[FFmpeg-devel] [PATCH]012v decoder, v2

Paul B Mahol onemda at gmail.com
Mon Jan 7 10:45:42 CET 2013


On 1/7/13, Carl Eugen Hoyos <cehoyos at ag.or.at> wrote:
> Paul B Mahol <onemda <at> gmail.com> writes:
>
>> > Slightly updated patch applied.
>>
>> With 16 bit pix fmt.
>
> Yes, I changed ffmpeg -i output since posting the patch.
> There should be no more dis-ambiguity.
>
>> Why we have non 16bit formats at all?
>
> My memory absolutely may be wrong but as far as I
> remember the reason is that x264 decided to
> define its input formats as "lsb-bound" (if that
> is the correct wording). Since x264 was the first
> user of >8 and <16 bit, we mostly had to follow.

Then why not stop using all of them then?

Currently this code is only one doing more shifting
then necessary.


More information about the ffmpeg-devel mailing list