/*
* 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
*/
#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"
#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))
};
#endif
+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 int16_t **lumSrc,
int lumFilterSize, const int16_t *chrFilter,
const int16_t **chrVSrc,
int chrFilterSize, const int16_t **alpSrc,
uint8_t *dest, uint8_t *uDest, uint8_t *vDest,
- uint8_t *aDest, int dstW, int chrDstW)
+ uint8_t *aDest, int dstW, int chrDstW,
+ const uint8_t *lumDither, const uint8_t *chrDither)
{
//FIXME Optimize (just quickly written not optimized..)
int i;
for (i=0; i<dstW; i++) {
- int val=1<<18;
+ int val = lumDither[i&7] << 12;
int j;
for (j=0; j<lumFilterSize; j++)
val += lumSrc[j][i] * lumFilter[j];
if (uDest)
for (i=0; i<chrDstW; i++) {
- int u=1<<18;
- int v=1<<18;
+ int u = chrDither[i&7] << 12;
+ int v = chrDither[(i+3)&7] << 12;
int j;
for (j=0; j<chrFilterSize; j++) {
u += chrUSrc[j][i] * chrFilter[j];
if (CONFIG_SWSCALE_ALPHA && aDest)
for (i=0; i<dstW; i++) {
- int val=1<<18;
+ int val = lumDither[i&7] << 12;
int j;
for (j=0; j<lumFilterSize; j++)
val += alpSrc[j][i] * lumFilter[j];
}
static void yuv2yuv1_c(SwsContext *c, const int16_t *lumSrc,
- const int16_t *chrUSrc, const int16_t *chrVSrc,
- const int16_t *alpSrc,
- uint8_t *dest, uint8_t *uDest, uint8_t *vDest,
- uint8_t *aDest, int dstW, int chrDstW)
+ const int16_t *chrUSrc, const int16_t *chrVSrc,
+ const int16_t *alpSrc,
+ uint8_t *dest, uint8_t *uDest, uint8_t *vDest,
+ uint8_t *aDest, int dstW, int chrDstW, const uint8_t *lumDither, const uint8_t *chrDither)
{
int i;
+
for (i=0; i<dstW; i++) {
- int val= (lumSrc[i]+64)>>7;
+ int val= (lumSrc[i]+lumDither[i&7])>>7;
dest[i]= av_clip_uint8(val);
}
if (uDest)
for (i=0; i<chrDstW; i++) {
- int u=(chrUSrc[i]+64)>>7;
- int v=(chrVSrc[i]+64)>>7;
+ int u=(chrUSrc[i]+chrDither[i&7])>>7;
+ int v=(chrVSrc[i]+chrDither[(i+3)&7])>>7;
uDest[i]= av_clip_uint8(u);
vDest[i]= av_clip_uint8(v);
}
if (CONFIG_SWSCALE_ALPHA && aDest)
for (i=0; i<dstW; i++) {
- int val= (alpSrc[i]+64)>>7;
+ int val= (alpSrc[i]+lumDither[i&7])>>7;
aDest[i]= av_clip_uint8(val);
}
}
const int16_t **chrVSrc, int chrFilterSize,
const int16_t **alpSrc, uint8_t *dest, uint8_t *uDest,
uint8_t *vDest, uint8_t *aDest,
- int dstW, int chrDstW)
+ int dstW, int chrDstW,
+ const uint8_t *lumDither, const uint8_t *chrDither)
{
enum PixelFormat dstFormat = c->dstFormat;
//FIXME Optimize (just quickly written not optimized..)
