Add 3 channel layout convenience macros.
[libav.git] / libavcodec / avcodec.h
1 /*
2 * copyright (c) 2001 Fabrice Bellard
3 *
4 * This file is part of FFmpeg.
5 *
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.
10 *
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.
15 *
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
19 */
20
21 #ifndef AVCODEC_AVCODEC_H
22 #define AVCODEC_AVCODEC_H
23
24 /**
25 * @file libavcodec/avcodec.h
26 * external API header
27 */
28
29 #include <errno.h>
30 #include "libavutil/avutil.h"
31
32 #define LIBAVCODEC_VERSION_MAJOR 52
33 #define LIBAVCODEC_VERSION_MINOR 27
34 #define LIBAVCODEC_VERSION_MICRO 0
35
36 #define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
37 LIBAVCODEC_VERSION_MINOR, \
38 LIBAVCODEC_VERSION_MICRO)
39 #define LIBAVCODEC_VERSION AV_VERSION(LIBAVCODEC_VERSION_MAJOR, \
40 LIBAVCODEC_VERSION_MINOR, \
41 LIBAVCODEC_VERSION_MICRO)
42 #define LIBAVCODEC_BUILD LIBAVCODEC_VERSION_INT
43
44 #define LIBAVCODEC_IDENT "Lavc" AV_STRINGIFY(LIBAVCODEC_VERSION)
45
46 #define AV_NOPTS_VALUE INT64_C(0x8000000000000000)
47 #define AV_TIME_BASE 1000000
48 #define AV_TIME_BASE_Q (AVRational){1, AV_TIME_BASE}
49
50 /**
51 * Identifies the syntax and semantics of the bitstream.
52 * The principle is roughly:
53 * Two decoders with the same ID can decode the same streams.
54 * Two encoders with the same ID can encode compatible streams.
55 * There may be slight deviations from the principle due to implementation
56 * details.
57 *
58 * If you add a codec ID to this list, add it so that
59 * 1. no value of a existing codec ID changes (that would break ABI),
60 * 2. it is as close as possible to similar codecs.
61 */
62 enum CodecID {
63 CODEC_ID_NONE,
64
65 /* video codecs */
66 CODEC_ID_MPEG1VIDEO,
67 CODEC_ID_MPEG2VIDEO, ///< preferred ID for MPEG-1/2 video decoding
68 CODEC_ID_MPEG2VIDEO_XVMC,
69 CODEC_ID_H261,
70 CODEC_ID_H263,
71 CODEC_ID_RV10,
72 CODEC_ID_RV20,
73 CODEC_ID_MJPEG,
74 CODEC_ID_MJPEGB,
75 CODEC_ID_LJPEG,
76 CODEC_ID_SP5X,
77 CODEC_ID_JPEGLS,
78 CODEC_ID_MPEG4,
79 CODEC_ID_RAWVIDEO,
80 CODEC_ID_MSMPEG4V1,
81 CODEC_ID_MSMPEG4V2,
82 CODEC_ID_MSMPEG4V3,
83 CODEC_ID_WMV1,
84 CODEC_ID_WMV2,
85 CODEC_ID_H263P,
86 CODEC_ID_H263I,
87 CODEC_ID_FLV1,
88 CODEC_ID_SVQ1,
89 CODEC_ID_SVQ3,
90 CODEC_ID_DVVIDEO,
91 CODEC_ID_HUFFYUV,
92 CODEC_ID_CYUV,
93 CODEC_ID_H264,
94 CODEC_ID_INDEO3,
95 CODEC_ID_VP3,
96 CODEC_ID_THEORA,
97 CODEC_ID_ASV1,
98 CODEC_ID_ASV2,
99 CODEC_ID_FFV1,
100 CODEC_ID_4XM,
101 CODEC_ID_VCR1,
102 CODEC_ID_CLJR,
103 CODEC_ID_MDEC,
104 CODEC_ID_ROQ,
105 CODEC_ID_INTERPLAY_VIDEO,
106 CODEC_ID_XAN_WC3,
107 CODEC_ID_XAN_WC4,
108 CODEC_ID_RPZA,
109 CODEC_ID_CINEPAK,
110 CODEC_ID_WS_VQA,
111 CODEC_ID_MSRLE,
112 CODEC_ID_MSVIDEO1,
113 CODEC_ID_IDCIN,
114 CODEC_ID_8BPS,
115 CODEC_ID_SMC,
116 CODEC_ID_FLIC,
117 CODEC_ID_TRUEMOTION1,
118 CODEC_ID_VMDVIDEO,
119 CODEC_ID_MSZH,
120 CODEC_ID_ZLIB,
121 CODEC_ID_QTRLE,
122 CODEC_ID_SNOW,
123 CODEC_ID_TSCC,
124 CODEC_ID_ULTI,
125 CODEC_ID_QDRAW,
126 CODEC_ID_VIXL,
127 CODEC_ID_QPEG,
128 CODEC_ID_XVID,
129 CODEC_ID_PNG,
130 CODEC_ID_PPM,
131 CODEC_ID_PBM,
132 CODEC_ID_PGM,
133 CODEC_ID_PGMYUV,
134 CODEC_ID_PAM,
135 CODEC_ID_FFVHUFF,
136 CODEC_ID_RV30,
137 CODEC_ID_RV40,
138 CODEC_ID_VC1,
139 CODEC_ID_WMV3,
140 CODEC_ID_LOCO,
141 CODEC_ID_WNV1,
142 CODEC_ID_AASC,
143 CODEC_ID_INDEO2,
144 CODEC_ID_FRAPS,
145 CODEC_ID_TRUEMOTION2,
146 CODEC_ID_BMP,
147 CODEC_ID_CSCD,
148 CODEC_ID_MMVIDEO,
149 CODEC_ID_ZMBV,
150 CODEC_ID_AVS,
151 CODEC_ID_SMACKVIDEO,
152 CODEC_ID_NUV,
153 CODEC_ID_KMVC,
154 CODEC_ID_FLASHSV,
155 CODEC_ID_CAVS,
156 CODEC_ID_JPEG2000,
157 CODEC_ID_VMNC,
158 CODEC_ID_VP5,
159 CODEC_ID_VP6,
160 CODEC_ID_VP6F,
161 CODEC_ID_TARGA,
162 CODEC_ID_DSICINVIDEO,
163 CODEC_ID_TIERTEXSEQVIDEO,
164 CODEC_ID_TIFF,
165 CODEC_ID_GIF,
166 CODEC_ID_FFH264,
167 CODEC_ID_DXA,
168 CODEC_ID_DNXHD,
169 CODEC_ID_THP,
170 CODEC_ID_SGI,
171 CODEC_ID_C93,
172 CODEC_ID_BETHSOFTVID,
173 CODEC_ID_PTX,
174 CODEC_ID_TXD,
175 CODEC_ID_VP6A,
176 CODEC_ID_AMV,
177 CODEC_ID_VB,
178 CODEC_ID_PCX,
