#define RV ( (int)(0.500*224/255*(1<<RGB2YUV_SHIFT)+0.5))
#define RU (-(int)(0.169*224/255*(1<<RGB2YUV_SHIFT)+0.5))
-static const double rgb2yuv_table[8][9]={
- {0.7152, 0.0722, 0.2126, -0.386, 0.5, -0.115, -0.454, -0.046, 0.5}, //ITU709
- {0.7152, 0.0722, 0.2126, -0.386, 0.5, -0.115, -0.454, -0.046, 0.5}, //ITU709
- {0.587 , 0.114 , 0.299 , -0.331, 0.5, -0.169, -0.419, -0.081, 0.5}, //DEFAULT / ITU601 / ITU624 / SMPTE 170M
- {0.587 , 0.114 , 0.299 , -0.331, 0.5, -0.169, -0.419, -0.081, 0.5}, //DEFAULT / ITU601 / ITU624 / SMPTE 170M
- {0.59 , 0.11 , 0.30 , -0.331, 0.5, -0.169, -0.421, -0.079, 0.5}, //FCC
- {0.587 , 0.114 , 0.299 , -0.331, 0.5, -0.169, -0.419, -0.081, 0.5}, //DEFAULT / ITU601 / ITU624 / SMPTE 170M
- {0.587 , 0.114 , 0.299 , -0.331, 0.5, -0.169, -0.419, -0.081, 0.5}, //DEFAULT / ITU601 / ITU624 / SMPTE 170M
- {0.701 , 0.087 , 0.212 , -0.384, 0.5, -0.116, -0.445, -0.055, 0.5}, //SMPTE 240M
-};
-
/*
NOTES
Special versions: fast Y 1:1 scaling (no interpolation in y direction)
{ 77, 23, 60, 15, 72, 21, 56, 14, },
};
#endif
+DECLARE_ALIGNED(8, const uint8_t, dither_8x8_128)[8][8] = {
+{ 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,},
+};
+DECLARE_ALIGNED(8, const uint8_t, ff_sws_pb_64)[8] =
+{ 64, 64, 64, 64, 64, 64, 64, 64 };
static av_always_inline void
-yuv2yuvX16_c_template(const int16_t *lumFilter, const int16_t **lumSrc,
+yuv2yuvX16_c_template(const int16_t *lumFilter, const int32_t **lumSrc,
int lumFilterSize, const int16_t *chrFilter,
- const int16_t **chrUSrc, const int16_t **chrVSrc,
- int chrFilterSize, const int16_t **alpSrc,
- uint16_t *dest, uint16_t *uDest, uint16_t *vDest,
- uint16_t *aDest, int dstW, int chrDstW,
+ const int32_t **chrUSrc, const int32_t **chrVSrc,
+ int chrFilterSize, const int32_t **alpSrc,
+ uint16_t *dest[4], int dstW, int chrDstW,
int big_endian, int output_bits)
{
//FIXME Optimize (just quickly written not optimized..)
int i;
- int shift = 11 + 16 - output_bits;
+ 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;
#define output_pixel(pos, val) \
if (big_endian) { \
} \
}
for (i = 0; i < dstW; i++) {
- int val = 1 << (26-output_bits);
+ int val = 1 << (30-output_bits - 1);
int j;
for (j = 0; j < lumFilterSize; j++)
- val += lumSrc[j][i] * lumFilter[j];
+ val += (lumSrc[j][i] * lumFilter[j]) >> 1;
- output_pixel(&dest[i], val);
+ output_pixel(&yDest[i], val);
}
if (uDest) {
for (i = 0; i < chrDstW; i++) {
- int u = 1 << (26-output_bits);
- int v = 1 << (26-output_bits);
+ int u = 1 << (30-output_bits - 1);
+ int v = 1 << (30-output_bits - 1);
int j;
for (j = 0; j < chrFilterSize; j++) {
- u += chrUSrc[j][i] * chrFilter[j];
- v += chrVSrc[j][i] * chrFilter[j];
+ u += (chrUSrc[j][i] * chrFilter[j]) >> 1;
+ v += (chrVSrc[j][i] * chrFilter[j]) >> 1;
}
output_pixel(&uDest[i], u);
if (CONFIG_SWSCALE_ALPHA && aDest) {
for (i = 0; i < dstW; i++) {
- int val = 1 << (26-output_bits);
+ int val = 1 << (30-output_bits - 1);
int j;
for (j = 0; j < lumFilterSize; j++)
- val += alpSrc[j][i] * lumFilter[j];
+ val += (alpSrc[j][i] * lumFilter[j]) >> 1;
output_pixel(&aDest[i], val);
}
#define yuv2NBPS(bits, BE_LE, is_be) \
static void yuv2yuvX ## bits ## BE_LE ## _c(SwsContext *c, const int16_t *lumFilter, \
- const int16_t **lumSrc, int lumFilterSize, \
- const int16_t *chrFilter, const int16_t **chrUSrc, \
- 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) \
+ const int16_t **_lumSrc, int lumFilterSize, \
+ const int16_t *chrFilter, const int16_t **_chrUSrc, \
+ const int16_t **_chrVSrc, \
+ int chrFilterSize, const int16_t **_alpSrc, \
+ uint8_t *_dest[4], int dstW, int chrDstW) \
{ \
- uint16_t *dest = (uint16_t *) _dest, *uDest = (uint16_t *) _uDest, \
- *vDest = (uint16_t *) _vDest, *aDest = (uint16_t *) _aDest; \
+ const int32_t **lumSrc = (const int32_t **) _lumSrc, \
+ **chrUSrc = (const int32_t **) _chrUSrc, \
+ **chrVSrc = (const int32_t **) _chrVSrc, \
+ **alpSrc = (const int32_t **) _alpSrc; \
yuv2yuvX16_c_template(lumFilter, lumSrc, lumFilterSize, \
chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
- alpSrc, \
- dest, uDest, vDest, aDest, \
+ alpSrc, (uint16_t **) _dest, \
dstW, chrDstW, is_be, bits); \
}
yuv2NBPS( 9, BE, 1);
const int16_t *chrFilter, const int16_t **chrUSrc,
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 *dest[4], int dstW, int chrDstW)
{
- //FIXME Optimize (just quickly written not optimized..)
+ uint8_t *yDest = dest[0], *uDest = dest[1], *vDest = dest[2],
+ *aDest = CONFIG_SWSCALE_ALPHA ? dest[3] : NULL;
int i;
+ const uint8_t *lumDither = c->lumDither8, *chrDither = c->chrDither8;
+
+ //FIXME Optimize (just quickly written not optimized..)
