[Ffmpeg-devel] [RFC] mp4 object types comments

[Michael Niedermayer]

Hi
On Sat, Mar 24, 2007 at 05:59:53PM +0000, M?ns Rullg?rd wrote:
> Michael Niedermayer <michaelni at gmx.at> writes:
> 
> > Hi
> >
> > On Sat, Mar 24, 2007 at 02:44:41PM +0000, M?ns Rullg?rd wrote:
> >> Baptiste Coudurier <baptiste.coudurier at smartjog.com> writes:
> >> 
> >> > Hi
> >> >
> >> > M?ns Rullg?rd wrote:
> >> >> Baptiste Coudurier <baptiste.coudurier at smartjog.com> writes:
> >> >> 
> >> >>> Hi
> >> >>>
> >> >>> Here are some modifications to mp4 object types Im not sure,
> >> >>> can someone confirm that 13818-3 can be decoded with MP2 id ?
> >> >> 
> >> >> 13818-3 extends 11172-3 (MPEG1 audio) with lower sampling rates and
> >> >> more channels.  An 11172-3 decoder will correctly decode basic stereo
> >> >> from a 13818-3 stream.  There are 3 defined layers corresponding to
> >> >> those in 11172-3.
> 
> [...]
> 
> >> >> This mixes addition of comments with actual changes.  The added
> >> >> comments are fine.  The changed codec IDs are not.  All MPEG audio
> >> >> should use CODEC_ID_MP3.  Please make it so, but add the other two
> >> >> comments separately.
> >> >
> >> > Yes. About codec_id, are you sure?
> >> 
> >> See the comment in avcodec.h, line 210.
> >> 
> >> > Different tags are needed if codec_id is MP2 or MP3, and when stream
> >> > copying codec_id is not changed/altered.
> >> 
> >> Both 11172-3 and 13818-3 define layers 1, 2 and 3.  The tag does not
> >> specify which layer is used, only which standard, so CODEC_ID_MP3 is
> >> the only safe choice.  As 13818-3 is an extension to 11172-3, it
> >> should really be assigned its own codec_id.  Until we have a decoder
> >> for these extensions, the best choice is to use CODEC_ID_MP3.
> >
> > our mp3 decoder should support these extensions, do you suggest that we
> > add a dummy codec_id for it? if so iam not opposed ...
> 
> I didn't know we supported the multichannel extensions.  Have a

oops, i didnt know there was multichannel support in 13818-3, i just
meant the low samplerate stuff ...


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

I know you won't believe me, but the highest form of Human Excellence is
to question oneself and others. -- Socrates
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20070324/44525147/attachment.pgp>