179 CODEC_ID_SUNRAST,
180 CODEC_ID_INDEO4,
181 CODEC_ID_INDEO5,
182 CODEC_ID_MIMIC,
183 CODEC_ID_RL2,
184 CODEC_ID_8SVX_EXP,
185 CODEC_ID_8SVX_FIB,
186 CODEC_ID_ESCAPE124,
187 CODEC_ID_DIRAC,
188 CODEC_ID_BFI,
189 CODEC_ID_CMV,
190 CODEC_ID_MOTIONPIXELS,
191 CODEC_ID_TGV,
192 CODEC_ID_TGQ,
193 CODEC_ID_TQI,
194 CODEC_ID_AURA,
195 CODEC_ID_AURA2,
196
197 /* various PCM "codecs" */
198 CODEC_ID_PCM_S16LE= 0x10000,
199 CODEC_ID_PCM_S16BE,
200 CODEC_ID_PCM_U16LE,
201 CODEC_ID_PCM_U16BE,
202 CODEC_ID_PCM_S8,
203 CODEC_ID_PCM_U8,
204 CODEC_ID_PCM_MULAW,
205 CODEC_ID_PCM_ALAW,
206 CODEC_ID_PCM_S32LE,
207 CODEC_ID_PCM_S32BE,
208 CODEC_ID_PCM_U32LE,
209 CODEC_ID_PCM_U32BE,
210 CODEC_ID_PCM_S24LE,
211 CODEC_ID_PCM_S24BE,
212 CODEC_ID_PCM_U24LE,
213 CODEC_ID_PCM_U24BE,
214 CODEC_ID_PCM_S24DAUD,
215 CODEC_ID_PCM_ZORK,
216 CODEC_ID_PCM_S16LE_PLANAR,
217 CODEC_ID_PCM_DVD,
218 CODEC_ID_PCM_F32BE,
219 CODEC_ID_PCM_F32LE,
220 CODEC_ID_PCM_F64BE,
221 CODEC_ID_PCM_F64LE,
222
223 /* various ADPCM codecs */
224 CODEC_ID_ADPCM_IMA_QT= 0x11000,
225 CODEC_ID_ADPCM_IMA_WAV,
226 CODEC_ID_ADPCM_IMA_DK3,
227 CODEC_ID_ADPCM_IMA_DK4,
228 CODEC_ID_ADPCM_IMA_WS,
229 CODEC_ID_ADPCM_IMA_SMJPEG,
230 CODEC_ID_ADPCM_MS,
231 CODEC_ID_ADPCM_4XM,
232 CODEC_ID_ADPCM_XA,
233 CODEC_ID_ADPCM_ADX,
234 CODEC_ID_ADPCM_EA,
235 CODEC_ID_ADPCM_G726,
236 CODEC_ID_ADPCM_CT,
237 CODEC_ID_ADPCM_SWF,
238 CODEC_ID_ADPCM_YAMAHA,
239 CODEC_ID_ADPCM_SBPRO_4,
240 CODEC_ID_ADPCM_SBPRO_3,
241 CODEC_ID_ADPCM_SBPRO_2,
242 CODEC_ID_ADPCM_THP,
243 CODEC_ID_ADPCM_IMA_AMV,
244 CODEC_ID_ADPCM_EA_R1,
245 CODEC_ID_ADPCM_EA_R3,
246 CODEC_ID_ADPCM_EA_R2,
247 CODEC_ID_ADPCM_IMA_EA_SEAD,
248 CODEC_ID_ADPCM_IMA_EA_EACS,
249 CODEC_ID_ADPCM_EA_XAS,
250 CODEC_ID_ADPCM_EA_MAXIS_XA,
251 CODEC_ID_ADPCM_IMA_ISS,
252
253 /* AMR */
254 CODEC_ID_AMR_NB= 0x12000,
255 CODEC_ID_AMR_WB,
256
257 /* RealAudio codecs*/
258 CODEC_ID_RA_144= 0x13000,
259 CODEC_ID_RA_288,
260
261 /* various DPCM codecs */
262 CODEC_ID_ROQ_DPCM= 0x14000,
263 CODEC_ID_INTERPLAY_DPCM,
264 CODEC_ID_XAN_DPCM,
265 CODEC_ID_SOL_DPCM,
266
267 /* audio codecs */
268 CODEC_ID_MP2= 0x15000,
269 CODEC_ID_MP3, ///< preferred ID for decoding MPEG audio layer 1, 2 or 3
270 CODEC_ID_AAC,
271 CODEC_ID_AC3,
272 CODEC_ID_DTS,
273 CODEC_ID_VORBIS,
274 CODEC_ID_DVAUDIO,
275 CODEC_ID_WMAV1,
276 CODEC_ID_WMAV2,
277 CODEC_ID_MACE3,
278 CODEC_ID_MACE6,
279 CODEC_ID_VMDAUDIO,
280 CODEC_ID_SONIC,
281 CODEC_ID_SONIC_LS,
282 CODEC_ID_FLAC,
283 CODEC_ID_MP3ADU,
284 CODEC_ID_MP3ON4,
285 CODEC_ID_SHORTEN,
286 CODEC_ID_ALAC,
287 CODEC_ID_WESTWOOD_SND1,
288 CODEC_ID_GSM, ///< as in Berlin toast format
289 CODEC_ID_QDM2,
290 CODEC_ID_COOK,
291 CODEC_ID_TRUESPEECH,
292 CODEC_ID_TTA,
293 CODEC_ID_SMACKAUDIO,
294 CODEC_ID_QCELP,
295 CODEC_ID_WAVPACK,
296 CODEC_ID_DSICINAUDIO,
297 CODEC_ID_IMC,
298 CODEC_ID_MUSEPACK7,
299 CODEC_ID_MLP,
300 CODEC_ID_GSM_MS, /* as found in WAV */
301 CODEC_ID_ATRAC3,
302 CODEC_ID_VOXWARE,
303 CODEC_ID_APE,
304 CODEC_ID_NELLYMOSER,
305 CODEC_ID_MUSEPACK8,
306 CODEC_ID_SPEEX,
307 CODEC_ID_WMAVOICE,
308 CODEC_ID_WMAPRO,
309 CODEC_ID_WMALOSSLESS,
310 CODEC_ID_ATRAC3P,
311 CODEC_ID_EAC3,
312 CODEC_ID_SIPR,
313 CODEC_ID_MP1,
314 CODEC_ID_TWINVQ,
315 CODEC_ID_TRUEHD,
316 CODEC_ID_MP4ALS,
317
318 /* subtitle codecs */
319 CODEC_ID_DVD_SUBTITLE= 0x17000,
320 CODEC_ID_DVB_SUBTITLE,
321 CODEC_ID_TEXT, ///< raw UTF-8 text
322 CODEC_ID_XSUB,
323 CODEC_ID_SSA,
324 CODEC_ID_MOV_TEXT,
325
326 /* other specific kind of codecs (generally used for attachments) */
327 CODEC_ID_TTF= 0x18000,
328
329 CODEC_ID_PROBE= 0x19000, ///< codec_id is not known (like CODEC_ID_NONE) but lavf should attempt to identify it
330
331 CODEC_ID_MPEG2TS= 0x20000, /**< _FAKE_ codec to indicate a raw MPEG-2 TS
332 * stream (only used by libavformat) */
333 };
334
335 enum CodecType {
336 CODEC_TYPE_UNKNOWN = -1,
337 CODEC_TYPE_VIDEO,
338 CODEC_TYPE_AUDIO,
339 CODEC_TYPE_DATA,
340 CODEC_TYPE_SUBTITLE,
341 CODEC_TYPE_ATTACHMENT,
342 CODEC_TYPE_NB
343 };
344
345 /**
346 * all in native-endian format
347 */
348 enum SampleFormat {
349 SAMPLE_FMT_NONE = -1,
