2 * Copyright (C) 2001-2003 Michael Niedermayer <michaelni@gmx.at>
4 * This file is part of Libav.
6 * Libav 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 * Libav 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 Libav; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21 #ifndef SWSCALE_SWSCALE_INTERNAL_H
22 #define SWSCALE_SWSCALE_INTERNAL_H
30 #include "libavutil/avutil.h"
31 #include "libavutil/log.h"
32 #include "libavutil/pixfmt.h"
34 #define STR(s) AV_TOSTRING(s) //AV_STRINGIFY is too long
36 #define FAST_BGR2YV12 //use 7-bit instead of 15-bit coefficients
38 #define MAX_FILTER_SIZE 256
41 #define ALT32_CORR (-1)
58 typedef int (*SwsFunc)(struct SwsContext *context, const uint8_t* src[],
59 int srcStride[], int srcSliceY, int srcSliceH,
60 uint8_t* dst[], int dstStride[]);
63 * Write one line of horizontally scaled Y/U/V/A to planar output
64 * without any additional vertical scaling (or point-scaling).
66 * @param c SWS scaling context
67 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
68 * 19-bit for 16bit output (in int32_t)
69 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
70 * 19-bit for 16bit output (in int32_t)
71 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
72 * 19-bit for 16bit output (in int32_t)
73 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
74 * 19-bit for 16bit output (in int32_t)
75 * @param dest pointer to the 4 output planes (Y/U/V/A). For >8bit
76 * output, this is in uint16_t
77 * @param dstW width of dest[0], dest[3], lumSrc and alpSrc in pixels
78 * @param chrDstW width of dest[1], dest[2], chrUSrc and chrVSrc
80 typedef void (*yuv2planar1_fn) (struct SwsContext *c,
81 const int16_t *lumSrc, const int16_t *chrUSrc,
82 const int16_t *chrVSrc, const int16_t *alpSrc,
83 uint8_t *dest[4], int dstW, int chrDstW);
85 * Write one line of horizontally scaled Y/U/V/A to planar output
86 * with multi-point vertical scaling between input pixels.
88 * @param c SWS scaling context
89 * @param lumFilter vertical luma/alpha scaling coefficients, 12bit [0,4096]
90 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
91 * 19-bit for 16bit output (in int32_t)
92 * @param lumFilterSize number of vertical luma/alpha input lines to scale
93 * @param chrFilter vertical chroma scaling coefficients, 12bit [0,4096]
94 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
95 * 19-bit for 16bit output (in int32_t)
96 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
97 * 19-bit for 16bit output (in int32_t)
98 * @param chrFilterSize number of vertical chroma input lines to scale
99 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
100 * 19-bit for 16bit output (in int32_t)
101 * @param dest pointer to the 4 output planes (Y/U/V/A). For >8bit
102 * output, this is in uint16_t
103 * @param dstW width of dest[0], dest[3], lumSrc and alpSrc in pixels
104 * @param chrDstW width of dest[1], dest[2], chrUSrc and chrVSrc
106 typedef void (*yuv2planarX_fn) (struct SwsContext *c, const int16_t *lumFilter,
107 const int16_t **lumSrc, int lumFilterSize,
108 const int16_t *chrFilter, const int16_t **chrUSrc,
109 const int16_t **chrVSrc, int chrFilterSize,
110 const int16_t **alpSrc, uint8_t *dest[4],
111 int dstW, int chrDstW);
113 * Write one line of horizontally scaled Y/U/V/A to packed-pixel YUV/RGB
114 * output without any additional vertical scaling (or point-scaling). Note
115 * that this function may do chroma scaling, see the "uvalpha" argument.
