A3A9-14E151BC01D0@gmail.com>
	<957E0E08-F3B2-45FD-848B-CA6DA9F2F90F@gmail.com>
Message-ID: <CAJmxhGUjybSnEaCbi8eT6goLcq+SYDH6Vcn2RG9nSwa4swQ80Q@mail.gmail.com>

TO Oleg
Sir the link you gave,i didn't find the download link and how to use,even i
didn't get the source fie as the whole text on site is not in english plz
do a favor tell how to use it

On Fri, Dec 21, 2012 at 5:00 PM, "Ren? J.V. Bertin" <rjvbertin at gmail.com>wrote:

> On Dec 21, 2012, at 12:22, Oleg wrote:
>
> > It's closed source, but it's a part of Windows OS. So you can use it
> free in your commercial application and even doesn't have h264 patenting
> problems as Microsoft has already payed for that.
>
> Very interesting, though I have some hopes to use this to improve
> QuickTime's handling of mpeg4 content on MS Windows O:-)
>
> > This framework has documentation where you can find all answers on your
> questions:
> >
> http://msdn.microsoft.com/ru-ru/library/windows/desktop/dd797815(v=vs.85).aspx
>
> http://msdn.microsoft.com/library/windows/desktop/dd797815(v=vs.85).aspxwill probably work for more readers ;)
>
>
> R
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/ee4d1d1d/attachment.html>

From srijith.vasudevan at gmail.com  Fri Dec 21 17:05:37 2012
From: srijith.vasudevan at gmail.com (srijith vasudevan)
Date: Fri, 21 Dec 2012 21:35:37 +0530
Subject: [Libav-user] facing an issue with avcodec_decode_subtitle2
Message-ID: <CAMo_p1mJqgXVuz0DWuiXPYiVdT3Mxg9ehG-c_2XVuUa3B+3+yw@mail.gmail.com>

Hello,

I am working on an assignment (using C++) to decode all the subtitle
streams present in a m2ts file using libav.

I have a m2ts file with one video stream (video stream index = 0), one
audio stream (audio stream index = 1) and 5 subtitle streams ( stream
indices = 2, 3, 4, 5).

I first noticed the problem when I compared the position (x, y, w, h) I got
from the decoded subtitle against the actual subtitle position from its
control file.

*The details of the actual events*
SUBTITLE_STREAM_INDEX_2
Event1 (X="607" Y="780" START_FRAME="764")

SUBTITLE_STREAM_INDEX_3
Event1 (X="575" Y="780" START_FRAME="764")

SUBTITLE_STREAM_INDEX_4
Event1 (X="701" Y="780" START_FRAME="764")

SUBTITLE_STREAM_INDEX_5
Event1 (X="483" Y="777" START_FRAME="765")

*
   Quote:
Please note Event1 of SUBTITLE_STREAM_INDEX_2, SUBTITLE_STREAM_INDEX_3,
SUBTITLE_STREAM_INDEX_4 start at frame no 764.
  *

When I ran the application, it first decoded Event1 of
SUBTITLE_STREAM_INDEX_4 and I got the position values from the decoded
subtitle as (701, 780, 507, 133).
Next it decoded Event1 of SUBTITLE_STREAM_INDEX_2 and the position values
decoded were as follows (701, 780, 507, 133). Please note the mismatch in
values.
The values were same for decoded subtitle of SUBTITLE_STREAM_INDEX_3.




AS the next step I am decoding the subtitle and generating a png file. This
is when I noticed the issue.

For example I am decoding SUBTITLE_STREAM_INDEX_4 at frame no = 764. I
generated the image using the subtitle->rects[0]->pict.data. This is
correct subtitle. (Assume the dialog to be "hi there").

Now I have another subtitle stream (index = 2) which also starts from the
same frame number (frame no = 764). Now when I generate a png using this
decoded subtitle it has the same subtitle text as the one decoded from
previous index. (the actual dialog is "hola", but the png generated had the
dialog as "hi there").
It is basically giving me the position and data of the subtitle decoded
from the first SUBTITLE_STREAM encountered for a particular frame number.



I am posting the part of code:

   *Code:*
AVPacket packet;
//looping till end of file
{
   av_init_packet(&packet);
    av_read_frame(pFormatContext, &packet);

if((packet.stream_index == m_subtitleStreamIndex[0]) ||
(packet.stream_index == m_subtitleStreamIndex[1]) || (packet.stream_index
== m_subtitleStreamIndex[2]) || (packet.stream_index ==
m_subtitleStreamIndex[3]))
{
   int has_subtitle = 0;
    AVSubtitle* subtitle = new AVSubtitle;
    avcodec_decode_subtitle2(pSubtitleCodecContext, subtitle,
&has_subtitle, &packet);

    if (has_subtitle != 0)
    {
        if (subtitle->rects)
            {
                struct SwsContext *mSwsContext =
sws_getContext(subtitle->rects[0]->w, subtitle->rects[0]->h,
pSubtitleCodecContext->pix_fmt,
                                        subtitle->rects[0]->w,
subtitle->rects[0]->h, PIX_FMT_RGBA,
                                        SWS_BICUBIC, NULL, NULL, NULL);
                AVFrame     *pPicture = AllocatePicture(PIX_FMT_RGBA,
subtitle->rects[0]->w, subtitle->rects[0]->h);

                /* convert to RGB */
                sws_scale(mSwsContext, (const uint8_t **)
subtitle->rects[0]->pict.data, subtitle->rects[0]->pict.linesize, 0,
subtitle->rects[0]->h,
                                        pPicture->data, pPicture->linesize);

                //code to generate png

                av_free(pPicture);
                sws_freeContext(mSwsContext);
            }
        }
   }
    avsubtitle_free(subtitle);
}
}


Could anyone please advice if I am doing something wrong here or suggest a
solution to this problem? I am expecting your priceless guidance and help.

Thanks,
Srijith
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121221/ec3fde53/attachment.html>

From cehoyos at ag.or.at  Fri Dec 21 17:41:35 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 21 Dec 2012 16:41:35 +0000 (UTC)
Subject: [Libav-user] Please help me will ffmpeg be able to decode the
	h264 data
References: <CAJmxhGVAAKYUEoKQdTSnxDGVtRHOn+SCVW8gfpKfeq3C3b-q6Q@mail.gmail.com>
	<51073774-9D3C-46BC-8857-E58E20E53262@gmail.com>
Message-ID: <loom.20121221T174018-978@post.gmane.org>

Oleg <mybrokenbeat at ...> writes:

> FFmpeg CAN'T be built in VS2010

FFmpeg can be built by VS2010, it is tested regularly, 
please see http://ffmpeg.org/platform.html#Microsoft-Visual-C_002b_002b 
and http://fate.ffmpeg.org/

Carl Eugen


From faeem.ali at gmail.com  Fri Dec 21 23:50:16 2012
From: faeem.ali at gmail.com (faeem)
Date: Sat, 22 Dec 2012 00:50:16 +0200
Subject: [Libav-user] Hardware decoding via libva
Message-ID: <50D4E7A8.7060608@gmail.com>

Hi All ...

I currently use a media player based on ffplay.c, but have a problem 
where HD video playing on an Intel Atom processor is far too slow, and 
uses too much CPU time. My understanding is that the problem may be 
resolved by using direct hardware decoding via VAAPI/libva. I have 
tested VLC compiled with VAAPI support and there is a huge performance 
improvement when playing HD video. I'm trying to accomplish something 
similar with libav.

Would someone please direct me to documentation and/or some other 
explanation of how to perform hardware decoding using VAAPI/libva with 
ffmpeg's libav libraries. Code examples would be great. I have no 
interest in hardware encoding, only decoding. Unfortunately, several 
hours on Google have not yielded any positive results thus far; I'm 
probably looking in the wrong places. I have compiled ffmpeg/libav with 
VAAPI support. The ffplay.c code has a line that states:

  // FIXME use direct rendering

I need to know how to fix that FIXME.

Additional info:
- My media player is based on the 0.11.1 version of ffmpeg's ffplay.c
- My media player has been altered to link against the new 1.0 version 
of ffmpeg/libav. However my code is still derived from the ffplay.c 
included in ffmpeg 0.11.1.
- The main modification I've made is to render the video via OpenGL 
instead of using an overlay. This is a requirement for the project on 
which I'm working. The only other major modification is removing the 
features of ffplay.c I don't require, which is pretty much everything 
aside from the basic decode and render code.
- The code is being developed on Linux. The graphics cards being used 
are NVIDIA and Intel i9xxx.

Thanks in advance.

Faeem


From cehoyos at ag.or.at  Sat Dec 22 01:53:11 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sat, 22 Dec 2012 00:53:11 +0000 (UTC)
Subject: [Libav-user] Hardware decoding via libva
References: <50D4E7A8.7060608@gmail.com>
Message-ID: <loom.20121222T014806-185@post.gmane.org>

faeem <faeem.ali at ...> writes:

> Would someone please direct me to documentation 
> and/or some other explanation of how to perform 
> hardware decoding using VAAPI/libva with ffmpeg's 
> libav libraries. Code examples would be great.

I know of two examples, the va-api code in vlc and 
the code in a patch for MPlayer, see for example 
http://thread.gmane.org/gmane.comp.video.mplayer.devel/61734/focus=61744

[...]

>   // FIXME use direct rendering
> 
> I need to know how to fix that FIXME.

This is not related to va-api at all.

> - The main modification I've made is to render the video via OpenGL 
> instead of using an overlay.

I sounds like a very welcome addition to the FFmpeg 
codebase, please consider sending a patch.

Carl Eugen


From klaussfreire at gmail.com  Sat Dec 22 02:35:05 2012
From: klaussfreire at gmail.com (Claudio Freire)
Date: Fri, 21 Dec 2012 22:35:05 -0300
Subject: [Libav-user] Hardware decoding via libva
In-Reply-To: <loom.20121222T014806-185@post.gmane.org>
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org>
Message-ID: <CAGTBQpbnJ1AfTZ576PhKVFrUhpybXRWkU1ef3X+3_4MinS3uvg@mail.gmail.com>

On Fri, Dec 21, 2012 at 9:53 PM, Carl Eugen Hoyos <cehoyos at ag.or.at> wrote:
>> - The main modification I've made is to render the video via OpenGL
>> instead of using an overlay.
>
> I sounds like a very welcome addition to the FFmpeg
> codebase, please consider sending a patch.

Have in mind that overlays tend to perform better than naively
implemented OpenGL rendering.
It's good to have the option though.

From faeem.ali at gmail.com  Sat Dec 22 13:00:19 2012
From: faeem.ali at gmail.com (faeem)
Date: Sat, 22 Dec 2012 14:00:19 +0200
Subject: [Libav-user] Hardware decoding via libva
In-Reply-To: <CAGTBQpbnJ1AfTZ576PhKVFrUhpybXRWkU1ef3X+3_4MinS3uvg@mail.gmail.com>
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org>
	<CAGTBQpbnJ1AfTZ576PhKVFrUhpybXRWkU1ef3X+3_4MinS3uvg@mail.gmail.com>
Message-ID: <50D5A0D3.2020108@gmail.com>

Hi Claudio ...

I'll look into patching ffplay.c with my OpenGL changes and submitting 
the code once I've gotten the direct rendering problem sorted out.

I am aware of the performance loss caused by OpenGL. It was chosen over 
overlays because it makes adding in special effects much easier, 
especially time critical stuff.

