[FFmpeg-devel] [PATCH] avformat: Add simple ACLR atom reading to set the color range

[wm4]

On Tue, 17 Feb 2015 12:40:52 +0000
Kevin Wheatley <kevin.j.wheatley at gmail.com> wrote:

> On Tue, Feb 17, 2015 at 12:25 PM, Michael Niedermayer <michaelni at gmx.at> wrote:
> > if the codec id doesnt match the expected, mov_read_extradata will
> > return 0 even without any truncation.
> > In this case the error message would be incorrect
> 
> So should I code the test against codec id against the files I have or
> the ones that might be valid, but I don't have samples for?

I think this really should be done in the decoder instead.

> It looks like AV_CODEC_ID_AVUI and AV_CODEC_ID_DNXHD are tested for in
> other parts of the code, so I'd be comfortable with those. However I'm
> thinking the call to mov_read_avid might be skipped/inlined if I only
> tested against the DNxHD codec, there is something that begins to look
> like a code smell if I duplicate the list of codecs, but I'm happy to
> go with whatever is preferred.
>