Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/FFmpeg/FFmpeg.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
path: root/libswscale/swscale.c
diff options
context:
space:
mode:
Diffstat (limited to 'libswscale/swscale.c')
-rw-r--r--libswscale/swscale.c354
1 files changed, 259 insertions, 95 deletions
diff --git a/libswscale/swscale.c b/libswscale/swscale.c
index dd9f4a108f..9d17868019 100644
--- a/libswscale/swscale.c
+++ b/libswscale/swscale.c
@@ -1,20 +1,20 @@
/*
* Copyright (C) 2001-2003 Michael Niedermayer <michaelni@gmx.at>
*
- * This file is part of Libav.
+ * This file is part of FFmpeg.
*
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
- * License along with Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
@@ -60,6 +60,7 @@ untested special converters
#include "swscale.h"
#include "swscale_internal.h"
#include "rgb2rgb.h"
+#include "libavutil/avassert.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/cpu.h"
#include "libavutil/avutil.h"
@@ -67,7 +68,6 @@ untested special converters
#include "libavutil/bswap.h"
#include "libavutil/pixdesc.h"
-#define DITHER1XBPP
#define RGB2YUV_SHIFT 15
#define BY ( (int)(0.114*219/255*(1<<RGB2YUV_SHIFT)+0.5))
@@ -195,6 +195,101 @@ DECLARE_ALIGNED(8, const uint8_t, dither_8x8_128)[8][8] = {
DECLARE_ALIGNED(8, const uint8_t, ff_sws_pb_64)[8] =
{ 64, 64, 64, 64, 64, 64, 64, 64 };
+DECLARE_ALIGNED(8, const uint8_t, dithers)[8][8][8]={
+{
+ { 0, 1, 0, 1, 0, 1, 0, 1,},
+ { 1, 0, 1, 0, 1, 0, 1, 0,},
+ { 0, 1, 0, 1, 0, 1, 0, 1,},
+ { 1, 0, 1, 0, 1, 0, 1, 0,},
+ { 0, 1, 0, 1, 0, 1, 0, 1,},
+ { 1, 0, 1, 0, 1, 0, 1, 0,},
+ { 0, 1, 0, 1, 0, 1, 0, 1,},
+ { 1, 0, 1, 0, 1, 0, 1, 0,},
+},{
+ { 1, 2, 1, 2, 1, 2, 1, 2,},
+ { 3, 0, 3, 0, 3, 0, 3, 0,},
+ { 1, 2, 1, 2, 1, 2, 1, 2,},
+ { 3, 0, 3, 0, 3, 0, 3, 0,},
+ { 1, 2, 1, 2, 1, 2, 1, 2,},
+ { 3, 0, 3, 0, 3, 0, 3, 0,},
+ { 1, 2, 1, 2, 1, 2, 1, 2,},
+ { 3, 0, 3, 0, 3, 0, 3, 0,},
+},{
+ { 2, 4, 3, 5, 2, 4, 3, 5,},
+ { 6, 0, 7, 1, 6, 0, 7, 1,},
+ { 3, 5, 2, 4, 3, 5, 2, 4,},
+ { 7, 1, 6, 0, 7, 1, 6, 0,},
+ { 2, 4, 3, 5, 2, 4, 3, 5,},
+ { 6, 0, 7, 1, 6, 0, 7, 1,},
+ { 3, 5, 2, 4, 3, 5, 2, 4,},
+ { 7, 1, 6, 0, 7, 1, 6, 0,},
+},{
+ { 4, 8, 7, 11, 4, 8, 7, 11,},
+ { 12, 0, 15, 3, 12, 0, 15, 3,},
+ { 6, 10, 5, 9, 6, 10, 5, 9,},
+ { 14, 2, 13, 1, 14, 2, 13, 1,},
+ { 4, 8, 7, 11, 4, 8, 7, 11,},
+ { 12, 0, 15, 3, 12, 0, 15, 3,},
+ { 6, 10, 5, 9, 6, 10, 5, 9,},
+ { 14, 2, 13, 1, 14, 2, 13, 1,},
+},{
+ { 9, 17, 15, 23, 8, 16, 14, 22,},
+ { 25, 1, 31, 7, 24, 0, 30, 6,},
+ { 13, 21, 11, 19, 12, 20, 10, 18,},
+ { 29, 5, 27, 3, 28, 4, 26, 2,},
+ { 8, 16, 14, 22, 9, 17, 15, 23,},
+ { 24, 0, 30, 6, 25, 1, 31, 7,},
+ { 12, 20, 10, 18, 13, 21, 11, 19,},
+ { 28, 4, 26, 2, 29, 5, 27, 3,},
+},{
+ { 18, 34, 30, 46, 17, 33, 29, 45,},
+ { 50, 2, 62, 14, 49, 1, 61, 13,},
+ { 26, 42, 22, 38, 25, 41, 21, 37,},
+ { 58, 10, 54, 6, 57, 9, 53, 5,},
+ { 16, 32, 28, 44, 19, 35, 31, 47,},
+ { 48, 0, 60, 12, 51, 3, 63, 15,},
+ { 24, 40, 20, 36, 27, 43, 23, 39,},
+ { 56, 8, 52, 4, 59, 11, 55, 7,},
+},{
+ { 18, 34, 30, 46, 17, 33, 29, 45,},
+ { 50, 2, 62, 14, 49, 1, 61, 13,},
+ { 26, 42, 22, 38, 25, 41, 21, 37,},
+ { 58, 10, 54, 6, 57, 9, 53, 5,},
+ { 16, 32, 28, 44, 19, 35, 31, 47,},
+ { 48, 0, 60, 12, 51, 3, 63, 15,},
+ { 24, 40, 20, 36, 27, 43, 23, 39,},
+ { 56, 8, 52, 4, 59, 11, 55, 7,},
+},{