Faeem

On 22/12/2012 03:35, Claudio Freire wrote:
> On Fri, Dec 21, 2012 at 9:53 PM, Carl Eugen Hoyos <cehoyos at ag.or.at> wrote:
>>> - The main modification I've made is to render the video via OpenGL
>>> instead of using an overlay.
>> I sounds like a very welcome addition to the FFmpeg
>> codebase, please consider sending a patch.
> Have in mind that overlays tend to perform better than naively
> implemented OpenGL rendering.
> It's good to have the option though.
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user


From faeem.ali at gmail.com  Sat Dec 22 13:15:07 2012
From: faeem.ali at gmail.com (faeem)
Date: Sat, 22 Dec 2012 14:15:07 +0200
Subject: [Libav-user] Hardware decoding via libva
In-Reply-To: <loom.20121222T014806-185@post.gmane.org>
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org>
Message-ID: <50D5A44B.5030408@gmail.com>

On 22/12/2012 02:53, Carl Eugen Hoyos wrote:
> faeem <faeem.ali at ...> writes:
>
> I know of two examples, the va-api code in vlc and the code in a patch 
> for MPlayer, see for example 
> http://thread.gmane.org/gmane.comp.video.mplayer.devel/61734/focus=61744 
> [...] 
Thanks. I'll be looking into those examples ASAP.

>>    // FIXME use direct rendering
>>
>> I need to know how to fix that FIXME.
> This is not related to va-api at all.
I selected that //FIXME because it specified "direct rendering", which I 
took to mean "handled by the GPU". It seems I was mistaken there.

My conceptual understanding of va-api thus far, within the ffmpeg 
framework, is that libavcodec will read an encoded video frame, then use 
va-api and the GPU to perform decoding of that frame instead of 
performing decoding in software.

The end result will probably be a frame in YUV. I'll need to run the YUV 
to RGB conversion on each frame if I'm running OpenGL and this will 
still be CPU intensive. I would benefit from the hardware frame decoding 
though, and that alone should make a significant difference.

Is this correct?

Faeem


From cehoyos at ag.or.at  Sat Dec 22 13:22:16 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sat, 22 Dec 2012 12:22:16 +0000 (UTC)
Subject: [Libav-user] Hardware decoding via libva
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org>
	<50D5A44B.5030408@gmail.com>
Message-ID: <loom.20121222T131903-681@post.gmane.org>

faeem <faeem.ali at ...> writes:

> I'll need to run the YUV to RGB conversion on each frame 
> if I'm running OpenGL and this will still be CPU intensive.

(I don't think this is correct but it is not related to 
va-api at all, consider testing mplayer -vo gl, it is 
very fast because it does not need CPU conversion of 
YUV->RGB.)

[...]

> Is this correct?

I am certainly no expert for va-api, but I wonder if 
you already realized that ffmpeg and ffplay (the 
applications) do not support va-api in any way, 
libavcodec contains code that application may use 
to decode using va-api, see vlc source code and 
the linked MPlayer patch for examples.

Carl Eugen


From faeem.ali at gmail.com  Sat Dec 22 13:28:54 2012
From: faeem.ali at gmail.com (faeem)
Date: Sat, 22 Dec 2012 14:28:54 +0200
Subject: [Libav-user] Hardware decoding via libva
In-Reply-To: <loom.20121222T131903-681@post.gmane.org>
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org>
	<50D5A44B.5030408@gmail.com>
	<loom.20121222T131903-681@post.gmane.org>
Message-ID: <50D5A786.5080900@gmail.com>

Thanks for the YUV->RGB tip. I'll look into that.

I think I read on wikipedia that ffmpeg/libavcodec has no va-api 
support, but does contain data structures that allow it to be used with 
va-api. I figured someone here would have some idea on what to do, and 
I've already got some very useful advice and code examples.

So, thanks :)

Faeem

On 22/12/2012 14:22, Carl Eugen Hoyos wrote:
> faeem <faeem.ali at ...> writes:
>
>> I'll need to run the YUV to RGB conversion on each frame
>> if I'm running OpenGL and this will still be CPU intensive.
> (I don't think this is correct but it is not related to
> va-api at all, consider testing mplayer -vo gl, it is
> very fast because it does not need CPU conversion of
> YUV->RGB.)
>
> [...]
>
>> Is this correct?
> I am certainly no expert for va-api, but I wonder if
> you already realized that ffmpeg and ffplay (the
> applications) do not support va-api in any way,
> libavcodec contains code that application may use
> to decode using va-api, see vlc source code and
> the linked MPlayer patch for examples.
>
> Carl Eugen
>
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user


From lixiaohui9219 at 126.com  Sat Dec 22 13:59:32 2012
From: lixiaohui9219 at 126.com (=?GBK?B?wOjQobvU?=)
Date: Sat, 22 Dec 2012 20:59:32 +0800 (CST)
Subject: [Libav-user] question about hls support in ffmpeg 0.8.5
Message-ID: <63b03b2e.2282b.13bc2b2725f.Coremail.lixiaohui9219@126.com>

Hi all,
Anybody can tell me whether ffmpeg can support HLS play,pause and seek for hsl VOD using Wowza stream server ? Thanks in advance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121222/7966608b/attachment.html>

From mybrokenbeat at gmail.com  Sat Dec 22 14:22:01 2012
From: mybrokenbeat at gmail.com (Oleg)
Date: Sat, 22 Dec 2012 15:22:01 +0200
Subject: [Libav-user] Hardware decoding via libva
In-Reply-To: <50D5A44B.5030408@gmail.com>
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org>
	<50D5A44B.5030408@gmail.com>
Message-ID: <F05F22E6-4972-4545-97FD-77F0C5886B60@gmail.com>

You're misunderstanding some concepts. There are two operations that can be accelerated by GPU: decoding and yuv->rgb converting. First thing can be achieved by vaapi as you mentioned. Second, by using OpenGL shader(I prefer OpenGL as it's cross-platform. Other option is to use DirectX on Win platform) that will convert YUV->RGB and draw converted frame immediately.


  
22.12.2012, ? 14:15, faeem ???????(?):

> On 22/12/2012 02:53, Carl Eugen Hoyos wrote:
>> faeem <faeem.ali at ...> writes:
>> 
>> I know of two examples, the va-api code in vlc and the code in a patch for MPlayer, see for example http://thread.gmane.org/gmane.comp.video.mplayer.devel/61734/focus=61744 [...] 
> Thanks. I'll be looking into those examples ASAP.
> 
>>>   // FIXME use direct rendering
>>> 
>>> I need to know how to fix that FIXME.
>> This is not related to va-api at all.
> I selected that //FIXME because it specified "direct rendering", which I took to mean "handled by the GPU". It seems I was mistaken there.
> 
> My conceptual understanding of va-api thus far, within the ffmpeg framework, is that libavcodec will read an encoded video frame, then use va-api and the GPU to perform decoding of that frame instead of performing decoding in software.
> 
> The end result will probably be a frame in YUV. I'll need to run the YUV to RGB conversion on each frame if I'm running OpenGL and this will still be CPU intensive. I would benefit from the hardware frame decoding though, and that alone should make a significant difference.
> 
> Is this correct?
> 
> Faeem
> 
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user


From haihao.xiang at intel.com  Mon Dec 24 03:07:33 2012
From: haihao.xiang at intel.com (Xiang, Haihao)
Date: Mon, 24 Dec 2012 10:07:33 +0800
Subject: [Libav-user] Hardware decoding via libva
In-Reply-To: <F05F22E6-4972-4545-97FD-77F0C5886B60@gmail.com>
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org> <50D5A44B.5030408@gmail.com>
	<F05F22E6-4972-4545-97FD-77F0C5886B60@gmail.com>
Message-ID: <1356314853.2652.22.camel@xhh-ilk32>


VA API has vaPutSurface() to directly render the decoded frame on a
drawable, you can refer to the code of mplayer vaapi
(git://gitorious.org/vaapi/mplayer.git)

Regards
Haihao


> You're misunderstanding some concepts. There are two operations that can be accelerated by GPU: decoding and yuv->rgb converting. First thing can be achieved by vaapi as you mentioned. Second, by using OpenGL shader(I prefer OpenGL as it's cross-platform. Other option is to use DirectX on Win platform) that will convert YUV->RGB and draw converted frame immediately.
> 
> 
>   
> 22.12.2012, ? 14:15, faeem ???????(?):
> 
> > On 22/12/2012 02:53, Carl Eugen Hoyos wrote:
> >> faeem <faeem.ali at ...> writes:
> >> 
> >> I know of two examples, the va-api code in vlc and the code in a patch for MPlayer, see for example http://thread.gmane.org/gmane.comp.video.mplayer.devel/61734/focus=61744 [...] 
> > Thanks. I'll be looking into those examples ASAP.
> > 
> >>>   // FIXME use direct rendering
> >>> 
> >>> I need to know how to fix that FIXME.
> >> This is not related to va-api at all.
> > I selected that //FIXME because it specified "direct rendering", which I took to mean "handled by the GPU". It seems I was mistaken there.
> > 
> > My conceptual understanding of va-api thus far, within the ffmpeg framework, is that libavcodec will read an encoded video frame, then use va-api and the GPU to perform decoding of that frame instead of performing decoding in software.
> > 
> > The end result will probably be a frame in YUV. I'll need to run the YUV to RGB conversion on each frame if I'm running OpenGL and this will still be CPU intensive. I would benefit from the hardware frame decoding though, and that alone should make a significant difference.
> > 
> > Is this correct?
> > 
> > Faeem
> > 
> > _______________________________________________
> > Libav-user mailing list
> > Libav-user at ffmpeg.org
> > http://ffmpeg.org/mailman/listinfo/libav-user
> 
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user



From klaussfreire at gmail.com  Mon Dec 24 03:42:28 2012
From: klaussfreire at gmail.com (Claudio Freire)
Date: Sun, 23 Dec 2012 23:42:28 -0300
Subject: [Libav-user] Hardware decoding via libva
In-Reply-To: <F05F22E6-4972-4545-97FD-77F0C5886B60@gmail.com>
References: <50D4E7A8.7060608@gmail.com>
	<loom.20121222T014806-185@post.gmane.org>
	<50D5A44B.5030408@gmail.com>
	<F05F22E6-4972-4545-97FD-77F0C5886B60@gmail.com>
Message-ID: <CAGTBQpY-hvCwWiez7EHJB+2LrX=4uisEpw=VwDr_PkVF2S=7Pw@mail.gmail.com>

On Sat, Dec 22, 2012 at 10:22 AM, Oleg <mybrokenbeat at gmail.com> wrote:
> You're misunderstanding some concepts. There are two operations that can be accelerated by GPU: decoding and yuv->rgb converting. First thing can be achieved by vaapi as you mentioned. Second, by using OpenGL shader(I prefer OpenGL as it's cross-platform. Other option is to use DirectX on Win platform) that will convert YUV->RGB and draw converted frame immediately.

If you don't use a shader, YUV-RGB conversion not only consumes
cycles. You can actually ignore the cycles it takes on most modern
CPUs. The problem is RGB requires a lot more bandwidth to upload back
into the CPU, and that usually becomes a bottleneck on lower-end
systems for HD content (which requires quite a lot of bandwidth).

Hence my "naively" comment earlier. You don't only need to get the
result in YUV to the shader, but also in downsampled YUV (ie: 4:1:1,
4:2:1, 4:2:2 or whatever format the codec is using internally).