117 * @param c SWS scaling context
118 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
119 * 19-bit for 16bit output (in int32_t)
120 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
121 * 19-bit for 16bit output (in int32_t)
122 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
123 * 19-bit for 16bit output (in int32_t)
124 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
125 * 19-bit for 16bit output (in int32_t)
126 * @param dest pointer to the output plane. For 16bit output, this is
128 * @param dstW width of lumSrc and alpSrc in pixels, number of pixels
129 * to write into dest[]
130 * @param uvalpha chroma scaling coefficient for the second line of chroma
131 * pixels, either 2048 or 0. If 0, one chroma input is used
132 * for 2 output pixels (or if the SWS_FLAG_FULL_CHR_INT flag
133 * is set, it generates 1 output pixel). If 2048, two chroma
134 * input pixels should be averaged for 2 output pixels (this
135 * only happens if SWS_FLAG_FULL_CHR_INT is not set)
136 * @param y vertical line number for this output. This does not need
137 * to be used to calculate the offset in the destination,
138 * but can be used to generate comfort noise using dithering
139 * for some output formats.
141 typedef void (*yuv2packed1_fn) (struct SwsContext *c, const int16_t *lumSrc,
142 const int16_t *chrUSrc[2], const int16_t *chrVSrc[2],
143 const int16_t *alpSrc, uint8_t *dest,
144 int dstW, int uvalpha, int y);
146 * Write one line of horizontally scaled Y/U/V/A to packed-pixel YUV/RGB
147 * output by doing bilinear scaling between two input lines.
149 * @param c SWS scaling context
150 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
151 * 19-bit for 16bit output (in int32_t)
152 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
153 * 19-bit for 16bit output (in int32_t)
154 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
155 * 19-bit for 16bit output (in int32_t)
156 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
157 * 19-bit for 16bit output (in int32_t)
158 * @param dest pointer to the output plane. For 16bit output, this is
160 * @param dstW width of lumSrc and alpSrc in pixels, number of pixels
161 * to write into dest[]
162 * @param yalpha luma/alpha scaling coefficients for the second input line.
163 * The first line's coefficients can be calculated by using
165 * @param uvalpha chroma scaling coefficient for the second input line. The
166 * first line's coefficients can be calculated by using
168 * @param y vertical line number for this output. This does not need
169 * to be used to calculate the offset in the destination,
170 * but can be used to generate comfort noise using dithering
171 * for some output formats.
173 typedef void (*yuv2packed2_fn) (struct SwsContext *c, const int16_t *lumSrc[2],
174 const int16_t *chrUSrc[2], const int16_t *chrVSrc[2],
175 const int16_t *alpSrc[2], uint8_t *dest,
176 int dstW, int yalpha, int uvalpha, int y);
178 * Write one line of horizontally scaled Y/U/V/A to packed-pixel YUV/RGB
179 * output by doing multi-point vertical scaling between input pixels.
181 * @param c SWS scaling context
182 * @param lumFilter vertical luma/alpha scaling coefficients, 12bit [0,4096]
183 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
184 * 19-bit for 16bit output (in int32_t)
185 * @param lumFilterSize number of vertical luma/alpha input lines to scale
186 * @param chrFilter vertical chroma scaling coefficients, 12bit [0,4096]
187 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
188 * 19-bit for 16bit output (in int32_t)
189 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
190 * 19-bit for 16bit output (in int32_t)
191 * @param chrFilterSize number of vertical chroma input lines to scale
192 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
193 * 19-bit for 16bit output (in int32_t)
194 * @param dest pointer to the output plane. For 16bit output, this is
196 * @param dstW width of lumSrc and alpSrc in pixels, number of pixels
197 * to write into dest[]
198 * @param y vertical line number for this output. This does not need
199 * to be used to calculate the offset in the destination,
200 * but can be used to generate comfort noise using dithering
201 * or some output formats.
203 typedef void (*yuv2packedX_fn) (struct SwsContext *c, const int16_t *lumFilter,
204 const int16_t **lumSrc, int lumFilterSize,
205 const int16_t *chrFilter, const int16_t **chrUSrc,
206 const int16_t **chrVSrc, int chrFilterSize,
207 const int16_t **alpSrc, uint8_t *dest,
210 /* This struct should be aligned on at least a 32-byte boundary. */
211 typedef struct SwsContext {
213 * info on struct for av_log
215 const AVClass *av_class;
218 * Note that src, dst, srcStride, dstStride will be copied in the
219 * sws_scale() wrapper so they can be freely modified here.
222 int srcW; ///< Width of source luma/alpha planes.