+ { 36, 68, 60, 92, 34, 66, 58, 90,},
+ { 100, 4,124, 28, 98, 2,122, 26,},
+ { 52, 84, 44, 76, 50, 82, 42, 74,},
+ { 116, 20,108, 12,114, 18,106, 10,},
+ { 32, 64, 56, 88, 38, 70, 62, 94,},
+ { 96, 0,120, 24,102, 6,126, 30,},
+ { 48, 80, 40, 72, 54, 86, 46, 78,},
+ { 112, 16,104, 8,118, 22,110, 14,},
+}};
+
+static const uint8_t flat64[8]={64,64,64,64,64,64,64,64};
+
+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 av_always_inline void
yuv2yuvX16_c_template(const int16_t *lumFilter, const int32_t **lumSrc,
int lumFilterSize, const int16_t *chrFilter,
@@ -205,9 +300,10 @@ yuv2yuvX16_c_template(const int16_t *lumFilter, const int32_t **lumSrc,
{
//FIXME Optimize (just quickly written not optimized..)
int i;
+ int dword= output_bits == 16;
uint16_t *yDest = dest[0], *uDest = dest[1], *vDest = dest[2],
*aDest = CONFIG_SWSCALE_ALPHA ? dest[3] : NULL;
- int shift = 15 + 16 - output_bits - 1;
+ int shift = 11 + 4*dword + 16 - output_bits - 1;
#define output_pixel(pos, val) \
if (big_endian) { \
@@ -224,24 +320,24 @@ yuv2yuvX16_c_template(const int16_t *lumFilter, const int32_t **lumSrc,
} \
}
for (i = 0; i < dstW; i++) {
- int val = 1 << (30-output_bits - 1);
+ int val = 1 << (26-output_bits + 4*dword - 1);
int j;
for (j = 0; j < lumFilterSize; j++)
- val += (lumSrc[j][i] * lumFilter[j]) >> 1;
+ val += ((dword ? lumSrc[j][i] : ((int16_t**)lumSrc)[j][i]) * lumFilter[j])>>1;
output_pixel(&yDest[i], val);
}
if (uDest) {
for (i = 0; i < chrDstW; i++) {
- int u = 1 << (30-output_bits - 1);
- int v = 1 << (30-output_bits - 1);
+ int u = 1 << (26-output_bits + 4*dword - 1);
+ int v = 1 << (26-output_bits + 4*dword - 1);
int j;
for (j = 0; j < chrFilterSize; j++) {
- u += (chrUSrc[j][i] * chrFilter[j]) >> 1;
- v += (chrVSrc[j][i] * chrFilter[j]) >> 1;
+ u += ((dword ? chrUSrc[j][i] : ((int16_t**)chrUSrc)[j][i]) * chrFilter[j]) >> 1;
+ v += ((dword ? chrVSrc[j][i] : ((int16_t**)chrVSrc)[j][i]) * chrFilter[j]) >> 1;
}
output_pixel(&uDest[i], u);
@@ -251,11 +347,11 @@ yuv2yuvX16_c_template(const int16_t *lumFilter, const int32_t **lumSrc,
if (CONFIG_SWSCALE_ALPHA && aDest) {
for (i = 0; i < dstW; i++) {
- int val = 1 << (30-output_bits - 1);
+ int val = 1 << (26-output_bits + 4*dword - 1);
int j;
for (j = 0; j < lumFilterSize; j++)
- val += (alpSrc[j][i] * lumFilter[j]) >> 1;
+ val += ((dword ? alpSrc[j][i] : ((int16_t**)alpSrc)[j][i]) * lumFilter[j]) >> 1;
output_pixel(&aDest[i], val);
}
@@ -483,8 +579,8 @@ yuv2gray16_1_c_template(SwsContext *c, const int32_t *buf0,
int i;
for (i = 0; i < (dstW >> 1); i++) {
- int Y1 = buf0[i * 2 ] << 1;
- int Y2 = buf0[i * 2 + 1] << 1;
+ int Y1 = (buf0[i * 2 ]+4)>>3;
+ int Y2 = (buf0[i * 2 + 1]+4)>>3;
output_pixel(&dest[i * 2 + 0], Y1);
output_pixel(&dest[i * 2 + 1], Y2);
@@ -932,8 +1028,8 @@ yuv2rgb48_1_c_template(SwsContext *c, const int32_t *buf0,
for (i = 0; i < (dstW >> 1); i++) {
int Y1 = (buf0[i * 2] ) >> 2;
int Y2 = (buf0[i * 2 + 1]) >> 2;
- int U = (ubuf0[i] + ubuf1[i] + (-128 << 11)) >> 3;
- int V = (vbuf0[i] + vbuf1[i] + (-128 << 11)) >> 3;
+ int U = (ubuf0[i] + ubuf1[i] + (-128 << 12)) >> 3;
+ int V = (vbuf0[i] + vbuf1[i] + (-128 << 12)) >> 3;
int R, G, B;
Y1 -= c->yuv2rgb_y_offset;
@@ -1004,6 +1100,7 @@ yuv2rgb_write(uint8_t *_dest, int i, int Y1, int Y2,
#define r_b ((target == PIX_FMT_RGB24) ? r : b)