From nkipe at tatapowersed.com  Mon Dec 24 06:26:25 2012
From: nkipe at tatapowersed.com (Navin)
Date: Mon, 24 Dec 2012 10:56:25 +0530
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
 array using Purify
In-Reply-To: <CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
Message-ID: <50D7E781.6080409@tatapowersed.com>

Actually I'm giving avipicture_fill a value. I didn't show that bit of 
the code.
/g_pFrameRGB = avcodec_alloc_frame_proc();/

But it's strange that purify is showing an error while reading from 
data[0][0]. At least the first value should've been read correctly. Am a 
bit worried about this, since bounds errors would give problems at 
completely unexpected times.
Could anyone help with this? Perhaps you'd get the same error if you try 
even a simple ffmpeg program with Purify.

Navin

On 12/18/2012 2:57 PM, Alex Cohn wrote:
>> So I reduced the code to just this
>> pAreaInMemory[0] = g_pFrameRGB->data[0][0];
>> and I'm still getting the error (which I've shown above).
>> The code goes something like this (I've shown many of the initializations in
>> GetFrame itself although they're supposed to be outside):
>>
>> bool GetFrame()
>> {
>>     ...blah...
>>     AVFrame*  g_pFrameRGB = NULL;
>>     avpicture_fill_proc((AVPicture *) g_pFrameRGB, pbuffer, PIX_FMT_RGB24,
>> g_metadata->WIDTH, g_metadata->HEIGHT);
> avipicture_fill does not expect NULL pointer to AVPicture as first
> parameter. The pointer must be initialized (see doc/examples/muxing.c
> for the correct usage).
>
> BR,
> Alex
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>


I've got a piece of code which extracts every rgb pixel from a video and 
displays it without any problem. But when I ran it with Purify, I got 
these errors:
*
[E] ABR: Array bounds read in GetFrame {197 occurrences}
         Reading 1 byte from 0x04ae0040 (1 byte at 0x04ae0040 illegal)
         Address 0x04ae0040 is 16 bytes before the beginning of a 
1769472 byte block at 0x04ae0050
         Address 0x04ae0040 points to a malloc'd block in heap 0x003a0000
         Thread ID: 0x17c4*

So I reduced the code to just this
*pAreaInMemory[0] = g_pFrameRGB->data[0][0];*
and I'm still getting the error (which I've shown above).
The code goes something like this (I've shown many of the 
initializations in GetFrame itself although they're supposed to be outside):

bool GetFrame()
{
    ...blah...
    AVFrame*  g_pFrameRGB = NULL;
    avpicture_fill_proc((AVPicture *) g_pFrameRGB, pbuffer, 
PIX_FMT_RGB24, g_metadata->WIDTH, g_metadata->HEIGHT);
    ...blah...
    sws_scale_proc(g_sws_ctx, (uint8_t const * const *)g_pFrame->data, 
g_pFrame->linesize, 0, g_metadata->HEIGHT, g_pFrameRGB->data, 
g_pFrameRGB->linesize );
    ...blah...
    unsigned char* pAreaInMemory = (unsigned char*) malloc( 
sizeof(unsigned char) * 1024 * 768 );
    ...get the video data here...
    pAreaInMemory[0] = g_pFrameRGB->data[0][0];
    ...blah...
    free(pAreaInMemory);
}

Why is there a bounds problem when reading from data[0][0] ? I noticed 
the way data is initialized in ffmpeg. Is this purify error something I 
can ignore (there's no memory leak. Just the bounds error) or can it 
lead to serious problems?

-- 
Navin


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121224/29dca980/attachment.html>

From alexcohn at netvision.net.il  Mon Dec 24 11:35:04 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Mon, 24 Dec 2012 12:35:04 +0200
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
 array using Purify
In-Reply-To: <50D7E781.6080409@tatapowersed.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
	<50D7E781.6080409@tatapowersed.com>
Message-ID: <CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>

On Mon, Dec 24, 2012 at 7:26 AM, Navin <nkipe at tatapowersed.com> wrote:
> Actually I'm giving avipicture_fill a value. I didn't show that bit of the code.
>
> g_pFrameRGB = avcodec_alloc_frame_proc();

Yes, this should be correct.

> But it's strange that purify is showing an error while reading from data[0][0].

Actually, after your call to avpicture_fill_proc(), data[0] is
pointing to pbuffer[0]. You didn't disclose how you initialized
pbuffer, so it's impossible to tell why Purify is unhappy with this
address.

By the way, I don't understand the rationale of allocating another
chunk of memory, pAreaInMemory. Whatever you do with these bytes, why
don't you use pbuffer directly?

BR,
Alex

From mtaha.ansari at gmail.com  Mon Dec 24 11:55:17 2012
From: mtaha.ansari at gmail.com (Taha Ansari)
Date: Mon, 24 Dec 2012 15:55:17 +0500
Subject: [Libav-user] Choppy dshow audio playback with ffplay
Message-ID: <CAK2cdgCb_Z0aqhfPgarfzvTCZORTpvfD=rfh_y+jW0pt=FyL5g@mail.gmail.com>

Hi!

I have a small test application that sends microphone audio over network.
But the audio playback is sometimes very choppy/lossy, and also I initially
need to 'seek' ffplay back to hear audio with minimum latency. I do this in
Windows, using dshow, zeranoe ffmpeg builds, MSVS; and here is custom code
of relevance (output file is in extension .mp2, and packets are sent on
udp. I tried AAC extension as well, but results are somewhat the same):

***********  +  Decoding part:   +  *************
if(this->packet.stream_index == this->audioStream)
    {
        unsigned int samples_size= 0;
        AVCodecContext *c = outputCodecCtxAudio;
        int finalPTS = 0;
        samples = (short *) av_fast_realloc(samples, &samples_size,
FFMAX(packet.size, AVCODEC_MAX_AUDIO_FRAME_SIZE));
        finalPTS =  packet.pts;
        audiobufsize = AVCODEC_MAX_AUDIO_FRAME_SIZE*2;
        avcodec_decode_audio3(pCodecCtxAudio, samples, &audiobufsize,
&packet);


        if(pCodecCtxAudio->sample_rate != c->sample_rate ||
pCodecCtxAudio->channels != c->channels )
        {
            if ( rs == NULL)
            {
                rs = av_audio_resample_init(c->channels,
pCodecCtxAudio->channels, c->sample_rate, pCodecCtxAudio->sample_rate,
AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_S16, 0,0,0,0);
            }
        }
        if(pCodecCtxAudio->sample_rate != c->sample_rate ||
pCodecCtxAudio->channels != c->channels)
        {
            int size_out = audio_resample(rs, (short *)buffer_resample,
samples, audiobufsize/ (pCodecCtxAudio->channels * 2) );
            av_fifo_generic_write(fifo, (uint8_t *)buffer_resample,
size_out * c->channels * 2, NULL );
        }
        else
        {
            av_fifo_generic_write(fifo, (uint8_t *)samples, audiobufsize,
NULL );
        }
    }
***********  -  Decoding part:   -  *************

***********  + Encoding part:   + ***************
if ( decoderData->audiobufsize )
    {
        AVPacket pkt;
        av_init_packet(&pkt);

        AVCodecContext* c = encoderData->audio_st->codec;

        int frame_bytes = c->frame_size * 2 * c->channels;

        while( av_fifo_size(decoderData->fifo) >= frame_bytes )
        {
            int ret = av_fifo_generic_read( decoderData->fifo, data_buf,
frame_bytes, NULL );
            /* encode the samples */
            pkt.size= avcodec_encode_audio(c, audio_out, frame_bytes
/*packet.size*/, (short *)data_buf);

            pkt.stream_index= encoderData->audio_st->index;
            pkt.data= audio_out;
            pkt.flags |= AV_PKT_FLAG_KEY;

            pkt.pts = pkt.dts = 0;
            /* write the compressed frame in the media file */
            if (av_interleaved_write_frame(encoderData->ocAud, &pkt) != 0)
            {
                fprintf(stderr, "Error while writing audio frame\n");
                exit(1);
            }
        }
    }
***********  - Encoding part:   - ***************

Other code is similar to the muxing.c example that comes with the builds. I
know the functions used above are kind of outdated, but that is the best
working source I could find from the internet.

Can anyone kindly highlight how I could improve my code, or do I need to
tweak ffplay somehow for better results?

Thanks for your time,

Best regards
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121224/732e4d94/attachment.html>

From michael at hinespot.com  Mon Dec 24 20:54:56 2012
From: michael at hinespot.com (Michael R. Hines)
Date: Mon, 24 Dec 2012 14:54:56 -0500
Subject: [Libav-user] decoding a duplicated packet in another process
Message-ID: <1356378896.2572.19.camel@chopin>

Hi,

My task is to send AVPackets over the network (to another process)
that does not have access to the original video for transcoding
purposes. To do this, I need to duplicate the packet data and the
AVPacket struct, like this:

Pseudocode (in a 2nd C program talking TCP/IP):

1. Allocate AVCodec with avcodec_find_decoder(mpeg2)
2. Allocate AVCodecContext with avcodec_alloc_context3()
3. Open the codec with avcodec_open2()
4. av_malloc() some space to get the duplicate packet data from network

No problems so far.... 

(Note: there's no AVFormatContext or AVStream here because I want to
copy the packet rather than read it from the original stream).

Then, I proceed to read in a copy of the AVPacket structure from the
network, along with the data portion of the packet that was pulled
previously from the original video stream. (side_data_elems == 0,
so I just ignored that part of the packet).

This all works fine -- all the AVPacket values are there, the
data is assigned to the AVPacket successfully.

But, then I try to decode the packet, with
avcodec_decode_video2(duplicated packet), and I get:

[mpeg2video @ 0xa9f3420] mpeg_decode_postinit() failure

I even ran GDB and found out that the failure comes from here:

file libavcodec/mpeg12.c:1277
	function static int mpeg_decode_postinit(AVCodecContext *avctx)
		line 1277:  if ((s->width == 0) || (s->height == 0))
			           return -2;

Ok, that makes sense - so the decoder is missing some information....

Which means I'm probably not performing steps #1, 2, 3 properly.

Are there are additional parameters (like width and height) in the
AVCodecContext or AVCodec structures that I'm supposed to change before
opening the codec for decoding before the MPEG2 code can start doing its
work properly?

What am I missing?

Thanks!

- Michael R. Hines


From michael at hinespot.com  Mon Dec 24 21:12:12 2012
From: michael at hinespot.com (Michael R. Hines)
Date: Mon, 24 Dec 2012 15:12:12 -0500
Subject: [Libav-user] decoding a duplicated packet in another process
Message-ID: <1356379932.28898.1.camel@chopin>

Hi,

Looks like I answered my own question: I was duplicating the packets out of order
as they came out of the AVStream on the other side of the network.

So the decoder was not decoding the same sequence of packets that it expected.

Thanks,
- Michael R. Hines




From michael at hinespot.com  Mon Dec 24 21:20:42 2012
From: michael at hinespot.com (Michael R. Hines)
Date: Mon, 24 Dec 2012 15:20:42 -0500
Subject: [Libav-user] segfault decoding a copied packet
Message-ID: <1356380442.28898.8.camel@chopin>

Hi all,

So, this is related to a previous post: I've written a program that duplicates an AVPacket over the network.