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];
- dest[i]= av_clip_uint8(val>>19);
+ yDest[i]= av_clip_uint8(val>>19);
}
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)
+ uint8_t *dest[4], int dstW, int chrDstW)
{
+ uint8_t *yDest = dest[0], *uDest = dest[1], *vDest = dest[2],
+ *aDest = CONFIG_SWSCALE_ALPHA ? dest[3] : NULL;
int i;
+ const uint8_t *lumDither = c->lumDither8, *chrDither = c->chrDither8;
+
for (i=0; i<dstW; i++) {
- int val= (lumSrc[i]+64)>>7;
- dest[i]= av_clip_uint8(val);
+ int val = (lumSrc[i]+ lumDither[i & 7]) >> 7;
+ yDest[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 **lumSrc, int lumFilterSize,
const int16_t *chrFilter, const int16_t **chrUSrc,
const int16_t **chrVSrc, int chrFilterSize,
- const int16_t **alpSrc, uint8_t *dest, uint8_t *uDest,
- uint8_t *vDest, uint8_t *aDest,
+ const int16_t **alpSrc, uint8_t *dest[4],
int dstW, int chrDstW)
{
+ uint8_t *yDest = dest[0], *uDest = dest[1];
enum PixelFormat dstFormat = c->dstFormat;
+ const uint8_t *lumDither = c->lumDither8, *chrDither = c->chrDither8;
//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];
- dest[i]= av_clip_uint8(val>>19);
+ yDest[i]= av_clip_uint8(val>>19);
}
if (!uDest)
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];
static av_always_inline void
yuv2gray16_X_c_template(SwsContext *c, const int16_t *lumFilter,
- const int16_t **lumSrc, int lumFilterSize,
- const int16_t *chrFilter, const int16_t **chrUSrc,
- const int16_t **chrVSrc, int chrFilterSize,
- const int16_t **alpSrc, uint8_t *dest, int dstW,
+ const int32_t **lumSrc, int lumFilterSize,
+ const int16_t *chrFilter, const int32_t **chrUSrc,
+ const int32_t **chrVSrc, int chrFilterSize,
+ const int32_t **alpSrc, uint16_t *dest, int dstW,
int y, enum PixelFormat target)
{
int i;
for (i = 0; i < (dstW >> 1); i++) {
int j;
- int Y1 = 1 << 18;
- int Y2 = 1 << 18;
- const int i2 = 2 * i;
+ int Y1 = 1 << 14;
+ int Y2 = 1 << 14;
for (j = 0; j < lumFilterSize; j++) {
- Y1 += lumSrc[j][i2] * lumFilter[j];
- Y2 += lumSrc[j][i2+1] * lumFilter[j];
+ Y1 += lumSrc[j][i * 2] * lumFilter[j];
+ Y2 += lumSrc[j][i * 2 + 1] * lumFilter[j];
}
- Y1 >>= 11;
- Y2 >>= 11;
+ Y1 >>= 15;
+ Y2 >>= 15;
if ((Y1 | Y2) & 0x10000) {
Y1 = av_clip_uint16(Y1);
Y2 = av_clip_uint16(Y2);
}
- output_pixel(&dest[2 * i2 + 0], Y1);
- output_pixel(&dest[2 * i2 + 2], Y2);
+ output_pixel(&dest[i * 2 + 0], Y1);
+ output_pixel(&dest[i * 2 + 1], Y2);
}
}
static av_always_inline void
-yuv2gray16_2_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *buf1, const uint16_t *ubuf0,
- const uint16_t *ubuf1, const uint16_t *vbuf0,
- const uint16_t *vbuf1, const uint16_t *abuf0,
- const uint16_t *abuf1, uint8_t *dest, int dstW,
+yuv2gray16_2_c_template(SwsContext *c, const int32_t *buf[2],
+ const int32_t *ubuf[2], const int32_t *vbuf[2],
+ const int32_t *abuf[2], uint16_t *dest, int dstW,
int yalpha, int uvalpha, int y,
enum PixelFormat target)
{
- int yalpha1 = 4095 - yalpha; \
+ int yalpha1 = 4095 - yalpha;
int i;
+ const int32_t *buf0 = buf[0], *buf1 = buf[1];
for (i = 0; i < (dstW >> 1); i++) {
- const int i2 = 2 * i;
- int Y1 = (buf0[i2 ] * yalpha1 + buf1[i2 ] * yalpha) >> 11;
- int Y2 = (buf0[i2+1] * yalpha1 + buf1[i2+1] * yalpha) >> 11;
+ int Y1 = (buf0[i * 2 ] * yalpha1 + buf1[i * 2 ] * yalpha) >> 15;
+ int Y2 = (buf0[i * 2 + 1] * yalpha1 + buf1[i * 2 + 1] * yalpha) >> 15;
- output_pixel(&dest[2 * i2 + 0], Y1);
- output_pixel(&dest[2 * i2 + 2], Y2);
+ output_pixel(&dest[i * 2 + 0], Y1);
+ output_pixel(&dest[i * 2 + 1], Y2);
}
}
static av_always_inline void
-yuv2gray16_1_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *ubuf0, const uint16_t *ubuf1,
- const uint16_t *vbuf0, const uint16_t *vbuf1,
- const uint16_t *abuf0, uint8_t *dest, int dstW,
- int uvalpha, enum PixelFormat dstFormat,
- int flags, int y, enum PixelFormat target)
+yuv2gray16_1_c_template(SwsContext *c, const int32_t *buf0,
+ const int32_t *ubuf[2], const int32_t *vbuf[2],
+ const int32_t *abuf0, uint16_t *dest, int dstW,
+ int uvalpha, int y, enum PixelFormat target)
{
int i;
for (i = 0; i < (dstW >> 1); i++) {
- const int i2 = 2 * i;
- int Y1 = buf0[i2 ] << 1;
- int Y2 = buf0[i2+1] << 1;
+ int Y1 = buf0[i * 2 ] << 1;
+ int Y2 = buf0[i * 2 + 1] << 1;
- output_pixel(&dest[2 * i2 + 0], Y1);
- output_pixel(&dest[2 * i2 + 2], Y2);
+ output_pixel(&dest[i * 2 + 0], Y1);
+ output_pixel(&dest[i * 2 + 1], Y2);
}
}
#undef output_pixel
-#define YUV2PACKEDWRAPPER(name, base, ext, fmt) \
+#define YUV2PACKED16WRAPPER(name, base, ext, fmt) \
static void name ## ext ## _X_c(SwsContext *c, const int16_t *lumFilter, \
- const int16_t **lumSrc, int lumFilterSize, \
- const int16_t *chrFilter, const int16_t **chrUSrc, \
- const int16_t **chrVSrc, int chrFilterSize, \
- const int16_t **alpSrc, uint8_t *dest, int dstW, \
+ const int16_t **_lumSrc, int lumFilterSize, \
+ const int16_t *chrFilter, const int16_t **_chrUSrc, \
+ const int16_t **_chrVSrc, int chrFilterSize, \
+ const int16_t **_alpSrc, uint8_t *_dest, int dstW, \
int y) \
{ \
+ const int32_t **lumSrc = (const int32_t **) _lumSrc, \
+ **chrUSrc = (const int32_t **) _chrUSrc, \
+ **chrVSrc = (const int32_t **) _chrVSrc, \
+ **alpSrc = (const int32_t **) _alpSrc; \
+ uint16_t *dest = (uint16_t *) _dest; \
name ## base ## _X_c_template(c, lumFilter, lumSrc, lumFilterSize, \
chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
alpSrc, dest, dstW, y, fmt); \
} \
\
-static void name ## ext ## _2_c(SwsContext *c, const uint16_t *buf0, \
- const uint16_t *buf1, const uint16_t *ubuf0, \
- const uint16_t *ubuf1, const uint16_t *vbuf0, \
- const uint16_t *vbuf1, const uint16_t *abuf0, \
- const uint16_t *abuf1, uint8_t *dest, int dstW, \
+static void name ## ext ## _2_c(SwsContext *c, const int16_t *_buf[2], \
+ const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
+ const int16_t *_abuf[2], uint8_t *_dest, int dstW, \
int yalpha, int uvalpha, int y) \
{ \
- name ## base ## _2_c_template(c, buf0, buf1, ubuf0, ubuf1, \
- vbuf0, vbuf1, abuf0, abuf1, \
+ const int32_t **buf = (const int32_t **) _buf, \
+ **ubuf = (const int32_t **) _ubuf, \
+ **vbuf = (const int32_t **) _vbuf, \
+ **abuf = (const int32_t **) _abuf; \
+ uint16_t *dest = (uint16_t *) _dest; \
+ name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