#define b_r ((target == PIX_FMT_RGB24) ? b : r)
+
dest[i * 6 + 0] = r_b[Y1];
dest[i * 6 + 1] = g[Y1];
dest[i * 6 + 2] = b_r[Y1];
@@ -1294,9 +1391,9 @@ yuv2rgb_full_X_c_template(SwsContext *c, const int16_t *lumFilter,
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;
@@ -1311,7 +1408,7 @@ yuv2rgb_full_X_c_template(SwsContext *c, const int16_t *lumFilter,
U >>= 10;
V >>= 10;
if (hasAlpha) {
- A = 1 << 21;
+ A = 1 << 18;
for (j = 0; j < lumFilterSize; j++) {
A += alpSrc[j][i] * lumFilter[j];
}
@@ -1354,7 +1451,6 @@ yuv2rgb_full_X_c_template(SwsContext *c, const int16_t *lumFilter,
dest[1] = B >> 22;
dest[2] = G >> 22;
dest[3] = R >> 22;
- dest += 4;
break;
case PIX_FMT_BGR24:
dest[0] = B >> 22;
@@ -1500,14 +1596,14 @@ rgb48funcs(bgr, BE, PIX_FMT_BGR48BE);
(isBE(origin) ? AV_RB16(&src[(i)*2]) : AV_RL16(&src[(i)*2])))
static av_always_inline void
-rgb16_32ToY_c_template(uint8_t *dst, const uint8_t *src,
+rgb16_32ToY_c_template(int16_t *dst, const uint8_t *src,
int width, enum PixelFormat origin,
int shr, int shg, int shb, int shp,
int maskr, int maskg, int maskb,
int rsh, int gsh, int bsh, int S)
{
const int ry = RY << rsh, gy = GY << gsh, by = BY << bsh,
- rnd = 33 << (S - 1);
+ rnd = (32<<((S)-1)) + (1<<(S-7));
int i;
for (i = 0; i < width; i++) {
@@ -1516,12 +1612,12 @@ rgb16_32ToY_c_template(uint8_t *dst, const uint8_t *src,
int g = (px & maskg) >> shg;
int r = (px & maskr) >> shr;
- dst[i] = (ry * r + gy * g + by * b + rnd) >> S;
+ dst[i] = (ry * r + gy * g + by * b + rnd) >> ((S)-6);
}
}
static av_always_inline void
-rgb16_32ToUV_c_template(uint8_t *dstU, uint8_t *dstV,
+rgb16_32ToUV_c_template(int16_t *dstU, int16_t *dstV,
const uint8_t *src, int width,
enum PixelFormat origin,
int shr, int shg, int shb, int shp,
@@ -1530,7 +1626,7 @@ rgb16_32ToUV_c_template(uint8_t *dstU, uint8_t *dstV,
{
const int ru = RU << rsh, gu = GU << gsh, bu = BU << bsh,
rv = RV << rsh, gv = GV << gsh, bv = BV << bsh,
- rnd = 257 << (S - 1);
+ rnd = (256<<((S)-1)) + (1<<(S-7));
int i;
for (i = 0; i < width; i++) {
@@ -1539,13 +1635,13 @@ rgb16_32ToUV_c_template(uint8_t *dstU, uint8_t *dstV,
int g = (px & maskg) >> shg;
int r = (px & maskr) >> shr;
- dstU[i] = (ru * r + gu * g + bu * b + rnd) >> S;
- dstV[i] = (rv * r + gv * g + bv * b + rnd) >> S;
+ dstU[i] = (ru * r + gu * g + bu * b + rnd) >> ((S)-6);
+ dstV[i] = (rv * r + gv * g + bv * b + rnd) >> ((S)-6);
}
}
static av_always_inline void
-rgb16_32ToUV_half_c_template(uint8_t *dstU, uint8_t *dstV,
+rgb16_32ToUV_half_c_template(int16_t *dstU, int16_t *dstV,
const uint8_t *src, int width,
enum PixelFormat origin,
int shr, int shg, int shb, int shp,
@@ -1554,7 +1650,7 @@ rgb16_32ToUV_half_c_template(uint8_t *dstU, uint8_t *dstV,
{
const int ru = RU << rsh, gu = GU << gsh, bu = BU << bsh,
rv = RV << rsh, gv = GV << gsh, bv = BV << bsh,
- rnd = 257 << S, maskgx = ~(maskr | maskb);
+ rnd = (256U<<(S)) + (1<<(S-6)), maskgx = ~(maskr | maskb);
int i;
maskr |= maskr << 1; maskb |= maskb << 1; maskg |= maskg << 1;
@@ -1573,8 +1669,8 @@ rgb16_32ToUV_half_c_template(uint8_t *dstU, uint8_t *dstV,
}
r = (rb & maskr) >> shr;
- dstU[i] = (ru * r + gu * g + bu * b + rnd) >> (S + 1);
- dstV[i] = (rv * r + gv * g + bv * b + rnd) >> (S + 1);
+ dstU[i] = (ru * r + gu * g + bu * b + (unsigned)rnd) >> ((S)-6+1);
+ dstV[i] = (rv * r + gv * g + bv * b + (unsigned)rnd) >> ((S)-6+1);