But, when it gets to the decoder on the other side (in another C program talking TCP/IP), 
I get this from my debugger:

Program received signal SIGSEGV, Segmentation fault.
avui_decode_frame (avctx=0x0, data=0x15f90, data_size=0x0, avpkt=0x180) at libavcodec/avuidec.c:114
114	                a[2 * k    ] = 0xFF - (transparent ? *srca++ : 0);
(gdb) bt
#0  avui_decode_frame (avctx=0x0, data=0x15f90, data_size=0x0, avpkt=0x180) at libavcodec/avuidec.c:114
#1  0x00015f90 in ?? ()
(gdb) 

The code up to this point was done with the following sequence of a events:

1. avcodec_find_decoder(mpeg2)
2. avcodec_alloc_context3()
3. avcodec_open2()
4. copy the AVPacket to duplicate AVPacket
5. av_malloc() some space to store the duplicate packet data from network
6. assigned the newly allocated memory to the duplicate AVPacket
7. read the packet data bytes from the network
8. avcodec_decode_video2(duplicated packet)

Is av_malloc() not sufficient to duplicate the packet? Do I need to align
the memory somehow before ?

- Michael R. Hines

	


From michael at hinespot.com  Mon Dec 24 21:32:58 2012
From: michael at hinespot.com (Michael R. Hines)
Date: Mon, 24 Dec 2012 15:32:58 -0500
Subject: [Libav-user] segfault decoding a copied packet
Message-ID: <1356381178.28898.9.camel@chopin>

Answered my own question again: av_new_packet()

=)

- Michael R. Hines

	

	


From beyond702 at gmail.com  Tue Dec 25 10:08:10 2012
From: beyond702 at gmail.com (Brian Chi)
Date: Tue, 25 Dec 2012 17:08:10 +0800
Subject: [Libav-user] There is no picture in the output file while encoding
	with libx264 as codec.
Message-ID: <CAO80oK2uvkmOQfpi-Q41qVBStJi-qr3Y6Nh5pY7_wZQkohGtaQ@mail.gmail.com>

Hi~,

I want to realize an application that firstly decode a multi-media
file(such as test.mp4 file, video codec id is H264), get a video stream and
an audio stream, then make some different in the audio stream, at last
encode the video stream(use libx264) and audio stream into a result
file(result.mp4). To promote the efficiency, i omitted the decode and
encode of video stream, i get the video packet via function
"av_read_frame", then output it directly into the result file via function
"av_write_frame". But there is no picture in the output file, and the size
of output file is fairly small.

I tracked the ffmpeg code and found that in the function
"av_write_frame->mov_write_packet->ff_mov_write_packet", it will call
function "ff_avc_parse_nal_units" to obtain the size of nal unit, but the
return value is very small(such as 208 bytes).

I can't find where the problem is. Does anybody know that? I'm very
appreciate it.

-- 
Best Regards,
Chi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121225/8462d3c0/attachment.html>

From 34973832 at qq.com  Tue Dec 25 13:39:19 2012
From: 34973832 at qq.com (=?ISO-8859-1?B?TXIuIFhpYW8=?=)
Date: Tue, 25 Dec 2012 20:39:19 +0800
Subject: [Libav-user] ffserver, snprintf not us corrctly!
Message-ID: <tencent_186ABB46183CA95748A3390E@qq.com>

hi, in ffserver.c, there is a error when usng snprintf. see the followed code:

    /* prepare http header */
    q = c->buffer;
    q += snprintf(q, q - (char *) c->buffer + c->buffer_size, "HTTP/1.0 200 OK\r\n");  // !!! error
    mime_type = c->stream->fmt->mime_type;
    if (!mime_type)
        mime_type = "application/x-octet-stream";
    q += snprintf(q, q - (char *) c->buffer + c->buffer_size, "Pragma: no-cache\r\n");   // !!! error

    /* for asf, we need extra headers */
    if (!strcmp(c->stream->fmt->name,"asf_stream")) {
        /* Need to allocate a client id */

        c->wmp_client_id = av_lfg_get(&random_state);

        q += snprintf(q, q - (char *) c->buffer + c->buffer_size, "Server: Cougar 4.1.0.3923\r\nCache-Control: no-cache\r\nPragma: client-id=%d\r\nPragma: features=\"broadcast\"\r\n", c->wmp_client_id);  // !!! error
    }
    q += snprintf(q, q - (char *) c->buffer + c->buffer_size, "Content-Type: %s\r\n", mime_type);  // !!! error
    q += snprintf(q, q - (char *) c->buffer + c->buffer_size, "\r\n");  // !!! error

 

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!should be like this:

    /* prepare http header */
    q = c->buffer;
    q += snprintf(q, (char *) c->buffer + c->buffer_size - q, "HTTP/1.0 200 OK\r\n");
    mime_type = c->stream->fmt->mime_type;
    if (!mime_type)
        mime_type = "application/x-octet-stream";
    q += snprintf(q, (char *) c->buffer + c->buffer_size - q, "Pragma: no-cache\r\n");

    /* for asf, we need extra headers */
    if (!strcmp(c->stream->fmt->name,"asf_stream")) {
        /* Need to allocate a client id */

        c->wmp_client_id = av_lfg_get(&random_state);

        q += snprintf(q, (char *) c->buffer + c->buffer_size - q, "Server: Cougar 4.1.0.3923\r\nCache-Control: no-cache\r\nPragma: client-id=%d\r\nPragma: features=\"broadcast\"\r\n", c->wmp_client_id);
    }
    q += snprintf(q, (char *) c->buffer + c->buffer_size - q, "Content-Type: %s\r\n", mime_type);
    q += snprintf(q, (char *) c->buffer + c->buffer_size - q, "\r\n");
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121225/8cc7df99/attachment.html>

From alexcohn at netvision.net.il  Tue Dec 25 16:55:19 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Tue, 25 Dec 2012 17:55:19 +0200
Subject: [Libav-user] There is no picture in the output file while
 encoding with libx264 as codec.
In-Reply-To: <CAO80oK2uvkmOQfpi-Q41qVBStJi-qr3Y6Nh5pY7_wZQkohGtaQ@mail.gmail.com>
References: <CAO80oK2uvkmOQfpi-Q41qVBStJi-qr3Y6Nh5pY7_wZQkohGtaQ@mail.gmail.com>
Message-ID: <CADiKgoiK4gnD2-js20WLtpUqcEsYz7PVk-vwrr39-zNUd9EfAw@mail.gmail.com>

On Tue, Dec 25, 2012 at 11:08 AM, Brian Chi <beyond702 at gmail.com> wrote:
> I tracked the ffmpeg code and found that in the function
> "av_write_frame->mov_write_packet->ff_mov_write_packet", it will call
> function "ff_avc_parse_nal_units" to obtain the size of nal unit, but the
> return value is very small(such as 208 bytes).

It's very natural for the first NAL unit to be small. An h264 stream
begins with configuration SPS/PPS untis that don't contain video data,
but let the decoder initialization to work. Did you continue reading
more video packets after the first one?

Alex

From cehoyos at ag.or.at  Tue Dec 25 17:03:28 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Tue, 25 Dec 2012 16:03:28 +0000 (UTC)
Subject: [Libav-user] ffserver, snprintf not us corrctly!
References: <tencent_186ABB46183CA95748A3390E@qq.com>
Message-ID: <loom.20121225T170136-923@post.gmane.org>

Mr. Xiao <34973832 at ...> writes:

> hi, in ffserver.c, there is a error when usng snprintf.

Please send patches made with git format-patch or git send-email 
to ffmpeg-devel (git diff >patch.diff is also acceptable).

Carl Eugen


From beyond702 at gmail.com  Wed Dec 26 05:35:21 2012
From: beyond702 at gmail.com (Brian Chi)
Date: Wed, 26 Dec 2012 12:35:21 +0800
Subject: [Libav-user] There is no picture in the output file while
 encoding with libx264 as codec.
In-Reply-To: <CADiKgoiK4gnD2-js20WLtpUqcEsYz7PVk-vwrr39-zNUd9EfAw@mail.gmail.com>
References: <CAO80oK2uvkmOQfpi-Q41qVBStJi-qr3Y6Nh5pY7_wZQkohGtaQ@mail.gmail.com>
	<CADiKgoiK4gnD2-js20WLtpUqcEsYz7PVk-vwrr39-zNUd9EfAw@mail.gmail.com>
Message-ID: <CAO80oK2hfAd8jy0xR=Eb74PGnFPALCEpjnxqN-if92BWJj_EVQ@mail.gmail.com>

Yes, I read all frames, most return value of function
"ff_avc_parse_nal_units" is 0.
i think maybe the problem is about NAL, will the NAL layer of H264 frame be
detached in the function "av_read_frame".
So in the "ff_avc_parse_nal_units", it can't find start code(0x000001), and
discard the frame.
Is that right?

2012/12/25 Alex Cohn <alexcohn at netvision.net.il>

> On Tue, Dec 25, 2012 at 11:08 AM, Brian Chi <beyond702 at gmail.com> wrote:
> > I tracked the ffmpeg code and found that in the function
> > "av_write_frame->mov_write_packet->ff_mov_write_packet", it will call
> > function "ff_avc_parse_nal_units" to obtain the size of nal unit, but the
> > return value is very small(such as 208 bytes).
>
> It's very natural for the first NAL unit to be small. An h264 stream
> begins with configuration SPS/PPS untis that don't contain video data,
> but let the decoder initialization to work. Did you continue reading
> more video packets after the first one?
>
> Alex
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>



-- 
Best Regards,
Chi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121226/851a3be1/attachment.html>

From nkipe at tatapowersed.com  Wed Dec 26 06:52:14 2012
From: nkipe at tatapowersed.com (Navin)
Date: Wed, 26 Dec 2012 11:22:14 +0530
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
 array using Purify
In-Reply-To: <CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
	<50D7E781.6080409@tatapowersed.com>
	<CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>
Message-ID: <50DA908E.6080504@tatapowersed.com>

Okay, so basically this is how the code goes. It's from the ffmpeg 
tutorials itself. Could the code be wrong because I used sws_getContext?
g_metadata is a struct I constructed, for holding some properties of the 
video.

AVFormatContext*   g_pFormatCtx = NULL;
AVCodecContext*    g_g_pCodecCtx = NULL;
int                g_videoStream;
int                g_numBytes;
AVCodec*           g_pCodec = NULL;
AVFrame*           g_pFrame = NULL;
AVFrame*           g_pFrameRGB = NULL;
AVDictionary*      g_optionsDict = NULL;
struct SwsContext* g_sws_ctx = NULL;


         av_register_all();
         if (avformat_open_input(&g_pFormatCtx, 
g_metadata->videoLocation.c_str(), NULL, NULL) != 0)    { 
freeVideoData(); return false; }
         if (avformat_find_stream_info(g_pFormatCtx, NULL) < 0)  { 
freeVideoData(); return false; }

         g_videoStream = -1;
         for(unsigned int i = 0; i < g_pFormatCtx->nb_streams; i++)
         {
             if (g_pFormatCtx->streams[i]->codec->codec_type == 
AVMEDIA_TYPE_VIDEO)
             {
                 g_videoStream=i;
                 break; //breaks out of for loop
             }
         }//for
         if (g_videoStream == -1) return false;

         g_memoryBuffer->size = unsigned int(g_metadata->WIDTH * 
g_metadata->HEIGHT * sizeof(RGBPixel));//encapsulating the calculation 
of memory size

         g_pCodec = avcodec_find_decoder(g_g_pCodecCtx->codec_id);
         if (NULL == g_pCodec) return false;
         if (avcodec_open2(g_g_pCodecCtx, g_pCodec, &g_optionsDict) < 0) 
return false;
         g_pFrame = avcodec_alloc_frame();
         if (NULL == g_pFrame) return false;

         g_pFrameRGB = avcodec_alloc_frame();
         if (NULL == g_pFrameRGB)  return false;

         uint8_t* pbuffer = NULL;
         g_numBytes = avpicture_get_size(PIX_FMT_RGB24, 
g_metadata->WIDTH, g_metadata->HEIGHT);
         pbuffer = (uint8_t *) av_malloc(g_numBytes * 
sizeof(uint8_t));//TODO: exception handling may be required for allocs

         g_sws_ctx = sws_getContext(g_metadata->WIDTH, 
g_metadata->HEIGHT, g_g_pCodecCtx->pix_fmt, \
                                         g_metadata->WIDTH, 
g_metadata->HEIGHT, PIX_FMT_RGB24, SWS_BILINEAR, NULL, NULL, NULL );

         avpicture_fill((AVPicture *) g_pFrameRGB, pbuffer, 
PIX_FMT_RGB24, g_metadata->WIDTH, g_metadata->HEIGHT);
         av_free(pbuffer);