dest, dstW, yalpha, uvalpha, y, fmt); \
} \
\
-static void name ## ext ## _1_c(SwsContext *c, const uint16_t *buf0, \
- const uint16_t *ubuf0, const uint16_t *ubuf1, \
- const uint16_t *vbuf0, const uint16_t *vbuf1, \
- const uint16_t *abuf0, uint8_t *dest, int dstW, \
- int uvalpha, enum PixelFormat dstFormat, \
- int flags, int y) \
+static void name ## ext ## _1_c(SwsContext *c, const int16_t *_buf0, \
+ const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
+ const int16_t *_abuf0, uint8_t *_dest, int dstW, \
+ int uvalpha, int y) \
{ \
- name ## base ## _1_c_template(c, buf0, ubuf0, ubuf1, vbuf0, \
- vbuf1, abuf0, dest, dstW, uvalpha, \
- dstFormat, flags, y, fmt); \
+ const int32_t *buf0 = (const int32_t *) _buf0, \
+ **ubuf = (const int32_t **) _ubuf, \
+ **vbuf = (const int32_t **) _vbuf, \
+ *abuf0 = (const int32_t *) _abuf0; \
+ uint16_t *dest = (uint16_t *) _dest; \
+ name ## base ## _1_c_template(c, buf0, ubuf, vbuf, abuf0, dest, \
+ dstW, uvalpha, y, fmt); \
}
-YUV2PACKEDWRAPPER(yuv2gray16,, LE, PIX_FMT_GRAY16LE);
-YUV2PACKEDWRAPPER(yuv2gray16,, BE, PIX_FMT_GRAY16BE);
+YUV2PACKED16WRAPPER(yuv2gray16,, LE, PIX_FMT_GRAY16LE);
+YUV2PACKED16WRAPPER(yuv2gray16,, BE, PIX_FMT_GRAY16BE);
#define output_pixel(pos, acc) \
if (target == PIX_FMT_MONOBLACK) { \
}
static av_always_inline void
-yuv2mono_2_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *buf1, const uint16_t *ubuf0,
- const uint16_t *ubuf1, const uint16_t *vbuf0,
- const uint16_t *vbuf1, const uint16_t *abuf0,
- const uint16_t *abuf1, uint8_t *dest, int dstW,
+yuv2mono_2_c_template(SwsContext *c, const int16_t *buf[2],
+ const int16_t *ubuf[2], const int16_t *vbuf[2],
+ const int16_t *abuf[2], uint8_t *dest, int dstW,
int yalpha, int uvalpha, int y,
enum PixelFormat target)
{
+ const int16_t *buf0 = buf[0], *buf1 = buf[1];
const uint8_t * const d128 = dither_8x8_220[y & 7];
uint8_t *g = c->table_gU[128] + c->table_gV[128];
int yalpha1 = 4095 - yalpha;
}
static av_always_inline void
-yuv2mono_1_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *ubuf0, const uint16_t *ubuf1,
- const uint16_t *vbuf0, const uint16_t *vbuf1,
- const uint16_t *abuf0, uint8_t *dest, int dstW,
- int uvalpha, enum PixelFormat dstFormat,
- int flags, int y, enum PixelFormat target)
+yuv2mono_1_c_template(SwsContext *c, const int16_t *buf0,
+ const int16_t *ubuf[2], const int16_t *vbuf[2],
+ const int16_t *abuf0, uint8_t *dest, int dstW,
+ int uvalpha, int y, enum PixelFormat target)
{
const uint8_t * const d128 = dither_8x8_220[y & 7];
uint8_t *g = c->table_gU[128] + c->table_gV[128];
#undef output_pixel
+#define YUV2PACKEDWRAPPER(name, base, ext, fmt) \
+static void name ## ext ## _X_c(SwsContext *c, const int16_t *lumFilter, \
+ const int16_t **lumSrc, int lumFilterSize, \
+ const int16_t *chrFilter, const int16_t **chrUSrc, \
+ const int16_t **chrVSrc, int chrFilterSize, \
+ const int16_t **alpSrc, uint8_t *dest, int dstW, \
+ int y) \
+{ \
+ name ## base ## _X_c_template(c, lumFilter, lumSrc, lumFilterSize, \
+ chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
+ alpSrc, dest, dstW, y, fmt); \
+} \
+ \
+static void name ## ext ## _2_c(SwsContext *c, const int16_t *buf[2], \
+ const int16_t *ubuf[2], const int16_t *vbuf[2], \
+ const int16_t *abuf[2], uint8_t *dest, int dstW, \
+ int yalpha, int uvalpha, int y) \
+{ \
+ name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
+ dest, dstW, yalpha, uvalpha, y, fmt); \
+} \
+ \
+static void name ## ext ## _1_c(SwsContext *c, const int16_t *buf0, \
+ const int16_t *ubuf[2], const int16_t *vbuf[2], \
+ const int16_t *abuf0, uint8_t *dest, int dstW, \
+ int uvalpha, int y) \
+{ \
+ name ## base ## _1_c_template(c, buf0, ubuf, vbuf, \
+ abuf0, dest, dstW, uvalpha, \
+ y, fmt); \
+}
+
YUV2PACKEDWRAPPER(yuv2mono,, white, PIX_FMT_MONOWHITE);
YUV2PACKEDWRAPPER(yuv2mono,, black, PIX_FMT_MONOBLACK);
}
static av_always_inline void
-yuv2422_2_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *buf1, const uint16_t *ubuf0,
- const uint16_t *ubuf1, const uint16_t *vbuf0,
- const uint16_t *vbuf1, const uint16_t *abuf0,
- const uint16_t *abuf1, uint8_t *dest, int dstW,
+yuv2422_2_c_template(SwsContext *c, const int16_t *buf[2],
+ const int16_t *ubuf[2], const int16_t *vbuf[2],
+ const int16_t *abuf[2], uint8_t *dest, int dstW,
int yalpha, int uvalpha, int y,
enum PixelFormat target)
{
+ const int16_t *buf0 = buf[0], *buf1 = buf[1],
+ *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
+ *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
int yalpha1 = 4095 - yalpha;
int uvalpha1 = 4095 - uvalpha;
int i;
}
static av_always_inline void
-yuv2422_1_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *ubuf0, const uint16_t *ubuf1,
- const uint16_t *vbuf0, const uint16_t *vbuf1,
- const uint16_t *abuf0, uint8_t *dest, int dstW,
- int uvalpha, enum PixelFormat dstFormat,
- int flags, int y, enum PixelFormat target)
+yuv2422_1_c_template(SwsContext *c, const int16_t *buf0,
+ const int16_t *ubuf[2], const int16_t *vbuf[2],
+ const int16_t *abuf0, uint8_t *dest, int dstW,
+ int uvalpha, int y, enum PixelFormat target)
{
+ const int16_t *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
+ *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
int i;
if (uvalpha < 2048) {
YUV2PACKEDWRAPPER(yuv2, 422, yuyv422, PIX_FMT_YUYV422);
YUV2PACKEDWRAPPER(yuv2, 422, uyvy422, PIX_FMT_UYVY422);
-#define r_b ((target == PIX_FMT_RGB48LE || target == PIX_FMT_RGB48BE) ? r : b)
-#define b_r ((target == PIX_FMT_RGB48LE || target == PIX_FMT_RGB48BE) ? b : r)
+#define R_B ((target == PIX_FMT_RGB48LE || target == PIX_FMT_RGB48BE) ? R : B)
+#define B_R ((target == PIX_FMT_RGB48LE || target == PIX_FMT_RGB48BE) ? B : R)
+#define output_pixel(pos, val) \
+ if (isBE(target)) { \
+ AV_WB16(pos, val); \
+ } else { \
+ AV_WL16(pos, val); \
+ }
static av_always_inline void
yuv2rgb48_X_c_template(SwsContext *c, const int16_t *lumFilter,
- const int16_t **lumSrc, int lumFilterSize,
- const int16_t *chrFilter, const int16_t **chrUSrc,
- const int16_t **chrVSrc, int chrFilterSize,
- const int16_t **alpSrc, uint8_t *dest, int dstW,
+ const int32_t **lumSrc, int lumFilterSize,
+ const int16_t *chrFilter, const int32_t **chrUSrc,
+ const int32_t **chrVSrc, int chrFilterSize,
