2 * Copyright (C) 2001-2003 Michael Niedermayer <michaelni@gmx.at>
4 * This file is part of FFmpeg.
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
23 #define _DEFAULT_SOURCE
24 #define _SVID_SOURCE // needed for MAP_ANONYMOUS
25 #define _DARWIN_C_SOURCE // needed for MAP_ANON
32 #if defined(MAP_ANON) && !defined(MAP_ANONYMOUS)
33 #define MAP_ANONYMOUS MAP_ANON
37 #define WIN32_LEAN_AND_MEAN
41 #include "libavutil/attributes.h"
42 #include "libavutil/avassert.h"
43 #include "libavutil/avutil.h"
44 #include "libavutil/bswap.h"
45 #include "libavutil/cpu.h"
46 #include "libavutil/imgutils.h"
47 #include "libavutil/intreadwrite.h"
48 #include "libavutil/libm.h"
49 #include "libavutil/mathematics.h"
50 #include "libavutil/opt.h"
51 #include "libavutil/pixdesc.h"
52 #include "libavutil/ppc/cpu.h"
53 #include "libavutil/x86/asm.h"
54 #include "libavutil/x86/cpu.h"
57 #include "swscale_internal.h"
59 static void handle_formats(SwsContext *c);
61 unsigned swscale_version(void)
63 av_assert0(LIBSWSCALE_VERSION_MICRO >= 100);
64 return LIBSWSCALE_VERSION_INT;
67 const char *swscale_configuration(void)
69 return FFMPEG_CONFIGURATION;
72 const char *swscale_license(void)
74 #define LICENSE_PREFIX "libswscale license: "
75 return LICENSE_PREFIX FFMPEG_LICENSE + sizeof(LICENSE_PREFIX) - 1;
78 typedef struct FormatEntry {
79 uint8_t is_supported_in :1;
80 uint8_t is_supported_out :1;
81 uint8_t is_supported_endianness :1;
84 static const FormatEntry format_entries[AV_PIX_FMT_NB] = {
85 [AV_PIX_FMT_YUV420P] = { 1, 1 },
86 [AV_PIX_FMT_YUYV422] = { 1, 1 },
87 [AV_PIX_FMT_RGB24] = { 1, 1 },
88 [AV_PIX_FMT_BGR24] = { 1, 1 },
89 [AV_PIX_FMT_YUV422P] = { 1, 1 },
90 [AV_PIX_FMT_YUV444P] = { 1, 1 },
91 [AV_PIX_FMT_YUV410P] = { 1, 1 },
92 [AV_PIX_FMT_YUV411P] = { 1, 1 },
93 [AV_PIX_FMT_GRAY8] = { 1, 1 },
94 [AV_PIX_FMT_MONOWHITE] = { 1, 1 },
95 [AV_PIX_FMT_MONOBLACK] = { 1, 1 },
96 [AV_PIX_FMT_PAL8] = { 1, 0 },
97 [AV_PIX_FMT_YUVJ420P] = { 1, 1 },
98 [AV_PIX_FMT_YUVJ411P] = { 1, 1 },
99 [AV_PIX_FMT_YUVJ422P] = { 1, 1 },
100 [AV_PIX_FMT_YUVJ444P] = { 1, 1 },
101 [AV_PIX_FMT_YVYU422] = { 1, 1 },
102 [AV_PIX_FMT_UYVY422] = { 1, 1 },
103 [AV_PIX_FMT_UYYVYY411] = { 0, 0 },
104 [AV_PIX_FMT_BGR8] = { 1, 1 },
105 [AV_PIX_FMT_BGR4] = { 0, 1 },
106 [AV_PIX_FMT_BGR4_BYTE] = { 1, 1 },
107 [AV_PIX_FMT_RGB8] = { 1, 1 },
108 [AV_PIX_FMT_RGB4] = { 0, 1 },
109 [AV_PIX_FMT_RGB4_BYTE] = { 1, 1 },
110 [AV_PIX_FMT_NV12] = { 1, 1 },
111 [AV_PIX_FMT_NV21] = { 1, 1 },
112 [AV_PIX_FMT_ARGB] = { 1, 1 },
113 [AV_PIX_FMT_RGBA] = { 1, 1 },
114 [AV_PIX_FMT_ABGR] = { 1, 1 },
115 [AV_PIX_FMT_BGRA] = { 1, 1 },
116 [AV_PIX_FMT_0RGB] = { 1, 1 },
117 [AV_PIX_FMT_RGB0] = { 1, 1 },
118 [AV_PIX_FMT_0BGR] = { 1, 1 },
119 [AV_PIX_FMT_BGR0] = { 1, 1 },
120 [AV_PIX_FMT_GRAY16BE] = { 1, 1 },
121 [AV_PIX_FMT_GRAY16LE] = { 1, 1 },
122 [AV_PIX_FMT_YUV440P] = { 1, 1 },
123 [AV_PIX_FMT_YUVJ440P] = { 1, 1 },
124 [AV_PIX_FMT_YUV440P10LE] = { 1, 1 },
125 [AV_PIX_FMT_YUV440P10BE] = { 1, 1 },
126 [AV_PIX_FMT_YUV440P12LE] = { 1, 1 },
127 [AV_PIX_FMT_YUV440P12BE] = { 1, 1 },
128 [AV_PIX_FMT_YUVA420P] = { 1, 1 },
129 [AV_PIX_FMT_YUVA422P] = { 1, 1 },
130 [AV_PIX_FMT_YUVA444P] = { 1, 1 },
131 [AV_PIX_FMT_YUVA420P9BE] = { 1, 1 },
132 [AV_PIX_FMT_YUVA420P9LE] = { 1, 1 },
133 [AV_PIX_FMT_YUVA422P9BE] = { 1, 1 },
134 [AV_PIX_FMT_YUVA422P9LE] = { 1, 1 },
135 [AV_PIX_FMT_YUVA444P9BE] = { 1, 1 },
136 [AV_PIX_FMT_YUVA444P9LE] = { 1, 1 },
137 [AV_PIX_FMT_YUVA420P10BE]= { 1, 1 },
138 [AV_PIX_FMT_YUVA420P10LE]= { 1, 1 },
139 [AV_PIX_FMT_YUVA422P10BE]= { 1, 1 },
140 [AV_PIX_FMT_YUVA422P10LE]= { 1, 1 },
141 [AV_PIX_FMT_YUVA444P10BE]= { 1, 1 },
142 [AV_PIX_FMT_YUVA444P10LE]= { 1, 1 },
143 [AV_PIX_FMT_YUVA420P16BE]= { 1, 1 },
144 [AV_PIX_FMT_YUVA420P16LE]= { 1, 1 },
145 [AV_PIX_FMT_YUVA422P16BE]= { 1, 1 },
146 [AV_PIX_FMT_YUVA422P16LE]= { 1, 1 },
147 [AV_PIX_FMT_YUVA444P16BE]= { 1, 1 },
148 [AV_PIX_FMT_YUVA444P16LE]= { 1, 1 },
149 [AV_PIX_FMT_RGB48BE] = { 1, 1 },
150 [AV_PIX_FMT_RGB48LE] = { 1, 1 },
151 [AV_PIX_FMT_RGBA64BE] = { 1, 1, 1 },
152 [AV_PIX_FMT_RGBA64LE] = { 1, 1, 1 },
153 [AV_PIX_FMT_RGB565BE] = { 1, 1 },
154 [AV_PIX_FMT_RGB565LE] = { 1, 1 },
155 [AV_PIX_FMT_RGB555BE] = { 1, 1 },
156 [AV_PIX_FMT_RGB555LE] = { 1, 1 },
157 [AV_PIX_FMT_BGR565BE] = { 1, 1 },
158 [AV_PIX_FMT_BGR565LE] = { 1, 1 },
159 [AV_PIX_FMT_BGR555BE] = { 1, 1 },
160 [AV_PIX_FMT_BGR555LE] = { 1, 1 },
161 [AV_PIX_FMT_YUV420P16LE] = { 1, 1 },
162 [AV_PIX_FMT_YUV420P16BE] = { 1, 1 },
163 [AV_PIX_FMT_YUV422P16LE] = { 1, 1 },
164 [AV_PIX_FMT_YUV422P16BE] = { 1, 1 },
165 [AV_PIX_FMT_YUV444P16LE] = { 1, 1 },
166 [AV_PIX_FMT_YUV444P16BE] = { 1, 1 },
167 [AV_PIX_FMT_RGB444LE] = { 1, 1 },
168 [AV_PIX_FMT_RGB444BE] = { 1, 1 },
169 [AV_PIX_FMT_BGR444LE] = { 1, 1 },
170 [AV_PIX_FMT_BGR444BE] = { 1, 1 },
171 [AV_PIX_FMT_YA8] = { 1, 1 },
172 [AV_PIX_FMT_YA16BE] = { 1, 0 },
173 [AV_PIX_FMT_YA16LE] = { 1, 0 },
174 [AV_PIX_FMT_BGR48BE] = { 1, 1 },
175 [AV_PIX_FMT_BGR48LE] = { 1, 1 },
176 [AV_PIX_FMT_BGRA64BE] = { 1, 1, 1 },
177 [AV_PIX_FMT_BGRA64LE] = { 1, 1, 1 },
178 [AV_PIX_FMT_YUV420P9BE] = { 1, 1 },
179 [AV_PIX_FMT_YUV420P9LE] = { 1, 1 },
180 [AV_PIX_FMT_YUV420P10BE] = { 1, 1 },
181 [AV_PIX_FMT_YUV420P10LE] = { 1, 1 },
182 [AV_PIX_FMT_YUV420P12BE] = { 1, 1 },
183 [AV_PIX_FMT_YUV420P12LE] = { 1, 1 },
184 [AV_PIX_FMT_YUV420P14BE] = { 1, 1 },
185 [AV_PIX_FMT_YUV420P14LE] = { 1, 1 },
186 [AV_PIX_FMT_YUV422P9BE] = { 1, 1 },
187 [AV_PIX_FMT_YUV422P9LE] = { 1, 1 },
188 [AV_PIX_FMT_YUV422P10BE] = { 1, 1 },
189 [AV_PIX_FMT_YUV422P10LE] = { 1, 1 },
190 [AV_PIX_FMT_YUV422P12BE] = { 1, 1 },
191 [AV_PIX_FMT_YUV422P12LE] = { 1, 1 },
192 [AV_PIX_FMT_YUV422P14BE] = { 1, 1 },
193 [AV_PIX_FMT_YUV422P14LE] = { 1, 1 },
194 [AV_PIX_FMT_YUV444P9BE] = { 1, 1 },
195 [AV_PIX_FMT_YUV444P9LE] = { 1, 1 },
196 [AV_PIX_FMT_YUV444P10BE] = { 1, 1 },
197 [AV_PIX_FMT_YUV444P10LE] = { 1, 1 },
198 [AV_PIX_FMT_YUV444P12BE] = { 1, 1 },
199 [AV_PIX_FMT_YUV444P12LE] = { 1, 1 },
200 [AV_PIX_FMT_YUV444P14BE] = { 1, 1 },
201 [AV_PIX_FMT_YUV444P14LE] = { 1, 1 },
202 [AV_PIX_FMT_GBRP] = { 1, 1 },
203 [AV_PIX_FMT_GBRP9LE] = { 1, 1 },
204 [AV_PIX_FMT_GBRP9BE] = { 1, 1 },
205 [AV_PIX_FMT_GBRP10LE] = { 1, 1 },
206 [AV_PIX_FMT_GBRP10BE] = { 1, 1 },
207 [AV_PIX_FMT_GBRP12LE] = { 1, 1 },
208 [AV_PIX_FMT_GBRP12BE] = { 1, 1 },
209 [AV_PIX_FMT_GBRP14LE] = { 1, 1 },
210 [AV_PIX_FMT_GBRP14BE] = { 1, 1 },
211 [AV_PIX_FMT_GBRP16LE] = { 1, 0 },
212 [AV_PIX_FMT_GBRP16BE] = { 1, 0 },
213 [AV_PIX_FMT_GBRAP] = { 1, 1 },
214 [AV_PIX_FMT_GBRAP16LE] = { 1, 0 },
215 [AV_PIX_FMT_GBRAP16BE] = { 1, 0 },
216 [AV_PIX_FMT_BAYER_BGGR8] = { 1, 0 },
217 [AV_PIX_FMT_BAYER_RGGB8] = { 1, 0 },
218 [AV_PIX_FMT_BAYER_GBRG8] = { 1, 0 },
219 [AV_PIX_FMT_BAYER_GRBG8] = { 1, 0 },
220 [AV_PIX_FMT_BAYER_BGGR16LE] = { 1, 0 },
221 [AV_PIX_FMT_BAYER_BGGR16BE] = { 1, 0 },
222 [AV_PIX_FMT_BAYER_RGGB16LE] = { 1, 0 },
223 [AV_PIX_FMT_BAYER_RGGB16BE] = { 1, 0 },
224 [AV_PIX_FMT_BAYER_GBRG16LE] = { 1, 0 },
225 [AV_PIX_FMT_BAYER_GBRG16BE] = { 1, 0 },
226 [AV_PIX_FMT_BAYER_GRBG16LE] = { 1, 0 },
227 [AV_PIX_FMT_BAYER_GRBG16BE] = { 1, 0 },
228 [AV_PIX_FMT_XYZ12BE] = { 1, 1, 1 },
229 [AV_PIX_FMT_XYZ12LE] = { 1, 1, 1 },
230 [AV_PIX_FMT_AYUV64LE] = { 1, 1},
233 int sws_isSupportedInput(enum AVPixelFormat pix_fmt)
235 return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
236 format_entries[pix_fmt].is_supported_in : 0;
239 int sws_isSupportedOutput(enum AVPixelFormat pix_fmt)
241 return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
242 format_entries[pix_fmt].is_supported_out : 0;
245 int sws_isSupportedEndiannessConversion(enum AVPixelFormat pix_fmt)
247 return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
248 format_entries[pix_fmt].is_supported_endianness : 0;
251 static double getSplineCoeff(double a, double b, double c, double d,
255 return ((d * dist + c) * dist + b) * dist + a;
257 return getSplineCoeff(0.0,
258 b + 2.0 * c + 3.0 * d,