Navin

On 12/24/2012 4:05 PM, Alex Cohn wrote:
> On Mon, Dec 24, 2012 at 7:26 AM, Navin<nkipe at tatapowersed.com>  wrote:
>> Actually I'm giving avipicture_fill a value. I didn't show that bit of the code.
>>
>> g_pFrameRGB = avcodec_alloc_frame_proc();
> Yes, this should be correct.
>
>> But it's strange that purify is showing an error while reading from data[0][0].
> Actually, after your call to avpicture_fill_proc(), data[0] is
> pointing to pbuffer[0]. You didn't disclose how you initialized
> pbuffer, so it's impossible to tell why Purify is unhappy with this
> address.
>
> By the way, I don't understand the rationale of allocating another
> chunk of memory, pAreaInMemory. Whatever you do with these bytes, why
> don't you use pbuffer directly?
>
> BR,
> Alex
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>

From alexcohn at netvision.net.il  Wed Dec 26 08:33:46 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Wed, 26 Dec 2012 09:33:46 +0200
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
 array using Purify
In-Reply-To: <50DA908E.6080504@tatapowersed.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
	<50D7E781.6080409@tatapowersed.com>
	<CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>
	<50DA908E.6080504@tatapowersed.com>
Message-ID: <CADiKgoiJ-=P673sBRVS_VmuKMRYVNFWM7jeit4rKHTva1+t-vQ@mail.gmail.com>

On Wed, Dec 26, 2012 at 7:52 AM, Navin <nkipe at tatapowersed.com> wrote:
>         avpicture_fill((AVPicture *) g_pFrameRGB, pbuffer, PIX_FMT_RGB24, g_metadata->WIDTH, g_metadata->HEIGHT);
.. so far so good
>         av_free(pbuffer);
oops, you still expect g_pFrameRGB->data[0] to be a valid array. But
you have released this memory!

BR,
Alex

From michael at hinespot.com  Wed Dec 26 11:15:36 2012
From: michael at hinespot.com (Michael R. Hines)
Date: Wed, 26 Dec 2012 05:15:36 -0500
Subject: [Libav-user] keyframe not a keyframe?
In-Reply-To: <50DA908E.6080504@tatapowersed.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
	<50D7E781.6080409@tatapowersed.com>
	<CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>
	<50DA908E.6080504@tatapowersed.com>
Message-ID: <1356516936.28898.16.camel@chopin>

Hi,

After decoding a packet containing AV_PKT_FLAG_KEY in the flags,
I was expecting to get I-frames, but instead I got P-frames:

The following statement:

	printf("packet flags: %d picture type: %c\n", inPacket.flags,
		av_get_picture_type_char(frame->pict_type));

Returns the output:

	packet flags: 1 picture type: P

When I was expecting:

	packet flags: 1 picture type: I

Where '1' == AV_PKT_FLAG_KEY.

Is this behavior correct? (I'm decoding MPEG2 video).

Thanks,

- Michael


From santhosh.kumar.bachu at gmail.com  Wed Dec 26 11:34:26 2012
From: santhosh.kumar.bachu at gmail.com (Santhosh Kumar)
Date: Wed, 26 Dec 2012 16:04:26 +0530
Subject: [Libav-user] test mail
Message-ID: <CAN_HgZk4W8hY0VBQemrEwx_zb+Gfq-3cvbC6MpNUu3Vwpn7ubA@mail.gmail.com>

-- 
Santhosh
-- Symbol of Smile:)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121226/99fc3052/attachment.html>

From alexcohn at netvision.net.il  Wed Dec 26 11:39:51 2012
From: alexcohn at netvision.net.il (Alex Cohn)
Date: Wed, 26 Dec 2012 12:39:51 +0200
Subject: [Libav-user] There is no picture in the output file while
 encoding with libx264 as codec.
In-Reply-To: <CAO80oK2hfAd8jy0xR=Eb74PGnFPALCEpjnxqN-if92BWJj_EVQ@mail.gmail.com>
References: <CAO80oK2uvkmOQfpi-Q41qVBStJi-qr3Y6Nh5pY7_wZQkohGtaQ@mail.gmail.com>
	<CADiKgoiK4gnD2-js20WLtpUqcEsYz7PVk-vwrr39-zNUd9EfAw@mail.gmail.com>
	<CAO80oK2hfAd8jy0xR=Eb74PGnFPALCEpjnxqN-if92BWJj_EVQ@mail.gmail.com>
Message-ID: <CADiKgohfwH5RsATzAjvRF4E24oB1NwxkR6KxP-SsdzeHg5p9Hg@mail.gmail.com>

On Wed, Dec 26, 2012 at 6:35 AM, Brian Chi <beyond702 at gmail.com> wrote:
> Yes, I read all frames, most return value of function "ff_avc_parse_nal_units" is 0.
> i think maybe the problem is about NAL, will the NAL layer of H264 frame be detached in the function "av_read_frame".
> So in the "ff_avc_parse_nal_units", it can't find start code(0x000001), and discard the frame.

Packets in mp4 are not stored in annex B format, you will never see
0x000001 start codes there. See
http://msdn.microsoft.com/en-us/library/windows/desktop/dd757808(v=vs.85).aspx

From nkipe at tatapowersed.com  Wed Dec 26 12:39:24 2012
From: nkipe at tatapowersed.com (Navin)
Date: Wed, 26 Dec 2012 17:09:24 +0530
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
 array using Purify
In-Reply-To: <CADiKgoiJ-=P673sBRVS_VmuKMRYVNFWM7jeit4rKHTva1+t-vQ@mail.gmail.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
	<50D7E781.6080409@tatapowersed.com>
	<CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>
	<50DA908E.6080504@tatapowersed.com>
	<CADiKgoiJ-=P673sBRVS_VmuKMRYVNFWM7jeit4rKHTva1+t-vQ@mail.gmail.com>
Message-ID: <50DAE1EC.8000707@tatapowersed.com>

Thank you! Thank you! Thank you! Thank you! Thank you! Thank you! Thank 
you! Thank you! Thank you!sooooo much! That was the problem. Solved now.

Navin

On 12/26/2012 1:03 PM, Alex Cohn wrote:
> On Wed, Dec 26, 2012 at 7:52 AM, Navin<nkipe at tatapowersed.com>  wrote:
>>          avpicture_fill((AVPicture *) g_pFrameRGB, pbuffer, PIX_FMT_RGB24, g_metadata->WIDTH, g_metadata->HEIGHT);
> .. so far so good
>>          av_free(pbuffer);
> oops, you still expect g_pFrameRGB->data[0] to be a valid array. But
> you have released this memory!
>
> BR,
> Alex
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>

From rjvbertin at gmail.com  Wed Dec 26 13:50:44 2012
From: rjvbertin at gmail.com (=?iso-8859-1?Q?=22Ren=E9_J=2EV=2E_Bertin=22?=)
Date: Wed, 26 Dec 2012 13:50:44 +0100
Subject: [Libav-user] Array bounds error while reading from ffmpeg data
	array using Purify
In-Reply-To: <50DA908E.6080504@tatapowersed.com>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
	<50D7E781.6080409@tatapowersed.com>
	<CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>
	<50DA908E.6080504@tatapowersed.com>
Message-ID: <EEDCA23A-A753-4F47-AA12-5E6E66CCB99C@gmail.com>


On Dec 26, 2012, at 06:52, Navin wrote:
>        g_pFrameRGB = avcodec_alloc_frame();
>        if (NULL == g_pFrameRGB)  return false;
> 
>        uint8_t* pbuffer = NULL;
>        g_numBytes = avpicture_get_size(PIX_FMT_RGB24, g_metadata->WIDTH, g_metadata->HEIGHT);
>        pbuffer = (uint8_t *) av_malloc(g_numBytes * sizeof(uint8_t));//TODO: exception handling may be required for allocs
> 
>        g_sws_ctx = sws_getContext(g_metadata->WIDTH, g_metadata->HEIGHT, g_g_pCodecCtx->pix_fmt, \
>                                        g_metadata->WIDTH, g_metadata->HEIGHT, PIX_FMT_RGB24, SWS_BILINEAR, NULL, NULL, NULL );
> 
>        avpicture_fill((AVPicture *) g_pFrameRGB, pbuffer, PIX_FMT_RGB24, g_metadata->WIDTH, g_metadata->HEIGHT);
>        av_free(pbuffer);
> 
> 
> Navin

>> oops, you still expect g_pFrameRGB->data[0] to be a valid array. But
>> you have released this memory!
> 

Oh, adorable, the sort of mess-up that's so easy to make and hard to trace. I cannot help but remark that avpicture_fill receives all the information required to allocate the working memory itself, making it more 'logical' to deallocate said memory in a subsequent call to a matching deconstructor function (e.g. avcodec_free_frame). Is there a reason for not having implemented such a default case ... or is it indeed possible to pass NULL as the second argument to avpicture_fill to have it do the allocation?

R.

From beyond702 at gmail.com  Thu Dec 27 04:23:20 2012
From: beyond702 at gmail.com (Brian Chi)
Date: Thu, 27 Dec 2012 11:23:20 +0800
Subject: [Libav-user] There is no picture in the output file while
 encoding with libx264 as codec.
In-Reply-To: <CADiKgohfwH5RsATzAjvRF4E24oB1NwxkR6KxP-SsdzeHg5p9Hg@mail.gmail.com>
References: <CAO80oK2uvkmOQfpi-Q41qVBStJi-qr3Y6Nh5pY7_wZQkohGtaQ@mail.gmail.com>
	<CADiKgoiK4gnD2-js20WLtpUqcEsYz7PVk-vwrr39-zNUd9EfAw@mail.gmail.com>
	<CAO80oK2hfAd8jy0xR=Eb74PGnFPALCEpjnxqN-if92BWJj_EVQ@mail.gmail.com>
	<CADiKgohfwH5RsATzAjvRF4E24oB1NwxkR6KxP-SsdzeHg5p9Hg@mail.gmail.com>
Message-ID: <CAO80oK1Fs0X94qcdBsSc86Aag9YhCUF9J_JapXUuJ_6ydUCYgg@mail.gmail.com>

I see, but the function
"av_write_frame->mov_write_packet->ff_mov_write_packet->ff_avc_parse_nal_units->ff_avc_find_startcode"
find the start code 0x000001, how can i resolve this problem to make it
works?
The first 4 bytes in the AVPacket is the size of this NALU, should i
replace the first 4 bytes with 0x000001 or 0x00000001? Is that ok?
Thanks a lot for your help.

