[FFmpeg-devel] [PATCH] avcodec/utvideodec: decode to GBR(A)P

Paul B Mahol onemda at gmail.com
Mon Jun 26 14:12:25 EEST 2017


On 6/26/17, wm4 <nfxjfg at googlemail.com> wrote:
> On Mon, 26 Jun 2017 12:46:29 +0200
> Paul B Mahol <onemda at gmail.com> wrote:
>
>> On 6/26/17, Clement Boesch <u at pkh.me> wrote:
>> > On Mon, Jun 26, 2017 at 11:33:31AM +0200, Paul B Mahol wrote:
>> >> This is actually internal utvideo format.
>> >> Allows to make use of SIMD for median prediction for rgb(a) formats,
>> >> thus speeding up decoding.
>> >> Simplifies code, eases further developement and maintenance.
>> >>
>> >> Signed-off-by: Paul B Mahol <onemda at gmail.com>
>> >> ---
>> >>  libavcodec/utvideodec.c                   | 283
>> >> +++---------------------------
>> >>  tests/ref/fate/utvideo_rgb_left           |   8 +-
>> >>  tests/ref/fate/utvideo_rgb_median         |  10 +-
>> >>  tests/ref/fate/utvideo_rgba_left          |  10 +-
>> >>  tests/ref/fate/utvideo_rgba_median        |  10 +-
>> >>  tests/ref/fate/utvideo_rgba_single_symbol |   2 +-
>> >>  6 files changed, 49 insertions(+), 274 deletions(-)
>> >>
>> >
>> > I think you can drop ff_ut_rgb_order
>>
>> Will do.
>>
>> >
>> > Can you also add an explanation about the FATE changes?
>>
>> FATE changes because it now decodes to GBR(A)P instead of packed formats.
>
> (I assume that you verified that the output is exactly the same when
> the output is reordered to the same pixfmt by swscale.)

Yes it is.


More information about the ffmpeg-devel mailing list