mbox series

[FFmpeg-devel,0/2] libswcale/input: fix incorrect rgbf32 yuv conversions

Message ID 20200912090714.66582-1-mindmark@gmail.com
Headers show
Series libswcale/input: fix incorrect rgbf32 yuv conversions | expand

Message

Mark Reid Sept. 12, 2020, 9:07 a.m. UTC
From: Mark Reid <mindmark@gmail.com>

Hi,
I'm not sure how I didn't notice this before but the f32 to yuv conversion
is slightly off. the folowinng 2 patches fix it.

In doing this I found that one of the exr tests probably wasn't testing the right 
layer, so I fixed that before.

Mark Reid (2):
  fate: use correct uint32 layer
  libswcale/input: fix incorrect rgbf32 yuv conversions

 libswscale/input.c                                   | 12 +++++-------
 tests/fate/image.mak                                 |  2 +-
 .../ref/fate/exr-rgb-scanline-pxr24-half-uint32-13x9 |  2 +-
 tests/ref/fate/filter-pixfmts-scale                  |  8 ++++----
 4 files changed, 11 insertions(+), 13 deletions(-)