223 int srcH; ///< Height of source luma/alpha planes.
224 int dstH; ///< Height of destination luma/alpha planes.
225 int chrSrcW; ///< Width of source chroma planes.
226 int chrSrcH; ///< Height of source chroma planes.
227 int chrDstW; ///< Width of destination chroma planes.
228 int chrDstH; ///< Height of destination chroma planes.
229 int lumXInc, chrXInc;
230 int lumYInc, chrYInc;
231 enum PixelFormat dstFormat; ///< Destination pixel format.
232 enum PixelFormat srcFormat; ///< Source pixel format.
233 int dstFormatBpp; ///< Number of bits per pixel of the destination pixel format.
234 int srcFormatBpp; ///< Number of bits per pixel of the source pixel format.
236 int chrSrcHSubSample; ///< Binary logarithm of horizontal subsampling factor between luma/alpha and chroma planes in source image.
237 int chrSrcVSubSample; ///< Binary logarithm of vertical subsampling factor between luma/alpha and chroma planes in source image.
238 int chrDstHSubSample; ///< Binary logarithm of horizontal subsampling factor between luma/alpha and chroma planes in destination image.
239 int chrDstVSubSample; ///< Binary logarithm of vertical subsampling factor between luma/alpha and chroma planes in destination image.
240 int vChrDrop; ///< Binary logarithm of extra vertical subsampling factor in source image chroma planes specified by user.
241 int sliceDir; ///< Direction that slices are fed to the scaler (1 = top-to-bottom, -1 = bottom-to-top).
242 double param[2]; ///< Input parameters for scaling algorithms that need them.
244 uint32_t pal_yuv[256];
245 uint32_t pal_rgb[256];
248 * @name Scaled horizontal lines ring buffer.
249 * The horizontal scaler keeps just enough scaled lines in a ring buffer
250 * so they may be passed to the vertical scaler. The pointers to the
251 * allocated buffers for each line are duplicated in sequence in the ring
252 * buffer to simplify indexing and avoid wrapping around between lines
253 * inside the vertical scaler code. The wrapping is done before the
254 * vertical scaler is called.
257 int16_t **lumPixBuf; ///< Ring buffer for scaled horizontal luma plane lines to be fed to the vertical scaler.
258 int16_t **chrUPixBuf; ///< Ring buffer for scaled horizontal chroma plane lines to be fed to the vertical scaler.
259 int16_t **chrVPixBuf; ///< Ring buffer for scaled horizontal chroma plane lines to be fed to the vertical scaler.
260 int16_t **alpPixBuf; ///< Ring buffer for scaled horizontal alpha plane lines to be fed to the vertical scaler.
261 int vLumBufSize; ///< Number of vertical luma/alpha lines allocated in the ring buffer.
262 int vChrBufSize; ///< Number of vertical chroma lines allocated in the ring buffer.
263 int lastInLumBuf; ///< Last scaled horizontal luma/alpha line from source in the ring buffer.
264 int lastInChrBuf; ///< Last scaled horizontal chroma line from source in the ring buffer.
265 int lumBufIndex; ///< Index in ring buffer of the last scaled horizontal luma/alpha line from source.
266 int chrBufIndex; ///< Index in ring buffer of the last scaled horizontal chroma line from source.
269 uint8_t *formatConvBuffer;
272 * @name Horizontal and vertical filters.
273 * To better understand the following fields, here is a pseudo-code of
274 * their usage in filtering a horizontal line:
276 * for (i = 0; i < width; i++) {
278 * for (j = 0; j < filterSize; j++)
279 * dst[i] += src[ filterPos[i] + j ] * filter[ filterSize * i + j ];
280 * dst[i] >>= FRAC_BITS; // The actual implementation is fixed-point.
285 int16_t *hLumFilter; ///< Array of horizontal filter coefficients for luma/alpha planes.
286 int16_t *hChrFilter; ///< Array of horizontal filter coefficients for chroma planes.
287 int16_t *vLumFilter; ///< Array of vertical filter coefficients for luma/alpha planes.
288 int16_t *vChrFilter; ///< Array of vertical filter coefficients for chroma planes.