260 -b - 3.0 * c - 6.0 * d,
264 static av_cold int get_local_pos(SwsContext *s, int chr_subsample, int pos, int dir)
266 if (pos == -1 || pos <= -513) {
267 pos = (128 << chr_subsample) - 128;
269 pos += 128; // relative to ideal left edge
270 return pos >> chr_subsample;
274 int flag; ///< flag associated to the algorithm
275 const char *description; ///< human-readable description
276 int size_factor; ///< size factor used when initing the filters
279 static const ScaleAlgorithm scale_algorithms[] = {
280 { SWS_AREA, "area averaging", 1 /* downscale only, for upscale it is bilinear */ },
281 { SWS_BICUBIC, "bicubic", 4 },
282 { SWS_BICUBLIN, "luma bicubic / chroma bilinear", -1 },
283 { SWS_BILINEAR, "bilinear", 2 },
284 { SWS_FAST_BILINEAR, "fast bilinear", -1 },
285 { SWS_GAUSS, "Gaussian", 8 /* infinite ;) */ },
286 { SWS_LANCZOS, "Lanczos", -1 /* custom */ },
287 { SWS_POINT, "nearest neighbor / point", -1 },
288 { SWS_SINC, "sinc", 20 /* infinite ;) */ },
289 { SWS_SPLINE, "bicubic spline", 20 /* infinite :)*/ },
290 { SWS_X, "experimental", 8 },
293 static av_cold int initFilter(int16_t **outFilter, int32_t **filterPos,
294 int *outFilterSize, int xInc, int srcW,
295 int dstW, int filterAlign, int one,
296 int flags, int cpu_flags,
297 SwsVector *srcFilter, SwsVector *dstFilter,
298 double param[2], int srcPos, int dstPos)
304 int64_t *filter = NULL;
305 int64_t *filter2 = NULL;
306 const int64_t fone = 1LL << (54 - FFMIN(av_log2(srcW/dstW), 8));
309 emms_c(); // FIXME should not be required but IS (even for non-MMX versions)
311 // NOTE: the +3 is for the MMX(+1) / SSE(+3) scaler which reads over the end
312 FF_ALLOC_ARRAY_OR_GOTO(NULL, *filterPos, (dstW + 3), sizeof(**filterPos), fail);
314 if (FFABS(xInc - 0x10000) < 10 && srcPos == dstPos) { // unscaled
317 FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter,
318 dstW, sizeof(*filter) * filterSize, fail);
320 for (i = 0; i < dstW; i++) {
321 filter[i * filterSize] = fone;
324 } else if (flags & SWS_POINT) { // lame looking point sampling mode
328 FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
329 dstW, sizeof(*filter) * filterSize, fail);
331 xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*0x8000LL)>>7);
332 for (i = 0; i < dstW; i++) {
333 int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;
335 (*filterPos)[i] = xx;
339 } else if ((xInc <= (1 << 16) && (flags & SWS_AREA)) ||
340 (flags & SWS_FAST_BILINEAR)) { // bilinear upscale
344 FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
345 dstW, sizeof(*filter) * filterSize, fail);
347 xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*0x8000LL)>>7);
348 for (i = 0; i < dstW; i++) {
349 int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;
352 (*filterPos)[i] = xx;
353 // bilinear upscale / linear interpolate / area averaging
354 for (j = 0; j < filterSize; j++) {
355 int64_t coeff= fone - FFABS(((int64_t)xx<<16) - xDstInSrc)*(fone>>16);
358 filter[i * filterSize + j] = coeff;
367 for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {
368 if (flags & scale_algorithms[i].flag && scale_algorithms[i].size_factor > 0) {
369 sizeFactor = scale_algorithms[i].size_factor;
373 if (flags & SWS_LANCZOS)
374 sizeFactor = param[0] != SWS_PARAM_DEFAULT ? ceil(2 * param[0]) : 6;
375 av_assert0(sizeFactor > 0);
378 filterSize = 1 + sizeFactor; // upscale
380 filterSize = 1 + (sizeFactor * srcW + dstW - 1) / dstW;
382 filterSize = FFMIN(filterSize, srcW - 2);
383 filterSize = FFMAX(filterSize, 1);
385 FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
386 dstW, sizeof(*filter) * filterSize, fail);
388 xDstInSrc = ((dstPos*(int64_t)xInc)>>7) - ((srcPos*0x10000LL)>>7);
389 for (i = 0; i < dstW; i++) {
390 int xx = (xDstInSrc - (filterSize - 2) * (1LL<<16)) / (1 << 17);
392 (*filterPos)[i] = xx;
393 for (j = 0; j < filterSize; j++) {
394 int64_t d = (FFABS(((int64_t)xx * (1 << 17)) - xDstInSrc)) << 13;
400 floatd = d * (1.0 / (1 << 30));
402 if (flags & SWS_BICUBIC) {
403 int64_t B = (param[0] != SWS_PARAM_DEFAULT ? param[0] : 0) * (1 << 24);
404 int64_t C = (param[1] != SWS_PARAM_DEFAULT ? param[1] : 0.6) * (1 << 24);
406 if (d >= 1LL << 31) {
409 int64_t dd = (d * d) >> 30;
410 int64_t ddd = (dd * d) >> 30;
413 coeff = (12 * (1 << 24) - 9 * B - 6 * C) * ddd +
414 (-18 * (1 << 24) + 12 * B + 6 * C) * dd +
415 (6 * (1 << 24) - 2 * B) * (1 << 30);
417 coeff = (-B - 6 * C) * ddd +
418 (6 * B + 30 * C) * dd +
419 (-12 * B - 48 * C) * d +
420 (8 * B + 24 * C) * (1 << 30);
422 coeff /= (1LL<<54)/fone;
425 else if (flags & SWS_X) {
426 double p = param ? param * 0.01 : 0.3;
427 coeff = d ? sin(d * M_PI) / (d * M_PI) : 1.0;
428 coeff *= pow(2.0, -p * d * d);
431 else if (flags & SWS_X) {
432 double A = param[0] != SWS_PARAM_DEFAULT ? param[0] : 1.0;
436 c = cos(floatd * M_PI);
443 coeff = (c * 0.5 + 0.5) * fone;
444 } else if (flags & SWS_AREA) {
445 int64_t d2 = d - (1 << 29);
446 if (d2 * xInc < -(1LL << (29 + 16)))
447 coeff = 1.0 * (1LL << (30 + 16));
448 else if (d2 * xInc < (1LL << (29 + 16)))
449 coeff = -d2 * xInc + (1LL << (29 + 16));
452 coeff *= fone >> (30 + 16);
453 } else if (flags & SWS_GAUSS) {
454 double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
455 coeff = exp2(-p * floatd * floatd) * fone;
456 } else if (flags & SWS_SINC) {
457 coeff = (d ? sin(floatd * M_PI) / (floatd * M_PI) : 1.0) * fone;
458 } else if (flags & SWS_LANCZOS) {
459 double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
460 coeff = (d ? sin(floatd * M_PI) * sin(floatd * M_PI / p) /
461 (floatd * floatd * M_PI * M_PI / p) : 1.0) * fone;
464 } else if (flags & SWS_BILINEAR) {
465 coeff = (1 << 30) - d;
469 } else if (flags & SWS_SPLINE) {
470 double p = -2.196152422706632;
471 coeff = getSplineCoeff(1.0, 0.0, p, -p - 1.0, floatd) * fone;
476 filter[i * filterSize + j] = coeff;
479 xDstInSrc += 2 * xInc;
483 /* apply src & dst Filter to filter -> filter2
486 av_assert0(filterSize > 0);
487 filter2Size = filterSize;
489 filter2Size += srcFilter->length - 1;
491 filter2Size += dstFilter->length - 1;
492 av_assert0(filter2Size > 0);
493 FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter2, dstW, filter2Size * sizeof(*filter2), fail);
495 for (i = 0; i < dstW; i++) {
499 for (k = 0; k < srcFilter->length; k++) {
500 for (j = 0; j < filterSize; j++)
501 filter2[i * filter2Size + k + j] +=
502 srcFilter->coeff[k] * filter[i * filterSize + j];
505 for (j = 0; j < filterSize; j++)
506 filter2[i * filter2Size + j] = filter[i * filterSize + j];
510 (*filterPos)[i] += (filterSize - 1) / 2 - (filter2Size - 1) / 2;
514 /* try to reduce the filter-size (step1 find size and shift left) */
515 // Assume it is near normalized (*0.5 or *2.0 is OK but * 0.001 is not).
517 for (i = dstW - 1; i >= 0; i--) {
518 int min = filter2Size;
520 int64_t cutOff = 0.0;
522 /* get rid of near zero elements on the left by shifting left */
523 for (j = 0; j < filter2Size; j++) {
525 cutOff += FFABS(filter2[i * filter2Size]);
527 if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)
530 /* preserve monotonicity because the core can't handle the
531 * filter otherwise */
532 if (i < dstW - 1 && (*filterPos)[i] >= (*filterPos)[i + 1])
535 // move filter coefficients left
536 for (k = 1; k < filter2Size; k++)
537 filter2[i * filter2Size + k - 1] = filter2[i * filter2Size + k];
538 filter2[i * filter2Size + k - 1] = 0;
543 /* count near zeros on the right */
544 for (j = filter2Size - 1; j > 0; j--) {
545 cutOff += FFABS(filter2[i * filter2Size + j]);
547 if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)
552 if (min > minFilterSize)
556 if (PPC_ALTIVEC(cpu_flags)) {
557 // we can handle the special case 4, so we don't want to go the full 8
558 if (minFilterSize < 5)
561 /* We really don't want to waste our time doing useless computation, so
562 * fall back on the scalar C code for very small filters.
