Quoting James Almer (2017-07-21 17:49:33)
> On 7/21/2017 9:19 AM, Anton Khirnov wrote:
> > Quoting Anton Khirnov (2017-07-21 13:59:41)
> >> CC: libav-sta...@libav.org
> >> ---
> >>  libavformat/caf.c | 1 +
> >>  1 file changed, 1 insertion(+)
> >>
> >> diff --git a/libavformat/caf.c b/libavformat/caf.c
> >> index cf128d5..c299cad 100644
> >> --- a/libavformat/caf.c
> >> +++ b/libavformat/caf.c
> >> @@ -49,6 +49,7 @@ const AVCodecTag ff_codec_caf_tags[] = {
> >>      { AV_CODEC_ID_QCELP,           MKBETAG('Q','c','l','p') },
> >>      { AV_CODEC_ID_QDM2,            MKBETAG('Q','D','M','2') },
> >>      { AV_CODEC_ID_QDM2,            MKBETAG('Q','D','M','C') },
> >> +    { AV_CODEC_ID_OPUS,            MKBETAG('o','p','u','s') },
> >>    /* currently unsupported codecs */
> >>    /*{ AC-3 over S/PDIF          MKBETAG('c','a','c','3') },*/
> >>    /*{ MPEG4CELP                 MKBETAG('c','e','l','p') },*/
> >> -- 
> >> 2.0.0
> >>
> > 
> > Please disregard, seems I didn't test this properly back when I wrote
> > it.
> > Sorry for the noise.
> 
> Were you able to figure out the full contents of the kuki chunk for Opus
> files?
> I looked at it briefly on two files, one stereo and one mono. First
> value was 2048 on both (no idea its meaning), second was sample rate
> (Apparently fixed to 48000 and not original sample rate as in OpusHead),
> third was the same value as frames_per_packet from the desc chunk,
> fourth was -1000 on both files as well (also no idea), and fifth was
> channel count.
> 
> Other formats would just dump the contents from the relevant ISOM box in
> the kuki chunk, but for some reason Opus doesn't. None of the values
> looked like pre_skip to me, so i don't think we could manually create a
> meaningful OpusHead to fill the stream's extradata.
> 
> The user that created the two files i checked also said he couldn't
> create a multichannel one, and demuxing/decoding seemed to work fine
> when i left extradata empty (the decoder fills it with defaults if
> channels <= 2).

Sounds about right. I think the reasonable thing to do would be to just
leave the extradata empty until we see some multichannel files.

-- 
Anton Khirnov
_______________________________________________
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to