350 SAMPLE_FMT_U8, ///< unsigned 8 bits
351 SAMPLE_FMT_S16, ///< signed 16 bits
352 SAMPLE_FMT_S32, ///< signed 32 bits
353 SAMPLE_FMT_FLT, ///< float
354 SAMPLE_FMT_DBL, ///< double
355 SAMPLE_FMT_NB ///< Number of sample formats. DO NOT USE if dynamically linking to libavcodec
356 };
357
358 /* Audio channel masks */
359 #define CH_FRONT_LEFT 0x00000001
360 #define CH_FRONT_RIGHT 0x00000002
361 #define CH_FRONT_CENTER 0x00000004
362 #define CH_LOW_FREQUENCY 0x00000008
363 #define CH_BACK_LEFT 0x00000010
364 #define CH_BACK_RIGHT 0x00000020
365 #define CH_FRONT_LEFT_OF_CENTER 0x00000040
366 #define CH_FRONT_RIGHT_OF_CENTER 0x00000080
367 #define CH_BACK_CENTER 0x00000100
368 #define CH_SIDE_LEFT 0x00000200
369 #define CH_SIDE_RIGHT 0x00000400
370 #define CH_TOP_CENTER 0x00000800
371 #define CH_TOP_FRONT_LEFT 0x00001000
372 #define CH_TOP_FRONT_CENTER 0x00002000
373 #define CH_TOP_FRONT_RIGHT 0x00004000
374 #define CH_TOP_BACK_LEFT 0x00008000
375 #define CH_TOP_BACK_CENTER 0x00010000
376 #define CH_TOP_BACK_RIGHT 0x00020000
377 #define CH_STEREO_LEFT 0x20000000 ///< Stereo downmix.
378 #define CH_STEREO_RIGHT 0x40000000 ///< See CH_STEREO_LEFT.
379
380 /* Audio channel convenience macros */
381 #define CH_LAYOUT_MONO (CH_FRONT_CENTER)
382 #define CH_LAYOUT_STEREO (CH_FRONT_LEFT|CH_FRONT_RIGHT)
383 #define CH_LAYOUT_2_1 (CH_LAYOUT_STEREO|CH_BACK_CENTER)
384 #define CH_LAYOUT_SURROUND (CH_LAYOUT_STEREO|CH_FRONT_CENTER)
385 #define CH_LAYOUT_4POINT0 (CH_LAYOUT_SURROUND|CH_BACK_CENTER)
386 #define CH_LAYOUT_2_2 (CH_LAYOUT_STEREO|CH_SIDE_LEFT|CH_SIDE_RIGHT)
387 #define CH_LAYOUT_QUAD (CH_LAYOUT_STEREO|CH_BACK_LEFT|CH_BACK_RIGHT)
388 #define CH_LAYOUT_5POINT0 (CH_LAYOUT_SURROUND|CH_SIDE_LEFT|CH_SIDE_RIGHT)
389 #define CH_LAYOUT_5POINT1 (CH_LAYOUT_5POINT0|CH_LOW_FREQUENCY)
390 #define CH_LAYOUT_5POINT0_BACK (CH_LAYOUT_SURROUND|CH_BACK_LEFT|CH_BACK_RIGHT)
391 #define CH_LAYOUT_5POINT1_BACK (CH_LAYOUT_5POINT0_BACK|CH_LOW_FREQUENCY)
392 #define CH_LAYOUT_7POINT1 (CH_LAYOUT_5POINT1|CH_BACK_LEFT|CH_BACK_RIGHT)
393 #define CH_LAYOUT_7POINT1_WIDE (CH_LAYOUT_5POINT1_BACK|\
394 CH_FRONT_LEFT_OF_CENTER|CH_FRONT_RIGHT_OF_CENTER)
395 #define CH_LAYOUT_STEREO_DOWNMIX (CH_STEREO_LEFT|CH_STEREO_RIGHT)
396
397 /* in bytes */
398 #define AVCODEC_MAX_AUDIO_FRAME_SIZE 192000 // 1 second of 48khz 32bit audio
399
400 /**
401 * Required number of additionally allocated bytes at the end of the input bitstream for decoding.
402 * This is mainly needed because some optimized bitstream readers read
403 * 32 or 64 bit at once and could read over the end.<br>
404 * Note: If the first 23 bits of the additional bytes are not 0, then damaged
405 * MPEG bitstreams could cause overread and segfault.
406 */
407 #define FF_INPUT_BUFFER_PADDING_SIZE 8
408
409 /**
410 * minimum encoding buffer size
411 * Used to avoid some checks during header writing.
412 */
413 #define FF_MIN_BUFFER_SIZE 16384
414
415
416 /**
417 * motion estimation type.
418 */
419 enum Motion_Est_ID {
420 ME_ZERO = 1, ///< no search, that is use 0,0 vector whenever one is needed
421 ME_FULL,
422 ME_LOG,
423 ME_PHODS,
424 ME_EPZS, ///< enhanced predictive zonal search
425 ME_X1, ///< reserved for experiments
426 ME_HEX, ///< hexagon based search
427 ME_UMH, ///< uneven multi-hexagon search
428 ME_ITER, ///< iterative search
429 ME_TESA, ///< transformed exhaustive search algorithm
430 };
431
432 enum AVDiscard{
433 /* We leave some space between them for extensions (drop some
434 * keyframes for intra-only or drop just some bidir frames). */
435 AVDISCARD_NONE =-16, ///< discard nothing
436 AVDISCARD_DEFAULT= 0, ///< discard useless packets like 0 size packets in avi
437 AVDISCARD_NONREF = 8, ///< discard all non reference
438 AVDISCARD_BIDIR = 16, ///< discard all bidirectional frames
439 AVDISCARD_NONKEY = 32, ///< discard all frames except keyframes
440 AVDISCARD_ALL = 48, ///< discard all
441 };
442
443 typedef struct RcOverride{
444 int start_frame;
445 int end_frame;
446 int qscale; // If this is 0 then quality_factor will be used instead.