563 * Vectorizing is worth it only if you have a decent-sized vector. */
564 if (minFilterSize < 3)
568 if (HAVE_MMX && cpu_flags & AV_CPU_FLAG_MMX) {
569 // special case for unscaled vertical filtering
570 if (minFilterSize == 1 && filterAlign == 2)
574 av_assert0(minFilterSize > 0);
575 filterSize = (minFilterSize + (filterAlign - 1)) & (~(filterAlign - 1));
576 av_assert0(filterSize > 0);
577 filter = av_malloc_array(dstW, filterSize * sizeof(*filter));
580 if (filterSize >= MAX_FILTER_SIZE * 16 /
581 ((flags & SWS_ACCURATE_RND) ? APCK_SIZE : 16)) {
582 ret = RETCODE_USE_CASCADE;
585 *outFilterSize = filterSize;
587 if (flags & SWS_PRINT_INFO)
588 av_log(NULL, AV_LOG_VERBOSE,
589 "SwScaler: reducing / aligning filtersize %d -> %d\n",
590 filter2Size, filterSize);
591 /* try to reduce the filter-size (step2 reduce it) */
592 for (i = 0; i < dstW; i++) {
595 for (j = 0; j < filterSize; j++) {
596 if (j >= filter2Size)
597 filter[i * filterSize + j] = 0;
599 filter[i * filterSize + j] = filter2[i * filter2Size + j];
600 if ((flags & SWS_BITEXACT) && j >= minFilterSize)
601 filter[i * filterSize + j] = 0;
605 // FIXME try to align filterPos if possible
608 for (i = 0; i < dstW; i++) {
610 if ((*filterPos)[i] < 0) {
611 // move filter coefficients left to compensate for filterPos
612 for (j = 1; j < filterSize; j++) {
613 int left = FFMAX(j + (*filterPos)[i], 0);
614 filter[i * filterSize + left] += filter[i * filterSize + j];
615 filter[i * filterSize + j] = 0;
620 if ((*filterPos)[i] + filterSize > srcW) {
621 int shift = (*filterPos)[i] + FFMIN(filterSize - srcW, 0);
624 for (j = filterSize - 1; j >= 0; j--) {
625 if ((*filterPos)[i] + j >= srcW) {
626 acc += filter[i * filterSize + j];
627 filter[i * filterSize + j] = 0;
630 for (j = filterSize - 1; j >= 0; j--) {
632 filter[i * filterSize + j] = 0;
634 filter[i * filterSize + j] = filter[i * filterSize + j - shift];
638 (*filterPos)[i]-= shift;
639 filter[i * filterSize + srcW - 1 - (*filterPos)[i]] += acc;
641 av_assert0((*filterPos)[i] >= 0);
642 av_assert0((*filterPos)[i] < srcW);
643 if ((*filterPos)[i] + filterSize > srcW) {
644 for (j = 0; j < filterSize; j++) {
645 av_assert0((*filterPos)[i] + j < srcW || !filter[i * filterSize + j]);
650 // Note the +1 is for the MMX scaler which reads over the end
651 /* align at 16 for AltiVec (needed by hScale_altivec_real) */
652 FF_ALLOCZ_ARRAY_OR_GOTO(NULL, *outFilter,
653 (dstW + 3), *outFilterSize * sizeof(int16_t), fail);
655 /* normalize & store in outFilter */
656 for (i = 0; i < dstW; i++) {
661 for (j = 0; j < filterSize; j++) {
662 sum += filter[i * filterSize + j];
664 sum = (sum + one / 2) / one;
666 av_log(NULL, AV_LOG_WARNING, "SwScaler: zero vector in scaling\n");
669 for (j = 0; j < *outFilterSize; j++) {
670 int64_t v = filter[i * filterSize + j] + error;
671 int intV = ROUNDED_DIV(v, sum);
672 (*outFilter)[i * (*outFilterSize) + j] = intV;
673 error = v - intV * sum;
677 (*filterPos)[dstW + 0] =
678 (*filterPos)[dstW + 1] =
679 (*filterPos)[dstW + 2] = (*filterPos)[dstW - 1]; /* the MMX/SSE scaler will
680 * read over the end */
681 for (i = 0; i < *outFilterSize; i++) {
682 int k = (dstW - 1) * (*outFilterSize) + i;
683 (*outFilter)[k + 1 * (*outFilterSize)] =
684 (*outFilter)[k + 2 * (*outFilterSize)] =
685 (*outFilter)[k + 3 * (*outFilterSize)] = (*outFilter)[k];
692 av_log(NULL, ret == RETCODE_USE_CASCADE ? AV_LOG_DEBUG : AV_LOG_ERROR, "sws: initFilter failed\n");
698 static void fill_rgb2yuv_table(SwsContext *c, const int table[4], int dstRange)
700 int64_t W, V, Z, Cy, Cu, Cv;
701 int64_t vr = table[0];
702 int64_t ub = table[1];
703 int64_t ug = -table[2];
704 int64_t vg = -table[3];
707 uint8_t *p = (uint8_t*)c->input_rgb2yuv_table;
709 static const int8_t map[] = {
710 BY_IDX, GY_IDX, -1 , BY_IDX, BY_IDX, GY_IDX, -1 , BY_IDX,
711 RY_IDX, -1 , GY_IDX, RY_IDX, RY_IDX, -1 , GY_IDX, RY_IDX,
712 RY_IDX, GY_IDX, -1 , RY_IDX, RY_IDX, GY_IDX, -1 , RY_IDX,
713 BY_IDX, -1 , GY_IDX, BY_IDX, BY_IDX, -1 , GY_IDX, BY_IDX,
714 BU_IDX, GU_IDX, -1 , BU_IDX, BU_IDX, GU_IDX, -1 , BU_IDX,
715 RU_IDX, -1 , GU_IDX, RU_IDX, RU_IDX, -1 , GU_IDX, RU_IDX,
716 RU_IDX, GU_IDX, -1 , RU_IDX, RU_IDX, GU_IDX, -1 , RU_IDX,
717 BU_IDX, -1 , GU_IDX, BU_IDX, BU_IDX, -1 , GU_IDX, BU_IDX,
718 BV_IDX, GV_IDX, -1 , BV_IDX, BV_IDX, GV_IDX, -1 , BV_IDX,
719 RV_IDX, -1 , GV_IDX, RV_IDX, RV_IDX, -1 , GV_IDX, RV_IDX,
720 RV_IDX, GV_IDX, -1 , RV_IDX, RV_IDX, GV_IDX, -1 , RV_IDX,
721 BV_IDX, -1 , GV_IDX, BV_IDX, BV_IDX, -1 , GV_IDX, BV_IDX,
722 RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX,
723 BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX,
724 GY_IDX, -1 , GY_IDX, -1 , GY_IDX, -1 , GY_IDX, -1 ,
725 -1 , GY_IDX, -1 , GY_IDX, -1 , GY_IDX, -1 , GY_IDX,
726 RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX,
727 BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX,
728 GU_IDX, -1 , GU_IDX, -1 , GU_IDX, -1 , GU_IDX, -1 ,
729 -1 , GU_IDX, -1 , GU_IDX, -1 , GU_IDX, -1 , GU_IDX,
730 RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX,
731 BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX,
732 GV_IDX, -1 , GV_IDX, -1 , GV_IDX, -1 , GV_IDX, -1 ,
733 -1 , GV_IDX, -1 , GV_IDX, -1 , GV_IDX, -1 , GV_IDX, //23
734 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //24
735 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //25
736 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //26
737 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //27
738 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //28
739 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //29
740 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //30
741 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //31
742 BY_IDX, GY_IDX, RY_IDX, -1 , -1 , -1 , -1 , -1 , //32
743 BU_IDX, GU_IDX, RU_IDX, -1 , -1 , -1 , -1 , -1 , //33
744 BV_IDX, GV_IDX, RV_IDX, -1 , -1 , -1 , -1 , -1 , //34
747 dstRange = 0; //FIXME range = 1 is handled elsewhere
757 W = ROUNDED_DIV(ONE*ONE*ug, ub);
758 V = ROUNDED_DIV(ONE*ONE*vg, vr);
761 Cy = ROUNDED_DIV(cy*Z, ONE);
762 Cu = ROUNDED_DIV(ub*Z, ONE);
763 Cv = ROUNDED_DIV(vr*Z, ONE);
765 c->input_rgb2yuv_table[RY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V , Cy);
766 c->input_rgb2yuv_table[GY_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE , Cy);
767 c->input_rgb2yuv_table[BY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W , Cy);
769 c->input_rgb2yuv_table[RU_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V , Cu);
770 c->input_rgb2yuv_table[GU_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE , Cu);
771 c->input_rgb2yuv_table[BU_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(Z+W) , Cu);
773 c->input_rgb2yuv_table[RV_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(V+Z) , Cv);
774 c->input_rgb2yuv_table[GV_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE , Cv);
775 c->input_rgb2yuv_table[BV_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W , Cv);
777 if(/*!dstRange && */!memcmp(table, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], sizeof(ff_yuv2rgb_coeffs[SWS_CS_DEFAULT]))) {
778 c->input_rgb2yuv_table[BY_IDX] = ((int)(0.114 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
779 c->input_rgb2yuv_table[BV_IDX] = (-(int)(0.081 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
780 c->input_rgb2yuv_table[BU_IDX] = ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
781 c->input_rgb2yuv_table[GY_IDX] = ((int)(0.587 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
782 c->input_rgb2yuv_table[GV_IDX] = (-(int)(0.419 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
783 c->input_rgb2yuv_table[GU_IDX] = (-(int)(0.331 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
784 c->input_rgb2yuv_table[RY_IDX] = ((int)(0.299 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
785 c->input_rgb2yuv_table[RV_IDX] = ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
786 c->input_rgb2yuv_table[RU_IDX] = (-(int)(0.169 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
788 for(i=0; i<FF_ARRAY_ELEMS(map); i++)
789 AV_WL16(p + 16*4 + 2*i, map[i] >= 0 ? c->input_rgb2yuv_table[map[i]] : 0);
792 static void fill_xyztables(struct SwsContext *c)
795 double xyzgamma = XYZ_GAMMA;
796 double rgbgamma = 1.0 / RGB_GAMMA;
797 double xyzgammainv = 1.0 / XYZ_GAMMA;
798 double rgbgammainv = RGB_GAMMA;
799 static const int16_t xyz2rgb_matrix[3][4] = {
800 {13270, -6295, -2041},