2012/12/26 Alex Cohn <alexcohn at netvision.net.il>

> On Wed, Dec 26, 2012 at 6:35 AM, Brian Chi <beyond702 at gmail.com> wrote:
> > Yes, I read all frames, most return value of function
> "ff_avc_parse_nal_units" is 0.
> > i think maybe the problem is about NAL, will the NAL layer of H264 frame
> be detached in the function "av_read_frame".
> > So in the "ff_avc_parse_nal_units", it can't find start code(0x000001),
> and discard the frame.
>
> Packets in mp4 are not stored in annex B format, you will never see
> 0x000001 start codes there. See
>
> http://msdn.microsoft.com/en-us/library/windows/desktop/dd757808(v=vs.85).aspx
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
>



-- 
Best Regards,
Chi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121227/c0e88b92/attachment.html>

From eng.m.behery at live.com  Thu Dec 27 15:55:13 2012
From: eng.m.behery at live.com (mustafa elbehery)
Date: Thu, 27 Dec 2012 16:55:13 +0200
Subject: [Libav-user] =?windows-1256?q?Help_to_convert_MP3_to_WMA_=28_16k_?=
 =?windows-1256?q?bit=2C_22k_Hz_=29=FE?=
Message-ID: <BAY148-W3014F1755AC7497790B3F3A3380@phx.gbl>


Dear sir, First of all I have to show my great respect for all the developers of "FFMPEG". I'm sending this e-mail to you specifically to help me solve my problem.I'm developing a JAVA module which converting from MP3 Format to many other formats, and I used (JAVE http://www.sauronsoftware.it/projects/jave/index.php), open src Java lib based on "FFMPEG", and I faced a problem in converting to "WMA (16k bit, 22k Hz) " format. I tried even the latest FFMPEG dist. which contains {wmalossless, wmapro, wmav1, wmav2, wmavoice} codecs, but unfortunately they only of them who can ENCODE to WMA are {wmav1, wmav2}, and they are not accepting my encoding parameters. I'm really confused and do not know what to do, the encoding is working fine from MP3 to WMA with (64k bit, 22k Hz) , but when I change to (16 k bit), it always broken.I'm waiting for your response because I really spent two working days without a result, I hope you can understand my situation, Looking forward to hear from you soon, Reagrds. 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121227/f61fad3b/attachment.html>

From cehoyos at ag.or.at  Thu Dec 27 23:10:41 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Thu, 27 Dec 2012 22:10:41 +0000 (UTC)
Subject: [Libav-user]
	=?utf-8?q?Help_to_convert_MP3_to_WMA_=28_16k_bit=2C_?=
	=?utf-8?b?MjJrIEh6ICnigI8=?=
References: <BAY148-W3014F1755AC7497790B3F3A3380@phx.gbl>
Message-ID: <loom.20121227T230959-288@post.gmane.org>

mustafa elbehery <eng.m.behery at ...> writes:

> I faced a problem in converting to "WMA (16k bit, 22k Hz)" format.

Do you have a 16k wma sample?

Carl Eugen


From eng.m.behery at live.com  Fri Dec 28 10:24:46 2012
From: eng.m.behery at live.com (mustafa elbehery)
Date: Fri, 28 Dec 2012 11:24:46 +0200
Subject: [Libav-user]
 =?windows-1256?q?Help_to_convert_MP3_to_WMA_=28_16k_?=
 =?windows-1256?q?bit=2C_22k_Hz_=29=FE?=
In-Reply-To: <loom.20121227T230959-288@post.gmane.org>
References: <BAY148-W3014F1755AC7497790B3F3A3380@phx.gbl>,
	<loom.20121227T230959-288@post.gmane.org>
Message-ID: <BAY148-W486D6E22DF9B2B057098DCA33F0@phx.gbl>


yes I do.
 

> To: libav-user at ffmpeg.org
> From: cehoyos at ag.or.at
> Date: Thu, 27 Dec 2012 22:10:41 +0000
> Subject: Re: [Libav-user] Help to convert MP3 to WMA ( 16k bit, 22k Hz )?
> 
> mustafa elbehery <eng.m.behery at ...> writes:
> 
> > I faced a problem in converting to "WMA (16k bit, 22k Hz)" format.
> 
> Do you have a 16k wma sample?
> 
> Carl Eugen
> 
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121228/a40dedbb/attachment.html>

From cehoyos at ag.or.at  Fri Dec 28 12:00:41 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Fri, 28 Dec 2012 11:00:41 +0000 (UTC)
Subject: [Libav-user]
	=?utf-8?q?Help_to_convert_MP3_to_WMA_=28_16k_bit=2C_?=
	=?utf-8?b?MjJrIEh6ICnigI8=?=
References: <BAY148-W3014F1755AC7497790B3F3A3380@phx.gbl>,
	<loom.20121227T230959-288@post.gmane.org>
	<BAY148-W486D6E22DF9B2B057098DCA33F0@phx.gbl>
Message-ID: <loom.20121228T115901-265@post.gmane.org>

mustafa elbehery <eng.m.behery at ...> writes:

?> > > I faced a problem in converting to "WMA (16k bit, 22k Hz)" format.
> > 
> > Do you have a 16k wma sample?
> 
> yes I do.?

Then please either post ffmpeg -i yoursample output or 
provide the file.

Please do not top-post here and please fix your mailer 
to use "text-only", Carl Eugen


From michael at hinespot.com  Fri Dec 28 19:08:51 2012
From: michael at hinespot.com (Michael R. Hines)
Date: Fri, 28 Dec 2012 13:08:51 -0500
Subject: [Libav-user] keyframe not a keyframe?
In-Reply-To: <1356516936.28898.16.camel@chopin>
References: <50CEF1CC.50308@tatapowersed.com>
	<CADiKgogpGDGKgqg6=xJfRBvO=yoXTbyaaTC2C0t2spwh2VFeDA@mail.gmail.com>
	<50D7E781.6080409@tatapowersed.com>
	<CADiKgoi674frf=thwtSQNE2_tVWRg4nZ+9ONYX1L7C2rVKWcDg@mail.gmail.com>
	<50DA908E.6080504@tatapowersed.com> <1356516936.28898.16.camel@chopin>
Message-ID: <1356718131.22135.0.camel@salieri>

I got a response from the ffmpeg bugtracker:

https://ffmpeg.org/trac/ffmpeg/ticket/2074


        I suspect your are missing the codec delay. The video frame you
        get out usually isn't based on the packet you just sent in at
        all. This is especially true with multithreaded decoding, but
        even without it's not generally possible due to how B-frames
        work.
        


And, indeed I print out the ->pict_type "picture type" 3 frames later,
there is an I-frame that pops up a few frames later with a "delay" of
sorts exactly a 3 frames later. In my case, I have a quad-core computer,
so perhaps there's a thread being allocated for each core on my computer
to do the encoding.

- Michael

On Wed, 2012-12-26 at 05:15 -0500, Michael R. Hines wrote:

> Hi,
> 
> After decoding a packet containing AV_PKT_FLAG_KEY in the flags,
> I was expecting to get I-frames, but instead I got P-frames:
> 
> The following statement:
> 
> 	printf("packet flags: %d picture type: %c\n", inPacket.flags,
> 		av_get_picture_type_char(frame->pict_type));
> 
> Returns the output:
> 
> 	packet flags: 1 picture type: P
> 
> When I was expecting:
> 
> 	packet flags: 1 picture type: I
> 
> Where '1' == AV_PKT_FLAG_KEY.
> 
> Is this behavior correct? (I'm decoding MPEG2 video).
> 
> Thanks,
> 
> - Michael
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121228/da2ff037/attachment.html>

From Mohamed.Salim at desay-svautomotive.com  Fri Dec 28 04:13:30 2012
From: Mohamed.Salim at desay-svautomotive.com (Salim, Mohamed)
Date: Fri, 28 Dec 2012 11:13:30 +0800
Subject: [Libav-user] Incorrect frame rate
Message-ID: <76C45F99342CDC46B7CE253C2343BDA11BA979@hzhe007a.v01.net>

HI,

 

I am using libav - 0.6.

I use AVCodecContext to read the framerate and bitrate of the video /
audio stream.

However, I consistently get wrong results on frame rate for wmv / asf
streams. The value is always 1000.

 

Is there a correction available in latest releases of libav?

 

Also in some cases (.mov, .mkv), bit rate is shown to be 0. Is it also
corrected in newer releases.

 

Kindest Regards,

Salim

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121228/95a8d014/attachment.html>

From savanevery at gmail.com  Sat Dec 29 10:16:23 2012
From: savanevery at gmail.com (Shawn Van Every)
Date: Sat, 29 Dec 2012 04:16:23 -0500
Subject: [Libav-user] Encoding and writing to a file with libavformat and
	libavcodec
Message-ID: <CADo2LyAd1S1de=BEfmqHcy6U3ZdHG1OaLWoRX2Bs8AfXv7m4ZA@mail.gmail.com>

Hi all,

Just getting started with the libraries and in need of a bit of help.

I have been building up a sample based off of the decoding_encoding example
along with some sample code using AVFormat.

Right now, I am running into issues which I am having a hard time finding
the cause of.   Here is the error along with the backtrace:

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000090
0x00000001000870b5 in ff_mov_write_packet ()
(gdb) bt
#0  0x00000001000870b5 in ff_mov_write_packet ()
#1  0x00007fff8e58cb28 in szone_malloc_should_clear ()
#2  0x00007fff8e57f71a in malloc_zone_memalign ()
#3  0x00007fff8e57ff4e in posix_memalign ()
#4  0x000000010000f9b5 in av_strdup ()

Full source is below.

Thanks for any help.

Best,
shawn

--

#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <math.h>

#include <libavcodec/avcodec.h>
#include <libavformat/avformat.h>
#include <libswscale/swscale.h>
#include <libavutil/avutil.h>

#undef exit

#define STREAM_DURATION   5.0 /* 5 seconds stream duration */
#define STREAM_FRAME_RATE 25 /* 25 images/s */
#define STREAM_NB_FRAMES  ((int)(STREAM_DURATION * STREAM_FRAME_RATE))
#define STREAM_PIX_FMT PIX_FMT_YUV420P /* default pix_fmt */

static int sws_flags = SWS_BICUBIC;

int main(int argc, char **argv)
{
    // Output file
    const char *outputfile;

    // Struct to hold the output format details, audio codec, video codec,
extension
    // http://ffmpeg.org/doxygen/trunk/structAVOutputFormat.html
    AVOutputFormat *avOutputFormat; // Was *fmt

    // http://ffmpeg.org/doxygen/trunk/structAVFormatContext.html
    AVFormatContext *avFormatContext; // was *oc

    // Individual AVStreams for audio and video
    // http://ffmpeg.org/doxygen/trunk/structAVStream.html
    AVStream *video_avstream; // was *video_st and *st

// Audio and Video Timestamps (presentation time stamp)
    double video_pts;
 int codec_id = AV_CODEC_ID_H264;
    AVCodec *codec;
    AVCodecContext *c= NULL;
    int i, ret, x, y, got_output;
    int write_ret;
    //FILE *f;
    AVFrame *frame;
    AVPacket pkt;
    uint8_t endcode[] = { 0, 0, 1, 0xb7 };

    printf("Encode video file\n");

    /* register all the codecs */
    //avcodec_register_all();

// Intialize libavcodec - register all codecs and formats??????????
    av_register_all();

// Set inpput filename and output filename
    outputfile = "outtestmov.mp4";

    // auto detect the output format from the name. default is mpeg.
    avOutputFormat = av_guess_format(NULL, outputfile, NULL);
    if (!avOutputFormat) {
        printf("Could not deduce output format from file extension: using
MPEG.\n");
        avOutputFormat = av_guess_format("mpeg", NULL, NULL);
    }
    if (!avOutputFormat) {
        fprintf(stderr, "Could not find suitable output format\n");
        exit(1);
    }

    avOutputFormat->video_codec = codec_id;