447 float quality_factor;
448 } RcOverride;
449
450 #define FF_MAX_B_FRAMES 16
451
452 /* encoding support
453 These flags can be passed in AVCodecContext.flags before initialization.
454 Note: Not everything is supported yet.
455 */
456
457 #define CODEC_FLAG_QSCALE 0x0002 ///< Use fixed qscale.
458 #define CODEC_FLAG_4MV 0x0004 ///< 4 MV per MB allowed / advanced prediction for H.263.
459 #define CODEC_FLAG_QPEL 0x0010 ///< Use qpel MC.
460 #define CODEC_FLAG_GMC 0x0020 ///< Use GMC.
461 #define CODEC_FLAG_MV0 0x0040 ///< Always try a MB with MV=<0,0>.
462 #define CODEC_FLAG_PART 0x0080 ///< Use data partitioning.
463 /**
464 * The parent program guarantees that the input for B-frames containing
465 * streams is not written to for at least s->max_b_frames+1 frames, if
466 * this is not set the input will be copied.
467 */
468 #define CODEC_FLAG_INPUT_PRESERVED 0x0100
469 #define CODEC_FLAG_PASS1 0x0200 ///< Use internal 2pass ratecontrol in first pass mode.
470 #define CODEC_FLAG_PASS2 0x0400 ///< Use internal 2pass ratecontrol in second pass mode.
471 #define CODEC_FLAG_EXTERN_HUFF 0x1000 ///< Use external Huffman table (for MJPEG).
472 #define CODEC_FLAG_GRAY 0x2000 ///< Only decode/encode grayscale.
473 #define CODEC_FLAG_EMU_EDGE 0x4000 ///< Don't draw edges.
474 #define CODEC_FLAG_PSNR 0x8000 ///< error[?] variables will be set during encoding.
475 #define CODEC_FLAG_TRUNCATED 0x00010000 /** Input bitstream might be truncated at a random
476 location instead of only at frame boundaries. */
477 #define CODEC_FLAG_NORMALIZE_AQP 0x00020000 ///< Normalize adaptive quantization.
478 #define CODEC_FLAG_INTERLACED_DCT 0x00040000 ///< Use interlaced DCT.
479 #define CODEC_FLAG_LOW_DELAY 0x00080000 ///< Force low delay.
480 #define CODEC_FLAG_ALT_SCAN 0x00100000 ///< Use alternate scan.
481 #define CODEC_FLAG_GLOBAL_HEADER 0x00400000 ///< Place global headers in extradata instead of every keyframe.
482 #define CODEC_FLAG_BITEXACT 0x00800000 ///< Use only bitexact stuff (except (I)DCT).
483 /* Fx : Flag for h263+ extra options */
484 #define CODEC_FLAG_AC_PRED 0x01000000 ///< H.263 advanced intra coding / MPEG-4 AC prediction
485 #define CODEC_FLAG_H263P_UMV 0x02000000 ///< unlimited motion vector
486 #define CODEC_FLAG_CBP_RD 0x04000000 ///< Use rate distortion optimization for cbp.
487 #define CODEC_FLAG_QP_RD 0x08000000 ///< Use rate distortion optimization for qp selectioon.
488 #define CODEC_FLAG_H263P_AIV 0x00000008 ///< H.263 alternative inter VLC
489 #define CODEC_FLAG_OBMC 0x00000001 ///< OBMC
490 #define CODEC_FLAG_LOOP_FILTER 0x00000800 ///< loop filter
491 #define CODEC_FLAG_H263P_SLICE_STRUCT 0x10000000
492 #define CODEC_FLAG_INTERLACED_ME 0x20000000 ///< interlaced motion estimation
493 #define CODEC_FLAG_SVCD_SCAN_OFFSET 0x40000000 ///< Will reserve space for SVCD scan offset user data.
494 #define CODEC_FLAG_CLOSED_GOP 0x80000000
495 #define CODEC_FLAG2_FAST 0x00000001 ///< Allow non spec compliant speedup tricks.
496 #define CODEC_FLAG2_STRICT_GOP 0x00000002 ///< Strictly enforce GOP size.
497 #define CODEC_FLAG2_NO_OUTPUT 0x00000004 ///< Skip bitstream encoding.
498 #define CODEC_FLAG2_LOCAL_HEADER 0x00000008 ///< Place global headers at every keyframe instead of in extradata.
499 #define CODEC_FLAG2_BPYRAMID 0x00000010 ///< H.264 allow B-frames to be used as references.
500 #define CODEC_FLAG2_WPRED 0x00000020 ///< H.264 weighted biprediction for B-frames
501 #define CODEC_FLAG2_MIXED_REFS 0x00000040 ///< H.264 one reference per partition, as opposed to one reference per macroblock
502 #define CODEC_FLAG2_8X8DCT 0x00000080 ///< H.264 high profile 8x8 transform
503 #define CODEC_FLAG2_FASTPSKIP 0x00000100 ///< H.264 fast pskip
504 #define CODEC_FLAG2_AUD 0x00000200 ///< H.264 access unit delimiters
505 #define CODEC_FLAG2_BRDO 0x00000400 ///< B-frame rate-distortion optimization
506 #define CODEC_FLAG2_INTRA_VLC 0x00000800 ///< Use MPEG-2 intra VLC table.
507 #define CODEC_FLAG2_MEMC_ONLY 0x00001000 ///< Only do ME/MC (I frames -> ref, P frame -> ME+MC).
508 #define CODEC_FLAG2_DROP_FRAME_TIMECODE 0x00002000 ///< timecode is in drop frame format.
509 #define CODEC_FLAG2_SKIP_RD 0x00004000 ///< RD optimal MB level residual skipping
510 #define CODEC_FLAG2_CHUNKS 0x00008000 ///< Input bitstream might be truncated at a packet boundaries instead of only at frame boundaries.
511 #define CODEC_FLAG2_NON_LINEAR_QUANT 0x00010000 ///< Use MPEG-2 nonlinear quantizer.
512 #define CODEC_FLAG2_BIT_RESERVOIR 0x00020000 ///< Use a bit reservoir when encoding if possible
513
514 /* Unsupported options :
515 * Syntax Arithmetic coding (SAC)
516 * Reference Picture Selection
517 * Independent Segment Decoding */
518 /* /Fx */
519 /* codec capabilities */
520
521 #define CODEC_CAP_DRAW_HORIZ_BAND 0x0001 ///< Decoder can use draw_horiz_band callback.