802 { 228, -835, 4329} };
803 static const int16_t rgb2xyz_matrix[3][4] = {
807 static int16_t xyzgamma_tab[4096], rgbgamma_tab[4096], xyzgammainv_tab[4096], rgbgammainv_tab[4096];
809 memcpy(c->xyz2rgb_matrix, xyz2rgb_matrix, sizeof(c->xyz2rgb_matrix));
810 memcpy(c->rgb2xyz_matrix, rgb2xyz_matrix, sizeof(c->rgb2xyz_matrix));
811 c->xyzgamma = xyzgamma_tab;
812 c->rgbgamma = rgbgamma_tab;
813 c->xyzgammainv = xyzgammainv_tab;
814 c->rgbgammainv = rgbgammainv_tab;
816 if (rgbgamma_tab[4095])
819 /* set gamma vectors */
820 for (i = 0; i < 4096; i++) {
821 xyzgamma_tab[i] = lrint(pow(i / 4095.0, xyzgamma) * 4095.0);
822 rgbgamma_tab[i] = lrint(pow(i / 4095.0, rgbgamma) * 4095.0);
823 xyzgammainv_tab[i] = lrint(pow(i / 4095.0, xyzgammainv) * 4095.0);
824 rgbgammainv_tab[i] = lrint(pow(i / 4095.0, rgbgammainv) * 4095.0);
828 int sws_setColorspaceDetails(struct SwsContext *c, const int inv_table[4],
829 int srcRange, const int table[4], int dstRange,
830 int brightness, int contrast, int saturation)
832 const AVPixFmtDescriptor *desc_dst;
833 const AVPixFmtDescriptor *desc_src;
837 desc_dst = av_pix_fmt_desc_get(c->dstFormat);
838 desc_src = av_pix_fmt_desc_get(c->srcFormat);
840 if(!isYUV(c->dstFormat) && !isGray(c->dstFormat))
842 if(!isYUV(c->srcFormat) && !isGray(c->srcFormat))
845 if (c->srcRange != srcRange ||
846 c->dstRange != dstRange ||
847 c->brightness != brightness ||
848 c->contrast != contrast ||
849 c->saturation != saturation ||
850 memcmp(c->srcColorspaceTable, inv_table, sizeof(int) * 4) ||
851 memcmp(c->dstColorspaceTable, table, sizeof(int) * 4)
855 memmove(c->srcColorspaceTable, inv_table, sizeof(int) * 4);
856 memmove(c->dstColorspaceTable, table, sizeof(int) * 4);
860 c->brightness = brightness;
861 c->contrast = contrast;
862 c->saturation = saturation;
863 c->srcRange = srcRange;
864 c->dstRange = dstRange;
866 //The srcBpc check is possibly wrong but we seem to lack a definitive reference to test this
867 //and what we have in ticket 2939 looks better with this check
868 if (need_reinit && (c->srcBpc == 8 || !isYUV(c->srcFormat)))
869 ff_sws_init_range_convert(c);
871 c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);
872 c->srcFormatBpp = av_get_bits_per_pixel(desc_src);
874 if (c->cascaded_context[c->cascaded_mainindex])
875 return sws_setColorspaceDetails(c->cascaded_context[c->cascaded_mainindex],inv_table, srcRange,table, dstRange, brightness, contrast, saturation);
880 if ((isYUV(c->dstFormat) || isGray(c->dstFormat)) && (isYUV(c->srcFormat) || isGray(c->srcFormat))) {
881 if (!c->cascaded_context[0] &&
882 memcmp(c->dstColorspaceTable, c->srcColorspaceTable, sizeof(int) * 4) &&
883 c->srcW && c->srcH && c->dstW && c->dstH) {
884 enum AVPixelFormat tmp_format;
885 int tmp_width, tmp_height;
891 av_log(c, AV_LOG_VERBOSE, "YUV color matrix differs for YUV->YUV, using intermediate RGB to convert\n");
893 if (isNBPS(c->dstFormat) || is16BPS(c->dstFormat)) {
894 if (isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) {
895 tmp_format = AV_PIX_FMT_BGRA64;
897 tmp_format = AV_PIX_FMT_BGR48;
900 if (isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) {
901 tmp_format = AV_PIX_FMT_BGRA;
903 tmp_format = AV_PIX_FMT_BGR24;
907 if (srcW*srcH > dstW*dstH) {
915 ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
916 tmp_width, tmp_height, tmp_format, 64);
920 c->cascaded_context[0] = sws_alloc_set_opts(srcW, srcH, c->srcFormat,
921 tmp_width, tmp_height, tmp_format,
923 if (!c->cascaded_context[0])
926 c->cascaded_context[0]->alphablend = c->alphablend;
927 ret = sws_init_context(c->cascaded_context[0], NULL , NULL);
930 //we set both src and dst depending on that the RGB side will be ignored
931 sws_setColorspaceDetails(c->cascaded_context[0], inv_table,
932 srcRange, table, dstRange,
933 brightness, contrast, saturation);
935 c->cascaded_context[1] = sws_getContext(tmp_width, tmp_height, tmp_format,
936 dstW, dstH, c->dstFormat,
937 c->flags, NULL, NULL, c->param);
938 if (!c->cascaded_context[1])
940 sws_setColorspaceDetails(c->cascaded_context[1], inv_table,
941 srcRange, table, dstRange,
942 0, 1 << 16, 1 << 16);
948 if (!isYUV(c->dstFormat) && !isGray(c->dstFormat)) {
949 ff_yuv2rgb_c_init_tables(c, inv_table, srcRange, brightness,
950 contrast, saturation);
954 ff_yuv2rgb_init_tables_ppc(c, inv_table, brightness,
955 contrast, saturation);
958 fill_rgb2yuv_table(c, table, dstRange);
963 int sws_getColorspaceDetails(struct SwsContext *c, int **inv_table,
964 int *srcRange, int **table, int *dstRange,
965 int *brightness, int *contrast, int *saturation)
970 *inv_table = c->srcColorspaceTable;
971 *table = c->dstColorspaceTable;
972 *srcRange = c->srcRange;
973 *dstRange = c->dstRange;
974 *brightness = c->brightness;
975 *contrast = c->contrast;
976 *saturation = c->saturation;
981 static int handle_jpeg(enum AVPixelFormat *format)
984 case AV_PIX_FMT_YUVJ420P:
985 *format = AV_PIX_FMT_YUV420P;
987 case AV_PIX_FMT_YUVJ411P:
988 *format = AV_PIX_FMT_YUV411P;
990 case AV_PIX_FMT_YUVJ422P:
991 *format = AV_PIX_FMT_YUV422P;
993 case AV_PIX_FMT_YUVJ444P:
994 *format = AV_PIX_FMT_YUV444P;
996 case AV_PIX_FMT_YUVJ440P:
997 *format = AV_PIX_FMT_YUV440P;
999 case AV_PIX_FMT_GRAY8:
1000 case AV_PIX_FMT_YA8:
1001 case AV_PIX_FMT_GRAY16LE:
1002 case AV_PIX_FMT_GRAY16BE:
1003 case AV_PIX_FMT_YA16BE:
1004 case AV_PIX_FMT_YA16LE:
1011 static int handle_0alpha(enum AVPixelFormat *format)
1014 case AV_PIX_FMT_0BGR : *format = AV_PIX_FMT_ABGR ; return 1;
1015 case AV_PIX_FMT_BGR0 : *format = AV_PIX_FMT_BGRA ; return 4;
1016 case AV_PIX_FMT_0RGB : *format = AV_PIX_FMT_ARGB ; return 1;
1017 case AV_PIX_FMT_RGB0 : *format = AV_PIX_FMT_RGBA ; return 4;
1022 static int handle_xyz(enum AVPixelFormat *format)
1025 case AV_PIX_FMT_XYZ12BE : *format = AV_PIX_FMT_RGB48BE; return 1;
1026 case AV_PIX_FMT_XYZ12LE : *format = AV_PIX_FMT_RGB48LE; return 1;
1031 static void handle_formats(SwsContext *c)
1033 c->src0Alpha |= handle_0alpha(&c->srcFormat);
1034 c->dst0Alpha |= handle_0alpha(&c->dstFormat);
1035 c->srcXYZ |= handle_xyz(&c->srcFormat);
1036 c->dstXYZ |= handle_xyz(&c->dstFormat);
1037 if (c->srcXYZ || c->dstXYZ)
1041 SwsContext *sws_alloc_context(void)
1043 SwsContext *c = av_mallocz(sizeof(SwsContext));
1045 av_assert0(offsetof(SwsContext, redDither) + DITHER32_INT == offsetof(SwsContext, dither32));
1048 c->av_class = &ff_sws_context_class;
1049 av_opt_set_defaults(c);
1055 static uint16_t * alloc_gamma_tbl(double e)
1059 tbl = (uint16_t*)av_malloc(sizeof(uint16_t) * 1 << 16);
1063 for (i = 0; i < 65536; ++i) {
1064 tbl[i] = pow(i / 65535.0, e) * 65535.0;
1069 static enum AVPixelFormat alphaless_fmt(enum AVPixelFormat fmt)
1072 case AV_PIX_FMT_ARGB: return AV_PIX_FMT_RGB24;
1073 case AV_PIX_FMT_RGBA: return AV_PIX_FMT_RGB24;
1074 case AV_PIX_FMT_ABGR: return AV_PIX_FMT_BGR24;
1075 case AV_PIX_FMT_BGRA: return AV_PIX_FMT_BGR24;
1076 case AV_PIX_FMT_YA8: return AV_PIX_FMT_GRAY8;
1078 case AV_PIX_FMT_YUVA420P: return AV_PIX_FMT_YUV420P;
1079 case AV_PIX_FMT_YUVA422P: return AV_PIX_FMT_YUV422P;
1080 case AV_PIX_FMT_YUVA444P: return AV_PIX_FMT_YUV444P;
1082 case AV_PIX_FMT_GBRAP: return AV_PIX_FMT_GBRP;
1084 case AV_PIX_FMT_GBRAP16LE: return AV_PIX_FMT_GBRP16;
1085 case AV_PIX_FMT_GBRAP16BE: return AV_PIX_FMT_GBRP16;
1087 case AV_PIX_FMT_RGBA64LE: return AV_PIX_FMT_RGB48;
1088 case AV_PIX_FMT_RGBA64BE: return AV_PIX_FMT_RGB48;
1089 case AV_PIX_FMT_BGRA64LE: return AV_PIX_FMT_BGR48;
1090 case AV_PIX_FMT_BGRA64BE: return AV_PIX_FMT_BGR48;
1092 case AV_PIX_FMT_YA16BE: return AV_PIX_FMT_GRAY16;
1093 case AV_PIX_FMT_YA16LE: return AV_PIX_FMT_GRAY16;
1095 case AV_PIX_FMT_YUVA420P9BE: return AV_PIX_FMT_YUV420P9;
1096 case AV_PIX_FMT_YUVA422P9BE: return AV_PIX_FMT_YUV422P9;
1097 case AV_PIX_FMT_YUVA444P9BE: return AV_PIX_FMT_YUV444P9;
1098 case AV_PIX_FMT_YUVA420P9LE: return AV_PIX_FMT_YUV420P9;
1099 case AV_PIX_FMT_YUVA422P9LE: return AV_PIX_FMT_YUV422P9;
1100 case AV_PIX_FMT_YUVA444P9LE: return AV_PIX_FMT_YUV444P9;
1101 case AV_PIX_FMT_YUVA420P10BE: return AV_PIX_FMT_YUV420P10;
1102 case AV_PIX_FMT_YUVA422P10BE: return AV_PIX_FMT_YUV422P10;
1103 case AV_PIX_FMT_YUVA444P10BE: return AV_PIX_FMT_YUV444P10;
1104 case AV_PIX_FMT_YUVA420P10LE: return AV_PIX_FMT_YUV420P10;
1105 case AV_PIX_FMT_YUVA422P10LE: return AV_PIX_FMT_YUV422P10;
1106 case AV_PIX_FMT_YUVA444P10LE: return AV_PIX_FMT_YUV444P10;
1107 case AV_PIX_FMT_YUVA420P16BE: return AV_PIX_FMT_YUV420P16;