289 int16_t *hLumFilterPos; ///< Array of horizontal filter starting positions for each dst[i] for luma/alpha planes.
290 int16_t *hChrFilterPos; ///< Array of horizontal filter starting positions for each dst[i] for chroma planes.
291 int16_t *vLumFilterPos; ///< Array of vertical filter starting positions for each dst[i] for luma/alpha planes.
292 int16_t *vChrFilterPos; ///< Array of vertical filter starting positions for each dst[i] for chroma planes.
293 int hLumFilterSize; ///< Horizontal filter size for luma/alpha pixels.
294 int hChrFilterSize; ///< Horizontal filter size for chroma pixels.
295 int vLumFilterSize; ///< Vertical filter size for luma/alpha pixels.
296 int vChrFilterSize; ///< Vertical filter size for chroma pixels.
299 int lumMmx2FilterCodeSize; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code size for luma/alpha planes.
300 int chrMmx2FilterCodeSize; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code size for chroma planes.
301 uint8_t *lumMmx2FilterCode; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code for luma/alpha planes.
302 uint8_t *chrMmx2FilterCode; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code for chroma planes.
306 int dstY; ///< Last destination vertical line output from last slice.
307 int flags; ///< Flags passed by the user to select scaler algorithm, optimizations, subsampling, etc...
308 void * yuvTable; // pointer to the yuv->rgb table start so it can be freed()
309 uint8_t * table_rV[256];
310 uint8_t * table_gU[256];
312 uint8_t * table_bU[256];
315 int contrast, brightness, saturation; // for sws_getColorspaceDetails
316 int srcColorspaceTable[4];
317 int dstColorspaceTable[4];
318 int srcRange; ///< 0 = MPG YUV range, 1 = JPG YUV range (source image).
319 int dstRange; ///< 0 = MPG YUV range, 1 = JPG YUV range (destination image).
320 int yuv2rgb_y_offset;
322 int yuv2rgb_v2r_coeff;
323 int yuv2rgb_v2g_coeff;
324 int yuv2rgb_u2g_coeff;
325 int yuv2rgb_u2b_coeff;
327 #define RED_DITHER "0*8"
328 #define GREEN_DITHER "1*8"
329 #define BLUE_DITHER "2*8"
330 #define Y_COEFF "3*8"
331 #define VR_COEFF "4*8"
332 #define UB_COEFF "5*8"
333 #define VG_COEFF "6*8"
334 #define UG_COEFF "7*8"
335 #define Y_OFFSET "8*8"
336 #define U_OFFSET "9*8"
337 #define V_OFFSET "10*8"
338 #define LUM_MMX_FILTER_OFFSET "11*8"
339 #define CHR_MMX_FILTER_OFFSET "11*8+4*4*256"
340 #define DSTW_OFFSET "11*8+4*4*256*2" //do not change, it is hardcoded in the ASM
341 #define ESP_OFFSET "11*8+4*4*256*2+8"
342 #define VROUNDER_OFFSET "11*8+4*4*256*2+16"
343 #define U_TEMP "11*8+4*4*256*2+24"
344 #define V_TEMP "11*8+4*4*256*2+32"
345 #define Y_TEMP "11*8+4*4*256*2+40"
346 #define ALP_MMX_FILTER_OFFSET "11*8+4*4*256*2+48"
347 #define UV_OFF_PX "11*8+4*4*256*3+48"
348 #define UV_OFF_BYTE "11*8+4*4*256*3+56"
349 #define DITHER16 "11*8+4*4*256*3+64"
350 #define DITHER32 "11*8+4*4*256*3+80"
352 DECLARE_ALIGNED(8, uint64_t, redDither);
353 DECLARE_ALIGNED(8, uint64_t, greenDither);
354 DECLARE_ALIGNED(8, uint64_t, blueDither);
356 DECLARE_ALIGNED(8, uint64_t, yCoeff);
357 DECLARE_ALIGNED(8, uint64_t, vrCoeff);
358 DECLARE_ALIGNED(8, uint64_t, ubCoeff);
359 DECLARE_ALIGNED(8, uint64_t, vgCoeff);
360 DECLARE_ALIGNED(8, uint64_t, ugCoeff);
361 DECLARE_ALIGNED(8, uint64_t, yOffset);
362 DECLARE_ALIGNED(8, uint64_t, uOffset);
363 DECLARE_ALIGNED(8, uint64_t, vOffset);
364 int32_t lumMmxFilter[4*MAX_FILTER_SIZE];
365 int32_t chrMmxFilter[4*MAX_FILTER_SIZE];
366 int dstW; ///< Width of destination luma/alpha planes.