522 /**
523 * Codec uses get_buffer() for allocating buffers.
524 * direct rendering method 1
525 */
526 #define CODEC_CAP_DR1 0x0002
527 /* If 'parse_only' field is true, then avcodec_parse_frame() can be used. */
528 #define CODEC_CAP_PARSE_ONLY 0x0004
529 #define CODEC_CAP_TRUNCATED 0x0008
530 /* Codec can export data for HW decoding (XvMC). */
531 #define CODEC_CAP_HWACCEL 0x0010
532 /**
533 * Codec has a nonzero delay and needs to be fed with NULL at the end to get the delayed data.
534 * If this is not set, the codec is guaranteed to never be fed with NULL data.
535 */
536 #define CODEC_CAP_DELAY 0x0020
537 /**
538 * Codec can be fed a final frame with a smaller size.
539 * This can be used to prevent truncation of the last audio samples.
540 */
541 #define CODEC_CAP_SMALL_LAST_FRAME 0x0040
542 /**
543 * Codec can export data for HW decoding (VDPAU).
544 */
545 #define CODEC_CAP_HWACCEL_VDPAU 0x0080
546
547 //The following defines may change, don't expect compatibility if you use them.
548 #define MB_TYPE_INTRA4x4 0x0001
549 #define MB_TYPE_INTRA16x16 0x0002 //FIXME H.264-specific
550 #define MB_TYPE_INTRA_PCM 0x0004 //FIXME H.264-specific
551 #define MB_TYPE_16x16 0x0008
552 #define MB_TYPE_16x8 0x0010
553 #define MB_TYPE_8x16 0x0020
554 #define MB_TYPE_8x8 0x0040
555 #define MB_TYPE_INTERLACED 0x0080
556 #define MB_TYPE_DIRECT2 0x0100 //FIXME
557 #define MB_TYPE_ACPRED 0x0200
558 #define MB_TYPE_GMC 0x0400
559 #define MB_TYPE_SKIP 0x0800
560 #define MB_TYPE_P0L0 0x1000
561 #define MB_TYPE_P1L0 0x2000
562 #define MB_TYPE_P0L1 0x4000
563 #define MB_TYPE_P1L1 0x8000
564 #define MB_TYPE_L0 (MB_TYPE_P0L0 | MB_TYPE_P1L0)
565 #define MB_TYPE_L1 (MB_TYPE_P0L1 | MB_TYPE_P1L1)
566 #define MB_TYPE_L0L1 (MB_TYPE_L0 | MB_TYPE_L1)
567 #define MB_TYPE_QUANT 0x00010000
568 #define MB_TYPE_CBP 0x00020000
569 //Note bits 24-31 are reserved for codec specific use (h264 ref0, mpeg1 0mv, ...)
570
571 /**
572 * Pan Scan area.
573 * This specifies the area which should be displayed.
574 * Note there may be multiple such areas for one frame.
575 */
576 typedef struct AVPanScan{
577 /**
578 * id
579 * - encoding: Set by user.
580 * - decoding: Set by libavcodec.
581 */
582 int id;
583
584 /**
585 * width and height in 1/16 pel
586 * - encoding: Set by user.
587 * - decoding: Set by libavcodec.
588 */
589 int width;
590 int height;
591
592 /**
593 * position of the top left corner in 1/16 pel for up to 3 fields/frames
594 * - encoding: Set by user.
595 * - decoding: Set by libavcodec.
596 */
597 int16_t position[3][2];
598 }AVPanScan;
599
600 #define FF_COMMON_FRAME \
601 /**\
602 * pointer to the picture planes.\
603 * This might be different from the first allocated byte\
604 * - encoding: \
605 * - decoding: \
606 */\
607 uint8_t *data[4];\
608 int linesize[4];\
609 /**\
610 * pointer to the first allocated byte of the picture. Can be used in get_buffer/release_buffer.\
611 * This isn't used by libavcodec unless the default get/release_buffer() is used.\
612 * - encoding: \
613 * - decoding: \
614 */\
615 uint8_t *base[4];\
616 /**\
617 * 1 -> keyframe, 0-> not\
618 * - encoding: Set by libavcodec.\
619 * - decoding: Set by libavcodec.\
620 */\
621 int key_frame;\
622 \
623 /**\
624 * Picture type of the frame, see ?_TYPE below.\
625 * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
626 * - decoding: Set by libavcodec.\
627 */\
628 int pict_type;\
629 \
630 /**\
631 * presentation timestamp in time_base units (time when frame should be shown to user)\
632 * If AV_NOPTS_VALUE then frame_rate = 1/time_base will be assumed.\
633 * - encoding: MUST be set by user.\
634 * - decoding: Set by libavcodec.\
635 */\
636 int64_t pts;\
637 \
638 /**\
639 * picture number in bitstream order\
640 * - encoding: set by\
641 * - decoding: Set by libavcodec.\
642 */\
643 int coded_picture_number;\
644 /**\
645 * picture number in display order\
646 * - encoding: set by\
647 * - decoding: Set by libavcodec.\
648 */\
649 int display_picture_number;\
650 \
651 /**\
652 * quality (between 1 (good) and FF_LAMBDA_MAX (bad)) \
653 * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
654 * - decoding: Set by libavcodec.\
655 */\
656 int quality; \
657 \
658 /**\
659 * buffer age (1->was last buffer and dint change, 2->..., ...).\
660 * Set to INT_MAX if the buffer has not been used yet.\
661 * - encoding: unused\
662 * - decoding: MUST be set by get_buffer().\
663 */\
664 int age;\
665 \
666 /**\
667 * is this picture used as reference\
668 * The values for this are the same as the MpegEncContext.picture_structure\
669 * variable, that is 1->top field, 2->bottom field, 3->frame/both fields.\
670 * Set to 4 for delayed, non-reference frames.\
671 * - encoding: unused\
672 * - decoding: Set by libavcodec. (before get_buffer() call)).\
673 */\
674 int reference;\
675 \
676 /**\
677 * QP table\
678 * - encoding: unused\
679 * - decoding: Set by libavcodec.\
680 */\
681 int8_t *qscale_table;\
682 /**\
683 * QP store stride\
684 * - encoding: unused\
685 * - decoding: Set by libavcodec.\
686 */\
687 int qstride;\
688 \
689 /**\
690 * mbskip_table[mb]>=1 if MB didn't change\
691 * stride= mb_width = (width+15)>>4\
692 * - encoding: unused\
693 * - decoding: Set by libavcodec.\
694 */\
695 uint8_t *mbskip_table;\
696 \
697 /**\
698 * motion vector table\
699 * @code\
700 * example:\
701 * int mv_sample_log2= 4 - motion_subsample_log2;\