1108 case AV_PIX_FMT_YUVA422P16BE: return AV_PIX_FMT_YUV422P16;
1109 case AV_PIX_FMT_YUVA444P16BE: return AV_PIX_FMT_YUV444P16;
1110 case AV_PIX_FMT_YUVA420P16LE: return AV_PIX_FMT_YUV420P16;
1111 case AV_PIX_FMT_YUVA422P16LE: return AV_PIX_FMT_YUV422P16;
1112 case AV_PIX_FMT_YUVA444P16LE: return AV_PIX_FMT_YUV444P16;
1114 // case AV_PIX_FMT_AYUV64LE:
1115 // case AV_PIX_FMT_AYUV64BE:
1116 // case AV_PIX_FMT_PAL8:
1117 default: return AV_PIX_FMT_NONE;
1121 av_cold int sws_init_context(SwsContext *c, SwsFilter *srcFilter,
1122 SwsFilter *dstFilter)
1125 int usesVFilter, usesHFilter;
1127 SwsFilter dummyFilter = { NULL, NULL, NULL, NULL };
1132 int dst_stride = FFALIGN(dstW * sizeof(int16_t) + 66, 16);
1133 int flags, cpu_flags;
1134 enum AVPixelFormat srcFormat = c->srcFormat;
1135 enum AVPixelFormat dstFormat = c->dstFormat;
1136 const AVPixFmtDescriptor *desc_src;
1137 const AVPixFmtDescriptor *desc_dst;
1139 enum AVPixelFormat tmpFmt;
1141 cpu_flags = av_get_cpu_flags();
1145 ff_sws_rgb2rgb_init();
1147 unscaled = (srcW == dstW && srcH == dstH);
1149 c->srcRange |= handle_jpeg(&c->srcFormat);
1150 c->dstRange |= handle_jpeg(&c->dstFormat);
1152 if(srcFormat!=c->srcFormat || dstFormat!=c->dstFormat)
1153 av_log(c, AV_LOG_WARNING, "deprecated pixel format used, make sure you did set range correctly\n");
1155 if (!c->contrast && !c->saturation && !c->dstFormatBpp)
1156 sws_setColorspaceDetails(c, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], c->srcRange,
1157 ff_yuv2rgb_coeffs[SWS_CS_DEFAULT],
1158 c->dstRange, 0, 1 << 16, 1 << 16);
1161 srcFormat = c->srcFormat;
1162 dstFormat = c->dstFormat;
1163 desc_src = av_pix_fmt_desc_get(srcFormat);
1164 desc_dst = av_pix_fmt_desc_get(dstFormat);
1166 // If the source has no alpha then disable alpha blendaway
1168 c->alphablend = SWS_ALPHA_BLEND_NONE;
1170 if (!(unscaled && sws_isSupportedEndiannessConversion(srcFormat) &&
1171 av_pix_fmt_swap_endianness(srcFormat) == dstFormat)) {
1172 if (!sws_isSupportedInput(srcFormat)) {
1173 av_log(c, AV_LOG_ERROR, "%s is not supported as input pixel format\n",
1174 av_get_pix_fmt_name(srcFormat));
1175 return AVERROR(EINVAL);
1177 if (!sws_isSupportedOutput(dstFormat)) {
1178 av_log(c, AV_LOG_ERROR, "%s is not supported as output pixel format\n",
1179 av_get_pix_fmt_name(dstFormat));
1180 return AVERROR(EINVAL);
1183 av_assert2(desc_src && desc_dst);
1185 i = flags & (SWS_POINT |
1197 /* provide a default scaler if not set by caller */
1199 if (dstW < srcW && dstH < srcH)
1200 flags |= SWS_BICUBIC;
1201 else if (dstW > srcW && dstH > srcH)
1202 flags |= SWS_BICUBIC;
1204 flags |= SWS_BICUBIC;
1206 } else if (i & (i - 1)) {
1207 av_log(c, AV_LOG_ERROR,
1208 "Exactly one scaler algorithm must be chosen, got %X\n", i);
1209 return AVERROR(EINVAL);
1212 if (srcW < 1 || srcH < 1 || dstW < 1 || dstH < 1) {
1213 /* FIXME check if these are enough and try to lower them after
1214 * fixing the relevant parts of the code */
1215 av_log(c, AV_LOG_ERROR, "%dx%d -> %dx%d is invalid scaling dimension\n",
1216 srcW, srcH, dstW, dstH);
1217 return AVERROR(EINVAL);
1219 if (flags & SWS_FAST_BILINEAR) {
1220 if (srcW < 8 || dstW < 8) {
1221 flags ^= SWS_FAST_BILINEAR | SWS_BILINEAR;
1227 dstFilter = &dummyFilter;
1229 srcFilter = &dummyFilter;
1231 c->lumXInc = (((int64_t)srcW << 16) + (dstW >> 1)) / dstW;
1232 c->lumYInc = (((int64_t)srcH << 16) + (dstH >> 1)) / dstH;
1233 c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);
1234 c->srcFormatBpp = av_get_bits_per_pixel(desc_src);
1235 c->vRounder = 4 * 0x0001000100010001ULL;
1237 usesVFilter = (srcFilter->lumV && srcFilter->lumV->length > 1) ||
1238 (srcFilter->chrV && srcFilter->chrV->length > 1) ||
1239 (dstFilter->lumV && dstFilter->lumV->length > 1) ||
1240 (dstFilter->chrV && dstFilter->chrV->length > 1);
1241 usesHFilter = (srcFilter->lumH && srcFilter->lumH->length > 1) ||
1242 (srcFilter->chrH && srcFilter->chrH->length > 1) ||
1243 (dstFilter->lumH && dstFilter->lumH->length > 1) ||
1244 (dstFilter->chrH && dstFilter->chrH->length > 1);
1246 av_pix_fmt_get_chroma_sub_sample(srcFormat, &c->chrSrcHSubSample, &c->chrSrcVSubSample);
1247 av_pix_fmt_get_chroma_sub_sample(dstFormat, &c->chrDstHSubSample, &c->chrDstVSubSample);
1249 if (isAnyRGB(dstFormat) && !(flags&SWS_FULL_CHR_H_INT)) {
1251 av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to odd output size\n");
1252 flags |= SWS_FULL_CHR_H_INT;
1256 if ( c->chrSrcHSubSample == 0
1257 && c->chrSrcVSubSample == 0
1258 && c->dither != SWS_DITHER_BAYER //SWS_FULL_CHR_H_INT is currently not supported with SWS_DITHER_BAYER
1259 && !(c->flags & SWS_FAST_BILINEAR)
1261 av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to input having non subsampled chroma\n");
1262 flags |= SWS_FULL_CHR_H_INT;
1267 if (c->dither == SWS_DITHER_AUTO) {
1268 if (flags & SWS_ERROR_DIFFUSION)
1269 c->dither = SWS_DITHER_ED;
1272 if(dstFormat == AV_PIX_FMT_BGR4_BYTE ||
1273 dstFormat == AV_PIX_FMT_RGB4_BYTE ||
1274 dstFormat == AV_PIX_FMT_BGR8 ||
1275 dstFormat == AV_PIX_FMT_RGB8) {
1276 if (c->dither == SWS_DITHER_AUTO)
1277 c->dither = (flags & SWS_FULL_CHR_H_INT) ? SWS_DITHER_ED : SWS_DITHER_BAYER;
1278 if (!(flags & SWS_FULL_CHR_H_INT)) {
1279 if (c->dither == SWS_DITHER_ED || c->dither == SWS_DITHER_A_DITHER || c->dither == SWS_DITHER_X_DITHER) {
1280 av_log(c, AV_LOG_DEBUG,
1281 "Desired dithering only supported in full chroma interpolation for destination format '%s'\n",
1282 av_get_pix_fmt_name(dstFormat));
1283 flags |= SWS_FULL_CHR_H_INT;
1287 if (flags & SWS_FULL_CHR_H_INT) {
1288 if (c->dither == SWS_DITHER_BAYER) {
1289 av_log(c, AV_LOG_DEBUG,
1290 "Ordered dither is not supported in full chroma interpolation for destination format '%s'\n",
1291 av_get_pix_fmt_name(dstFormat));
1292 c->dither = SWS_DITHER_ED;
1296 if (isPlanarRGB(dstFormat)) {
1297 if (!(flags & SWS_FULL_CHR_H_INT)) {
1298 av_log(c, AV_LOG_DEBUG,
1299 "%s output is not supported with half chroma resolution, switching to full\n",
1300 av_get_pix_fmt_name(dstFormat));
1301 flags |= SWS_FULL_CHR_H_INT;
1306 /* reuse chroma for 2 pixels RGB/BGR unless user wants full
1307 * chroma interpolation */
1308 if (flags & SWS_FULL_CHR_H_INT &&
1309 isAnyRGB(dstFormat) &&
1310 !isPlanarRGB(dstFormat) &&
1311 dstFormat != AV_PIX_FMT_RGBA64LE &&
1312 dstFormat != AV_PIX_FMT_RGBA64BE &&
1313 dstFormat != AV_PIX_FMT_BGRA64LE &&
1314 dstFormat != AV_PIX_FMT_BGRA64BE &&
1315 dstFormat != AV_PIX_FMT_RGB48LE &&
1316 dstFormat != AV_PIX_FMT_RGB48BE &&
1317 dstFormat != AV_PIX_FMT_BGR48LE &&
1318 dstFormat != AV_PIX_FMT_BGR48BE &&
1319 dstFormat != AV_PIX_FMT_RGBA &&
1320 dstFormat != AV_PIX_FMT_ARGB &&
1321 dstFormat != AV_PIX_FMT_BGRA &&
1322 dstFormat != AV_PIX_FMT_ABGR &&
1323 dstFormat != AV_PIX_FMT_RGB24 &&
1324 dstFormat != AV_PIX_FMT_BGR24 &&
1325 dstFormat != AV_PIX_FMT_BGR4_BYTE &&
1326 dstFormat != AV_PIX_FMT_RGB4_BYTE &&
1327 dstFormat != AV_PIX_FMT_BGR8 &&
1328 dstFormat != AV_PIX_FMT_RGB8
1330 av_log(c, AV_LOG_WARNING,
1331 "full chroma interpolation for destination format '%s' not yet implemented\n",
1332 av_get_pix_fmt_name(dstFormat));
1333 flags &= ~SWS_FULL_CHR_H_INT;
1336 if (isAnyRGB(dstFormat) && !(flags & SWS_FULL_CHR_H_INT))
1337 c->chrDstHSubSample = 1;
1339 // drop some chroma lines if the user wants it
1340 c->vChrDrop = (flags & SWS_SRC_V_CHR_DROP_MASK) >>
1341 SWS_SRC_V_CHR_DROP_SHIFT;
1342 c->chrSrcVSubSample += c->vChrDrop;
1344 /* drop every other pixel for chroma calculation unless user
1345 * wants full chroma */
1346 if (isAnyRGB(srcFormat) && !(flags & SWS_FULL_CHR_H_INP) &&
1347 srcFormat != AV_PIX_FMT_RGB8 && srcFormat != AV_PIX_FMT_BGR8 &&
1348 srcFormat != AV_PIX_FMT_RGB4 && srcFormat != AV_PIX_FMT_BGR4 &&
1349 srcFormat != AV_PIX_FMT_RGB4_BYTE && srcFormat != AV_PIX_FMT_BGR4_BYTE &&
1350 srcFormat != AV_PIX_FMT_GBRP9BE && srcFormat != AV_PIX_FMT_GBRP9LE &&
1351 srcFormat != AV_PIX_FMT_GBRP10BE && srcFormat != AV_PIX_FMT_GBRP10LE &&
1352 srcFormat != AV_PIX_FMT_GBRP12BE && srcFormat != AV_PIX_FMT_GBRP12LE &&
1353 srcFormat != AV_PIX_FMT_GBRP14BE && srcFormat != AV_PIX_FMT_GBRP14LE &&
1354 srcFormat != AV_PIX_FMT_GBRP16BE && srcFormat != AV_PIX_FMT_GBRP16LE &&
1355 ((dstW >> c->chrDstHSubSample) <= (srcW >> 1) ||
1356 (flags & SWS_FAST_BILINEAR)))
1357 c->chrSrcHSubSample = 1;
1359 // Note the FF_CEIL_RSHIFT is so that we always round toward +inf.