    /// allocate the output media context, avFormatContext
    avFormatContext = avformat_alloc_context();
    if (!avFormatContext) {
        fprintf(stderr, "Memory error\n");
        exit(1);
    }
    // Set the output format of the newly allocated avFormatContext to our
avOutputFormat
    avFormatContext->oformat = avOutputFormat;

    // Set the output filename to the outputfile
    snprintf(avFormatContext->filename, sizeof(avFormatContext->filename),
"%s", outputfile);

    /* find the mpeg1 video encoder */
    //codec = avcodec_find_encoder(codec_id);
    codec = avcodec_find_encoder(avOutputFormat->video_codec);
    if (!codec) {
        fprintf(stderr, "Codec not found\n");
        exit(1);
    }

    // add the video stream using the default format codec and initialize
the codecs
    video_avstream = avformat_new_stream(avFormatContext, codec);
    if (!video_avstream) {
        fprintf(stderr, "Could not alloc stream\n");
        exit(1);
    }

    if (video_avstream->codec == NULL) {
    fprintf(stderr, "AVStream codec is NULL\n");
    exit(1);
    }

    //c = avcodec_alloc_context3(codec);
    c = video_avstream->codec;
    //c->codec_id = avOutputFormat->video_codec;
    if (!c) {
        fprintf(stderr, "Could not allocate video codec context\n");
        exit(1);
    }

    /* put sample parameters */
    c->bit_rate = 400000;
    /* resolution must be a multiple of two */
    c->width = 352;
    c->height = 288;
    /* 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 = AV_PIX_FMT_YUV420P;

    if(codec_id == AV_CODEC_ID_H264)
        av_opt_set(c->priv_data, "preset", "slow", 0);

    /* open it */
    if (avcodec_open2(c, codec, NULL) < 0) {
        fprintf(stderr, "Could not open codec\n");
        exit(1);
    }
 //video_avstream->codec = c;
    frame = avcodec_alloc_frame();
    if (!frame) {
        fprintf(stderr, "Could not allocate video frame\n");
        exit(1);
    }
    frame->format = c->pix_fmt;
    frame->width  = c->width;
    frame->height = c->height;

    /* the image can be allocated by any means and av_image_alloc() is
     * just the most convenient way if av_malloc() is to be used */
    ret = av_image_alloc(frame->data, frame->linesize, c->width, c->height,
                         c->pix_fmt, 32);
    if (ret < 0) {
        fprintf(stderr, "Could not allocate raw picture buffer\n");
        exit(1);
    }

    // some formats want stream headers to be separate
    if(avFormatContext->oformat->flags & AVFMT_GLOBALHEADER)
        c->flags |= CODEC_FLAG_GLOBAL_HEADER;

avformat_write_header(avFormatContext, NULL);

    /* encode 1 second of video */
    for(i=0;i<25;i++) {
        av_init_packet(&pkt);
        pkt.data = NULL;    // packet data will be allocated by the encoder
        pkt.size = 0;

        fflush(stdout);
        /* prepare a dummy image */
        /* Y */
        for(y=0;y<c->height;y++) {
            for(x=0;x<c->width;x++) {
                frame->data[0][y * frame->linesize[0] + x] = x + y + i * 3;
            }
        }

        /* Cb and Cr */
        for(y=0;y<c->height/2;y++) {
            for(x=0;x<c->width/2;x++) {
                frame->data[1][y * frame->linesize[1] + x] = 128 + y + i *
2;
                frame->data[2][y * frame->linesize[2] + x] = 64 + x + i * 5;
            }
        }

frame->pts = i;

        /* encode the image */
        ret = avcodec_encode_video2(c, &pkt, frame, &got_output);
        if (ret < 0) {
            fprintf(stderr, "Error encoding frame\n");
            exit(1);
        }

        if (got_output) {
            /* write the compressed frame in the media file */
            write_ret = av_write_frame(avFormatContext, &pkt);
            if (write_ret < 0) {
            fprintf(stderr, "Error writing frame\n");
            exit(1);
            }

            av_free_packet(&pkt);
        } else {
        printf("got_output false: %d", i);
        }
    }

    /* write the trailer, if any.  the trailer must be written
     * before you close the CodecContexts open when you wrote the
     * header; otherwise write_trailer may try to use memory that
     * was freed on av_codec_close() */
    av_write_trailer(avFormatContext);

    avcodec_close(c);
    av_free(c);
    av_freep(&frame->data[0]);
    avcodec_free_frame(&frame);

    /* free the streams */
    for(i = 0; i < avFormatContext->nb_streams; i++) {
        av_freep(&avFormatContext->streams[i]->codec);
        av_freep(&avFormatContext->streams[i]);
    }

    if (!(avOutputFormat->flags & AVFMT_NOFILE)) {
        /* close the output file */
        avio_close(avFormatContext->pb);
    }

    /* free the stream */
    av_free(avFormatContext);

    printf("\n");
    return 0;
}
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121229/b89f6bee/attachment.html>

From cehoyos at ag.or.at  Sat Dec 29 12:17:23 2012
From: cehoyos at ag.or.at (Carl Eugen Hoyos)
Date: Sat, 29 Dec 2012 11:17:23 +0000 (UTC)
Subject: [Libav-user] Encoding and writing to a file with libavformat
	and libavcodec
References: <CADo2LyAd1S1de=BEfmqHcy6U3ZdHG1OaLWoRX2Bs8AfXv7m4ZA@mail.gmail.com>
Message-ID: <loom.20121229T121626-592@post.gmane.org>

Shawn Van Every <savanevery at ...> writes:

> Program received signal EXC_BAD_ACCESS, Could not access memory.
> Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000090
> 0x00000001000870b5 in ff_mov_write_packet ()
> (gdb) bt
> #0 ?0x00000001000870b5 in ff_mov_write_packet ()
> #1 ?0x00007fff8e58cb28 in szone_malloc_should_clear ()
> #2 ?0x00007fff8e57f71a in malloc_zone_memalign ()
> #3 ?0x00007fff8e57ff4e in posix_memalign ()
> #4 ?0x000000010000f9b5 in av_strdup ()

The backtrace does not look very helpful, 
either use static linking or use --disable-stripping.

Carl Eugen


From savanevery at gmail.com  Sat Dec 29 16:50:38 2012
From: savanevery at gmail.com (Shawn Van Every)
Date: Sat, 29 Dec 2012 10:50:38 -0500
Subject: [Libav-user] Encoding and writing to a file with libavformat
	and libavcodec
In-Reply-To: <loom.20121229T121626-592@post.gmane.org>
References: <CADo2LyAd1S1de=BEfmqHcy6U3ZdHG1OaLWoRX2Bs8AfXv7m4ZA@mail.gmail.com>
	<loom.20121229T121626-592@post.gmane.org>
Message-ID: <CADo2LyB5ErvfPsG2G0pFU=ygPeLzXjT8EFqHWs2K-60+9nWCHA@mail.gmail.com>

On Sat, Dec 29, 2012 at 6:17 AM, Carl Eugen Hoyos <cehoyos at ag.or.at> wrote:

> Shawn Van Every <savanevery at ...> writes:
>
> > Program received signal EXC_BAD_ACCESS, Could not access memory.
> > Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000090
> > 0x00000001000870b5 in ff_mov_write_packet ()
> > (gdb) bt
> > #0  0x00000001000870b5 in ff_mov_write_packet ()
> > #1  0x00007fff8e58cb28 in szone_malloc_should_clear ()
> > #2  0x00007fff8e57f71a in malloc_zone_memalign ()
> > #3  0x00007fff8e57ff4e in posix_memalign ()
> > #4  0x000000010000f9b5 in av_strdup ()
>
> The backtrace does not look very helpful,
> either use static linking or use --disable-stripping.
>
> Carl Eugen
>

Thanks Carl, here is an updated backtrace:

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000090
0x0000000100090095 in mov_write_header (s=0x7fff5fbff410) at movenc.c:3499
3499    if (!s->pb->seekable &&

 It looks like I either haven't instantiated AVFormatContext (s in
movenc.c:3499) or AVIOContext (s->pb in movenc.c 3499) correctly.
 Unfortunately, I haven't seen AVIOContext used in any of the examples.

Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121229/d65bc95b/attachment.html>

From savanevery at gmail.com  Sat Dec 29 20:29:44 2012
From: savanevery at gmail.com (Shawn Van Every)
Date: Sat, 29 Dec 2012 14:29:44 -0500
Subject: [Libav-user] Encoding and writing to a file with libavformat
	and libavcodec
In-Reply-To: <CADo2LyB5ErvfPsG2G0pFU=ygPeLzXjT8EFqHWs2K-60+9nWCHA@mail.gmail.com>
References: <CADo2LyAd1S1de=BEfmqHcy6U3ZdHG1OaLWoRX2Bs8AfXv7m4ZA@mail.gmail.com>
	<loom.20121229T121626-592@post.gmane.org>
	<CADo2LyB5ErvfPsG2G0pFU=ygPeLzXjT8EFqHWs2K-60+9nWCHA@mail.gmail.com>
Message-ID: <CADo2LyA2+ifv7PHj63xrevXCfL+JuYhVWkRVspt7wkZ2FDn3sw@mail.gmail.com>

On Sat, Dec 29, 2012 at 10:50 AM, Shawn Van Every <savanevery at gmail.com>wrote:

>
> Program received signal EXC_BAD_ACCESS, Could not access memory.
> Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000090
> 0x0000000100090095 in mov_write_header (s=0x7fff5fbff410) at movenc.c:3499
> 3499    if (!s->pb->seekable &&
>
>  It looks like I either haven't instantiated AVFormatContext (s in
> movenc.c:3499) or AVIOContext (s->pb in movenc.c 3499) correctly.
>  Unfortunately, I haven't seen AVIOContext used in any of the examples.
>
> Thanks
>

Ahh..  I missed actually opening the output file.  Doing the following
fixes that error:

    /* open the output file, if needed */
    if (!(avOutputFormat->flags & AVFMT_NOFILE)) {
        if (avio_open(&avFormatContext->pb, outputfile, AVIO_FLAG_WRITE) <
0) {
            fprintf(stderr, "Could not open '%s'\n", outputfile);
            return 1;
        }
    }

Now I have this:

mine(1228,0x7fff70ca4180) malloc: *** error for object 0x10180f800: pointer
being freed was not allocated
*** set a breakpoint in malloc_error_break to debug

Program received signal SIGABRT, Aborted.
0x00007fff83cab212 in __pthread_kill ()
(gdb) bt
#0  0x00007fff83cab212 in __pthread_kill ()
#1  0x00007fff8332bb34 in pthread_kill ()
#2  0x00007fff8336fdfa in abort ()
#3  0x00007fff83343989 in free ()
#4  0x000000010000f6dc in av_freep (arg=0x101111628) at mem.c:185
#5  0x00007fff855ba7e1 in start ()

Here is the updated full code if anyone feels like having a look:

#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <math.h>

#include <libavcodec/avcodec.h>
#include <libavformat/avformat.h>
#include <libswscale/swscale.h>
#include <libavutil/avutil.h>

#undef exit

#define STREAM_DURATION   5.0 /* 5 seconds stream duration */
#define STREAM_FRAME_RATE 25 /* 25 images/s */
#define STREAM_NB_FRAMES  ((int)(STREAM_DURATION * STREAM_FRAME_RATE))
#define STREAM_PIX_FMT PIX_FMT_YUV420P /* default pix_fmt */

static int sws_flags = SWS_BICUBIC;

int main(int argc, char **argv)
{
    // Output file
    const char *outputfile;