702 * int mb_width= (width+15)>>4;\
703 * int mv_stride= (mb_width << mv_sample_log2) + 1;\
704 * motion_val[direction][x + y*mv_stride][0->mv_x, 1->mv_y];\
705 * @endcode\
706 * - encoding: Set by user.\
707 * - decoding: Set by libavcodec.\
708 */\
709 int16_t (*motion_val[2])[2];\
710 \
711 /**\
712 * macroblock type table\
713 * mb_type_base + mb_width + 2\
714 * - encoding: Set by user.\
715 * - decoding: Set by libavcodec.\
716 */\
717 uint32_t *mb_type;\
718 \
719 /**\
720 * log2 of the size of the block which a single vector in motion_val represents: \
721 * (4->16x16, 3->8x8, 2-> 4x4, 1-> 2x2)\
722 * - encoding: unused\
723 * - decoding: Set by libavcodec.\
724 */\
725 uint8_t motion_subsample_log2;\
726 \
727 /**\
728 * for some private data of the user\
729 * - encoding: unused\
730 * - decoding: Set by user.\
731 */\
732 void *opaque;\
733 \
734 /**\
735 * error\
736 * - encoding: Set by libavcodec. if flags&CODEC_FLAG_PSNR.\
737 * - decoding: unused\
738 */\
739 uint64_t error[4];\
740 \
741 /**\
742 * type of the buffer (to keep track of who has to deallocate data[*])\
743 * - encoding: Set by the one who allocates it.\
744 * - decoding: Set by the one who allocates it.\
745 * Note: User allocated (direct rendering) & internal buffers cannot coexist currently.\
746 */\
747 int type;\
748 \
749 /**\
750 * When decoding, this signals how much the picture must be delayed.\
751 * extra_delay = repeat_pict / (2*fps)\
752 * - encoding: unused\
753 * - decoding: Set by libavcodec.\
754 */\
755 int repeat_pict;\
756 \
757 /**\
758 * \
759 */\
760 int qscale_type;\
761 \
762 /**\
763 * The content of the picture is interlaced.\
764 * - encoding: Set by user.\
765 * - decoding: Set by libavcodec. (default 0)\
766 */\
767 int interlaced_frame;\
768 \
769 /**\
770 * If the content is interlaced, is top field displayed first.\
771 * - encoding: Set by user.\
772 * - decoding: Set by libavcodec.\
773 */\
774 int top_field_first;\
775 \
776 /**\
777 * Pan scan.\
778 * - encoding: Set by user.\
779 * - decoding: Set by libavcodec.\
780 */\
781 AVPanScan *pan_scan;\
782 \
783 /**\
784 * Tell user application that palette has changed from previous frame.\
785 * - encoding: ??? (no palette-enabled encoder yet)\
786 * - decoding: Set by libavcodec. (default 0).\
787 */\
788 int palette_has_changed;\
789 \
790 /**\
791 * codec suggestion on buffer type if != 0\
792 * - encoding: unused\
793 * - decoding: Set by libavcodec. (before get_buffer() call)).\
794 */\
795 int buffer_hints;\
796 \
797 /**\
798 * DCT coefficients\
799 * - encoding: unused\
800 * - decoding: Set by libavcodec.\
801 */\
802 short *dct_coeff;\
803 \
804 /**\
805 * motion referece frame index\
806 * - encoding: Set by user.\
807 * - decoding: Set by libavcodec.\
808 */\
809 int8_t *ref_index[2];\
810 \
811 /**\
812 * reordered opaque 64bit number (generally a PTS) from AVCodecContext.reordered_opaque\
813 * output in AVFrame.reordered_opaque\
814 * - encoding: unused\
815 * - decoding: Read by user.\
816 */\
817 int64_t reordered_opaque;\
818 \
819 /**\
820 * hardware accelerator private data (FFmpeg allocated)\
821 * - encoding: unused\
822 * - decoding: Set by libavcodec\
823 */\
824 void *hwaccel_picture_private;\
825
826
827 #define FF_QSCALE_TYPE_MPEG1 0
828 #define FF_QSCALE_TYPE_MPEG2 1
829 #define FF_QSCALE_TYPE_H264 2
830
831 #define FF_BUFFER_TYPE_INTERNAL 1
832 #define FF_BUFFER_TYPE_USER 2 ///< direct rendering buffers (image is (de)allocated by user)
833 #define FF_BUFFER_TYPE_SHARED 4 ///< Buffer from somewhere else; don't deallocate image (data/base), all other tables are not shared.
834 #define FF_BUFFER_TYPE_COPY 8 ///< Just a (modified) copy of some other buffer, don't deallocate anything.
835
836
837 #define FF_I_TYPE 1 ///< Intra
838 #define FF_P_TYPE 2 ///< Predicted
839 #define FF_B_TYPE 3 ///< Bi-dir predicted
840 #define FF_S_TYPE 4 ///< S(GMC)-VOP MPEG4
841 #define FF_SI_TYPE 5 ///< Switching Intra
842 #define FF_SP_TYPE 6 ///< Switching Predicted
843 #define FF_BI_TYPE 7
844
845 #define FF_BUFFER_HINTS_VALID 0x01 // Buffer hints value is meaningful (if 0 ignore).
846 #define FF_BUFFER_HINTS_READABLE 0x02 // Codec will read from buffer.
847 #define FF_BUFFER_HINTS_PRESERVE 0x04 // User must not alter buffer content.
848 #define FF_BUFFER_HINTS_REUSABLE 0x08 // Codec will reuse the buffer (update).
849
850 typedef struct AVPacket {
851 /**
852 * Presentation timestamp in AVStream->time_base units; the time at which
853 * the decompressed packet will be presented to the user.
854 * Can be AV_NOPTS_VALUE if it is not stored in the file.
855 * pts MUST be larger or equal to dts as presentation cannot happen before
856 * decompression, unless one wants to view hex dumps. Some formats misuse
857 * the terms dts and pts/cts to mean something different. Such timestamps
858 * must be converted to true pts/dts before they are stored in AVPacket.
859 */
860 int64_t pts;
861 /**
862 * Decompression timestamp in AVStream->time_base units; the time at which
863 * the packet is decompressed.
864 * Can be AV_NOPTS_VALUE if it is not stored in the file.
865 */
866 int64_t dts;
867 uint8_t *data;
868 int size;
869 int stream_index;
870 int flags;
871 /**
872 * Duration of this packet in AVStream->time_base units, 0 if unknown.
873 * Equals next_pts - this_pts in presentation order.