}
}
@@ -1618,65 +1714,83 @@ rgb16_32_wrapper(PIX_FMT_BGR555BE, bgr15be, 0, 0, 0, 0, 0x001F, 0x03E0, 0x7
rgb16_32_wrapper(PIX_FMT_RGB565BE, rgb16be, 0, 0, 0, 0, 0xF800, 0x07E0, 0x001F, 0, 5, 11, RGB2YUV_SHIFT+8);
rgb16_32_wrapper(PIX_FMT_RGB555BE, rgb15be, 0, 0, 0, 0, 0x7C00, 0x03E0, 0x001F, 0, 5, 10, RGB2YUV_SHIFT+7);
-static void abgrToA_c(uint8_t *dst, const uint8_t *src, int width, uint32_t *unused)
+static void abgrToA_c(int16_t *dst, const uint8_t *src, int width, uint32_t *unused)
+{
+ int i;
+ for (i=0; i<width; i++) {
+ dst[i]= src[4*i]<<6;
+ }
+}
+
+static void rgbaToA_c(int16_t *dst, const uint8_t *src, int width, uint32_t *unused)
{
int i;
for (i=0; i<width; i++) {
- dst[i]= src[4*i];
+ dst[i]= src[4*i+3]<<6;
}
}
-static void rgbaToA_c(uint8_t *dst, const uint8_t *src, int width, uint32_t *unused)
+static void palToA_c(int16_t *dst, const uint8_t *src, int width, uint32_t *pal)
{
int i;
for (i=0; i<width; i++) {
- dst[i]= src[4*i+3];
+ int d= src[i];
+
+ dst[i]= (pal[d] >> 24)<<6;
}
}
-static void palToY_c(uint8_t *dst, const uint8_t *src, int width, uint32_t *pal)
+static void palToY_c(int16_t *dst, const uint8_t *src, long width, uint32_t *pal)
{
int i;
for (i=0; i<width; i++) {
int d= src[i];
- dst[i]= pal[d] & 0xFF;
+ dst[i]= (pal[d] & 0xFF)<<6;
}
}
-static void palToUV_c(uint8_t *dstU, uint8_t *dstV,
- const uint8_t *src1, const uint8_t *src2,
- int width, uint32_t *pal)
+static void palToUV_c(uint16_t *dstU, int16_t *dstV,
+ const uint8_t *src1, const uint8_t *src2,
+ int width, uint32_t *pal)
{
int i;
assert(src1 == src2);
for (i=0; i<width; i++) {
int p= pal[src1[i]];
- dstU[i]= p>>8;
- dstV[i]= p>>16;
+ dstU[i]= (uint8_t)(p>> 8)<<6;
+ dstV[i]= (uint8_t)(p>>16)<<6;
}
}
-static void monowhite2Y_c(uint8_t *dst, const uint8_t *src,
- int width, uint32_t *unused)
+static void monowhite2Y_c(int16_t *dst, const uint8_t *src, int width, uint32_t *unused)
{
int i, j;
for (i=0; i<width/8; i++) {
int d= ~src[i];
for(j=0; j<8; j++)
- dst[8*i+j]= ((d>>(7-j))&1)*255;
+ dst[8*i+j]= ((d>>(7-j))&1)*16383;
+ }
+ if(width&7){
+ int d= ~src[i];
+ for(j=0; j<(width&7); j++)
+ dst[8*i+j]= ((d>>(7-j))&1)*16383;
}
}
-static void monoblack2Y_c(uint8_t *dst, const uint8_t *src,
- int width, uint32_t *unused)
+static void monoblack2Y_c(int16_t *dst, const uint8_t *src, int width, uint32_t *unused)
{
int i, j;
for (i=0; i<width/8; i++) {
int d= src[i];
for(j=0; j<8; j++)
- dst[8*i+j]= ((d>>(7-j))&1)*255;
+ dst[8*i+j]= ((d>>(7-j))&1)*16383;
+ }
+ if(width&7){
+ int d= src[i];
+ for(j=0; j<(width&7); j++)
+ dst[8*i+j]= ((d>>(7-j))&1)*16383;
}
}
@@ -1771,7 +1885,7 @@ static void nv21ToUV_c(uint8_t *dstU, uint8_t *dstV,
#define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))
-static void bgr24ToY_c(uint8_t *dst, const uint8_t *src,
+static void bgr24ToY_c(int16_t *dst, const uint8_t *src,
int width, uint32_t *unused)
{
int i;
@@ -1780,11 +1894,11 @@ static void bgr24ToY_c(uint8_t *dst, const uint8_t *src,
int g= src[i*3+1];
int r= src[i*3+2];
- dst[i]= ((RY*r + GY*g + BY*b + (33<<(RGB2YUV_SHIFT-1)))>>RGB2YUV_SHIFT);
+ dst[i]= ((RY*r + GY*g + BY*b + (32<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6));
}
}
-static void bgr24ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
+static void bgr24ToUV_c(int16_t *dstU, int16_t *dstV, const uint8_t *src1,