    // Struct to hold the output format details, audio codec, video codec,
extension
    // http://ffmpeg.org/doxygen/trunk/structAVOutputFormat.html
    AVOutputFormat *avOutputFormat; // Was *fmt

    // http://ffmpeg.org/doxygen/trunk/structAVFormatContext.html
    AVFormatContext *avFormatContext; // was *oc

    // Individual AVStreams for audio and video
    // http://ffmpeg.org/doxygen/trunk/structAVStream.html
    AVStream *video_avstream; // was *video_st and *st

// Audio and Video Timestamps (presentation time stamp)
    double audio_pts, video_pts;

///////////////////

int codec_id = AV_CODEC_ID_H264;
    AVCodec *codec;
    AVCodecContext *c= NULL;
    int i, ret, x, y, got_output;
    int write_ret;
    //FILE *f;
    AVFrame *frame;
    AVPacket pkt;
    //uint8_t endcode[] = { 0, 0, 1, 0xb7 };

    printf("Encode video file\n");

    /* register all the codecs */
    //avcodec_register_all();

///////
// Intialize libavcodec - register all codecs and formats??????????
    av_register_all();

// Set inpput filename and output filename
    outputfile = "outtestmov.mp4";

    // auto detect the output format from the name. default is mpeg.
    avOutputFormat = av_guess_format(NULL, outputfile, NULL);
    if (!avOutputFormat) {
        printf("Could not deduce output format from file extension: using
MPEG.\n");
        avOutputFormat = av_guess_format("mpeg", NULL, NULL);
    }
    if (!avOutputFormat) {
        fprintf(stderr, "Could not find suitable output format\n");
        exit(1);
    }

    fprintf(stdout,"Format Name: %s\n",avOutputFormat->name);
    fprintf(stdout,"Format Video Codec: %d\n",avOutputFormat->video_codec);
    avOutputFormat->video_codec = codec_id;
    fprintf(stdout,"Format Video Codec: %d\n",avOutputFormat->video_codec);

    /// allocate the output media context, avFormatContext
    avFormatContext = avformat_alloc_context();
    if (!avFormatContext) {
        fprintf(stderr, "Memory error\n");
        exit(1);
    }
    // Set the output format of the newly allocated avFormatContext to our
avOutputFormat
    avFormatContext->oformat = avOutputFormat;

    // Set the output filename to the outputfile
    snprintf(avFormatContext->filename, sizeof(avFormatContext->filename),
"%s", outputfile);

    /* find the mpeg1 video encoder */
    //codec = avcodec_find_encoder(codec_id);
    codec = avcodec_find_encoder(avOutputFormat->video_codec);
    if (!codec) {
        fprintf(stderr, "Codec not found\n");
        exit(1);
    }

    // add the video stream using the default format codec and initialize
the codecs
    video_avstream = avformat_new_stream(avFormatContext, codec);
    if (!video_avstream) {
        fprintf(stderr, "Could not alloc stream\n");
        exit(1);
    }

    if (video_avstream->codec == NULL) {
    fprintf(stderr, "AVStream codec is NULL\n");
    exit(1);
    }

    //c = avcodec_alloc_context3(codec);
    c = video_avstream->codec;
    //c->codec_id = avOutputFormat->video_codec;
    if (!c) {
        fprintf(stderr, "Could not allocate video codec context\n");
        exit(1);
    }

    /* put sample parameters */
    c->bit_rate = 400000;
    /* resolution must be a multiple of two */
    c->width = 352;
    c->height = 288;
    /* 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 = AV_PIX_FMT_YUV420P;

    if(codec_id == AV_CODEC_ID_H264)
        av_opt_set(c->priv_data, "preset", "slow", 0);

    /* open it */
    if (avcodec_open2(c, codec, NULL) < 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);
    }
*/
 //video_avstream->codec = c;

//////////////////////////////////
    frame = avcodec_alloc_frame();
    if (!frame) {
        fprintf(stderr, "Could not allocate video frame\n");
        exit(1);
    }
    frame->format = c->pix_fmt;
    frame->width  = c->width;
    frame->height = c->height;

    /* the image can be allocated by any means and av_image_alloc() is
     * just the most convenient way if av_malloc() is to be used */
    ret = av_image_alloc(frame->data, frame->linesize, c->width, c->height,
                         c->pix_fmt, 32);
    if (ret < 0) {
        fprintf(stderr, "Could not allocate raw picture buffer\n");
        exit(1);
    }

    /* open the output file, if needed */
    if (!(avOutputFormat->flags & AVFMT_NOFILE)) {
        if (avio_open(&avFormatContext->pb, outputfile, AVIO_FLAG_WRITE) <
0) {
            fprintf(stderr, "Could not open '%s'\n", outputfile);
            return 1;
        }
    }

    // some formats want stream headers to be separate
    if(avFormatContext->oformat->flags & AVFMT_GLOBALHEADER)
        c->flags |= CODEC_FLAG_GLOBAL_HEADER;

avformat_write_header(avFormatContext, NULL);

    /* encode 1 second of video */
    for(i=0;i<25;i++) {
        av_init_packet(&pkt);
        pkt.data = NULL;    // packet data will be allocated by the encoder
        pkt.size = 0;

        fflush(stdout);
        /* prepare a dummy image */
        /* Y */
        for(y=0;y<c->height;y++) {
            for(x=0;x<c->width;x++) {
                frame->data[0][y * frame->linesize[0] + x] = x + y + i * 3;
            }
        }

        /* Cb and Cr */
        for(y=0;y<c->height/2;y++) {
            for(x=0;x<c->width/2;x++) {
                frame->data[1][y * frame->linesize[1] + x] = 128 + y + i *
2;
                frame->data[2][y * frame->linesize[2] + x] = 64 + x + i * 5;
            }
        }

frame->pts = i;

        /* encode the image */
        ret = avcodec_encode_video2(c, &pkt, frame, &got_output);
        if (ret < 0) {
            fprintf(stderr, "Error encoding frame\n");
            exit(1);
        }

        if (got_output) {

        /*
            if (c->coded_frame->pts != AV_NOPTS_VALUE) {
                pkt.pts = av_rescale_q(c->coded_frame->pts, c->time_base,
video_avstream->time_base);
            } else {
            printf("pkt.pts AV_NOPTS_VALUE");
            }

            pkt.pts = i;


        printf("pkt.pts:%5d\n",pkt.pts);

            printf("Write frame %3d (size=%5d)\n", i, pkt.size);
            */

            /* write the compressed frame in the media file */
            write_ret = av_write_frame(avFormatContext, &pkt);
            if (write_ret < 0) {
            fprintf(stderr, "Error writing frame\n");
            exit(1);
            }

//            fwrite(pkt.data, 1, pkt.size, f);
            av_free_packet(&pkt);
        } else {
        printf("got_output false: %d", i);
        }
    }

    /* get the delayed frames */
/*
    for (got_output = 1; got_output; i++) {
        fflush(stdout);

ret = avcodec_encode_video2(c, &pkt, NULL, &got_output);
        if (ret < 0) {
            fprintf(stderr, "Error encoding frame\n");
            exit(1);
        }

        if (got_output) {
            printf("Write frame %3d (size=%5d)\n", i, pkt.size);
//            fwrite(pkt.data, 1, pkt.size, f);

            // write the compressed frame in the media file
// ret = av_interleaved_write_frame(avFormatContext, &pkt);

            av_free_packet(&pkt);
        }
    }
*/

    /* add sequence end code to have a real mpeg file */
    //fwrite(endcode, 1, sizeof(endcode), f);
    //fclose(f);

    /* write the trailer, if any.  the trailer must be written
     * before you close the CodecContexts open when you wrote the
     * header; otherwise write_trailer may try to use memory that
     * was freed on av_codec_close() */
    //av_write_trailer(avFormatContext);

    avcodec_close(c);
    av_free(c);
    av_freep(&frame->data[0]);
    avcodec_free_frame(&frame);


    /* free the streams */
    for(i = 0; i < avFormatContext->nb_streams; i++) {
        av_freep(&avFormatContext->streams[i]->codec);
        av_freep(&avFormatContext->streams[i]);
    }

    if (!(avOutputFormat->flags & AVFMT_NOFILE)) {
        /* close the output file */
        avio_close(avFormatContext->pb);
    }

    /* free the stream */
    av_free(avFormatContext);

    printf("\n");
    return 0;
}
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121229/960d455e/attachment.html>

From chandrashekar01 at gmail.com  Mon Dec 31 08:02:21 2012
From: chandrashekar01 at gmail.com (chandrashekar reddy)
Date: Mon, 31 Dec 2012 12:32:21 +0530
Subject: [Libav-user] Regarding integrating Media Parcer of ffmepg:
Message-ID: <CAEHR9pZJ1=V8__0GTa0Y2VccFfV+XCZ0iLEHt2TLk6HUnOR9ZA@mail.gmail.com>

Hi All,

I would like to use ffmpeg media parcers like ( AVI, Asf,MP4 etc..) in my
application.Can u sujject me what is the procedure I should fallow to
integarte ffmpeg media parcers in application.



Regards,
Chan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20121231/90ac550c/attachment.html>

From funkyfactory at gmx.at  Sat Dec 29 12:54:36 2012
From: funkyfactory at gmx.at (Funky Factory Development)
Date: Sat, 29 Dec 2012 12:54:36 +0100
Subject: [Libav-user] AAC decoding problems
Message-ID: <50DED9FC.8070208@gmx.at>

Hi,

I'm having some issues with AAC decoding with recent version of ffmpeg:

I'm using this code to decode AAC Frames (from a stream).
I had some troubles with the decoding - everything runs fine but the 
after a few seconds playing
the decoded buffer (m_avAACFrame->data) only contains zeros.

         int gotframe=0;
         if(!m_avAACFrame)
             m_avAACFrame=avcodec_alloc_frame();
         else
             avcodec_get_frame_defaults(m_avAACFrame);
         int buff=avcodec_decode_audio4(pAACCodecCtx, m_avAACFrame, 
&gotframe, &avAACpkt);
         if (gotframe==0 || buff<0)
         {
             *outputsize=0;
             return;
         }
         *outputsize=av_samples_get_buffer_size(NULL, 2, 
m_avAACFrame->nb_samples, pAACCodecCtx->sample_fmt, 1);
memcpy(outbuffer,*(byte**)&m_avAACFrame->data,*outputsize);

So i switched to the latest version of ffmpeg (Zeranoe FFmpeg for 
Windows) and now i have the problem that the output sample format is 
always AV_SAMPLE_FMT_FLT instead of AV_SAMPLE_FMT_S16 ?

     pAACCodecCtx->bit_rate = 64000;
     pAACCodecCtx->sample_rate = 44100;
     pAACCodecCtx->channels = 2;
     pAACCodecCtx->channel_layout = 3;
     pAACCodecCtx->sample_fmt = AV_SAMPLE_FMT_S16;

     //pAACCodecCtx->bit_rate = 64000;
     //pAACCodecCtx->sample_rate = fmtp[11];
     //pAACCodecCtx->channels = fmtp[7];
     //c->sample_fmt = AV_SAMPLE_FMT_DBL;
     //c->profile = FF_PROFILE_AAC_SSR;

     /* open it */

     if (avcodec_open2(pAACCodecCtx, AACCodec,NULL) < 0)
     {
         return false;
     }

Is AAC not properly supported on ffmpeg ?

Thanks for your help