+ const int32_t **alpSrc, uint16_t *dest, int dstW,
int y, enum PixelFormat target)
{
int i;
for (i = 0; i < (dstW >> 1); i++) {
int j;
- int Y1 = 1 << 18;
- int Y2 = 1 << 18;
- int U = 1 << 18;
- int V = 1 << 18;
- const uint8_t *r, *g, *b;
+ int Y1 = 0;
+ int Y2 = 0;
+ int U = -128 << 23; // 19
+ int V = -128 << 23;
+ int R, G, B;
for (j = 0; j < lumFilterSize; j++) {
Y1 += lumSrc[j][i * 2] * lumFilter[j];
U += chrUSrc[j][i] * chrFilter[j];
V += chrVSrc[j][i] * chrFilter[j];
}
- Y1 >>= 19;
- Y2 >>= 19;
- U >>= 19;
- V >>= 19;
- if ((Y1 | Y2 | U | V) & 0x100) {
- Y1 = av_clip_uint8(Y1);
- Y2 = av_clip_uint8(Y2);
- U = av_clip_uint8(U);
- V = av_clip_uint8(V);
- }
- /* FIXME fix tables so that clipping is not needed and then use _NOCLIP*/
- r = (const uint8_t *) c->table_rV[V];
- g = (const uint8_t *)(c->table_gU[U] + c->table_gV[V]);
- b = (const uint8_t *) c->table_bU[U];
+ // 8bit: 12+15=27; 16-bit: 12+19=31
+ Y1 >>= 14; // 10
+ Y2 >>= 14;
+ U >>= 14;
+ V >>= 14;
- dest[ 0] = dest[ 1] = r_b[Y1];
- dest[ 2] = dest[ 3] = g[Y1];
- dest[ 4] = dest[ 5] = b_r[Y1];
- dest[ 6] = dest[ 7] = r_b[Y2];
- dest[ 8] = dest[ 9] = g[Y2];
- dest[10] = dest[11] = b_r[Y2];
- dest += 12;
+ // 8bit: 27 -> 17bit, 16bit: 31 - 14 = 17bit
+ Y1 -= c->yuv2rgb_y_offset;
+ Y2 -= c->yuv2rgb_y_offset;
+ Y1 *= c->yuv2rgb_y_coeff;
+ Y2 *= c->yuv2rgb_y_coeff;
+ Y1 += 1 << 13; // 21
+ Y2 += 1 << 13;
+ // 8bit: 17 + 13bit = 30bit, 16bit: 17 + 13bit = 30bit
+
+ R = V * c->yuv2rgb_v2r_coeff;
+ G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
+ B = U * c->yuv2rgb_u2b_coeff;
+
+ // 8bit: 30 - 22 = 8bit, 16bit: 30bit - 14 = 16bit
+ output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
+ output_pixel(&dest[1], av_clip_uintp2( G + Y1, 30) >> 14);
+ output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
+ output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
+ output_pixel(&dest[4], av_clip_uintp2( G + Y2, 30) >> 14);
+ output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
+ dest += 6;
}
}
static av_always_inline void
-yuv2rgb48_2_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *buf1, const uint16_t *ubuf0,
- const uint16_t *ubuf1, const uint16_t *vbuf0,
- const uint16_t *vbuf1, const uint16_t *abuf0,
- const uint16_t *abuf1, uint8_t *dest, int dstW,
+yuv2rgb48_2_c_template(SwsContext *c, const int32_t *buf[2],
+ const int32_t *ubuf[2], const int32_t *vbuf[2],
+ const int32_t *abuf[2], uint16_t *dest, int dstW,
int yalpha, int uvalpha, int y,
enum PixelFormat target)
{
+ const int32_t *buf0 = buf[0], *buf1 = buf[1],
+ *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
+ *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
int yalpha1 = 4095 - yalpha;
int uvalpha1 = 4095 - uvalpha;
int i;
for (i = 0; i < (dstW >> 1); i++) {
- int Y1 = (buf0[i * 2] * yalpha1 + buf1[i * 2] * yalpha) >> 19;
- int Y2 = (buf0[i * 2 + 1] * yalpha1 + buf1[i * 2 + 1] * yalpha) >> 19;
- int U = (ubuf0[i] * uvalpha1 + ubuf1[i] * uvalpha) >> 19;
- int V = (vbuf0[i] * uvalpha1 + vbuf1[i] * uvalpha) >> 19;
- const uint8_t *r = (const uint8_t *) c->table_rV[V],
- *g = (const uint8_t *)(c->table_gU[U] + c->table_gV[V]),
- *b = (const uint8_t *) c->table_bU[U];
+ int Y1 = (buf0[i * 2] * yalpha1 + buf1[i * 2] * yalpha) >> 14;
+ int Y2 = (buf0[i * 2 + 1] * yalpha1 + buf1[i * 2 + 1] * yalpha) >> 14;
+ int U = (ubuf0[i] * uvalpha1 + ubuf1[i] * uvalpha + (-128 << 23)) >> 14;
+ int V = (vbuf0[i] * uvalpha1 + vbuf1[i] * uvalpha + (-128 << 23)) >> 14;
+ int R, G, B;
- dest[ 0] = dest[ 1] = r_b[Y1];
- dest[ 2] = dest[ 3] = g[Y1];
- dest[ 4] = dest[ 5] = b_r[Y1];
- dest[ 6] = dest[ 7] = r_b[Y2];
- dest[ 8] = dest[ 9] = g[Y2];
- dest[10] = dest[11] = b_r[Y2];
- dest += 12;
+ Y1 -= c->yuv2rgb_y_offset;
+ Y2 -= c->yuv2rgb_y_offset;
+ Y1 *= c->yuv2rgb_y_coeff;
+ Y2 *= c->yuv2rgb_y_coeff;
+ Y1 += 1 << 13;
+ Y2 += 1 << 13;
+
+ R = V * c->yuv2rgb_v2r_coeff;
+ G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
+ B = U * c->yuv2rgb_u2b_coeff;
+
+ output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
+ output_pixel(&dest[1], av_clip_uintp2( G + Y1, 30) >> 14);
+ output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
+ output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
+ output_pixel(&dest[4], av_clip_uintp2( G + Y2, 30) >> 14);
+ output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
+ dest += 6;
}
}
static av_always_inline void
-yuv2rgb48_1_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *ubuf0, const uint16_t *ubuf1,
- const uint16_t *vbuf0, const uint16_t *vbuf1,
- const uint16_t *abuf0, uint8_t *dest, int dstW,
- int uvalpha, enum PixelFormat dstFormat,
- int flags, int y, enum PixelFormat target)
+yuv2rgb48_1_c_template(SwsContext *c, const int32_t *buf0,
+ const int32_t *ubuf[2], const int32_t *vbuf[2],
+ const int32_t *abuf0, uint16_t *dest, int dstW,
+ int uvalpha, int y, enum PixelFormat target)
{
+ const int32_t *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
+ *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
int i;
if (uvalpha < 2048) {
for (i = 0; i < (dstW >> 1); i++) {
- int Y1 = buf0[i * 2] >> 7;
- int Y2 = buf0[i * 2 + 1] >> 7;
- int U = ubuf1[i] >> 7;
- int V = vbuf1[i] >> 7;
- const uint8_t *r = (const uint8_t *) c->table_rV[V],
- *g = (const uint8_t *)(c->table_gU[U] + c->table_gV[V]),
- *b = (const uint8_t *) c->table_bU[U];
-
- dest[ 0] = dest[ 1] = r_b[Y1];
- dest[ 2] = dest[ 3] = g[Y1];
- dest[ 4] = dest[ 5] = b_r[Y1];
- dest[ 6] = dest[ 7] = r_b[Y2];
- dest[ 8] = dest[ 9] = g[Y2];
- dest[10] = dest[11] = b_r[Y2];
- dest += 12;
+ int Y1 = (buf0[i * 2] ) >> 2;
+ int Y2 = (buf0[i * 2 + 1]) >> 2;
+ int U = (ubuf0[i] + (-128 << 11)) >> 2;
+ int V = (vbuf0[i] + (-128 << 11)) >> 2;
+ int R, G, B;
+
+ Y1 -= c->yuv2rgb_y_offset;
+ Y2 -= c->yuv2rgb_y_offset;
+ Y1 *= c->yuv2rgb_y_coeff;
+ Y2 *= c->yuv2rgb_y_coeff;
+ Y1 += 1 << 13;
+ Y2 += 1 << 13;
+
+ R = V * c->yuv2rgb_v2r_coeff;
+ G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
+ B = U * c->yuv2rgb_u2b_coeff;
+
+ output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
+ output_pixel(&dest[1], av_clip_uintp2( G + Y1, 30) >> 14);
+ output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
+ output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
+ output_pixel(&dest[4], av_clip_uintp2( G + Y2, 30) >> 14);