const uint8_t *src2, int width, uint32_t *unused)
{
int i;
@@ -1793,13 +1907,13 @@ static void bgr24ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
int g= src1[3*i + 1];
int r= src1[3*i + 2];
- dstU[i]= (RU*r + GU*g + BU*b + (257<<(RGB2YUV_SHIFT-1)))>>RGB2YUV_SHIFT;
- dstV[i]= (RV*r + GV*g + BV*b + (257<<(RGB2YUV_SHIFT-1)))>>RGB2YUV_SHIFT;
+ dstU[i]= (RU*r + GU*g + BU*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
+ dstV[i]= (RV*r + GV*g + BV*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
}
assert(src1 == src2);
}
-static void bgr24ToUV_half_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
+static void bgr24ToUV_half_c(int16_t *dstU, int16_t *dstV, const uint8_t *src1,
const uint8_t *src2, int width, uint32_t *unused)
{
int i;
@@ -1808,13 +1922,13 @@ static void bgr24ToUV_half_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
int g= src1[6*i + 1] + src1[6*i + 4];
int r= src1[6*i + 2] + src1[6*i + 5];
- dstU[i]= (RU*r + GU*g + BU*b + (257<<RGB2YUV_SHIFT))>>(RGB2YUV_SHIFT+1);
- dstV[i]= (RV*r + GV*g + BV*b + (257<<RGB2YUV_SHIFT))>>(RGB2YUV_SHIFT+1);
+ dstU[i]= (RU*r + GU*g + BU*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
+ dstV[i]= (RV*r + GV*g + BV*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
}
assert(src1 == src2);
}
-static void rgb24ToY_c(uint8_t *dst, const uint8_t *src, int width,
+static void rgb24ToY_c(int16_t *dst, const uint8_t *src, int width,
uint32_t *unused)
{
int i;
@@ -1823,11 +1937,11 @@ static void rgb24ToY_c(uint8_t *dst, const uint8_t *src, int width,
int g= src[i*3+1];
int b= src[i*3+2];
- dst[i]= ((RY*r + GY*g + BY*b + (33<<(RGB2YUV_SHIFT-1)))>>RGB2YUV_SHIFT);
+ dst[i]= ((RY*r + GY*g + BY*b + (32<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6));
}
}
-static void rgb24ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
+static void rgb24ToUV_c(int16_t *dstU, int16_t *dstV, const uint8_t *src1,
const uint8_t *src2, int width, uint32_t *unused)
{
int i;
@@ -1837,13 +1951,13 @@ static void rgb24ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
int g= src1[3*i + 1];
int b= src1[3*i + 2];
- dstU[i]= (RU*r + GU*g + BU*b + (257<<(RGB2YUV_SHIFT-1)))>>RGB2YUV_SHIFT;
- dstV[i]= (RV*r + GV*g + BV*b + (257<<(RGB2YUV_SHIFT-1)))>>RGB2YUV_SHIFT;
+ dstU[i]= (RU*r + GU*g + BU*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
+ dstV[i]= (RV*r + GV*g + BV*b + (256<<(RGB2YUV_SHIFT-1)) + (1<<(RGB2YUV_SHIFT-7)))>>(RGB2YUV_SHIFT-6);
}
}
-static void rgb24ToUV_half_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
- const uint8_t *src2, int width, uint32_t *unused)
+static void rgb24ToUV_half_c(int16_t *dstU, int16_t *dstV, const uint8_t *src1,
+ const uint8_t *src2, int width, uint32_t *unused)
{
int i;
assert(src1==src2);
@@ -1852,8 +1966,8 @@ static void rgb24ToUV_half_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
int g= src1[6*i + 1] + src1[6*i + 4];
int b= src1[6*i + 2] + src1[6*i + 5];
- dstU[i]= (RU*r + GU*g + BU*b + (257<<RGB2YUV_SHIFT))>>(RGB2YUV_SHIFT+1);
- dstV[i]= (RV*r + GV*g + BV*b + (257<<RGB2YUV_SHIFT))>>(RGB2YUV_SHIFT+1);
+ dstU[i]= (RU*r + GU*g + BU*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
+ dstV[i]= (RV*r + GV*g + BV*b + (256<<RGB2YUV_SHIFT) + (1<<(RGB2YUV_SHIFT-6)))>>(RGB2YUV_SHIFT-5);
}
}
@@ -1867,6 +1981,9 @@ static void hScale16_c(SwsContext *c, int16_t *_dst, int dstW, const uint8_t *_s
