[FFmpeg-devel] [PATCH] als in mp4

[Jai Menon]

On 4/17/09, Baptiste Coudurier <baptiste.coudurier at gmail.com> wrote:
> Hi Jai,
>
>
>  On 4/17/2009 9:52 AM, Jai Menon wrote:
>  > On 4/9/09, Jai Menon <jmenon86 at gmail.com> wrote:
>  >> On 4/9/09, Baptiste Coudurier <baptiste.coudurier at gmail.com> wrote:
>  >>  > Hi Jai,
>  >>  >
>  >>  >
>  >>  >  On 4/9/2009 12:06 AM, Jai Menon wrote:
>  >>  >  > On Wed, Apr 8, 2009 at 10:38 AM, Baptiste Coudurier
>  >>  >  > <baptiste.coudurier at gmail.com> wrote:
>  >>  >  >> Hi Jai,
>  >>  >  >>
>  >>  >  >> On 4/6/2009 5:12 AM, Jai Menon wrote:
>  >>  >  >>> On Fri, Apr 3, 2009 at 11:28 PM, Baptiste Coudurier
>  >
>  > [...]
>  >
>  >> So how should we do this in your opinion? There should be some way we
>  >>  can instruct the demuxer that chan_config is an absolute value and not
>  >>  an index. The patch does this using the absolute_channels flag. As for
>  >>  the bits_per_sample field, I will remove it in the next patch.
>
>
> Well, if specs says channels is in AOT specific config for ALS, then we
>  should be what spec says.
>
>  IMHO just add "channels" field to MPEG4AudioConfig and set it for ALS.
>  I think it could be set for others AOT too, it could avoid accessing
>  chan_config outside and if chan_config is wrong it would be set to -1.

Okay, then I'll add a channels field. This will be set to correct
value (from aot specific config) whenever chan_config == 0 (as spec
mandates).

>  I'd like to hear comments from others.
>
>
>  > Okay, I'm attaching the bare minimum patchset required. This should be
>  > pretty uncontroversial. I have no idea how to simplify this further.
>  > Decoder will be required for setting channels. I haven't attached the
>  > patch for muxing (which is another hack in itself). Is this fine?
>
>
> Ok for the part I maintain.

Michael, is the avcodec.h part okay to apply?

-- 
Regards,

Jai