int i;
for (i=0; i<dstW; i++) {
- int val=1<<18;
+ int val = lumDither[i&7]<<12;
int j;
for (j=0; j<lumFilterSize; j++)
val += lumSrc[j][i] * lumFilter[j];
if (dstFormat == PIX_FMT_NV12)
for (i=0; i<chrDstW; i++) {
- int u=1<<18;
- int v=1<<18;
+ int u = chrDither[i&7]<<12;
+ int v = chrDither[(i+3)&7]<<12;
int j;
for (j=0; j<chrFilterSize; j++) {
u += chrUSrc[j][i] * chrFilter[j];
}
else
for (i=0; i<chrDstW; i++) {
- int u=1<<18;
- int v=1<<18;
+ int u = chrDither[i&7]<<12;
+ int v = chrDither[(i+3)&7]<<12;
int j;
for (j=0; j<chrFilterSize; j++) {
u += chrUSrc[j][i] * chrFilter[j];
#define YSCALE_YUV_2_RGBX_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;\
\
U >>=10;\
V >>=10;\
if (alpha) {\
- A = rnd;\
+ A = rnd>>3;\
for (j=0; j<lumFilterSize; j++)\
A += alpSrc[j][i ] * lumFilter[j];\
A >>=19;\
}
static av_always_inline void
-rgb48ToY_c_template(uint8_t *dst, const uint8_t *src, int width,
+rgb48ToY_c_template(int16_t *dst, const uint16_t *src, int width,
enum PixelFormat origin)
{
int i;
for (i = 0; i < width; i++) {
#define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))
- int a = input_pixel(&src[i*6+0]) >> 8;
- int g = input_pixel(&src[i*6+2]) >> 8;
- int c = input_pixel(&src[i*6+4]) >> 8;
+ int a = input_pixel(&src[i*3+0]);
+ int g = input_pixel(&src[i*3+1]);
+ int c = input_pixel(&src[i*3+2]);
#define r ((origin == PIX_FMT_BGR48BE || origin == PIX_FMT_BGR48LE) ? c : a)
#define b ((origin == PIX_FMT_BGR48BE || origin == PIX_FMT_BGR48LE) ? a : c)
- 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+8)) + (1<<(RGB2YUV_SHIFT-7+8))) >> (RGB2YUV_SHIFT-6+8);
}
}
static av_always_inline void
-rgb48ToUV_c_template(uint8_t *dstU, uint8_t *dstV,
- const uint8_t *src1, const uint8_t *src2,
+rgb48ToUV_c_template(int16_t *dstU, int16_t *dstV,
+ const uint16_t *src1, const uint16_t *src2,
int width, enum PixelFormat origin)
{
int i;
assert(src1==src2);
for (i = 0; i < width; i++) {
- int a = input_pixel(&src1[6*i + 0]) >> 8;
- int g = input_pixel(&src1[6*i + 2]) >> 8;
- int c = input_pixel(&src1[6*i + 4]) >> 8;
+ int a = input_pixel(&src1[3*i + 0]);
+ int g = input_pixel(&src1[3*i + 1]);
+ int c = input_pixel(&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+8)) + (1<<(RGB2YUV_SHIFT-7+8))) >> (RGB2YUV_SHIFT-6+8);
+ dstV[i] = (RV*r + GV*g + BV*b + (256<<(RGB2YUV_SHIFT-1+8)) + (1<<(RGB2YUV_SHIFT-7+8))) >> (RGB2YUV_SHIFT-6+8);
}
}
static av_always_inline void
-rgb48ToUV_half_c_template(uint8_t *dstU, uint8_t *dstV,
- const uint8_t *src1, const uint8_t *src2,
+rgb48ToUV_half_c_template(int16_t *dstU, int16_t *dstV,
+ const uint16_t *src1, const uint16_t *src2,
int width, enum PixelFormat origin)
{
int i;
assert(src1==src2);
for (i = 0; i < width; i++) {
- int a = (input_pixel(&src1[12*i + 0]) >> 8) + (input_pixel(&src1[12*i + 6]) >> 8);
- int g = (input_pixel(&src1[12*i + 2]) >> 8) + (input_pixel(&src1[12*i + 8]) >> 8);
- int c = (input_pixel(&src1[12*i + 4]) >> 8) + (input_pixel(&src1[12*i + 10]) >> 8);
+ int a = (input_pixel(&src1[6*i + 0])) + (input_pixel(&src1[6*i + 3]));
+ int g = (input_pixel(&src1[6*i + 1])) + (input_pixel(&src1[6*i + 4]));
+ int c = (input_pixel(&src1[6*i + 2])) + (input_pixel(&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 + (256U<<(RGB2YUV_SHIFT+8)) + (1<<(RGB2YUV_SHIFT-6+8))) >> (RGB2YUV_SHIFT-5+8);
+ dstV[i]= (RV*r + GV*g + BV*b + (256U<<(RGB2YUV_SHIFT+8)) + (1<<(RGB2YUV_SHIFT-6+8))) >> (RGB2YUV_SHIFT-5+8);
}
#undef r
#undef b
rgb48funcs(bgr, BE, PIX_FMT_BGR48BE);
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++) {
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,
{
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++) {
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,
{
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;
}
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);
}
#undef input_pixel
}
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];
+ dst[i]= src[4*i]<<6;
}
}
-static void rgbaToA_c(uint8_t *dst, const uint8_t *src, int width, uint32_t *unused)
+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+3];
+ dst[i]= src[4*i+3]<<6;
}
}
-static void palToY_c(uint8_t *dst, const uint8_t *src, int width, uint32_t *pal)
+static void palToA_c(int16_t *dst, const uint8_t *src, int 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] >> 24)<<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 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)<<6;
+ }
+}
+
+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;
}
}
-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;
}
}
}
}
-static void nv12ToUV_c(uint8_t *dstU, uint8_t *dstV,
- const uint8_t *src1, const uint8_t *src2,
- int width, uint32_t *unused)
-{
- nvXXtoUV_c(dstU, dstV, src1, width);
-}
-
-static void nv21ToUV_c(uint8_t *dstU, uint8_t *dstV,
- const uint8_t *src1, const uint8_t *src2,
- int width, uint32_t *unused)
-{
- nvXXtoUV_c(dstV, dstU, src1, width);
-}
-
// FIXME Maybe dither instead.