int bits = av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1;
int sh = (bits <= 7) ? 11 : (bits - 4);
+ if((isAnyRGB(c->srcFormat) || c->srcFormat==PIX_FMT_PAL8) && av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1<15)
+ sh= 9;
+
for (i = 0; i < dstW; i++) {
int j;
int srcPos = filterPos[i];
@@ -1899,6 +2016,35 @@ static void hScale_c(SwsContext *c, int16_t *dst, int dstW, const uint8_t *src,
}
}
+static inline void hScale16N_c(int16_t *dst, int dstW, const uint16_t *src, int srcW, int xInc,
+ const int16_t *filter, const int16_t *filterPos, long filterSize, int shift)
+{
+ int i, j;
+
+ for (i=0; i<dstW; i++) {
+ int srcPos= filterPos[i];
+ int val=0;
+ for (j=0; j<filterSize; j++) {
+ val += ((int)src[srcPos + j])*filter[filterSize*i + j];
+ }
+ dst[i] = FFMIN(val>>shift, (1<<15)-1); // the cubic equation does overflow ...
+ }
+}
+
+static inline void hScale16NX_c(int16_t *dst, int dstW, const uint16_t *src, int srcW, int xInc,
+ const int16_t *filter, const int16_t *filterPos, long filterSize, int shift)
+{
+ int i, j;
+ for (i=0; i<dstW; i++) {
+ int srcPos= filterPos[i];
+ int val=0;
+ for (j=0; j<filterSize; j++) {
+ val += ((int)av_bswap16(src[srcPos + j]))*filter[filterSize*i + j];
+ }
+ dst[i] = FFMIN(val>>shift, (1<<15)-1); // the cubic equation does overflow ...
+ }
+}
+
//FIXME all pal and rgb srcFormats could do this convertion as well
//FIXME all scalers more complex than bilinear could do half of this transform
static void chrRangeToJpeg_c(int16_t *dstU, int16_t *dstV, int width)
@@ -1962,7 +2108,7 @@ static void lumRangeFromJpeg16_c(int16_t *_dst, int width)
int i;
int32_t *dst = (int32_t *) _dst;
for (i = 0; i < width; i++)
- dst[i] = (dst[i]*14071 + (33561947<<4))>>14;
+ dst[i] = (dst[i]*(14071/4) + (33561947<<4)/4)>>12;
}
static void hyscale_fast_c(SwsContext *c, int16_t *dst, int dstWidth,
@@ -1976,6 +2122,8 @@ static void hyscale_fast_c(SwsContext *c, int16_t *dst, int dstWidth,
dst[i]= (src[xx]<<7) + (src[xx+1] - src[xx])*xalpha;
xpos+=xInc;
}
+ for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--)
+ dst[i] = src[srcW-1]*128;
}
static void scale8To16Rv_c(uint16_t *_dst, const uint8_t *src, int len)
@@ -2011,12 +2159,15 @@ static av_always_inline void hyscale(SwsContext *c, int16_t *dst, int dstWidth,
src= formatConvBuffer;
}
- if (av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16) {
+ if (av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16 && !isAnyRGB(c->srcFormat)) {
c->scale8To16Rv((uint16_t *) formatConvBuffer, src, srcW);
src = formatConvBuffer;
}
- if (!c->hyscale_fast) {
+ if (c->hScale16) {
+ int shift= isAnyRGB(c->srcFormat) || c->srcFormat==PIX_FMT_PAL8 ? 13 : av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1;
+ c->hScale16(dst, dstWidth, (const uint16_t*)src, srcW, xInc, hLumFilter, hLumFilterPos, hLumFilterSize, shift);
+ } else if (!c->hyscale_fast) {
c->hScale(c, dst, dstWidth, src, hLumFilter, hLumFilterPos, hLumFilterSize);
} else { // fast bilinear upscale / crap downscale
c->hyscale_fast(c, dst, dstWidth, src, srcW, xInc);
@@ -2025,7 +2176,7 @@ static av_always_inline void hyscale(SwsContext *c, int16_t *dst, int dstWidth,
if (convertRange)
convertRange(dst, dstWidth);
- if (av_pix_fmt_descriptors[c->dstFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16) {