+ output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
+ dest += 6;
}
} else {
for (i = 0; i < (dstW >> 1); i++) {
- int Y1 = buf0[i * 2] >> 7;
- int Y2 = buf0[i * 2 + 1] >> 7;
- int U = (ubuf0[i] + ubuf1[i]) >> 8;
- int V = (vbuf0[i] + vbuf1[i]) >> 8;
- const uint8_t *r = (const uint8_t *) c->table_rV[V],
- *g = (const uint8_t *)(c->table_gU[U] + c->table_gV[V]),
- *b = (const uint8_t *) c->table_bU[U];
-
- dest[ 0] = dest[ 1] = r_b[Y1];
- dest[ 2] = dest[ 3] = g[Y1];
- dest[ 4] = dest[ 5] = b_r[Y1];
- dest[ 6] = dest[ 7] = r_b[Y2];
- dest[ 8] = dest[ 9] = g[Y2];
- dest[10] = dest[11] = b_r[Y2];
- dest += 12;
+ 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 R, G, B;
+
+ Y1 -= c->yuv2rgb_y_offset;
+ Y2 -= c->yuv2rgb_y_offset;
+ Y1 *= c->yuv2rgb_y_coeff;
+ Y2 *= c->yuv2rgb_y_coeff;
+ Y1 += 1 << 13;
+ Y2 += 1 << 13;
+
+ R = V * c->yuv2rgb_v2r_coeff;
+ G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
+ B = U * c->yuv2rgb_u2b_coeff;
+
+ output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
+ output_pixel(&dest[1], av_clip_uintp2( G + Y1, 30) >> 14);
+ output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
+ output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
+ output_pixel(&dest[4], av_clip_uintp2( G + Y2, 30) >> 14);
+ output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
+ dest += 6;
}
}
}
+#undef output_pixel
#undef r_b
#undef b_r
-YUV2PACKEDWRAPPER(yuv2, rgb48, rgb48be, PIX_FMT_RGB48BE);
-//YUV2PACKEDWRAPPER(yuv2, rgb48, rgb48le, PIX_FMT_RGB48LE);
-YUV2PACKEDWRAPPER(yuv2, rgb48, bgr48be, PIX_FMT_BGR48BE);
-//YUV2PACKEDWRAPPER(yuv2, rgb48, bgr48le, PIX_FMT_BGR48LE);
+YUV2PACKED16WRAPPER(yuv2, rgb48, rgb48be, PIX_FMT_RGB48BE);
+YUV2PACKED16WRAPPER(yuv2, rgb48, rgb48le, PIX_FMT_RGB48LE);
+YUV2PACKED16WRAPPER(yuv2, rgb48, bgr48be, PIX_FMT_BGR48BE);
+YUV2PACKED16WRAPPER(yuv2, rgb48, bgr48le, PIX_FMT_BGR48LE);
static av_always_inline void
yuv2rgb_write(uint8_t *_dest, int i, int Y1, int Y2,
const uint32_t *b = (const uint32_t *) _b;
#if CONFIG_SMALL
- int sh = hasAlpha ? ((fmt == PIX_FMT_RGB32_1 || fmt == PIX_FMT_BGR32_1) ? 0 : 24) : 0;
+ int sh = hasAlpha ? ((target == PIX_FMT_RGB32_1 || target == PIX_FMT_BGR32_1) ? 0 : 24) : 0;
dest[i * 2 + 0] = r[Y1] + g[Y1] + b[Y1] + (hasAlpha ? A1 << sh : 0);
dest[i * 2 + 1] = r[Y2] + g[Y2] + b[Y2] + (hasAlpha ? A2 << sh : 0);
U = av_clip_uint8(U);
V = av_clip_uint8(V);
}
- if (hasAlpha) {\
+ if (hasAlpha) {
A1 = 1 << 18;
A2 = 1 << 18;
for (j = 0; j < lumFilterSize; j++) {
}
static av_always_inline void
-yuv2rgb_2_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *buf1, const uint16_t *ubuf0,
- const uint16_t *ubuf1, const uint16_t *vbuf0,
- const uint16_t *vbuf1, const uint16_t *abuf0,
- const uint16_t *abuf1, uint8_t *dest, int dstW,
+yuv2rgb_2_c_template(SwsContext *c, const int16_t *buf[2],
+ const int16_t *ubuf[2], const int16_t *vbuf[2],
+ const int16_t *abuf[2], uint8_t *dest, int dstW,
int yalpha, int uvalpha, int y,
enum PixelFormat target, int hasAlpha)
{
+ const int16_t *buf0 = buf[0], *buf1 = buf[1],
+ *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
+ *vbuf0 = vbuf[0], *vbuf1 = vbuf[1],
+ *abuf0 = hasAlpha ? abuf[0] : NULL,
+ *abuf1 = hasAlpha ? abuf[1] : NULL;
int yalpha1 = 4095 - yalpha;
int uvalpha1 = 4095 - uvalpha;
int i;
}
static av_always_inline void
-yuv2rgb_1_c_template(SwsContext *c, const uint16_t *buf0,
- const uint16_t *ubuf0, const uint16_t *ubuf1,
- const uint16_t *vbuf0, const uint16_t *vbuf1,
- const uint16_t *abuf0, uint8_t *dest, int dstW,
- int uvalpha, enum PixelFormat dstFormat,
- int flags, int y, enum PixelFormat target,
+yuv2rgb_1_c_template(SwsContext *c, const int16_t *buf0,
+ const int16_t *ubuf[2], const int16_t *vbuf[2],
+ const int16_t *abuf0, uint8_t *dest, int dstW,
+ int uvalpha, int y, enum PixelFormat target,
int hasAlpha)
{
+ const int16_t *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
+ *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
int i;
if (uvalpha < 2048) {
}
#define YUV2RGBWRAPPER(name, base, ext, fmt, hasAlpha) \
YUV2RGBWRAPPERX(name, base, ext, fmt, hasAlpha) \
-static void name ## ext ## _2_c(SwsContext *c, const uint16_t *buf0, \
- const uint16_t *buf1, const uint16_t *ubuf0, \
- const uint16_t *ubuf1, const uint16_t *vbuf0, \
- const uint16_t *vbuf1, const uint16_t *abuf0, \
- const uint16_t *abuf1, uint8_t *dest, int dstW, \
+static void name ## ext ## _2_c(SwsContext *c, const int16_t *buf[2], \
+ const int16_t *ubuf[2], const int16_t *vbuf[2], \
+ const int16_t *abuf[2], uint8_t *dest, int dstW, \
int yalpha, int uvalpha, int y) \
{ \
- name ## base ## _2_c_template(c, buf0, buf1, ubuf0, ubuf1, \
- vbuf0, vbuf1, abuf0, abuf1, \
+ name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
dest, dstW, yalpha, uvalpha, y, fmt, hasAlpha); \
} \
\
-static void name ## ext ## _1_c(SwsContext *c, const uint16_t *buf0, \
- const uint16_t *ubuf0, const uint16_t *ubuf1, \
- const uint16_t *vbuf0, const uint16_t *vbuf1, \
- const uint16_t *abuf0, uint8_t *dest, int dstW, \
- int uvalpha, enum PixelFormat dstFormat, \
- int flags, int y) \
+static void name ## ext ## _1_c(SwsContext *c, const int16_t *buf0, \
+ const int16_t *ubuf[2], const int16_t *vbuf[2], \
+ const int16_t *abuf0, uint8_t *dest, int dstW, \
+ int uvalpha, int y) \
{ \
- name ## base ## _1_c_template(c, buf0, ubuf0, ubuf1, vbuf0, \
- vbuf1, abuf0, dest, dstW, uvalpha, \
- dstFormat, flags, y, fmt, hasAlpha); \
+ name ## base ## _1_c_template(c, buf0, ubuf, vbuf, abuf0, dest, \
+ dstW, uvalpha, y, fmt, hasAlpha); \
}
#if CONFIG_SMALL
#define b ((origin == PIX_FMT_BGR48BE || origin == PIX_FMT_BGR48LE) ? r_b : b_r)
static av_always_inline void
-rgb48ToY_c_template(uint8_t *dst, const uint8_t *src, int width,
+rgb48ToY_c_template(uint16_t *dst, const uint16_t *src, int width,
enum PixelFormat origin)
{
int i;
for (i = 0; i < width; i++) {
- int r_b = input_pixel(&src[i*6+0]) >> 8;
- int g = input_pixel(&src[i*6+2]) >> 8;
- int b_r = input_pixel(&src[i*6+4]) >> 8;
+ unsigned int r_b = input_pixel(&src[i*3+0]);
+ unsigned int g = input_pixel(&src[i*3+1]);
+ unsigned int b_r = input_pixel(&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 + (0x2001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
}
}
