You've already forked FFmpeg
mirror of
https://github.com/FFmpeg/FFmpeg.git
synced 2025-12-25 22:17:24 +02:00
c0f49db53d
This information is now pre-filled automatically for SWS_OP_READ when
relevant.
yuv444p10msbbe -> rgb24:
[u16 XXXX -> +++X] SWS_OP_READ : 3 elem(s) planar >> 0
[u16 ...X -> +++X] SWS_OP_SWAP_BYTES
[u16 ...X -> +++X] SWS_OP_RSHIFT : >> 6
[u16 ...X -> +++X] SWS_OP_CONVERT : u16 -> f32
[f32 ...X -> ...X] SWS_OP_LINEAR : matrix3+off3 [...]
[f32 ...X -> ...X] SWS_OP_DITHER : 16x16 matrix + {0 3 2 5}
[f32 ...X -> ...X] SWS_OP_MAX : {0 0 0 0} <= x
+ [f32 ...X -> ...X] SWS_OP_MIN : x <= {255 255 255 _}
[f32 ...X -> +++X] SWS_OP_CONVERT : f32 -> u8
[ u8 ...X -> +++X] SWS_OP_WRITE : 3 elem(s) packed >> 0
(X = unused, + = exact, 0 = zero)
(This clamp is needed and was incorrectly optimized away before, because the
`SWS_OP_RSHIFT` incorrectly distorted the value range assertion)