1360 c->chrSrcW = FF_CEIL_RSHIFT(srcW, c->chrSrcHSubSample);
1361 c->chrSrcH = FF_CEIL_RSHIFT(srcH, c->chrSrcVSubSample);
1362 c->chrDstW = FF_CEIL_RSHIFT(dstW, c->chrDstHSubSample);
1363 c->chrDstH = FF_CEIL_RSHIFT(dstH, c->chrDstVSubSample);
1365 FF_ALLOCZ_OR_GOTO(c, c->formatConvBuffer, FFALIGN(srcW*2+78, 16) * 2, fail);
1367 c->srcBpc = desc_src->comp[0].depth;
1370 c->dstBpc = desc_dst->comp[0].depth;
1373 if (isAnyRGB(srcFormat) || srcFormat == AV_PIX_FMT_PAL8)
1375 if (c->dstBpc == 16)
1378 if (INLINE_MMXEXT(cpu_flags) && c->srcBpc == 8 && c->dstBpc <= 14) {
1379 c->canMMXEXTBeUsed = dstW >= srcW && (dstW & 31) == 0 &&
1380 c->chrDstW >= c->chrSrcW &&
1382 if (!c->canMMXEXTBeUsed && dstW >= srcW && c->chrDstW >= c->chrSrcW && (srcW & 15) == 0
1384 && (flags & SWS_FAST_BILINEAR)) {
1385 if (flags & SWS_PRINT_INFO)
1386 av_log(c, AV_LOG_INFO,
1387 "output width is not a multiple of 32 -> no MMXEXT scaler\n");
1389 if (usesHFilter || isNBPS(c->srcFormat) || is16BPS(c->srcFormat) || isAnyRGB(c->srcFormat))
1390 c->canMMXEXTBeUsed = 0;
1392 c->canMMXEXTBeUsed = 0;
1394 c->chrXInc = (((int64_t)c->chrSrcW << 16) + (c->chrDstW >> 1)) / c->chrDstW;
1395 c->chrYInc = (((int64_t)c->chrSrcH << 16) + (c->chrDstH >> 1)) / c->chrDstH;
1397 /* Match pixel 0 of the src to pixel 0 of dst and match pixel n-2 of src
1398 * to pixel n-2 of dst, but only for the FAST_BILINEAR mode otherwise do
1400 * n-2 is the last chrominance sample available.
1401 * This is not perfect, but no one should notice the difference, the more
1402 * correct variant would be like the vertical one, but that would require
1403 * some special code for the first and last pixel */
1404 if (flags & SWS_FAST_BILINEAR) {
1405 if (c->canMMXEXTBeUsed) {
1409 // we don't use the x86 asm scaler if MMX is available
1410 else if (INLINE_MMX(cpu_flags) && c->dstBpc <= 14) {
1411 c->lumXInc = ((int64_t)(srcW - 2) << 16) / (dstW - 2) - 20;
1412 c->chrXInc = ((int64_t)(c->chrSrcW - 2) << 16) / (c->chrDstW - 2) - 20;
1416 // hardcoded for now
1417 c->gamma_value = 2.2;
1418 tmpFmt = AV_PIX_FMT_RGBA64LE;
1421 if (!unscaled && c->gamma_flag && (srcFormat != tmpFmt || dstFormat != tmpFmt)) {
1423 c->cascaded_context[0] = NULL;
1425 ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
1426 srcW, srcH, tmpFmt, 64);
1430 c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,
1432 flags, NULL, NULL, c->param);
1433 if (!c->cascaded_context[0]) {
1437 c->cascaded_context[1] = sws_getContext(srcW, srcH, tmpFmt,
1439 flags, srcFilter, dstFilter, c->param);
1441 if (!c->cascaded_context[1])
1444 c2 = c->cascaded_context[1];
1445 c2->is_internal_gamma = 1;
1446 c2->gamma = alloc_gamma_tbl( c->gamma_value);
1447 c2->inv_gamma = alloc_gamma_tbl(1.f/c->gamma_value);
1448 if (!c2->gamma || !c2->inv_gamma)
1449 return AVERROR(ENOMEM);
1451 // is_internal_flag is set after creating the context
1452 // to properly create the gamma convert FilterDescriptor
1453 // we have to re-initialize it
1454 ff_free_filters(c2);
1455 if (ff_init_filters(c2) < 0) {
1456 sws_freeContext(c2);
1460 c->cascaded_context[2] = NULL;
1461 if (dstFormat != tmpFmt) {
1462 ret = av_image_alloc(c->cascaded1_tmp, c->cascaded1_tmpStride,
1463 dstW, dstH, tmpFmt, 64);
1467 c->cascaded_context[2] = sws_getContext(dstW, dstH, tmpFmt,
1468 dstW, dstH, dstFormat,
1469 flags, NULL, NULL, c->param);
1470 if (!c->cascaded_context[2])
1476 if (isBayer(srcFormat)) {
1478 (dstFormat != AV_PIX_FMT_RGB24 && dstFormat != AV_PIX_FMT_YUV420P)) {
1479 enum AVPixelFormat tmpFormat = AV_PIX_FMT_RGB24;
1481 ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
1482 srcW, srcH, tmpFormat, 64);
1486 c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,
1487 srcW, srcH, tmpFormat,
1488 flags, srcFilter, NULL, c->param);
1489 if (!c->cascaded_context[0])
1492 c->cascaded_context[1] = sws_getContext(srcW, srcH, tmpFormat,
1493 dstW, dstH, dstFormat,
1494 flags, NULL, dstFilter, c->param);
1495 if (!c->cascaded_context[1])
1501 if (CONFIG_SWSCALE_ALPHA && isALPHA(srcFormat) && !isALPHA(dstFormat)) {
1502 enum AVPixelFormat tmpFormat = alphaless_fmt(srcFormat);
1504 if (tmpFormat != AV_PIX_FMT_NONE && c->alphablend != SWS_ALPHA_BLEND_NONE)
1506 dstFormat != tmpFormat ||
1507 usesHFilter || usesVFilter ||
1508 c->srcRange != c->dstRange
1510 c->cascaded_mainindex = 1;
1511 ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
1512 srcW, srcH, tmpFormat, 64);
1516 c->cascaded_context[0] = sws_alloc_set_opts(srcW, srcH, srcFormat,
1517 srcW, srcH, tmpFormat,
1519 if (!c->cascaded_context[0])
1521 c->cascaded_context[0]->alphablend = c->alphablend;
1522 ret = sws_init_context(c->cascaded_context[0], NULL , NULL);
1526 c->cascaded_context[1] = sws_alloc_set_opts(srcW, srcH, tmpFormat,
1527 dstW, dstH, dstFormat,
1529 if (!c->cascaded_context[1])
1532 c->cascaded_context[1]->srcRange = c->srcRange;
1533 c->cascaded_context[1]->dstRange = c->dstRange;
1534 ret = sws_init_context(c->cascaded_context[1], srcFilter , dstFilter);
1542 #define USE_MMAP (HAVE_MMAP && HAVE_MPROTECT && defined MAP_ANONYMOUS)
1544 /* precalculate horizontal scaler filter coefficients */
1546 #if HAVE_MMXEXT_INLINE
1547 // can't downscale !!!
1548 if (c->canMMXEXTBeUsed && (flags & SWS_FAST_BILINEAR)) {
1549 c->lumMmxextFilterCodeSize = ff_init_hscaler_mmxext(dstW, c->lumXInc, NULL,
1551 c->chrMmxextFilterCodeSize = ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc,
1552 NULL, NULL, NULL, 4);
1555 c->lumMmxextFilterCode = mmap(NULL, c->lumMmxextFilterCodeSize,
1556 PROT_READ | PROT_WRITE,
1557 MAP_PRIVATE | MAP_ANONYMOUS,
1559 c->chrMmxextFilterCode = mmap(NULL, c->chrMmxextFilterCodeSize,
1560 PROT_READ | PROT_WRITE,
1561 MAP_PRIVATE | MAP_ANONYMOUS,
1563 #elif HAVE_VIRTUALALLOC
1564 c->lumMmxextFilterCode = VirtualAlloc(NULL,
1565 c->lumMmxextFilterCodeSize,
1567 PAGE_EXECUTE_READWRITE);
1568 c->chrMmxextFilterCode = VirtualAlloc(NULL,
1569 c->chrMmxextFilterCodeSize,
1571 PAGE_EXECUTE_READWRITE);
1573 c->lumMmxextFilterCode = av_malloc(c->lumMmxextFilterCodeSize);
1574 c->chrMmxextFilterCode = av_malloc(c->chrMmxextFilterCodeSize);
1577 #ifdef MAP_ANONYMOUS
1578 if (c->lumMmxextFilterCode == MAP_FAILED || c->chrMmxextFilterCode == MAP_FAILED)
1580 if (!c->lumMmxextFilterCode || !c->chrMmxextFilterCode)
1583 av_log(c, AV_LOG_ERROR, "Failed to allocate MMX2FilterCode\n");
1584 return AVERROR(ENOMEM);
1587 FF_ALLOCZ_OR_GOTO(c, c->hLumFilter, (dstW / 8 + 8) * sizeof(int16_t), fail);
1588 FF_ALLOCZ_OR_GOTO(c, c->hChrFilter, (c->chrDstW / 4 + 8) * sizeof(int16_t), fail);
1589 FF_ALLOCZ_OR_GOTO(c, c->hLumFilterPos, (dstW / 2 / 8 + 8) * sizeof(int32_t), fail);
1590 FF_ALLOCZ_OR_GOTO(c, c->hChrFilterPos, (c->chrDstW / 2 / 4 + 8) * sizeof(int32_t), fail);
1592 ff_init_hscaler_mmxext( dstW, c->lumXInc, c->lumMmxextFilterCode,
1593 c->hLumFilter, (uint32_t*)c->hLumFilterPos, 8);
1594 ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc, c->chrMmxextFilterCode,
1595 c->hChrFilter, (uint32_t*)c->hChrFilterPos, 4);
1598 if ( mprotect(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1
1599 || mprotect(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1) {
1600 av_log(c, AV_LOG_ERROR, "mprotect failed, cannot use fast bilinear scaler\n");
1605 #endif /* HAVE_MMXEXT_INLINE */
1607 const int filterAlign = X86_MMX(cpu_flags) ? 4 :
1608 PPC_ALTIVEC(cpu_flags) ? 8 : 1;
1610 if ((ret = initFilter(&c->hLumFilter, &c->hLumFilterPos,
1611 &c->hLumFilterSize, c->lumXInc,
1612 srcW, dstW, filterAlign, 1 << 14,
1613 (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,
1614 cpu_flags, srcFilter->lumH, dstFilter->lumH,
1616 get_local_pos(c, 0, 0, 0),
1617 get_local_pos(c, 0, 0, 0))) < 0)
1619 if ((ret = initFilter(&c->hChrFilter, &c->hChrFilterPos,
1620 &c->hChrFilterSize, c->chrXInc,
1621 c->chrSrcW, c->chrDstW, filterAlign, 1 << 14,
1622 (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,
1623 cpu_flags, srcFilter->chrH, dstFilter->chrH,
1625 get_local_pos(c, c->chrSrcHSubSample, c->src_h_chr_pos, 0),
1626 get_local_pos(c, c->chrDstHSubSample, c->dst_h_chr_pos, 0))) < 0)
1629 } // initialize horizontal stuff
1631 /* precalculate vertical scaler filter coefficients */
1633 const int filterAlign = X86_MMX(cpu_flags) ? 2 :
1634 PPC_ALTIVEC(cpu_flags) ? 8 : 1;
1636 if ((ret = initFilter(&c->vLumFilter, &c->vLumFilterPos, &c->vLumFilterSize,
1637 c->lumYInc, srcH, dstH, filterAlign, (1 << 12),
1638 (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,
1639 cpu_flags, srcFilter->lumV, dstFilter->lumV,
1641 get_local_pos(c, 0, 0, 1),
1642 get_local_pos(c, 0, 0, 1))) < 0)
1644 if ((ret = initFilter(&c->vChrFilter, &c->vChrFilterPos, &c->vChrFilterSize,
1645 c->chrYInc, c->chrSrcH, c->chrDstH,
