[FFmpeg-devel] [PATCH] libswscale/swscale_unscaled: fix DITHER_COPY macro

Mateusz mateuszb at poczta.onet.pl
Mon Sep 4 10:33:34 EEST 2017


If ffmpeg reduces bit-depth to 8-bit or more, it uses DITHER_COPY macro.
The problem is DITHER_COPY macro make images darker (on all planes).

In x265 project there is issue #365 which is caused by this DITHER_COPY bug.
I think it is time to fix -- there are more and more high bit-depth sources.

Please review.

 libswscale/swscale_unscaled.c | 44 +++++++++++++------------------------------
 1 file changed, 13 insertions(+), 31 deletions(-)

diff --git a/libswscale/swscale_unscaled.c b/libswscale/swscale_unscaled.c
index ef36aec..3b7a5a9 100644
--- a/libswscale/swscale_unscaled.c
+++ b/libswscale/swscale_unscaled.c
@@ -110,24 +110,6 @@ DECLARE_ALIGNED(8, static const uint8_t, dithers)[8][8][8]={
   { 112, 16,104,  8,118, 22,110, 14,},
 }};
 
-static const uint16_t dither_scale[15][16]={
-{    2,    3,    3,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,},
-{    2,    3,    7,    7,   13,   13,   25,   25,   25,   25,   25,   25,   25,   25,   25,   25,},
-{    3,    3,    4,   15,   15,   29,   57,   57,   57,  113,  113,  113,  113,  113,  113,  113,},
-{    3,    4,    4,    5,   31,   31,   61,  121,  241,  241,  241,  241,  481,  481,  481,  481,},
-{    3,    4,    5,    5,    6,   63,   63,  125,  249,  497,  993,  993,  993,  993,  993, 1985,},
-{    3,    5,    6,    6,    6,    7,  127,  127,  253,  505, 1009, 2017, 4033, 4033, 4033, 4033,},
-{    3,    5,    6,    7,    7,    7,    8,  255,  255,  509, 1017, 2033, 4065, 8129,16257,16257,},
-{    3,    5,    6,    8,    8,    8,    8,    9,  511,  511, 1021, 2041, 4081, 8161,16321,32641,},
-{    3,    5,    7,    8,    9,    9,    9,    9,   10, 1023, 1023, 2045, 4089, 8177,16353,32705,},
-{    3,    5,    7,    8,   10,   10,   10,   10,   10,   11, 2047, 2047, 4093, 8185,16369,32737,},
-{    3,    5,    7,    8,   10,   11,   11,   11,   11,   11,   12, 4095, 4095, 8189,16377,32753,},
-{    3,    5,    7,    9,   10,   12,   12,   12,   12,   12,   12,   13, 8191, 8191,16381,32761,},
-{    3,    5,    7,    9,   10,   12,   13,   13,   13,   13,   13,   13,   14,16383,16383,32765,},
-{    3,    5,    7,    9,   10,   12,   14,   14,   14,   14,   14,   14,   14,   15,32767,32767,},
-{    3,    5,    7,    9,   11,   12,   14,   15,   15,   15,   15,   15,   15,   15,   16,65535,},
-};
-
 
 static void fillPlane(uint8_t *plane, int stride, int width, int height, int y,
                       uint8_t val)
@@ -1502,22 +1484,22 @@ static int packedCopyWrapper(SwsContext *c, const uint8_t *src[],
 }
 
 #define DITHER_COPY(dst, dstStride, src, srcStride, bswap, dbswap)\
-    uint16_t scale= dither_scale[dst_depth-1][src_depth-1];\
-    int shift= src_depth-dst_depth + dither_scale[src_depth-2][dst_depth-1];\
+    unsigned shift= src_depth-dst_depth, tmp;\
     for (i = 0; i < height; i++) {\
-        const uint8_t *dither= dithers[src_depth-9][i&7];\
+        const uint8_t *dither= dithers[shift-1][i&7];\
         for (j = 0; j < length-7; j+=8){\
-            dst[j+0] = dbswap((bswap(src[j+0]) + dither[0])*scale>>shift);\
-            dst[j+1] = dbswap((bswap(src[j+1]) + dither[1])*scale>>shift);\
-            dst[j+2] = dbswap((bswap(src[j+2]) + dither[2])*scale>>shift);\
-            dst[j+3] = dbswap((bswap(src[j+3]) + dither[3])*scale>>shift);\
-            dst[j+4] = dbswap((bswap(src[j+4]) + dither[4])*scale>>shift);\
-            dst[j+5] = dbswap((bswap(src[j+5]) + dither[5])*scale>>shift);\
-            dst[j+6] = dbswap((bswap(src[j+6]) + dither[6])*scale>>shift);\
-            dst[j+7] = dbswap((bswap(src[j+7]) + dither[7])*scale>>shift);\
+            tmp = (bswap(src[j+0]) + dither[0])>>shift; dst[j+0] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+1]) + dither[1])>>shift; dst[j+1] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+2]) + dither[2])>>shift; dst[j+2] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+3]) + dither[3])>>shift; dst[j+3] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+4]) + dither[4])>>shift; dst[j+4] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+5]) + dither[5])>>shift; dst[j+5] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+6]) + dither[6])>>shift; dst[j+6] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+7]) + dither[7])>>shift; dst[j+7] = dbswap(tmp - (tmp>>dst_depth));\
+        }\
+        for (; j < length; j++){\
+            tmp = (bswap(src[j]) + dither[j&7])>>shift; dst[j] = dbswap(tmp - (tmp>>dst_depth));\
         }\
-        for (; j < length; j++)\
-            dst[j] = dbswap((bswap(src[j]) + dither[j&7])*scale>>shift);\
         dst += dstStride;\
         src += srcStride;\
     }

