[FFmpeg-cvslog] swscale: More accurate rounding in YSCALE_YUV_2_PACKEDX_FULL_C()

Michael Niedermayer git at videolan.org
Wed Jun 1 05:15:55 CEST 2011


ffmpeg | branch: master | Michael Niedermayer <michaelni at gmx.at> | Wed Jun  1 04:15:13 2011 +0200| [201549d1a95e5ea81a97257368668cda0afcb2f8] | committer: Michael Niedermayer

swscale: More accurate rounding in YSCALE_YUV_2_PACKEDX_FULL_C()

Signed-off-by: Michael Niedermayer <michaelni at gmx.at>

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=201549d1a95e5ea81a97257368668cda0afcb2f8
---

 libswscale/swscale.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/libswscale/swscale.c b/libswscale/swscale.c
index e7e81d3..6bf7fcf 100644
--- a/libswscale/swscale.c
+++ b/libswscale/swscale.c
@@ -557,9 +557,9 @@ static inline void yuv2nv12XinC(const int16_t *lumFilter, const int16_t **lumSrc
 #define YSCALE_YUV_2_PACKEDX_FULL_C(rnd,alpha) \
     for (i=0; i<dstW; i++) {\
         int j;\
-        int Y = 0;\
-        int U = -128<<19;\
-        int V = -128<<19;\
+        int Y = 1<<9;\
+        int U = (1<<9)-(128<<19);\
+        int V = (1<<9)-(128<<19);\
         int av_unused A;\
         int R,G,B;\
         \



More information about the ffmpeg-cvslog mailing list