+ if (av_pix_fmt_descriptors[c->dstFormat].comp[0].depth_minus1 < 15 && c->scalingBpp == 16) {
c->scale19To15Fw(dst, (int32_t *) dst, dstWidth);
}
}
@@ -2043,6 +2194,10 @@ static void hcscale_fast_c(SwsContext *c, int16_t *dst1, int16_t *dst2,
dst2[i]=(src2[xx]*(xalpha^127)+src2[xx+1]*xalpha);
xpos+=xInc;
}
+ for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--) {
+ dst1[i] = src1[srcW-1]*128;
+ dst2[i] = src2[srcW-1]*128;
+ }
}
static av_always_inline void hcscale(SwsContext *c, int16_t *dst1, int16_t *dst2, int dstWidth,
@@ -2052,21 +2207,25 @@ static av_always_inline void hcscale(SwsContext *c, int16_t *dst1, int16_t *dst2
uint8_t *formatConvBuffer, uint32_t *pal)
{
if (c->chrToYV12) {
- uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW * c->scalingBpp >> 3, 16);
+ uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW*2+78, 16);
c->chrToYV12(formatConvBuffer, buf2, src1, src2, srcW, pal);
src1= formatConvBuffer;
src2= buf2;
}
- if (av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16) {
- uint8_t *buf2 = (formatConvBuffer + FFALIGN(srcW * 2, 16));
+ if (av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16 && !isAnyRGB(c->srcFormat)) {
+ uint8_t *buf2 = (formatConvBuffer + FFALIGN(srcW * 2+78, 16));
c->scale8To16Rv((uint16_t *) formatConvBuffer, src1, srcW);
c->scale8To16Rv((uint16_t *) buf2, src2, srcW);
src1 = formatConvBuffer;
src2 = buf2;
}
- if (!c->hcscale_fast) {
+ if (c->hScale16) {
+ int shift= isAnyRGB(c->srcFormat) || c->srcFormat==PIX_FMT_PAL8 ? 13 : av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1;
+ c->hScale16(dst1, dstWidth, (const uint16_t*)src1, srcW, xInc, hChrFilter, hChrFilterPos, hChrFilterSize, shift);
+ c->hScale16(dst2, dstWidth, (const uint16_t*)src2, srcW, xInc, hChrFilter, hChrFilterPos, hChrFilterSize, shift);
+ } else if (!c->hcscale_fast) {
c->hScale(c, dst1, dstWidth, src1, hChrFilter, hChrFilterPos, hChrFilterSize);
c->hScale(c, dst2, dstWidth, src2, hChrFilter, hChrFilterPos, hChrFilterSize);
} else { // fast bilinear upscale / crap downscale
@@ -2076,7 +2235,7 @@ static av_always_inline void hcscale(SwsContext *c, int16_t *dst1, int16_t *dst2
if (c->chrConvertRange)
c->chrConvertRange(dst1, dst2, dstWidth);
- if (av_pix_fmt_descriptors[c->dstFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16) {
+ if (av_pix_fmt_descriptors[c->dstFormat].comp[0].depth_minus1 < 15 && c->scalingBpp == 16) {
c->scale19To15Fw(dst1, (int32_t *) dst1, dstWidth);
c->scale19To15Fw(dst2, (int32_t *) dst2, dstWidth);
}
@@ -2169,7 +2328,10 @@ find_c_packed_planar_out_funcs(SwsContext *c,
*yuv2packedX = yuv2bgr24_full_X_c;
break;
}
+ if(!*yuv2packedX)
+ goto YUV_PACKED;
} else {
+ YUV_PACKED:
switch (dstFormat) {
case PIX_FMT_GRAY16BE:
*yuv2packed1 = yuv2gray16BE_1_c;
@@ -2362,12 +2524,13 @@ static int swScale(SwsContext *c, const uint8_t* src[],
const int chrSrcSliceH= -((-srcSliceH) >> c->chrSrcVSubSample);
int lastDstY;
uint32_t *pal=c->pal_yuv;
+
+ int should_dither= isNBPS(c->srcFormat) || is16BPS(c->srcFormat);
yuv2planar1_fn yuv2yuv1 = c->yuv2yuv1;
yuv2planarX_fn yuv2yuvX = c->yuv2yuvX;
yuv2packed1_fn yuv2packed1 = c->yuv2packed1;
yuv2packed2_fn yuv2packed2 = c->yuv2packed2;
yuv2packedX_fn yuv2packedX = c->yuv2packedX;
- int should_dither = is9_OR_10BPS(c->srcFormat) || is16BPS(c->srcFormat);