874 */
875 int duration;
876 void (*destruct)(struct AVPacket *);
877 void *priv;
878 int64_t pos; ///< byte position in stream, -1 if unknown
879
880 /**
881 * Time difference in AVStream->time_base units from the pts of this
882 * packet to the point at which the output from the decoder has converged
883 * independent from the availability of previous frames. That is, the
884 * frames are virtually identical no matter if decoding started from
885 * the very first frame or from this keyframe.
886 * Is AV_NOPTS_VALUE if unknown.
887 * This field is not the display duration of the current packet.
888 *
889 * The purpose of this field is to allow seeking in streams that have no
890 * keyframes in the conventional sense. It corresponds to the
891 * recovery point SEI in H.264 and match_time_delta in NUT. It is also
892 * essential for some types of subtitle streams to ensure that all
893 * subtitles are correctly displayed after seeking.
894 */
895 int64_t convergence_duration;
896 } AVPacket;
897 #define PKT_FLAG_KEY 0x0001
898
899 /**
900 * Audio Video Frame.
901 * New fields can be added to the end of FF_COMMON_FRAME with minor version
902 * bumps.
903 * Removal, reordering and changes to existing fields require a major
904 * version bump. No fields should be added into AVFrame before or after
905 * FF_COMMON_FRAME!
906 * sizeof(AVFrame) must not be used outside libav*.
907 */
908 typedef struct AVFrame {
909 FF_COMMON_FRAME
910 } AVFrame;
911
912 /**
913 * main external API structure.
914 * New fields can be added to the end with minor version bumps.
915 * Removal, reordering and changes to existing fields require a major
916 * version bump.
917 * sizeof(AVCodecContext) must not be used outside libav*.
918 */
919 typedef struct AVCodecContext {
920 /**
921 * information on struct for av_log
922 * - set by avcodec_alloc_context
923 */
924 const AVClass *av_class;
925 /**
926 * the average bitrate
927 * - encoding: Set by user; unused for constant quantizer encoding.
928 * - decoding: Set by libavcodec. 0 or some bitrate if this info is available in the stream.
929 */
930 int bit_rate;
931
932 /**
933 * number of bits the bitstream is allowed to diverge from the reference.
934 * the reference can be CBR (for CBR pass1) or VBR (for pass2)
935 * - encoding: Set by user; unused for constant quantizer encoding.
936 * - decoding: unused
937 */
938 int bit_rate_tolerance;
939
940 /**
941 * CODEC_FLAG_*.
942 * - encoding: Set by user.
943 * - decoding: Set by user.
944 */
945 int flags;
946
947 /**
948 * Some codecs need additional format info. It is stored here.
949 * If any muxer uses this then ALL demuxers/parsers AND encoders for the
950 * specific codec MUST set it correctly otherwise stream copy breaks.
951 * In general use of this field by muxers is not recommanded.
952 * - encoding: Set by libavcodec.
953 * - decoding: Set by libavcodec. (FIXME: Is this OK?)
954 */
955 int sub_id;
956
957 /**
958 * Motion estimation algorithm used for video coding.
959 * 1 (zero), 2 (full), 3 (log), 4 (phods), 5 (epzs), 6 (x1), 7 (hex),
960 * 8 (umh), 9 (iter), 10 (tesa) [7, 8, 10 are x264 specific, 9 is snow specific]
961 * - encoding: MUST be set by user.
962 * - decoding: unused
963 */
964 int me_method;
965
966 /**
967 * some codecs need / can use extradata like Huffman tables.
968 * mjpeg: Huffman tables
969 * rv10: additional flags
970 * mpeg4: global headers (they can be in the bitstream or here)
971 * The allocated memory should be FF_INPUT_BUFFER_PADDING_SIZE bytes larger
972 * than extradata_size to avoid prolems if it is read with the bitstream reader.
973 * The bytewise contents of extradata must not depend on the architecture or CPU endianness.
974 * - encoding: Set/allocated/freed by libavcodec.
975 * - decoding: Set/allocated/freed by user.
976 */
977 uint8_t *extradata;
978 int extradata_size;
979
980 /**
981 * This is the fundamental unit of time (in seconds) in terms
982 * of which frame timestamps are represented. For fixed-fps content,
983 * timebase should be 1/framerate and timestamp increments should be
984 * identically 1.
985 * - encoding: MUST be set by user.
986 * - decoding: Set by libavcodec.
987 */
988 AVRational time_base;
989
990 /* video only */
991 /**
992 * picture width / height.
993 * - encoding: MUST be set by user.
994 * - decoding: Set by libavcodec.
995 * Note: For compatibility it is possible to set this instead of
996 * coded_width/height before decoding.
997 */
998 int width, height;
999
1000 #define FF_ASPECT_EXTENDED 15
1001
1002 /**
1003 * the number of pictures in a group of pictures, or 0 for intra_only
1004 * - encoding: Set by user.
1005 * - decoding: unused
1006 */
1007 int gop_size;
1008
1009 /**
1010 * Pixel format, see PIX_FMT_xxx.
1011 * - encoding: Set by user.
1012 * - decoding: Set by libavcodec.
1013 */
1014 enum PixelFormat pix_fmt;
1015
1016 /**
1017 * Frame rate emulation. If not zero, the lower layer (i.e. format handler)
1018 * has to read frames at native frame rate.
1019 * - encoding: Set by user.
1020 * - decoding: unused
1021 */
1022 int rate_emu;
1023
1024 /**
1025 * If non NULL, 'draw_horiz_band' is called by the libavcodec
1026 * decoder to draw a horizontal band. It improves cache usage. Not
1027 * all codecs can do that. You must check the codec capabilities
1028 * beforehand.
1029 * The function is also used by hardware acceleration APIs.
1030 * It is called at least once during frame decoding to pass
1031 * the data needed for hardware render.
1032 * In that mode instead of pixel data, AVFrame points to
1033 * a structure specific to the acceleration API. The application
1034 * reads the structure and can change some fields to indicate progress
1035 * or mark state.
1036 * - encoding: unused
1037 * - decoding: Set by user.
1038 * @param height the height of the slice
1039 * @param y the y position of the slice
1040 * @param type 1->top field, 2->bottom field, 3->frame
1041 * @param offset offset into the AVFrame.data from which the slice should be read
1042 */
1043 void (*draw_horiz_band)(struct AVCodecContext *s,
1044 const AVFrame *src, int offset[4],
1045 int y, int type, int height);
1046
1047 /* audio only */
1048 int sample_rate; ///< samples per second
1049 int channels; ///< number of audio channels
1050
1051 /**
1052 * audio sample format
1053 * - encoding: Set by user.
1054 * - decoding: Set by libavcodec.
1055 */
1056 enum SampleFormat sample_fmt; ///< sample format, currently unused
1057
1058 /* The following data should not be initialized. */
1059 /**
1060 * Samples per packet, initialized when calling 'init'.
1061 */
1062 int frame_size;
1063 int frame_number; ///< audio or video frame number
1064 int real_pict_num; ///< Returns the real picture number of previous encoded frame.