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(uint16_t *dstU, uint16_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 r_b = input_pixel(&src1[i*6+0]) >> 8;
- int g = input_pixel(&src1[i*6+2]) >> 8;
- int b_r = input_pixel(&src1[i*6+4]) >> 8;
+ int r_b = input_pixel(&src1[i*3+0]);
+ int g = input_pixel(&src1[i*3+1]);
+ int b_r = input_pixel(&src1[i*3+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 + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
+ dstV[i] = (RV*r + GV*g + BV*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
}
}
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(uint16_t *dstU, uint16_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 r_b = (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 b_r = (input_pixel(&src1[12*i + 4]) >> 8) + (input_pixel(&src1[12*i + 10]) >> 8);
+ int r_b = (input_pixel(&src1[6 * i + 0]) + input_pixel(&src1[6 * i + 3]) + 1) >> 1;
+ int g = (input_pixel(&src1[6 * i + 1]) + input_pixel(&src1[6 * i + 4]) + 1) >> 1;
+ int b_r = (input_pixel(&src1[6 * i + 2]) + input_pixel(&src1[6 * i + 5]) + 1) >> 1;
- 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 + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
+ dstV[i]= (RV*r + GV*g + BV*b + (0x10001<<(RGB2YUV_SHIFT-1))) >> RGB2YUV_SHIFT;
}
}
#undef input_pixel
#define rgb48funcs(pattern, BE_LE, origin) \
-static void pattern ## 48 ## BE_LE ## ToY_c(uint8_t *dst, const uint8_t *src, \
+static void pattern ## 48 ## BE_LE ## ToY_c(uint8_t *_dst, const uint8_t *_src, \
int width, uint32_t *unused) \
{ \
+ const uint16_t *src = (const uint16_t *) _src; \
+ uint16_t *dst = (uint16_t *) _dst; \
rgb48ToY_c_template(dst, src, width, origin); \
} \
\
-static void pattern ## 48 ## BE_LE ## ToUV_c(uint8_t *dstU, uint8_t *dstV, \
- const uint8_t *src1, const uint8_t *src2, \
+static void pattern ## 48 ## BE_LE ## ToUV_c(uint8_t *_dstU, uint8_t *_dstV, \
+ const uint8_t *_src1, const uint8_t *_src2, \
int width, uint32_t *unused) \
{ \
+ const uint16_t *src1 = (const uint16_t *) _src1, \
+ *src2 = (const uint16_t *) _src2; \
+ uint16_t *dstU = (uint16_t *) _dstU, *dstV = (uint16_t *) _dstV; \
rgb48ToUV_c_template(dstU, dstV, src1, src2, width, origin); \
} \
\
-static void pattern ## 48 ## BE_LE ## ToUV_half_c(uint8_t *dstU, uint8_t *dstV, \
- const uint8_t *src1, const uint8_t *src2, \
+static void pattern ## 48 ## BE_LE ## ToUV_half_c(uint8_t *_dstU, uint8_t *_dstV, \
+ const uint8_t *_src1, const uint8_t *_src2, \
int width, uint32_t *unused) \
{ \
+ const uint16_t *src1 = (const uint16_t *) _src1, \
+ *src2 = (const uint16_t *) _src2; \
+ uint16_t *dstU = (uint16_t *) _dstU, *dstV = (uint16_t *) _dstV; \
rgb48ToUV_half_c_template(dstU, dstV, src1, src2, width, origin); \
}
assert(src1 == src2);
}
-static void LEToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
- const uint8_t *src2, int width, uint32_t *unused)
+static void bswap16Y_c(uint8_t *_dst, const uint8_t *_src, int width, uint32_t *unused)
{
int i;
+ const uint16_t *src = (const uint16_t *) _src;
+ uint16_t *dst = (uint16_t *) _dst;
for (i=0; i<width; i++) {
- dstU[i]= src1[2*i + 1];
- dstV[i]= src2[2*i + 1];
+ dst[i] = av_bswap16(src[i]);
+ }
+}
+
+static void bswap16UV_c(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *_src1,
+ const uint8_t *_src2, int width, uint32_t *unused)
+{
+ int i;
+ const uint16_t *src1 = (const uint16_t *) _src1,
+ *src2 = (const uint16_t *) _src2;
+ uint16_t *dstU = (uint16_t *) _dstU, *dstV = (uint16_t *) _dstV;
+ for (i=0; i<width; i++) {
+ dstU[i] = av_bswap16(src1[i]);
+ dstV[i] = av_bswap16(src2[i]);
}
}
assert(src1 == src2);
}
-static void BEToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
- const uint8_t *src2, int width, uint32_t *unused)
-{
- int i;
- for (i=0; i<width; i++) {
- dstU[i]= src1[2*i];
- dstV[i]= src2[2*i];
- }
-}
-
static av_always_inline void nvXXtoUV_c(uint8_t *dst1, uint8_t *dst2,
const uint8_t *src, int width)
{
#define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))
-// FIXME Maybe dither instead.
-static av_always_inline void
-yuv9_OR_10ToUV_c_template(uint8_t *dstU, uint8_t *dstV,
- const uint8_t *_srcU, const uint8_t *_srcV,
- int width, enum PixelFormat origin, int depth)
-{
- int i;
- const uint16_t *srcU = (const uint16_t *) _srcU;
- const uint16_t *srcV = (const uint16_t *) _srcV;
-
- for (i = 0; i < width; i++) {
- dstU[i] = input_pixel(&srcU[i]) >> (depth - 8);
- dstV[i] = input_pixel(&srcV[i]) >> (depth - 8);
- }
-}
-
-static av_always_inline void
-yuv9_or_10ToY_c_template(uint8_t *dstY, const uint8_t *_srcY,
- int width, enum PixelFormat origin, int depth)
-{
- int i;
- const uint16_t *srcY = (const uint16_t*)_srcY;
-
- for (i = 0; i < width; i++)
- dstY[i] = input_pixel(&srcY[i]) >> (depth - 8);
-}
-
-#undef input_pixel
-
-#define YUV_NBPS(depth, BE_LE, origin) \
-static void BE_LE ## depth ## ToUV_c(uint8_t *dstU, uint8_t *dstV, \
- const uint8_t *srcU, const uint8_t *srcV, \
- int width, uint32_t *unused) \
-{ \
- yuv9_OR_10ToUV_c_template(dstU, dstV, srcU, srcV, width, origin, depth); \
-} \
-static void BE_LE ## depth ## ToY_c(uint8_t *dstY, const uint8_t *srcY, \
- int width, uint32_t *unused) \
-{ \
- yuv9_or_10ToY_c_template(dstY, srcY, width, origin, depth); \
-}
-
-YUV_NBPS( 9, LE, PIX_FMT_YUV420P9LE);
-YUV_NBPS( 9, BE, PIX_FMT_YUV420P9BE);
-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,
int width, uint32_t *unused)
{
}
}
+static void hScale16_c(SwsContext *c, int16_t *_dst, int dstW, const uint8_t *_src,
+ const int16_t *filter,
+ const int16_t *filterPos, int filterSize)
+{
+ int i;
+ int32_t *dst = (int32_t *) _dst;
+ const uint16_t *src = (const uint16_t *) _src;
+ int bits = av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1;
+ int sh = (bits <= 7) ? 11 : (bits - 4);
+
+ for (i = 0; i < dstW; i++) {
+ int j;
+ int srcPos = filterPos[i];
+ int val = 0;
+
+ for (j = 0; j < filterSize; j++) {
+ val += src[srcPos + j] * filter[filterSize * i + j];
+ }
+ // filter=14 bit, input=16 bit, output=30 bit, >> 11 makes 19 bit
+ dst[i] = FFMIN(val >> sh, (1 << 19) - 1);
+ }
+}
+
// bilinear / bicubic scaling
-static void hScale_c(int16_t *dst, int dstW, const uint8_t *src,
+static void hScale_c(SwsContext *c, int16_t *dst, int dstW, const uint8_t *src,
const int16_t *filter, const int16_t *filterPos,