static av_always_inline void
yuv9_OR_10ToUV_c_template(uint8_t *dstU, uint8_t *dstV,
YUV_NBPS(10, LE, PIX_FMT_YUV420P10LE);
YUV_NBPS(10, BE, PIX_FMT_YUV420P10BE);
-static void bgr24ToY_c(uint8_t *dst, const uint8_t *src,
+static inline void nv12ToUV_c(uint8_t *dstU, uint8_t *dstV,
+ const uint8_t *src1, const uint8_t *src2,
+ int width, uint32_t *unused)
+{
+ nvXXtoUV_c(dstU, dstV, src1, width);
+}
+
+static inline void nv21ToUV_c(uint8_t *dstU, uint8_t *dstV,
+ const uint8_t *src1, const uint8_t *src2,
+ int width, uint32_t *unused)
+{
+ nvXXtoUV_c(dstV, dstU, src1, width);
+}
+
+static void bgr24ToY_c(int16_t *dst, const uint8_t *src,
int width, uint32_t *unused)
{
int i;
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;
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;
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;
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;
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);
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);
}
}
}
}
+static inline void hScale16_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 hScale16X_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(uint16_t *dstU, uint16_t *dstV, int width)
+static void chrRangeToJpeg_c(int16_t *dstU, int16_t *dstV, int width)
{
int i;
for (i = 0; i < width; i++) {
dstV[i] = (FFMIN(dstV[i],30775)*4663 - 9289992)>>12; //-264
}
}
-static void chrRangeFromJpeg_c(uint16_t *dstU, uint16_t *dstV, int width)
+static void chrRangeFromJpeg_c(int16_t *dstU, int16_t *dstV, int width)
{
int i;
for (i = 0; i < width; i++) {
dstV[i] = (dstV[i]*1799 + 4081085)>>11; //1469
}
}
-static void lumRangeToJpeg_c(uint16_t *dst, int width)
+static void lumRangeToJpeg_c(int16_t *dst, int width)
{
int i;
for (i = 0; i < width; i++)
dst[i] = (FFMIN(dst[i],30189)*19077 - 39057361)>>14;
}
-static void lumRangeFromJpeg_c(uint16_t *dst, int width)
+static void lumRangeFromJpeg_c(int16_t *dst, int width)
{
int i;
for (i = 0; i < width; i++)
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;
}
// *** horizontal scale Y line to temp buffer
uint32_t *pal, int isAlpha)
{
void (*toYV12)(uint8_t *, const uint8_t *, int, uint32_t *) = isAlpha ? c->alpToYV12 : c->lumToYV12;
- void (*convertRange)(uint16_t *, int) = isAlpha ? NULL : c->lumConvertRange;
+ void (*convertRange)(int16_t *, int) = isAlpha ? NULL : c->lumConvertRange;
if (toYV12) {
toYV12(formatConvBuffer, src, srcW, pal);
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(dst, dstWidth, src, srcW, xInc, hLumFilter, hLumFilterPos, hLumFilterSize);
} else { // fast bilinear upscale / crap downscale
c->hyscale_fast(c, dst, dstWidth, src, srcW, xInc);
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, uint16_t *dst1, uint16_t *dst2, int dstWidth,
uint8_t *formatConvBuffer, uint32_t *pal)
{
if (c->chrToYV12) {
- uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW, 16);
+ uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW*2+78, 16);
c->chrToYV12(formatConvBuffer, buf2, src1, src2, srcW, pal);
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(dst1, dstWidth, src1, srcW, xInc, hChrFilter, hChrFilterPos, hChrFilterSize);
c->hScale(dst2, dstWidth, src2, srcW, xInc, hChrFilter, hChrFilterPos, hChrFilterSize);
} else { // fast bilinear upscale / crap downscale
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;
unsigned char *uDest=dst[1]+dstStride[1]*chrDstY;