367 DECLARE_ALIGNED(8, uint64_t, esp);
368 DECLARE_ALIGNED(8, uint64_t, vRounder);
369 DECLARE_ALIGNED(8, uint64_t, u_temp);
370 DECLARE_ALIGNED(8, uint64_t, v_temp);
371 DECLARE_ALIGNED(8, uint64_t, y_temp);
372 int32_t alpMmxFilter[4*MAX_FILTER_SIZE];
373 // alignment of these values is not necessary, but merely here
374 // to maintain the same offset across x8632 and x86-64. Once we
375 // use proper offset macros in the asm, they can be removed.
376 DECLARE_ALIGNED(8, ptrdiff_t, uv_off_px); ///< offset (in pixels) between u and v planes
377 DECLARE_ALIGNED(8, ptrdiff_t, uv_off_byte); ///< offset (in bytes) between u and v planes
378 DECLARE_ALIGNED(8, uint16_t, dither16)[8];
379 DECLARE_ALIGNED(8, uint32_t, dither32)[8];
381 const uint8_t *chrDither8, *lumDither8;
384 vector signed short CY;
385 vector signed short CRV;
386 vector signed short CBU;
387 vector signed short CGU;
388 vector signed short CGV;
389 vector signed short OY;
390 vector unsigned short CSHIFT;
391 vector signed short *vYCoeffsBank, *vCCoeffsBank;
395 DECLARE_ALIGNED(4, uint32_t, oy);
396 DECLARE_ALIGNED(4, uint32_t, oc);
397 DECLARE_ALIGNED(4, uint32_t, zero);
398 DECLARE_ALIGNED(4, uint32_t, cy);
399 DECLARE_ALIGNED(4, uint32_t, crv);
400 DECLARE_ALIGNED(4, uint32_t, rmask);
401 DECLARE_ALIGNED(4, uint32_t, cbu);
402 DECLARE_ALIGNED(4, uint32_t, bmask);
403 DECLARE_ALIGNED(4, uint32_t, cgu);
404 DECLARE_ALIGNED(4, uint32_t, cgv);
405 DECLARE_ALIGNED(4, uint32_t, gmask);
409 DECLARE_ALIGNED(8, uint64_t, sparc_coeffs)[10];
412 /* function pointers for swScale() */
413 yuv2planar1_fn yuv2yuv1;
414 yuv2planarX_fn yuv2yuvX;
415 yuv2packed1_fn yuv2packed1;
416 yuv2packed2_fn yuv2packed2;
417 yuv2packedX_fn yuv2packedX;
419 void (*lumToYV12)(uint8_t *dst, const uint8_t *src,
420 int width, uint32_t *pal); ///< Unscaled conversion of luma plane to YV12 for horizontal scaler.
421 void (*alpToYV12)(uint8_t *dst, const uint8_t *src,
422 int width, uint32_t *pal); ///< Unscaled conversion of alpha plane to YV12 for horizontal scaler.
423 void (*chrToYV12)(uint8_t *dstU, uint8_t *dstV,
424 const uint8_t *src1, const uint8_t *src2,
425 int width, uint32_t *pal); ///< Unscaled conversion of chroma planes to YV12 for horizontal scaler.
427 * Scale one horizontal line of input data using a bilinear filter
428 * to produce one line of output data. Compared to SwsContext->hScale(),
429 * please take note of the following caveats when using these:
430 * - Scaling is done using only 7bit instead of 14bit coefficients.
431 * - You can use no more than 5 input pixels to produce 4 output
432 * pixels. Therefore, this filter should not be used for downscaling
433 * by more than ~20% in width (because that equals more than 5/4th
434 * downscaling and thus more than 5 pixels input per 4 pixels output).