-------------- next part --------------
 libswscale/swscale_unscaled.c | 44 +++++++++++++------------------------------
 1 file changed, 13 insertions(+), 31 deletions(-)

diff --git a/libswscale/swscale_unscaled.c b/libswscale/swscale_unscaled.c
index ef36aec..3b7a5a9 100644
--- a/libswscale/swscale_unscaled.c
+++ b/libswscale/swscale_unscaled.c
@@ -110,24 +110,6 @@ DECLARE_ALIGNED(8, static const uint8_t, dithers)[8][8][8]={
   { 112, 16,104,  8,118, 22,110, 14,},
 }};
 
-static const uint16_t dither_scale[15][16]={
-{    2,    3,    3,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,    5,},
-{    2,    3,    7,    7,   13,   13,   25,   25,   25,   25,   25,   25,   25,   25,   25,   25,},
-{    3,    3,    4,   15,   15,   29,   57,   57,   57,  113,  113,  113,  113,  113,  113,  113,},
-{    3,    4,    4,    5,   31,   31,   61,  121,  241,  241,  241,  241,  481,  481,  481,  481,},
-{    3,    4,    5,    5,    6,   63,   63,  125,  249,  497,  993,  993,  993,  993,  993, 1985,},
-{    3,    5,    6,    6,    6,    7,  127,  127,  253,  505, 1009, 2017, 4033, 4033, 4033, 4033,},
-{    3,    5,    6,    7,    7,    7,    8,  255,  255,  509, 1017, 2033, 4065, 8129,16257,16257,},
-{    3,    5,    6,    8,    8,    8,    8,    9,  511,  511, 1021, 2041, 4081, 8161,16321,32641,},
-{    3,    5,    7,    8,    9,    9,    9,    9,   10, 1023, 1023, 2045, 4089, 8177,16353,32705,},
-{    3,    5,    7,    8,   10,   10,   10,   10,   10,   11, 2047, 2047, 4093, 8185,16369,32737,},
-{    3,    5,    7,    8,   10,   11,   11,   11,   11,   11,   12, 4095, 4095, 8189,16377,32753,},
-{    3,    5,    7,    9,   10,   12,   12,   12,   12,   12,   12,   13, 8191, 8191,16381,32761,},
-{    3,    5,    7,    9,   10,   12,   13,   13,   13,   13,   13,   13,   14,16383,16383,32765,},
-{    3,    5,    7,    9,   10,   12,   14,   14,   14,   14,   14,   14,   14,   15,32767,32767,},
-{    3,    5,    7,    9,   11,   12,   14,   15,   15,   15,   15,   15,   15,   15,   16,65535,},
-};
-
 
 static void fillPlane(uint8_t *plane, int stride, int width, int height, int y,
                       uint8_t val)
@@ -1502,22 +1484,22 @@ static int packedCopyWrapper(SwsContext *c, const uint8_t *src[],
 }
 
 #define DITHER_COPY(dst, dstStride, src, srcStride, bswap, dbswap)\
-    uint16_t scale= dither_scale[dst_depth-1][src_depth-1];\
-    int shift= src_depth-dst_depth + dither_scale[src_depth-2][dst_depth-1];\
+    unsigned shift= src_depth-dst_depth, tmp;\
     for (i = 0; i < height; i++) {\
-        const uint8_t *dither= dithers[src_depth-9][i&7];\
+        const uint8_t *dither= dithers[shift-1][i&7];\
         for (j = 0; j < length-7; j+=8){\
-            dst[j+0] = dbswap((bswap(src[j+0]) + dither[0])*scale>>shift);\
-            dst[j+1] = dbswap((bswap(src[j+1]) + dither[1])*scale>>shift);\
-            dst[j+2] = dbswap((bswap(src[j+2]) + dither[2])*scale>>shift);\
-            dst[j+3] = dbswap((bswap(src[j+3]) + dither[3])*scale>>shift);\
-            dst[j+4] = dbswap((bswap(src[j+4]) + dither[4])*scale>>shift);\
-            dst[j+5] = dbswap((bswap(src[j+5]) + dither[5])*scale>>shift);\
-            dst[j+6] = dbswap((bswap(src[j+6]) + dither[6])*scale>>shift);\
-            dst[j+7] = dbswap((bswap(src[j+7]) + dither[7])*scale>>shift);\
+            tmp = (bswap(src[j+0]) + dither[0])>>shift; dst[j+0] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+1]) + dither[1])>>shift; dst[j+1] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+2]) + dither[2])>>shift; dst[j+2] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+3]) + dither[3])>>shift; dst[j+3] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+4]) + dither[4])>>shift; dst[j+4] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+5]) + dither[5])>>shift; dst[j+5] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+6]) + dither[6])>>shift; dst[j+6] = dbswap(tmp - (tmp>>dst_depth));\
+            tmp = (bswap(src[j+7]) + dither[7])>>shift; dst[j+7] = dbswap(tmp - (tmp>>dst_depth));\
+        }\
+        for (; j < length; j++){\
+            tmp = (bswap(src[j]) + dither[j&7])>>shift; dst[j] = dbswap(tmp - (tmp>>dst_depth));\
         }\
-        for (; j < length; j++)\
-            dst[j] = dbswap((bswap(src[j]) + dither[j&7])*scale>>shift);\
         dst += dstStride;\
         src += srcStride;\
     }


More information about the ffmpeg-devel mailing list