[MPlayer-dev-eng] [PATCH] Add support for 9-bit/10-bit H.264

[Reimar Döffinger]

Should be ok, though I hadn't time to look at it much.
Though I'd consider it better to add all of 444, 422 and 420 formats at once for consistency.

On 8 May 2011, at 22:22, Arne Bochem <arneb.mp at ccan.de> wrote:

> Since ffv210 was changed to output 422P10 too, I have updated its
> codecs.conf entry accordingly.
> 
>> Could you please provide some samples to test?
> Of course, here are some samples encoded with recent x264:
> http://ps-auxw.de/10bit-h264-sample/
> 
> For convenience's sake, v210:
> http://samples.multimedia.cx/V-codecs/v210.mov
> 
>> The idea was to just count up, that's why it's 0x51 for the 16 bit formats,
>> you should just continue with 0x52, 0x53...
> I see. Changed.
> 
>> With IMGFMT_IS_YUVP16_* covering these formats vo_gl will incorrectly
>> claim to support them but it might still be best.
> Since IMGFMT_IS_YUVP16_* doesn't seem to be used elsewhere, not having
> it apply to P10/P9 probably wouldn't hurt much, though.
> 
>> You should be able to use
>> #define IMGFMT_IS_YUVP16_LE(fmt) (((fmt - 0x51000034) & 0xfc0000ff) == 0)
>> #define IMGFMT_IS_YUVP16_BE(fmt) (((fmt - 0x34000051) & 0xff0000fc) == 0)
> Switched to this for now.
> 
>> It might make sense to extend mp_get_chroma_shift to also return something
>> like y_bits or so.
> Where y_bits would be 9/10 for 420P9/10? I suppose I could add something
> like that, but I'm not sure where and how the result should be used.
> 
> 
> Regards,
> Arne Bochem
> <MPlayer-33447-colorspaces.diff>
> _______________________________________________
> MPlayer-dev-eng mailing list
> MPlayer-dev-eng at mplayerhq.hu
> https://lists.mplayerhq.hu/mailman/listinfo/mplayer-dev-eng