435 * - In general, bilinear filters create artifacts during downscaling
436 * (even when <20%), because one output pixel will span more than one
437 * input pixel, and thus some pixels will need edges of both neighbor
438 * pixels to interpolate the output pixel. Since you can use at most
439 * two input pixels per output pixel in bilinear scaling, this is
440 * impossible and thus downscaling by any size will create artifacts.
441 * To enable this type of scaling, set SWS_FLAG_FAST_BILINEAR
442 * in SwsContext->flags.
445 void (*hyscale_fast)(struct SwsContext *c,
446 int16_t *dst, int dstWidth,
447 const uint8_t *src, int srcW, int xInc);
448 void (*hcscale_fast)(struct SwsContext *c,
449 int16_t *dst1, int16_t *dst2, int dstWidth,
450 const uint8_t *src1, const uint8_t *src2,
455 * Scale one horizontal line of input data using a filter over the input
456 * lines, to produce one (differently sized) line of output data.
458 * @param dst pointer to destination buffer for horizontally scaled
459 * data. If the number of bits per component of one
460 * destination pixel (SwsContext->dstBpc) is <= 10, data
461 * will be 15bpc in 16bits (int16_t) width. Else (i.e.
462 * SwsContext->dstBpc == 16), data will be 19bpc in
463 * 32bits (int32_t) width.
464 * @param dstW width of destination image
465 * @param src pointer to source data to be scaled. If the number of
466 * bits per component of a source pixel (SwsContext->srcBpc)
467 * is 8, this is 8bpc in 8bits (uint8_t) width. Else
468 * (i.e. SwsContext->dstBpc > 8), this is native depth
469 * in 16bits (uint16_t) width. In other words, for 9-bit
470 * YUV input, this is 9bpc, for 10-bit YUV input, this is
471 * 10bpc, and for 16-bit RGB or YUV, this is 16bpc.
472 * @param filter filter coefficients to be used per output pixel for
473 * scaling. This contains 14bpp filtering coefficients.
474 * Guaranteed to contain dstW * filterSize entries.
475 * @param filterPos position of the first input pixel to be used for
476 * each output pixel during scaling. Guaranteed to
477 * contain dstW entries.
478 * @param filterSize the number of input coefficients to be used (and
479 * thus the number of input pixels to be used) for
480 * creating a single output pixel. Is aligned to 4
481 * (and input coefficients thus padded with zeroes)
482 * to simplify creating SIMD code.
485 void (*hyScale)(struct SwsContext *c, int16_t *dst, int dstW, const uint8_t *src,
486 const int16_t *filter, const int16_t *filterPos,
488 void (*hcScale)(struct SwsContext *c, int16_t *dst, int dstW, const uint8_t *src,
489 const int16_t *filter, const int16_t *filterPos,
493 void (*lumConvertRange)(int16_t *dst, int width); ///< Color range conversion function for luma plane if needed.
494 void (*chrConvertRange)(int16_t *dst1, int16_t *dst2, int width); ///< Color range conversion function for chroma planes if needed.
496 int needs_hcscale; ///< Set if there are chroma planes to be converted.