unsigned char *vDest=dst[2]+dstStride[2]*chrDstY;
unsigned char *aDest=(CONFIG_SWSCALE_ALPHA && alpPixBuf) ? dst[3]+dstStride[3]*dstY : NULL;
+ const uint8_t *lumDither= should_dither ? dithers[7][dstY &7] : flat64;
+ const uint8_t *chrDither= should_dither ? dithers[7][chrDstY&7] : flat64;
const int firstLumSrcY= vLumFilterPos[dstY]; //First line needed as input
const int firstLumSrcY2= vLumFilterPos[FFMIN(dstY | ((1<<c->chrDstVSubSample) - 1), dstH-1)];
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)) uDest=vDest= NULL; //FIXME split functions in lumi / chromi
const int16_t *chrVBuf= chrVSrcPtr[0];
const int16_t *alpBuf= (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? alpSrcPtr[0] : NULL;
yuv2yuv1(c, lumBuf, chrUBuf, chrVBuf, alpBuf, dest,
- uDest, vDest, aDest, dstW, chrDstW);
+ uDest, vDest, aDest, dstW, chrDstW, lumDither, chrDither);
} else { //General YV12
yuv2yuvX(c,
vLumFilter+dstY*vLumFilterSize , lumSrcPtr, vLumFilterSize,
vChrFilter+chrDstY*vChrFilterSize, chrUSrcPtr,
chrVSrcPtr, vChrFilterSize,
- alpSrcPtr, dest, uDest, vDest, aDest, dstW, chrDstW);
+ alpSrcPtr, dest, uDest, vDest, aDest, dstW, chrDstW, lumDither, chrDither);
}
} else {
assert(lumSrcPtr + vLumFilterSize - 1 < lumPixBuf + vLumBufSize*2);
case PIX_FMT_PAL8 :
case PIX_FMT_BGR4_BYTE:
case PIX_FMT_RGB4_BYTE: c->chrToYV12 = palToUV_c; break;
- case PIX_FMT_YUV420P9BE: c->chrToYV12 = BE9ToUV_c; break;
- case PIX_FMT_YUV420P9LE: c->chrToYV12 = LE9ToUV_c; break;
- case PIX_FMT_YUV420P10BE: c->chrToYV12 = BE10ToUV_c; break;
- case PIX_FMT_YUV420P10LE: c->chrToYV12 = LE10ToUV_c; break;
+ case PIX_FMT_GRAY16BE :
+ case PIX_FMT_YUV420P9BE:
+ case PIX_FMT_YUV422P10BE:
+ case PIX_FMT_YUV420P10BE:
case PIX_FMT_YUV420P16BE:
case PIX_FMT_YUV422P16BE:
- case PIX_FMT_YUV444P16BE: c->chrToYV12 = BEToUV_c; break;
+ case PIX_FMT_YUV444P16BE: c->hScale16= HAVE_BIGENDIAN ? hScale16_c : hScale16X_c; break;
+ case PIX_FMT_GRAY16LE :
+ case PIX_FMT_YUV420P9LE:
+ case PIX_FMT_YUV422P10LE:
+ case PIX_FMT_YUV420P10LE:
case PIX_FMT_YUV420P16LE:
case PIX_FMT_YUV422P16LE:
- case PIX_FMT_YUV444P16LE: c->chrToYV12 = LEToUV_c; break;
+ case PIX_FMT_YUV444P16LE: c->hScale16= HAVE_BIGENDIAN ? hScale16X_c : hScale16_c; break;
}
if (c->chrSrcHSubSample) {
switch(srcFormat) {
c->lumToYV12 = NULL;
c->alpToYV12 = NULL;
switch (srcFormat) {
- case PIX_FMT_YUV420P9BE: c->lumToYV12 = BE9ToY_c; break;
- case PIX_FMT_YUV420P9LE: c->lumToYV12 = LE9ToY_c; break;
- case PIX_FMT_YUV420P10BE: c->lumToYV12 = BE10ToY_c; break;
- case PIX_FMT_YUV420P10LE: c->lumToYV12 = LE10ToY_c; break;
case PIX_FMT_YUYV422 :
- case PIX_FMT_YUV420P16BE:
- case PIX_FMT_YUV422P16BE:
- case PIX_FMT_YUV444P16BE:
- case PIX_FMT_Y400A :
- case PIX_FMT_GRAY16BE : c->lumToYV12 = yuy2ToY_c; break;
+ case PIX_FMT_GRAY8A :
+ c->lumToYV12 = yuy2ToY_c; break;
case PIX_FMT_UYVY422 :
- case PIX_FMT_YUV420P16LE:
- case PIX_FMT_YUV422P16LE:
- case PIX_FMT_YUV444P16LE:
- case PIX_FMT_GRAY16LE : c->lumToYV12 = uyvyToY_c; break;
+ c->lumToYV12 = uyvyToY_c; break;
case PIX_FMT_BGR24 : c->lumToYV12 = bgr24ToY_c; break;
case PIX_FMT_BGR565LE : c->lumToYV12 = bgr16leToY_c; break;
case PIX_FMT_BGR565BE : c->lumToYV12 = bgr16beToY_c; break;
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) || c->srcFormat == PIX_FMT_PAL8)
+ c->hScale16= hScale16_c;
+
if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) {
if (c->srcRange) {
c->lumConvertRange = lumRangeFromJpeg_c;