int filterSize)
{
dst[i] = (dst[i]*14071 + 33561947)>>14;
}
+static void chrRangeToJpeg16_c(int16_t *_dstU, int16_t *_dstV, int width)
+{
+ int i;
+ int32_t *dstU = (int32_t *) _dstU;
+ int32_t *dstV = (int32_t *) _dstV;
+ for (i = 0; i < width; i++) {
+ dstU[i] = (FFMIN(dstU[i],30775<<4)*4663 - (9289992<<4))>>12; //-264
+ dstV[i] = (FFMIN(dstV[i],30775<<4)*4663 - (9289992<<4))>>12; //-264
+ }
+}
+static void chrRangeFromJpeg16_c(int16_t *_dstU, int16_t *_dstV, int width)
+{
+ int i;
+ int32_t *dstU = (int32_t *) _dstU;
+ int32_t *dstV = (int32_t *) _dstV;
+ for (i = 0; i < width; i++) {
+ dstU[i] = (dstU[i]*1799 + (4081085<<4))>>11; //1469
+ dstV[i] = (dstV[i]*1799 + (4081085<<4))>>11; //1469
+ }
+}
+static void lumRangeToJpeg16_c(int16_t *_dst, int width)
+{
+ int i;
+ int32_t *dst = (int32_t *) _dst;
+ for (i = 0; i < width; i++)
+ dst[i] = (FFMIN(dst[i],30189<<4)*4769 - (39057361<<2))>>12;
+}
+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;
+}
+
static void hyscale_fast_c(SwsContext *c, int16_t *dst, int dstWidth,
const uint8_t *src, int srcW, int xInc)
{
}
}
+static void scale8To16Rv_c(uint16_t *_dst, const uint8_t *src, int len)
+{
+ int i;
+ uint8_t *dst = (uint8_t *) _dst;
+ for (i = len - 1; i >= 0; i--) {
+ dst[i * 2] = dst[i * 2 + 1] = src[i];
+ }
+}
+
+static void scale19To15Fw_c(int16_t *dst, const int32_t *src, int len)
+{
+ int i;
+ for (i = 0; i < len; i++) {
+ dst[i] = src[i] >> 4;
+ }
+}
+
// *** horizontal scale Y line to temp buffer
-static av_always_inline void hyscale(SwsContext *c, uint16_t *dst, int dstWidth,
+static av_always_inline void hyscale(SwsContext *c, int16_t *dst, int dstWidth,
const uint8_t *src, int srcW, int xInc,
const int16_t *hLumFilter,
const int16_t *hLumFilterPos, int hLumFilterSize,
src= formatConvBuffer;
}
+ if (av_pix_fmt_descriptors[c->srcFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16) {
+ c->scale8To16Rv((uint16_t *) formatConvBuffer, src, srcW);
+ src = formatConvBuffer;
+ }
+
if (!c->hyscale_fast) {
- c->hScale(dst, dstWidth, src, hLumFilter, hLumFilterPos, hLumFilterSize);
+ c->hScale(c, dst, dstWidth, src, hLumFilter, hLumFilterPos, hLumFilterSize);
} else { // fast bilinear upscale / crap downscale
c->hyscale_fast(c, dst, dstWidth, src, srcW, xInc);
}
if (convertRange)
convertRange(dst, dstWidth);
+
+ if (av_pix_fmt_descriptors[c->dstFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16) {
+ c->scale19To15Fw(dst, (int32_t *) dst, dstWidth);
+ }
}
static void hcscale_fast_c(SwsContext *c, int16_t *dst1, int16_t *dst2,
}
}
-static av_always_inline void hcscale(SwsContext *c, uint16_t *dst1, uint16_t *dst2, int dstWidth,
+static av_always_inline void hcscale(SwsContext *c, int16_t *dst1, int16_t *dst2, int dstWidth,
const uint8_t *src1, const uint8_t *src2,
int srcW, int xInc, const int16_t *hChrFilter,
const int16_t *hChrFilterPos, int hChrFilterSize,
uint8_t *formatConvBuffer, uint32_t *pal)
{
if (c->chrToYV12) {
- uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW, 16);
+ uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW * c->scalingBpp >> 3, 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));
+ c->scale8To16Rv((uint16_t *) formatConvBuffer, src1, srcW);
+ c->scale8To16Rv((uint16_t *) buf2, src2, srcW);
+ src1 = formatConvBuffer;
+ src2 = buf2;
+ }
+
if (!c->hcscale_fast) {
- c->hScale(dst1, dstWidth, src1, hChrFilter, hChrFilterPos, hChrFilterSize);
- c->hScale(dst2, dstWidth, src2, hChrFilter, hChrFilterPos, hChrFilterSize);
+ c->hScale(c, dst1, dstWidth, src1, hChrFilter, hChrFilterPos, hChrFilterSize);
+ c->hScale(c, dst2, dstWidth, src2, hChrFilter, hChrFilterPos, hChrFilterSize);
} else { // fast bilinear upscale / crap downscale
c->hcscale_fast(c, dst1, dst2, dstWidth, src1, src2, srcW, xInc);
}
if (c->chrConvertRange)
c->chrConvertRange(dst1, dst2, dstWidth);
+
+ if (av_pix_fmt_descriptors[c->dstFormat].comp[0].depth_minus1 < 8 && c->scalingBpp == 16) {
+ c->scale19To15Fw(dst1, (int32_t *) dst1, dstWidth);
+ c->scale19To15Fw(dst2, (int32_t *) dst2, dstWidth);
+ }
}
static av_always_inline void
*yuv2packedX = yuv2uyvy422_X_c;
break;
case PIX_FMT_RGB48LE:
- //*yuv2packed1 = yuv2rgb48le_1_c;
- //*yuv2packed2 = yuv2rgb48le_2_c;
- //*yuv2packedX = yuv2rgb48le_X_c;
- //break;
+ *yuv2packed1 = yuv2rgb48le_1_c;
+ *yuv2packed2 = yuv2rgb48le_2_c;
+ *yuv2packedX = yuv2rgb48le_X_c;
+ break;
case PIX_FMT_RGB48BE:
*yuv2packed1 = yuv2rgb48be_1_c;
*yuv2packed2 = yuv2rgb48be_2_c;
*yuv2packedX = yuv2rgb48be_X_c;
break;
case PIX_FMT_BGR48LE:
- //*yuv2packed1 = yuv2bgr48le_1_c;
- //*yuv2packed2 = yuv2bgr48le_2_c;
- //*yuv2packedX = yuv2bgr48le_X_c;
- //break;
+ *yuv2packed1 = yuv2bgr48le_1_c;
+ *yuv2packed2 = yuv2bgr48le_2_c;
+ *yuv2packedX = yuv2bgr48le_X_c;
+ break;
case PIX_FMT_BGR48BE:
*yuv2packed1 = yuv2bgr48be_1_c;
*yuv2packed2 = yuv2bgr48be_2_c;
*yuv2packed2 = yuv2bgr24_2_c;
*yuv2packedX = yuv2bgr24_X_c;
break;
- case PIX_FMT_RGB565:
- case PIX_FMT_BGR565:
+ case PIX_FMT_RGB565LE:
+ case PIX_FMT_RGB565BE:
+ case PIX_FMT_BGR565LE:
+ case PIX_FMT_BGR565BE:
*yuv2packed1 = yuv2rgb16_1_c;
*yuv2packed2 = yuv2rgb16_2_c;
*yuv2packedX = yuv2rgb16_X_c;
break;
- case PIX_FMT_RGB555:
- case PIX_FMT_BGR555:
+ case PIX_FMT_RGB555LE:
+ case PIX_FMT_RGB555BE:
+ case PIX_FMT_BGR555LE:
+ case PIX_FMT_BGR555BE:
*yuv2packed1 = yuv2rgb15_1_c;
*yuv2packed2 = yuv2rgb15_2_c;
*yuv2packedX = yuv2rgb15_X_c;
break;
- case PIX_FMT_RGB444:
- case PIX_FMT_BGR444:
+ case PIX_FMT_RGB444LE:
+ case PIX_FMT_RGB444BE:
+ case PIX_FMT_BGR444LE:
+ case PIX_FMT_BGR444BE:
*yuv2packed1 = yuv2rgb12_1_c;
*yuv2packed2 = yuv2rgb12_2_c;
*yuv2packedX = yuv2rgb12_X_c;
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;
lastInChrBuf= -1;
}
+ if (!should_dither) {
+ c->chrDither8 = c->lumDither8 = ff_sws_pb_64;
+ }
lastDstY= dstY;
for (;dstY < dstH; dstY++) {
- unsigned char *dest =dst[0]+dstStride[0]*dstY;
const int chrDstY= dstY>>c->chrDstVSubSample;
- 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;
+ uint8_t *dest[4] = {
+ dst[0] + dstStride[0] * dstY,
+ dst[1] + dstStride[1] * chrDstY,
+ dst[2] + dstStride[2] * chrDstY,
+ (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? dst[3] + dstStride[3] * dstY : NULL,
+ };
const int firstLumSrcY= vLumFilterPos[dstY]; //First line needed as input
