[FFmpeg-devel] [PATCH] Read extradata Quicktime-Atom "FFVH" for ffvhuff files from libquicktime

Michael Niedermayer michaelni
Mon Oct 29 03:33:45 CET 2007


Hi

On Sun, Oct 28, 2007 at 10:14:26PM +0100, Baptiste Coudurier wrote:
> Baptiste Coudurier wrote:
> > Hi,
> > 
> > Alexander Strange wrote:
> > 
> >>On Oct 24, 2007, at 5:34 AM, Baptiste Coudurier wrote:
> >>
> >>>Concerning extradata, I think It is a good opportunity to use the new
> >>>design defined to store Xiph codecs:
> >>>
> >>>http://wiki.xiph.org/index.php/ 
> >>>Oggless#Example_and_Discussion_of_the_Mov_container
> >>>
> >>>"A single glbl atom shall be placed in the stsd atom in which the the
> >>>global header shall be stored."
> >>
> >>The traditional way to do it is to add an atom with the codec's  
> >>fourcc to the stsd atom.
> > 
> > 
> > 'avcC' for 'avc1', 'SMI ' for 'SVQ3', 'jp2h' for 'mjp2', 'esds' for
> > 'mp4a', 'QDM2' uses another extradata atom. Well Im not sure at all this
> > is the traditional way for mov. AVID uses his own scheme too.
> > 
> > 'alac' use this design, but that's the only one I know, MPEG-4/3GP are
> > using different names than fourcc for extradata atoms, or the generic
> > esds if 'mp4a' is used as fourcc.
> > 
> > 
> >>This is what Xiph's QT component actually does, no matter what it  
> >>says there.
> >>
> >>Also, Perian implicitly defines MOV formats for everything it can  
> >>play; for avi formats it just copies the video packets
> >>and puts the extradata at the end of stsd like that.
> >>
> >>(I uploaded ogg-audio.mov and huffyuv.mov to the mplayerhq incoming)
> > 
> > 
> > Like Loren pointed, it would be way more generic and simple to use a
> > single 'glbl' atom at the end of 'stsd' and I agree with him.
> > 
> > If we can agree to a common recommended practice I would be very happy
> > to modify mov muxer to support that generic design.
> > 
> 
> I'd like to add that generic mechanism to libavformat, Michael are you
> ok with it ? I need your blessing for ffvh.

yes, of course

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Complexity theory is the science of finding the exact solution to an
approximation. Benchmarking OTOH is finding an approximation of the exact
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20071029/1fc08060/attachment.pgp>



More information about the ffmpeg-devel mailing list