[FFmpeg-devel] [PATCH 2/6] swscale: add swscale support for 422/444 decode outputs
Fu, Linjie
linjie.fu at intel.com
Wed Sep 11 08:35:39 EEST 2019
> -----Original Message-----
> From: ffmpeg-devel [mailto:ffmpeg-devel-bounces at ffmpeg.org] On Behalf
> Of Carl Eugen Hoyos
> Sent: Wednesday, September 11, 2019 03:32
> To: FFmpeg development discussions and patches <ffmpeg-
> devel at ffmpeg.org>
> Subject: Re: [FFmpeg-devel] [PATCH 2/6] swscale: add swscale support for
> 422/444 decode outputs
>
> Am Di., 10. Sept. 2019 um 18:09 Uhr schrieb Linjie Fu <linjie.fu at intel.com>:
>
> > --- a/libswscale/output.c
> > +++ b/libswscale/output.c
> > @@ -2403,6 +2403,53 @@ yuv2ya8_X_c(SwsContext *c, const int16_t
> *lumFilter,
> > }
> >
> > static void
> > +yuv2ayuv_X_c(SwsContext *c, const int16_t *lumFilter,
> > + const int16_t **lumSrc, int lumFilterSize,
> > + const int16_t *chrFilter, const int16_t **chrUSrc,
> > + const int16_t **chrVSrc, int chrFilterSize,
> > + const int16_t **alpSrc, uint8_t *dest, int dstW, int y)
>
> The commit message does not match this change.
> Do I understand correctly that encoding is expected from AYUV
> but not from the other formats?
> Or is this the result of an earlier comment?
+ [AV_PIX_FMT_Y210BE] = { 1, 0 },
+ [AV_PIX_FMT_Y210LE] = { 1, 0 },
+ [AV_PIX_FMT_AYUV] = { 1, 1 },
This patch aims at adding swscale input support for Y210/AYUV and
output/encoding support for AYUV only.
IMHO swscale input support for these packed pixel format is good
at current stage.(since there is no requirement for software
CSC to these packed pixel formats)
For Y410, it's kind of different:
https://docs.microsoft.com/en-us/windows/win32/medfound/10-bit-and-16-bit-yuv-video-formats#y410
Bits 0-9 contain the U sample, bits 10-19 contain the Y sample, bits 20-29
contain the V sample, and bits 30-31 contain the alpha value.
Some attempts have been made locally to support swscale input for Y410:
https://github.com/fulinjie/ffmpeg/commits/pr-y410
And I think the swscale input support for Y410 could be held.
> I don't know if splitting the patch would make it clearer.
Sure, will split it for each pixel format.
Regards,
Linjie
More information about the ffmpeg-devel
mailing list