/* vars which will change and which we need to store back in the context */
int dstY= c->dstY;
@@ -2525,6 +2688,7 @@ static int swScale(SwsContext *c, const uint8_t* src[],
const int16_t **chrUSrcPtr= (const int16_t **) chrUPixBuf + chrBufIndex + firstChrSrcY - lastInChrBuf + vChrBufSize;
const int16_t **chrVSrcPtr= (const int16_t **) chrVPixBuf + chrBufIndex + firstChrSrcY - lastInChrBuf + vChrBufSize;
const int16_t **alpSrcPtr= (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? (const int16_t **) alpPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize : NULL;
+
if (isPlanarYUV(dstFormat) || dstFormat==PIX_FMT_GRAY8) { //YV12 like
const int chrSkipMask= (1<<c->chrDstVSubSample)-1;
if ((dstY&chrSkipMask) || isGray(dstFormat))
@@ -2607,21 +2771,21 @@ static av_cold void sws_init_swScale_c(SwsContext *c)
case PIX_FMT_PAL8 :
case PIX_FMT_BGR4_BYTE:
case PIX_FMT_RGB4_BYTE: c->chrToYV12 = palToUV_c; break;
-#if HAVE_BIGENDIAN
+ case PIX_FMT_YUV444P9BE:
+ case PIX_FMT_YUV420P9BE:
+ case PIX_FMT_YUV444P10BE:
+ case PIX_FMT_YUV422P10BE:
+ case PIX_FMT_YUV420P10BE: c->hScale16= HAVE_BIGENDIAN ? hScale16N_c : hScale16NX_c; break;
case PIX_FMT_YUV444P9LE:
case PIX_FMT_YUV420P9LE:
case PIX_FMT_YUV422P10LE:
- case PIX_FMT_YUV444P10LE:
case PIX_FMT_YUV420P10LE:
+ case PIX_FMT_YUV444P10LE: c->hScale16= HAVE_BIGENDIAN ? hScale16NX_c : hScale16N_c; break;
+#if HAVE_BIGENDIAN
case PIX_FMT_YUV420P16LE:
case PIX_FMT_YUV422P16LE:
case PIX_FMT_YUV444P16LE: c->chrToYV12 = bswap16UV_c; break;
#else
- case PIX_FMT_YUV444P9BE:
- case PIX_FMT_YUV420P9BE:
- case PIX_FMT_YUV444P10BE:
- case PIX_FMT_YUV422P10BE:
- case PIX_FMT_YUV420P10BE:
case PIX_FMT_YUV420P16BE:
case PIX_FMT_YUV422P16BE:
case PIX_FMT_YUV444P16BE: c->chrToYV12 = bswap16UV_c; break;
@@ -2675,21 +2839,11 @@ static av_cold void sws_init_swScale_c(SwsContext *c)
c->alpToYV12 = NULL;
switch (srcFormat) {
#if HAVE_BIGENDIAN
- case PIX_FMT_YUV444P9LE:
- case PIX_FMT_YUV420P9LE:
- case PIX_FMT_YUV444P10LE:
- case PIX_FMT_YUV422P10LE:
- case PIX_FMT_YUV420P10LE:
case PIX_FMT_YUV420P16LE:
case PIX_FMT_YUV422P16LE:
case PIX_FMT_YUV444P16LE:
case PIX_FMT_GRAY16LE: c->lumToYV12 = bswap16Y_c; break;
#else
- case PIX_FMT_YUV444P9BE:
- case PIX_FMT_YUV420P9BE:
- case PIX_FMT_YUV444P10BE:
- case PIX_FMT_YUV422P10BE:
- case PIX_FMT_YUV420P10BE:
case PIX_FMT_YUV420P16BE:
case PIX_FMT_YUV422P16BE:
case PIX_FMT_YUV444P16BE:
@@ -2731,9 +2885,14 @@ static av_cold void sws_init_swScale_c(SwsContext *c)
case PIX_FMT_ABGR:
case PIX_FMT_ARGB: c->alpToYV12 = abgrToA_c; break;
case PIX_FMT_Y400A: c->alpToYV12 = uyvyToY_c; break;
+ case PIX_FMT_PAL8 : c->alpToYV12 = palToA_c; break;
}
}
+ if((isAnyRGB(c->srcFormat) && av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1<15)
+ || c->srcFormat == PIX_FMT_PAL8)
+ c->hScale16= hScale16N_c;
+
if (c->scalingBpp == 8) {
c->hScale = hScale_c;
if (c->flags & SWS_FAST_BILINEAR) {
@@ -2751,6 +2910,11 @@ static av_cold void sws_init_swScale_c(SwsContext *c)
}
}
} else {
+ if(c->hScale16 == hScale16NX_c && !isAnyRGB(c->srcFormat)){
+ c->chrToYV12 = bswap16UV_c;
+ c->lumToYV12 = bswap16Y_c;
+ }
+ c->hScale16 = NULL;
c->hScale = hScale16_c;
c->scale19To15Fw = scale19To15Fw_c;
c->scale8To16Rv = scale8To16Rv_c;
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-01 11:01:47 +0300