1646 filterAlign, (1 << 12),
1647 (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,
1648 cpu_flags, srcFilter->chrV, dstFilter->chrV,
1650 get_local_pos(c, c->chrSrcVSubSample, c->src_v_chr_pos, 1),
1651 get_local_pos(c, c->chrDstVSubSample, c->dst_v_chr_pos, 1))) < 0)
1656 FF_ALLOC_OR_GOTO(c, c->vYCoeffsBank, sizeof(vector signed short) * c->vLumFilterSize * c->dstH, fail);
1657 FF_ALLOC_OR_GOTO(c, c->vCCoeffsBank, sizeof(vector signed short) * c->vChrFilterSize * c->chrDstH, fail);
1659 for (i = 0; i < c->vLumFilterSize * c->dstH; i++) {
1661 short *p = (short *)&c->vYCoeffsBank[i];
1662 for (j = 0; j < 8; j++)
1663 p[j] = c->vLumFilter[i];
1666 for (i = 0; i < c->vChrFilterSize * c->chrDstH; i++) {
1668 short *p = (short *)&c->vCCoeffsBank[i];
1669 for (j = 0; j < 8; j++)
1670 p[j] = c->vChrFilter[i];
1675 // calculate buffer sizes so that they won't run out while handling these damn slices
1676 c->vLumBufSize = c->vLumFilterSize;
1677 c->vChrBufSize = c->vChrFilterSize;
1678 for (i = 0; i < dstH; i++) {
1679 int chrI = (int64_t)i * c->chrDstH / dstH;
1680 int nextSlice = FFMAX(c->vLumFilterPos[i] + c->vLumFilterSize - 1,
1681 ((c->vChrFilterPos[chrI] + c->vChrFilterSize - 1)
1682 << c->chrSrcVSubSample));
1684 nextSlice >>= c->chrSrcVSubSample;
1685 nextSlice <<= c->chrSrcVSubSample;
1686 if (c->vLumFilterPos[i] + c->vLumBufSize < nextSlice)
1687 c->vLumBufSize = nextSlice - c->vLumFilterPos[i];
1688 if (c->vChrFilterPos[chrI] + c->vChrBufSize <
1689 (nextSlice >> c->chrSrcVSubSample))
1690 c->vChrBufSize = (nextSlice >> c->chrSrcVSubSample) -
1691 c->vChrFilterPos[chrI];
1694 for (i = 0; i < 4; i++)
1695 FF_ALLOCZ_OR_GOTO(c, c->dither_error[i], (c->dstW+2) * sizeof(int), fail);
1697 /* Allocate pixbufs (we use dynamic allocation because otherwise we would
1698 * need to allocate several megabytes to handle all possible cases) */
1699 FF_ALLOCZ_OR_GOTO(c, c->lumPixBuf, c->vLumBufSize * 3 * sizeof(int16_t *), fail);
1700 FF_ALLOCZ_OR_GOTO(c, c->chrUPixBuf, c->vChrBufSize * 3 * sizeof(int16_t *), fail);
1701 FF_ALLOCZ_OR_GOTO(c, c->chrVPixBuf, c->vChrBufSize * 3 * sizeof(int16_t *), fail);
1702 if (CONFIG_SWSCALE_ALPHA && isALPHA(c->srcFormat) && isALPHA(c->dstFormat))
1703 FF_ALLOCZ_OR_GOTO(c, c->alpPixBuf, c->vLumBufSize * 3 * sizeof(int16_t *), fail);
1704 /* Note we need at least one pixel more at the end because of the MMX code
1705 * (just in case someone wants to replace the 4000/8000). */
1706 /* align at 16 bytes for AltiVec */
1707 for (i = 0; i < c->vLumBufSize; i++) {
1708 FF_ALLOCZ_OR_GOTO(c, c->lumPixBuf[i + c->vLumBufSize],
1709 dst_stride + 16, fail);
1710 c->lumPixBuf[i] = c->lumPixBuf[i + c->vLumBufSize];
1712 // 64 / c->scalingBpp is the same as 16 / sizeof(scaling_intermediate)
1713 c->uv_off = (dst_stride>>1) + 64 / (c->dstBpc &~ 7);
1714 c->uv_offx2 = dst_stride + 16;
1715 for (i = 0; i < c->vChrBufSize; i++) {
1716 FF_ALLOC_OR_GOTO(c, c->chrUPixBuf[i + c->vChrBufSize],
1717 dst_stride * 2 + 32, fail);
1718 c->chrUPixBuf[i] = c->chrUPixBuf[i + c->vChrBufSize];
1719 c->chrVPixBuf[i] = c->chrVPixBuf[i + c->vChrBufSize]
1720 = c->chrUPixBuf[i] + (dst_stride >> 1) + 8;
1722 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf)
1723 for (i = 0; i < c->vLumBufSize; i++) {
1724 FF_ALLOCZ_OR_GOTO(c, c->alpPixBuf[i + c->vLumBufSize],
1725 dst_stride + 16, fail);
1726 c->alpPixBuf[i] = c->alpPixBuf[i + c->vLumBufSize];
1729 // try to avoid drawing green stuff between the right end and the stride end
1730 for (i = 0; i < c->vChrBufSize; i++)
1731 if(desc_dst->comp[0].depth == 16){
1732 av_assert0(c->dstBpc > 14);
1733 for(j=0; j<dst_stride/2+1; j++)
1734 ((int32_t*)(c->chrUPixBuf[i]))[j] = 1<<18;
1736 for(j=0; j<dst_stride+1; j++)
1737 ((int16_t*)(c->chrUPixBuf[i]))[j] = 1<<14;
1739 av_assert0(c->chrDstH <= dstH);
1741 if (flags & SWS_PRINT_INFO) {
1742 const char *scaler = NULL, *cpucaps;
1744 for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {
1745 if (flags & scale_algorithms[i].flag) {
1746 scaler = scale_algorithms[i].description;
1751 scaler = "ehh flags invalid?!";
1752 av_log(c, AV_LOG_INFO, "%s scaler, from %s to %s%s ",
1754 av_get_pix_fmt_name(srcFormat),
1756 dstFormat == AV_PIX_FMT_BGR555 || dstFormat == AV_PIX_FMT_BGR565 ||
1757 dstFormat == AV_PIX_FMT_RGB444BE || dstFormat == AV_PIX_FMT_RGB444LE ||
1758 dstFormat == AV_PIX_FMT_BGR444BE || dstFormat == AV_PIX_FMT_BGR444LE ?
1763 av_get_pix_fmt_name(dstFormat));
1765 if (INLINE_MMXEXT(cpu_flags))
1767 else if (INLINE_AMD3DNOW(cpu_flags))
1769 else if (INLINE_MMX(cpu_flags))
1771 else if (PPC_ALTIVEC(cpu_flags))
1772 cpucaps = "AltiVec";
1776 av_log(c, AV_LOG_INFO, "using %s\n", cpucaps);
1778 av_log(c, AV_LOG_VERBOSE, "%dx%d -> %dx%d\n", srcW, srcH, dstW, dstH);
1779 av_log(c, AV_LOG_DEBUG,
1780 "lum srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
1781 c->srcW, c->srcH, c->dstW, c->dstH, c->lumXInc, c->lumYInc);
1782 av_log(c, AV_LOG_DEBUG,
1783 "chr srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
1784 c->chrSrcW, c->chrSrcH, c->chrDstW, c->chrDstH,
1785 c->chrXInc, c->chrYInc);
1788 /* alpha blend special case, note this has been split via cascaded contexts if its scaled */
1789 if (unscaled && !usesHFilter && !usesVFilter &&
1790 c->alphablend != SWS_ALPHA_BLEND_NONE &&
1791 isALPHA(srcFormat) &&
1792 (c->srcRange == c->dstRange || isAnyRGB(dstFormat)) &&
1793 alphaless_fmt(srcFormat) == dstFormat
1795 c->swscale = ff_sws_alphablendaway;
1797 if (flags & SWS_PRINT_INFO)
1798 av_log(c, AV_LOG_INFO,
1799 "using alpha blendaway %s -> %s special converter\n",
1800 av_get_pix_fmt_name(srcFormat), av_get_pix_fmt_name(dstFormat));
1804 /* unscaled special cases */
1805 if (unscaled && !usesHFilter && !usesVFilter &&
1806 (c->srcRange == c->dstRange || isAnyRGB(dstFormat))) {
1807 ff_get_unscaled_swscale(c);
1810 if (flags & SWS_PRINT_INFO)
1811 av_log(c, AV_LOG_INFO,
1812 "using unscaled %s -> %s special converter\n",
1813 av_get_pix_fmt_name(srcFormat), av_get_pix_fmt_name(dstFormat));
1818 c->swscale = ff_getSwsFunc(c);
1819 return ff_init_filters(c);
1820 fail: // FIXME replace things by appropriate error codes
1821 if (ret == RETCODE_USE_CASCADE) {
1822 int tmpW = sqrt(srcW * (int64_t)dstW);
1823 int tmpH = sqrt(srcH * (int64_t)dstH);
1824 enum AVPixelFormat tmpFormat = AV_PIX_FMT_YUV420P;
1826 if (isALPHA(srcFormat))
1827 tmpFormat = AV_PIX_FMT_YUVA420P;
1829 if (srcW*(int64_t)srcH <= 4LL*dstW*dstH)
1830 return AVERROR(EINVAL);
1832 ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
1833 tmpW, tmpH, tmpFormat, 64);
1837 c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,
1838 tmpW, tmpH, tmpFormat,
1839 flags, srcFilter, NULL, c->param);
1840 if (!c->cascaded_context[0])
1843 c->cascaded_context[1] = sws_getContext(tmpW, tmpH, tmpFormat,
1844 dstW, dstH, dstFormat,
1845 flags, NULL, dstFilter, c->param);
1846 if (!c->cascaded_context[1])
1853 SwsContext *sws_alloc_set_opts(int srcW, int srcH, enum AVPixelFormat srcFormat,
1854 int dstW, int dstH, enum AVPixelFormat dstFormat,
1855 int flags, const double *param)
1859 if (!(c = sws_alloc_context()))
1867 c->srcFormat = srcFormat;
1868 c->dstFormat = dstFormat;
1871 c->param[0] = param[0];
1872 c->param[1] = param[1];
1878 SwsContext *sws_getContext(int srcW, int srcH, enum AVPixelFormat srcFormat,
1879 int dstW, int dstH, enum AVPixelFormat dstFormat,
1880 int flags, SwsFilter *srcFilter,
1881 SwsFilter *dstFilter, const double *param)
1885 c = sws_alloc_set_opts(srcW, srcH, srcFormat,
1886 dstW, dstH, dstFormat,
1891 if (sws_init_context(c, srcFilter, dstFilter) < 0) {
1899 static int isnan_vec(SwsVector *a)
1902 for (i=0; i<a->length; i++)
1903 if (isnan(a->coeff[i]))
1908 static void makenan_vec(SwsVector *a)
1911 for (i=0; i<a->length; i++)
1915 SwsFilter *sws_getDefaultFilter(float lumaGBlur, float chromaGBlur,
1916 float lumaSharpen, float chromaSharpen,
1917 float chromaHShift, float chromaVShift,
1920 SwsFilter *filter = av_malloc(sizeof(SwsFilter));
1924 if (lumaGBlur != 0.0) {
1925 filter->lumH = sws_getGaussianVec(lumaGBlur, 3.0);
1926 filter->lumV = sws_getGaussianVec(lumaGBlur, 3.0);
1928 filter->lumH = sws_getIdentityVec();
1929 filter->lumV = sws_getIdentityVec();
1932 if (chromaGBlur != 0.0) {
1933 filter->chrH = sws_getGaussianVec(chromaGBlur, 3.0);
1934 filter->chrV = sws_getGaussianVec(chromaGBlur, 3.0);
1936 filter->chrH = sws_getIdentityVec();
1937 filter->chrV = sws_getIdentityVec();
1940 if (!filter->lumH || !filter->lumV || !filter->chrH || !filter->chrV)
1943 if (chromaSharpen != 0.0) {
1944 SwsVector *id = sws_getIdentityVec();
1947 sws_scaleVec(filter->chrH, -chromaSharpen);
1948 sws_scaleVec(filter->chrV, -chromaSharpen);