499 //FIXME check init (where 0)
501 SwsFunc ff_yuv2rgb_get_func_ptr(SwsContext *c);
502 int ff_yuv2rgb_c_init_tables(SwsContext *c, const int inv_table[4],
503 int fullRange, int brightness,
504 int contrast, int saturation);
506 void ff_yuv2rgb_init_tables_altivec(SwsContext *c, const int inv_table[4],
507 int brightness, int contrast, int saturation);
508 void updateMMXDitherTables(SwsContext *c, int dstY, int lumBufIndex, int chrBufIndex,
509 int lastInLumBuf, int lastInChrBuf);
511 SwsFunc ff_yuv2rgb_init_mmx(SwsContext *c);
512 SwsFunc ff_yuv2rgb_init_vis(SwsContext *c);
513 SwsFunc ff_yuv2rgb_init_mlib(SwsContext *c);
514 SwsFunc ff_yuv2rgb_init_altivec(SwsContext *c);
515 SwsFunc ff_yuv2rgb_get_func_ptr_bfin(SwsContext *c);
516 void ff_bfin_get_unscaled_swscale(SwsContext *c);
518 const char *sws_format_name(enum PixelFormat format);
520 //FIXME replace this with something faster
521 #define is16BPS(x) ( \
522 (x)==PIX_FMT_GRAY16BE \
523 || (x)==PIX_FMT_GRAY16LE \
524 || (x)==PIX_FMT_BGR48BE \
525 || (x)==PIX_FMT_BGR48LE \
526 || (x)==PIX_FMT_RGB48BE \
527 || (x)==PIX_FMT_RGB48LE \
528 || (x)==PIX_FMT_YUV420P16LE \
529 || (x)==PIX_FMT_YUV422P16LE \
530 || (x)==PIX_FMT_YUV444P16LE \
531 || (x)==PIX_FMT_YUV420P16BE \
532 || (x)==PIX_FMT_YUV422P16BE \
533 || (x)==PIX_FMT_YUV444P16BE \
535 #define is9_OR_10BPS(x) ( \
536 (x)==PIX_FMT_YUV420P9LE \
537 || (x)==PIX_FMT_YUV420P9BE \
538 || (x)==PIX_FMT_YUV422P9LE \
539 || (x)==PIX_FMT_YUV422P9BE \
540 || (x)==PIX_FMT_YUV444P9BE \
541 || (x)==PIX_FMT_YUV444P9LE \
542 || (x)==PIX_FMT_YUV422P10BE \
543 || (x)==PIX_FMT_YUV422P10LE \
544 || (x)==PIX_FMT_YUV444P10BE \
545 || (x)==PIX_FMT_YUV444P10LE \
546 || (x)==PIX_FMT_YUV420P10LE \
547 || (x)==PIX_FMT_YUV420P10BE \
549 #define isBE(x) ((x)&1)
550 #define isPlanar8YUV(x) ( \
551 (x)==PIX_FMT_YUV410P \
552 || (x)==PIX_FMT_YUV420P \
553 || (x)==PIX_FMT_YUVA420P \
554 || (x)==PIX_FMT_YUV411P \
555 || (x)==PIX_FMT_YUV422P \
556 || (x)==PIX_FMT_YUV444P \
557 || (x)==PIX_FMT_YUV440P \
558 || (x)==PIX_FMT_NV12 \
559 || (x)==PIX_FMT_NV21 \
561 #define isPlanarYUV(x) ( \
563 || (x)==PIX_FMT_YUV420P9LE \
564 || (x)==PIX_FMT_YUV422P9LE \
565 || (x)==PIX_FMT_YUV444P9LE \
566 || (x)==PIX_FMT_YUV420P10LE \
567 || (x)==PIX_FMT_YUV422P10LE \
568 || (x)==PIX_FMT_YUV444P10LE \
569 || (x)==PIX_FMT_YUV420P16LE \
570 || (x)==PIX_FMT_YUV422P16LE \
571 || (x)==PIX_FMT_YUV444P16LE \
572 || (x)==PIX_FMT_YUV420P9BE \
573 || (x)==PIX_FMT_YUV422P9BE \
574 || (x)==PIX_FMT_YUV444P9BE \
575 || (x)==PIX_FMT_YUV420P10BE \
576 || (x)==PIX_FMT_YUV422P10BE \
577 || (x)==PIX_FMT_YUV444P10BE \
578 || (x)==PIX_FMT_YUV420P16BE \
579 || (x)==PIX_FMT_YUV422P16BE \
580 || (x)==PIX_FMT_YUV444P16BE \
