[FFmpeg-devel] [PATCH] matroskaenc: implement CueRelativePosition

Michael Niedermayer michaelni at gmx.at
Wed Jul 24 14:09:19 CEST 2013


On Wed, Jul 24, 2013 at 09:30:27AM +0200, wm4 wrote:
> On Tue, 23 Jul 2013 23:21:41 +0200
> Hendrik Leppkes <h.leppkes at gmail.com> wrote:
> 
> > On Tue, Jul 23, 2013 at 11:15 PM, Michael Niedermayer <michaelni at gmx.at> wrote:
> > > On Tue, Jul 23, 2013 at 02:08:51PM -0700, Bernie Habermeier wrote:
> > >> wm4 did some rudimentary testing, and I'm currently using it for my own player implementation, but that is still in its early stages.  It looks to be correct and working (ie: I use it to seek to the proper place in the file, and things don't blow up on me regarding file parsing after I do the seek).
> > >>
> > >> I'd be happy to test it, but not sure how many players today actually rely on this somewhat new feature.
> > >
> > > did you test your player with some reference file or output of some
> > > other muxer that sets these ?
> > > Iam trying to avoid here that new muxer code is tested by new demuxer
> > > code as that could both thne have a off by 1 error and it wouldnt be
> > > detected.
> > > or is there also no other muxer writing these ?
> > >
> > 
> > mkvmerge can write these, and in recent development versions even correctly.
> 
> Until a week ago mkvmerge wrote broken CueRelativePosition elements
> (pointing to the wrong EBML element: Block instead of BlockGroup), and
> nobody noticed until now. So I can't imagine there are many demuxers
> which use this element - if there are any at all.
> 
> Personally I did test some patches from Bernie Habermeier, and the last
> one he sent to me seemed to produce the same results as newer versions
> of mkvmerge (with the bug mentioned above fixed). (I'm not 100% sure
> though - the offset calculations involved with CueRelativePosition are
> a bit tricky...)

ok, thanks for testing

patch applied

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

You can kill me, but you cannot change the truth.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20130724/b8243a7a/attachment.asc>


More information about the ffmpeg-devel mailing list