const int firstLumSrcY2= vLumFilterPos[FFMIN(dstY | ((1<<c->chrDstVSubSample) - 1), dstH-1)];
#if HAVE_MMX
updateMMXDitherTables(c, dstY, lumBufIndex, chrBufIndex, lastInLumBuf, lastInChrBuf);
#endif
+ if (should_dither) {
+ c->chrDither8 = dither_8x8_128[chrDstY & 7];
+ c->lumDither8 = dither_8x8_128[dstY & 7];
+ }
if (dstY >= dstH-2) {
// hmm looks like we can't use MMX here without overwriting this array's tail
find_c_packed_planar_out_funcs(c, &yuv2yuv1, &yuv2yuvX,
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
+ if ((dstY&chrSkipMask) || isGray(dstFormat))
+ dest[1] = dest[2] = NULL; //FIXME split functions in lumi / chromi
if (c->yuv2yuv1 && vLumFilterSize == 1 && vChrFilterSize == 1) { // unscaled YV12
- const int16_t *lumBuf = lumSrcPtr[0];
- const int16_t *chrUBuf= chrUSrcPtr[0];
- 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);
+ yuv2yuv1(c, lumSrcPtr[0], chrUSrcPtr[0], chrVSrcPtr[0], alpBuf,
+ dest, dstW, chrDstW);
} else { //General YV12
- yuv2yuvX(c,
- vLumFilter+dstY*vLumFilterSize , lumSrcPtr, vLumFilterSize,
- vChrFilter+chrDstY*vChrFilterSize, chrUSrcPtr,
- chrVSrcPtr, vChrFilterSize,
- alpSrcPtr, dest, uDest, vDest, aDest, dstW, chrDstW);
+ yuv2yuvX(c, vLumFilter + dstY * vLumFilterSize,
+ lumSrcPtr, vLumFilterSize,
+ vChrFilter + chrDstY * vChrFilterSize,
+ chrUSrcPtr, chrVSrcPtr, vChrFilterSize,
+ alpSrcPtr, dest, dstW, chrDstW);
}
} else {
assert(lumSrcPtr + vLumFilterSize - 1 < lumPixBuf + vLumBufSize*2);
assert(chrUSrcPtr + vChrFilterSize - 1 < chrUPixBuf + vChrBufSize*2);
if (c->yuv2packed1 && vLumFilterSize == 1 && vChrFilterSize == 2) { //unscaled RGB
- int chrAlpha= vChrFilter[2*dstY+1];
- yuv2packed1(c, *lumSrcPtr, *chrUSrcPtr, *(chrUSrcPtr+1),
- *chrVSrcPtr, *(chrVSrcPtr+1),
- alpPixBuf ? *alpSrcPtr : NULL,
- dest, dstW, chrAlpha, dstFormat, flags, dstY);
+ int chrAlpha = vChrFilter[2 * dstY + 1];
+ yuv2packed1(c, *lumSrcPtr, chrUSrcPtr, chrVSrcPtr,
+ alpPixBuf ? *alpSrcPtr : NULL,
+ dest[0], dstW, chrAlpha, dstY);
} else if (c->yuv2packed2 && vLumFilterSize == 2 && vChrFilterSize == 2) { //bilinear upscale RGB
- int lumAlpha= vLumFilter[2*dstY+1];
- int chrAlpha= vChrFilter[2*dstY+1];
- lumMmxFilter[2]=
- lumMmxFilter[3]= vLumFilter[2*dstY ]*0x10001;
- chrMmxFilter[2]=
- chrMmxFilter[3]= vChrFilter[2*chrDstY]*0x10001;
- yuv2packed2(c, *lumSrcPtr, *(lumSrcPtr+1), *chrUSrcPtr, *(chrUSrcPtr+1),
- *chrVSrcPtr, *(chrVSrcPtr+1),
- alpPixBuf ? *alpSrcPtr : NULL, alpPixBuf ? *(alpSrcPtr+1) : NULL,
- dest, dstW, lumAlpha, chrAlpha, dstY);
+ int lumAlpha = vLumFilter[2 * dstY + 1];
+ int chrAlpha = vChrFilter[2 * dstY + 1];
+ lumMmxFilter[2] =
+ lumMmxFilter[3] = vLumFilter[2 * dstY ] * 0x10001;
+ chrMmxFilter[2] =
+ chrMmxFilter[3] = vChrFilter[2 * chrDstY] * 0x10001;
+ yuv2packed2(c, lumSrcPtr, chrUSrcPtr, chrVSrcPtr,
+ alpPixBuf ? alpSrcPtr : NULL,
+ dest[0], dstW, lumAlpha, chrAlpha, dstY);
} else { //general RGB
- yuv2packedX(c,
- vLumFilter+dstY*vLumFilterSize, lumSrcPtr, vLumFilterSize,
- vChrFilter+dstY*vChrFilterSize, chrUSrcPtr, chrVSrcPtr, vChrFilterSize,
- alpSrcPtr, dest, dstW, dstY);
+ yuv2packedX(c, vLumFilter + dstY * vLumFilterSize,
+ lumSrcPtr, vLumFilterSize,
+ vChrFilter + dstY * vChrFilterSize,
+ chrUSrcPtr, chrVSrcPtr, vChrFilterSize,
+ alpSrcPtr, dest[0], dstW, dstY);
}
}
}
&c->yuv2packed1, &c->yuv2packed2,
&c->yuv2packedX);
- c->hScale = hScale_c;
-
- if (c->flags & SWS_FAST_BILINEAR) {
- c->hyscale_fast = hyscale_fast_c;
- c->hcscale_fast = hcscale_fast_c;
- }
-
c->chrToYV12 = NULL;
switch(srcFormat) {
case PIX_FMT_YUYV422 : c->chrToYV12 = yuy2ToUV_c; break;
case PIX_FMT_PAL8 :
case PIX_FMT_BGR4_BYTE:
case PIX_FMT_RGB4_BYTE: c->chrToYV12 = palToUV_c; break;
- case PIX_FMT_YUV444P9BE:
- case PIX_FMT_YUV420P9BE: c->chrToYV12 = BE9ToUV_c; break;
+#if HAVE_BIGENDIAN
case PIX_FMT_YUV444P9LE:
- case PIX_FMT_YUV420P9LE: c->chrToYV12 = LE9ToUV_c; break;
- case PIX_FMT_YUV444P10BE:
- case PIX_FMT_YUV422P10BE:
- case PIX_FMT_YUV420P10BE: c->chrToYV12 = BE10ToUV_c; break;
+ case PIX_FMT_YUV420P9LE:
case PIX_FMT_YUV422P10LE:
case PIX_FMT_YUV444P10LE:
- case PIX_FMT_YUV420P10LE: c->chrToYV12 = LE10ToUV_c; break;
- case PIX_FMT_YUV420P16BE:
- case PIX_FMT_YUV422P16BE:
- case PIX_FMT_YUV444P16BE: c->chrToYV12 = BEToUV_c; break;
+ 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->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;
+#endif
}
if (c->chrSrcHSubSample) {
switch(srcFormat) {
c->lumToYV12 = NULL;
c->alpToYV12 = NULL;
switch (srcFormat) {
- case PIX_FMT_YUV444P9BE:
- case PIX_FMT_YUV420P9BE: c->lumToYV12 = BE9ToY_c; break;
+#if HAVE_BIGENDIAN
case PIX_FMT_YUV444P9LE:
- case PIX_FMT_YUV420P9LE: c->lumToYV12 = LE9ToY_c; break;
- case PIX_FMT_YUV444P10BE:
- case PIX_FMT_YUV422P10BE:
- case PIX_FMT_YUV420P10BE: c->lumToYV12 = BE10ToY_c; break;
+ case PIX_FMT_YUV420P9LE:
case PIX_FMT_YUV444P10LE:
case PIX_FMT_YUV422P10LE:
- 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_UYVY422 :
+ case PIX_FMT_YUV420P10LE:
case PIX_FMT_YUV420P16LE:
case PIX_FMT_YUV422P16LE:
case PIX_FMT_YUV444P16LE:
- case PIX_FMT_GRAY16LE : c->lumToYV12 = uyvyToY_c; break;
+ 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:
+ case PIX_FMT_GRAY16BE: c->lumToYV12 = bswap16Y_c; break;
+#endif
+ case PIX_FMT_YUYV422 :
+ case PIX_FMT_Y400A : c->lumToYV12 = yuy2ToY_c; break;
+ case PIX_FMT_UYVY422 : 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;
}
}
+ if (c->scalingBpp == 8) {
+ c->hScale = hScale_c;
+ if (c->flags & SWS_FAST_BILINEAR) {
+ c->hyscale_fast = hyscale_fast_c;
+ c->hcscale_fast = hcscale_fast_c;
+ }
+
if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) {
if (c->srcRange) {
c->lumConvertRange = lumRangeFromJpeg_c;
c->chrConvertRange = chrRangeToJpeg_c;
}
}
+ } else {
+ c->hScale = hScale16_c;
+ c->scale19To15Fw = scale19To15Fw_c;
+ c->scale8To16Rv = scale8To16Rv_c;
+
+ if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) {
+ if (c->srcRange) {
+ c->lumConvertRange = lumRangeFromJpeg16_c;
+ c->chrConvertRange = chrRangeFromJpeg16_c;
+ } else {
+ c->lumConvertRange = lumRangeToJpeg16_c;
+ c->chrConvertRange = chrRangeToJpeg16_c;
+ }
+ }
+ }
if (!(isGray(srcFormat) || isGray(c->dstFormat) ||
srcFormat == PIX_FMT_MONOBLACK || srcFormat == PIX_FMT_MONOWHITE))
projects / ffmpeg / blobdiff