583 (x)==PIX_FMT_UYVY422 \
584 || (x)==PIX_FMT_YUYV422 \
587 #define isGray(x) ( \
589 || (x)==PIX_FMT_Y400A \
590 || (x)==PIX_FMT_GRAY16BE \
591 || (x)==PIX_FMT_GRAY16LE \
593 #define isGray16(x) ( \
594 (x)==PIX_FMT_GRAY16BE \
595 || (x)==PIX_FMT_GRAY16LE \
597 #define isRGBinInt(x) ( \
598 (x)==PIX_FMT_RGB48BE \
599 || (x)==PIX_FMT_RGB48LE \
600 || (x)==PIX_FMT_RGB32 \
601 || (x)==PIX_FMT_RGB32_1 \
602 || (x)==PIX_FMT_RGB24 \
603 || (x)==PIX_FMT_RGB565BE \
604 || (x)==PIX_FMT_RGB565LE \
605 || (x)==PIX_FMT_RGB555BE \
606 || (x)==PIX_FMT_RGB555LE \
607 || (x)==PIX_FMT_RGB444BE \
608 || (x)==PIX_FMT_RGB444LE \
609 || (x)==PIX_FMT_RGB8 \
610 || (x)==PIX_FMT_RGB4 \
611 || (x)==PIX_FMT_RGB4_BYTE \
612 || (x)==PIX_FMT_MONOBLACK \
613 || (x)==PIX_FMT_MONOWHITE \
615 #define isBGRinInt(x) ( \
616 (x)==PIX_FMT_BGR48BE \
617 || (x)==PIX_FMT_BGR48LE \
618 || (x)==PIX_FMT_BGR32 \
619 || (x)==PIX_FMT_BGR32_1 \
620 || (x)==PIX_FMT_BGR24 \
621 || (x)==PIX_FMT_BGR565BE \
622 || (x)==PIX_FMT_BGR565LE \
623 || (x)==PIX_FMT_BGR555BE \
624 || (x)==PIX_FMT_BGR555LE \
625 || (x)==PIX_FMT_BGR444BE \
626 || (x)==PIX_FMT_BGR444LE \
627 || (x)==PIX_FMT_BGR8 \
628 || (x)==PIX_FMT_BGR4 \
629 || (x)==PIX_FMT_BGR4_BYTE \
630 || (x)==PIX_FMT_MONOBLACK \
631 || (x)==PIX_FMT_MONOWHITE \
633 #define isRGBinBytes(x) ( \
634 (x)==PIX_FMT_RGB48BE \
635 || (x)==PIX_FMT_RGB48LE \
636 || (x)==PIX_FMT_RGBA \
637 || (x)==PIX_FMT_ARGB \
638 || (x)==PIX_FMT_RGB24 \
640 #define isBGRinBytes(x) ( \
641 (x)==PIX_FMT_BGR48BE \
642 || (x)==PIX_FMT_BGR48LE \
643 || (x)==PIX_FMT_BGRA \
644 || (x)==PIX_FMT_ABGR \
645 || (x)==PIX_FMT_BGR24 \
647 #define isAnyRGB(x) ( \
651 #define isALPHA(x) ( \
653 || (x)==PIX_FMT_BGR32_1 \
654 || (x)==PIX_FMT_RGB32 \
655 || (x)==PIX_FMT_RGB32_1 \
656 || (x)==PIX_FMT_Y400A \
657 || (x)==PIX_FMT_YUVA420P \
659 #define isPacked(x) ( \
661 || (x)==PIX_FMT_YUYV422 \
662 || (x)==PIX_FMT_UYVY422 \
663 || (x)==PIX_FMT_Y400A \
666 #define usePal(x) ((av_pix_fmt_descriptors[x].flags & PIX_FMT_PAL) || (x) == PIX_FMT_Y400A)
668 extern const uint64_t ff_dither4[2];
669 extern const uint64_t ff_dither8[2];
671 extern const AVClass sws_context_class;
674 * Sets c->swScale to an unscaled converter if one exists for the specific
675 * source and destination formats, bit depths, flags, etc.
677 void ff_get_unscaled_swscale(SwsContext *c);
679 void ff_swscale_get_unscaled_altivec(SwsContext *c);
682 * Returns function pointer to fastest main scaler path function depending
683 * on architecture and available optimizations.
685 SwsFunc ff_getSwsFunc(SwsContext *c);
687 void ff_sws_init_swScale_altivec(SwsContext *c);
688 void ff_sws_init_swScale_mmx(SwsContext *c);
690 #endif /* SWSCALE_SWSCALE_INTERNAL_H */