1949 sws_addVec(filter->chrH, id);
1950 sws_addVec(filter->chrV, id);
1954 if (lumaSharpen != 0.0) {
1955 SwsVector *id = sws_getIdentityVec();
1958 sws_scaleVec(filter->lumH, -lumaSharpen);
1959 sws_scaleVec(filter->lumV, -lumaSharpen);
1960 sws_addVec(filter->lumH, id);
1961 sws_addVec(filter->lumV, id);
1965 if (chromaHShift != 0.0)
1966 sws_shiftVec(filter->chrH, (int)(chromaHShift + 0.5));
1968 if (chromaVShift != 0.0)
1969 sws_shiftVec(filter->chrV, (int)(chromaVShift + 0.5));
1971 sws_normalizeVec(filter->chrH, 1.0);
1972 sws_normalizeVec(filter->chrV, 1.0);
1973 sws_normalizeVec(filter->lumH, 1.0);
1974 sws_normalizeVec(filter->lumV, 1.0);
1976 if (isnan_vec(filter->chrH) ||
1977 isnan_vec(filter->chrV) ||
1978 isnan_vec(filter->lumH) ||
1979 isnan_vec(filter->lumV))
1983 sws_printVec2(filter->chrH, NULL, AV_LOG_DEBUG);
1985 sws_printVec2(filter->lumH, NULL, AV_LOG_DEBUG);
1990 sws_freeVec(filter->lumH);
1991 sws_freeVec(filter->lumV);
1992 sws_freeVec(filter->chrH);
1993 sws_freeVec(filter->chrV);
1998 SwsVector *sws_allocVec(int length)
2002 if(length <= 0 || length > INT_MAX/ sizeof(double))
2005 vec = av_malloc(sizeof(SwsVector));
2008 vec->length = length;
2009 vec->coeff = av_malloc(sizeof(double) * length);
2015 SwsVector *sws_getGaussianVec(double variance, double quality)
2017 const int length = (int)(variance * quality + 0.5) | 1;
2019 double middle = (length - 1) * 0.5;
2022 if(variance < 0 || quality < 0)
2025 vec = sws_allocVec(length);
2030 for (i = 0; i < length; i++) {
2031 double dist = i - middle;
2032 vec->coeff[i] = exp(-dist * dist / (2 * variance * variance)) /
2033 sqrt(2 * variance * M_PI);
2036 sws_normalizeVec(vec, 1.0);
2041 SwsVector *sws_getConstVec(double c, int length)
2044 SwsVector *vec = sws_allocVec(length);
2049 for (i = 0; i < length; i++)
2055 SwsVector *sws_getIdentityVec(void)
2057 return sws_getConstVec(1.0, 1);
2060 static double sws_dcVec(SwsVector *a)
2065 for (i = 0; i < a->length; i++)
2071 void sws_scaleVec(SwsVector *a, double scalar)
2075 for (i = 0; i < a->length; i++)
2076 a->coeff[i] *= scalar;
2079 void sws_normalizeVec(SwsVector *a, double height)
2081 sws_scaleVec(a, height / sws_dcVec(a));
2084 static SwsVector *sws_getConvVec(SwsVector *a, SwsVector *b)
2086 int length = a->length + b->length - 1;
2088 SwsVector *vec = sws_getConstVec(0.0, length);
2093 for (i = 0; i < a->length; i++) {
2094 for (j = 0; j < b->length; j++) {
2095 vec->coeff[i + j] += a->coeff[i] * b->coeff[j];
2102 static SwsVector *sws_sumVec(SwsVector *a, SwsVector *b)
2104 int length = FFMAX(a->length, b->length);
2106 SwsVector *vec = sws_getConstVec(0.0, length);
2111 for (i = 0; i < a->length; i++)
2112 vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];
2113 for (i = 0; i < b->length; i++)
2114 vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] += b->coeff[i];
2119 static SwsVector *sws_diffVec(SwsVector *a, SwsVector *b)
2121 int length = FFMAX(a->length, b->length);
2123 SwsVector *vec = sws_getConstVec(0.0, length);
2128 for (i = 0; i < a->length; i++)
2129 vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];
2130 for (i = 0; i < b->length; i++)
2131 vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] -= b->coeff[i];
2136 /* shift left / or right if "shift" is negative */
2137 static SwsVector *sws_getShiftedVec(SwsVector *a, int shift)
2139 int length = a->length + FFABS(shift) * 2;
2141 SwsVector *vec = sws_getConstVec(0.0, length);
2146 for (i = 0; i < a->length; i++) {
2147 vec->coeff[i + (length - 1) / 2 -
2148 (a->length - 1) / 2 - shift] = a->coeff[i];
2154 void sws_shiftVec(SwsVector *a, int shift)
2156 SwsVector *shifted = sws_getShiftedVec(a, shift);
2162 a->coeff = shifted->coeff;
2163 a->length = shifted->length;
2167 void sws_addVec(SwsVector *a, SwsVector *b)
2169 SwsVector *sum = sws_sumVec(a, b);
2175 a->coeff = sum->coeff;
2176 a->length = sum->length;
2180 void sws_subVec(SwsVector *a, SwsVector *b)
2182 SwsVector *diff = sws_diffVec(a, b);
2188 a->coeff = diff->coeff;
2189 a->length = diff->length;
2193 void sws_convVec(SwsVector *a, SwsVector *b)
2195 SwsVector *conv = sws_getConvVec(a, b);
2201 a->coeff = conv->coeff;
2202 a->length = conv->length;
2206 SwsVector *sws_cloneVec(SwsVector *a)
2208 SwsVector *vec = sws_allocVec(a->length);
2213 memcpy(vec->coeff, a->coeff, a->length * sizeof(*a->coeff));
2218 void sws_printVec2(SwsVector *a, AVClass *log_ctx, int log_level)
2225 for (i = 0; i < a->length; i++)
2226 if (a->coeff[i] > max)
2229 for (i = 0; i < a->length; i++)
2230 if (a->coeff[i] < min)
2235 for (i = 0; i < a->length; i++) {
2236 int x = (int)((a->coeff[i] - min) * 60.0 / range + 0.5);
2237 av_log(log_ctx, log_level, "%1.3f ", a->coeff[i]);
2239 av_log(log_ctx, log_level, " ");
2240 av_log(log_ctx, log_level, "|\n");
2244 void sws_freeVec(SwsVector *a)
2248 av_freep(&a->coeff);
2253 void sws_freeFilter(SwsFilter *filter)
2258 sws_freeVec(filter->lumH);
2259 sws_freeVec(filter->lumV);
2260 sws_freeVec(filter->chrH);
2261 sws_freeVec(filter->chrV);
2265 void sws_freeContext(SwsContext *c)
2272 for (i = 0; i < c->vLumBufSize; i++)
2273 av_freep(&c->lumPixBuf[i]);
2274 av_freep(&c->lumPixBuf);
2277 if (c->chrUPixBuf) {
2278 for (i = 0; i < c->vChrBufSize; i++)
2279 av_freep(&c->chrUPixBuf[i]);
2280 av_freep(&c->chrUPixBuf);
2281 av_freep(&c->chrVPixBuf);
2284 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
2285 for (i = 0; i < c->vLumBufSize; i++)
2286 av_freep(&c->alpPixBuf[i]);
2287 av_freep(&c->alpPixBuf);
2290 for (i = 0; i < 4; i++)
2291 av_freep(&c->dither_error[i]);
2293 av_freep(&c->vLumFilter);
2294 av_freep(&c->vChrFilter);
2295 av_freep(&c->hLumFilter);
2296 av_freep(&c->hChrFilter);
2298 av_freep(&c->vYCoeffsBank);
2299 av_freep(&c->vCCoeffsBank);
2302 av_freep(&c->vLumFilterPos);
2303 av_freep(&c->vChrFilterPos);
2304 av_freep(&c->hLumFilterPos);
2305 av_freep(&c->hChrFilterPos);
2309 if (c->lumMmxextFilterCode)
2310 munmap(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize);
2311 if (c->chrMmxextFilterCode)
2312 munmap(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize);
2313 #elif HAVE_VIRTUALALLOC
2314 if (c->lumMmxextFilterCode)
2315 VirtualFree(c->lumMmxextFilterCode, 0, MEM_RELEASE);
2316 if (c->chrMmxextFilterCode)
2317 VirtualFree(c->chrMmxextFilterCode, 0, MEM_RELEASE);
2319 av_free(c->lumMmxextFilterCode);
2320 av_free(c->chrMmxextFilterCode);
2322 c->lumMmxextFilterCode = NULL;
2323 c->chrMmxextFilterCode = NULL;
2324 #endif /* HAVE_MMX_INLINE */
2326 av_freep(&c->yuvTable);
2327 av_freep(&c->formatConvBuffer);
2329 sws_freeContext(c->cascaded_context[0]);
2330 sws_freeContext(c->cascaded_context[1]);
2331 sws_freeContext(c->cascaded_context[2]);
2332 memset(c->cascaded_context, 0, sizeof(c->cascaded_context));
2333 av_freep(&c->cascaded_tmp[0]);
2334 av_freep(&c->cascaded1_tmp[0]);
2336 av_freep(&c->gamma);
2337 av_freep(&c->inv_gamma);
2344 struct SwsContext *sws_getCachedContext(struct SwsContext *context, int srcW,
2345 int srcH, enum AVPixelFormat srcFormat,
2347 enum AVPixelFormat dstFormat, int flags,
2348 SwsFilter *srcFilter,
2349 SwsFilter *dstFilter,
2350 const double *param)
2352 static const double default_param[2] = { SWS_PARAM_DEFAULT,
2353 SWS_PARAM_DEFAULT };
2354 int64_t src_h_chr_pos = -513, dst_h_chr_pos = -513,
2355 src_v_chr_pos = -513, dst_v_chr_pos = -513;
2358 param = default_param;
2361 (context->srcW != srcW ||
2362 context->srcH != srcH ||
2363 context->srcFormat != srcFormat ||
2364 context->dstW != dstW ||
2365 context->dstH != dstH ||
2366 context->dstFormat != dstFormat ||
2367 context->flags != flags ||
2368 context->param[0] != param[0] ||
2369 context->param[1] != param[1])) {
2371 av_opt_get_int(context, "src_h_chr_pos", 0, &src_h_chr_pos);
2372 av_opt_get_int(context, "src_v_chr_pos", 0, &src_v_chr_pos);
2373 av_opt_get_int(context, "dst_h_chr_pos", 0, &dst_h_chr_pos);
2374 av_opt_get_int(context, "dst_v_chr_pos", 0, &dst_v_chr_pos);
2375 sws_freeContext(context);
2380 if (!(context = sws_alloc_context()))
2382 context->srcW = srcW;
2383 context->srcH = srcH;
2384 context->srcFormat = srcFormat;
2385 context->dstW = dstW;
2386 context->dstH = dstH;
2387 context->dstFormat = dstFormat;
2388 context->flags = flags;
2389 context->param[0] = param[0];
2390 context->param[1] = param[1];
2392 av_opt_set_int(context, "src_h_chr_pos", src_h_chr_pos, 0);
2393 av_opt_set_int(context, "src_v_chr_pos", src_v_chr_pos, 0);
2394 av_opt_set_int(context, "dst_h_chr_pos", dst_h_chr_pos, 0);
2395 av_opt_set_int(context, "dst_v_chr_pos", dst_v_chr_pos, 0);
2397 if (sws_init_context(context, srcFilter, dstFilter) < 0) {
2398 sws_freeContext(context);