[FFmpeg-devel] [PATCH] Video decoder and demuxer for AMV files
Wed Sep 26 16:48:53 CEST 2007
2007/9/26, Aurelien Jacobs <aurel at gnuage.org>:
> On Wed, 26 Sep 2007 17:40:21 +0700
> "Vladimir Voroshilov" <voroshil at gmail.com> wrote:
> > 2007/9/25, Vladimir Voroshilov <voroshil at gmail.com>:
> > [...]
> > > Fake fourcc code removed.
> > > I've also changed CODEC_ID_ADPCM_IMA_WS to CODEC_ID_ADPCM_IMA_AMV,
> > > thus Vitor's patch is required too (his codec provides better sound).
> > Small nit: AMV files are currently reported as "avi", but strictly
> > speaking they are broken AVI files.
> > Thus, i suggest to add separate AVInputFormat structure especially for
> > AMV files (with apropriate name and comment) and
> > register it with AVI's config variable (IOW, REGISTER_DEMUXER (AVI, amv))
> I don't think it's worth. I much prefer your previous patch.
> Note that you will have to update it (simplify it) according to
> my recent commit on avidec.c.
Attached file is previous patch, synced with latest svn.
I suppose it is latest version.
If separate format will be really required it can be easily
implemented sometime in the future.
> > > I haven't checked the surrounding code but if stream_index can
> > > only be 0 or 1, you can replace your whole switch() by:
> > >
> > > tag1 = stream_index ? MKTAG('a','u','d','s') : MKTAG('v','i','d','s');
> > I have no info about streams count, so kept as is
> I gave a deeper look at the source. My proposition is safe.
> tag1 will always be video for the first stream and audio for
> the second stream (and all subsequent stream if they happen
> to exist).
Ok. Got suggested line.
P.S. Can anybody commit all three AMV patches ? Michael?
Vladimir Voroshilov mailto:voroshil at gmail.com
JID: voroshil at gmail.com, voroshil at jabber.ru
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 3184 bytes
Desc: not available
More information about the ffmpeg-devel