[FFmpeg-devel] [PATCH] Fix bug in voc demuxer that was introduced in SVN r26309.

[Justin Ruggles]

On 01/18/2011 07:50 AM, Carl Eugen Hoyos wrote:

> Ronald S. Bultje <rsbultje <at> gmail.com> writes:
> 
>> @@ -117,7 +117,7 @@ voc_get_packet(AVFormatContext *s, AVPacket *pkt,
>> AVStream *st, int max_size)
>>              dec->sample_rate = get_le32(pb);
>>              dec->bits_per_coded_sample = get_byte(pb);
>>              dec->channels = get_byte(pb);
>> -            tmp_codec = ff_codec_get_id(ff_voc_codec_tags, get_byte(pb));
>> +            tmp_codec = ff_codec_get_id(ff_voc_codec_tags, get_le16(pb));
>>
>> Does this have (a) fate test(s)?
> 
> No, since there is a native muxer, this should be a regression test.


That might be a good idea, especially considering our demuxer apparently
can't read a stereo s16 file created by our muxer.  I'll try to fix it,
then send a patch for adding a regression test for voc stereo s16.

-Justin