1065
1066 /**
1067 * Number of frames the decoded output will be delayed relative to
1068 * the encoded input.
1069 * - encoding: Set by libavcodec.
1070 * - decoding: unused
1071 */
1072 int delay;
1073
1074 /* - encoding parameters */
1075 float qcompress; ///< amount of qscale change between easy & hard scenes (0.0-1.0)
1076 float qblur; ///< amount of qscale smoothing over time (0.0-1.0)
1077
1078 /**
1079 * minimum quantizer
1080 * - encoding: Set by user.
1081 * - decoding: unused
1082 */
1083 int qmin;
1084
1085 /**
1086 * maximum quantizer
1087 * - encoding: Set by user.
1088 * - decoding: unused
1089 */
1090 int qmax;
1091
1092 /**
1093 * maximum quantizer difference between frames
1094 * - encoding: Set by user.
1095 * - decoding: unused
1096 */
1097 int max_qdiff;
1098
1099 /**
1100 * maximum number of B-frames between non-B-frames
1101 * Note: The output will be delayed by max_b_frames+1 relative to the input.
1102 * - encoding: Set by user.
1103 * - decoding: unused
1104 */
1105 int max_b_frames;
1106
1107 /**
1108 * qscale factor between IP and B-frames
1109 * If > 0 then the last P-frame quantizer will be used (q= lastp_q*factor+offset).
1110 * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
1111 * - encoding: Set by user.
1112 * - decoding: unused
1113 */
1114 float b_quant_factor;
1115
1116 /** obsolete FIXME remove */
1117 int rc_strategy;
1118 #define FF_RC_STRATEGY_XVID 1
1119
1120 int b_frame_strategy;
1121
1122 /**
1123 * hurry up amount
1124 * - encoding: unused
1125 * - decoding: Set by user. 1-> Skip B-frames, 2-> Skip IDCT/dequant too, 5-> Skip everything except header
1126 * @deprecated Deprecated in favor of skip_idct and skip_frame.
1127 */
1128 int hurry_up;
1129
1130 struct AVCodec *codec;
1131
1132 void *priv_data;
1133
1134 int rtp_payload_size; /* The size of the RTP payload: the coder will */
1135 /* do its best to deliver a chunk with size */
1136 /* below rtp_payload_size, the chunk will start */
1137 /* with a start code on some codecs like H.263. */
1138 /* This doesn't take account of any particular */
1139 /* headers inside the transmitted RTP payload. */
1140
1141
1142 /* The RTP callback: This function is called */
1143 /* every time the encoder has a packet to send. */
1144 /* It depends on the encoder if the data starts */
1145 /* with a Start Code (it should). H.263 does. */
1146 /* mb_nb contains the number of macroblocks */
1147 /* encoded in the RTP payload. */
1148 void (*rtp_callback)(struct AVCodecContext *avctx, void *data, int size, int mb_nb);
1149
1150 /* statistics, used for 2-pass encoding */
1151 int mv_bits;
1152 int header_bits;
1153 int i_tex_bits;
1154 int p_tex_bits;
1155 int i_count;
1156 int p_count;
1157 int skip_count;
1158 int misc_bits;
1159
1160 /**
1161 * number of bits used for the previously encoded frame
1162 * - encoding: Set by libavcodec.
1163 * - decoding: unused
1164 */
1165 int frame_bits;
1166
1167 /**
1168 * Private data of the user, can be used to carry app specific stuff.
1169 * - encoding: Set by user.
1170 * - decoding: Set by user.
1171 */
1172 void *opaque;
1173
1174 char codec_name[32];
1175 enum CodecType codec_type; /* see CODEC_TYPE_xxx */
1176 enum CodecID codec_id; /* see CODEC_ID_xxx */
1177
1178 /**
1179 * fourcc (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
1180 * This is used to work around some encoder bugs.
1181 * A demuxer should set this to what is stored in the field used to identify the codec.
1182 * If there are multiple such fields in a container then the demuxer should choose the one
1183 * which maximizes the information about the used codec.
1184 * If the codec tag field in a container is larger then 32 bits then the demuxer should
1185 * remap the longer ID to 32 bits with a table or other structure. Alternatively a new
1186 * extra_codec_tag + size could be added but for this a clear advantage must be demonstrated
1187 * first.
1188 * - encoding: Set by user, if not then the default based on codec_id will be used.
1189 * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
1190 */
1191 unsigned int codec_tag;
1192
1193 /**
1194 * Work around bugs in encoders which sometimes cannot be detected automatically.
1195 * - encoding: Set by user
1196 * - decoding: Set by user
1197 */
1198 int workaround_bugs;
1199 #define FF_BUG_AUTODETECT 1 ///< autodetection
1200 #define FF_BUG_OLD_MSMPEG4 2
1201 #define FF_BUG_XVID_ILACE 4
1202 #define FF_BUG_UMP4 8
1203 #define FF_BUG_NO_PADDING 16
1204 #define FF_BUG_AMV 32
1205 #define FF_BUG_AC_VLC 0 ///< Will be removed, libavcodec can now handle these non-compliant files by default.
1206 #define FF_BUG_QPEL_CHROMA 64
1207 #define FF_BUG_STD_QPEL 128
1208 #define FF_BUG_QPEL_CHROMA2 256
1209 #define FF_BUG_DIRECT_BLOCKSIZE 512
1210 #define FF_BUG_EDGE 1024
1211 #define FF_BUG_HPEL_CHROMA 2048
1212 #define FF_BUG_DC_CLIP 4096
1213 #define FF_BUG_MS 8192 ///< Work around various bugs in Microsoft's broken decoders.
1214 //#define FF_BUG_FAKE_SCALABILITY 16 //Autodetection should work 100%.
1215
1216 /**
1217 * luma single coefficient elimination threshold
1218 * - encoding: Set by user.
1219 * - decoding: unused
1220 */
1221 int luma_elim_threshold;
1222
1223 /**
1224 * chroma single coeff elimination threshold
1225 * - encoding: Set by user.
1226 * - decoding: unused
1227 */
1228 int chroma_elim_threshold;
1229
1230 /**
1231 * strictly follow the standard (MPEG4, ...).
1232 * - encoding: Set by user.
1233 * - decoding: Set by user.
1234 * Setting this to STRICT or higher means the encoder and decoder will
1235 * generally do stupid things. While setting it to inofficial or lower
1236 * will mean the encoder might use things that are not supported by all
1237 * spec compliant decoders. Decoders make no difference between normal,
1238 * inofficial and experimental, that is they always try to decode things
1239 * when they can unless they are explicitly asked to behave stupid
1240 * (=strictly conform to the specs)
1241 */
1242 int strict_std_compliance;
1243 #define FF_COMPLIANCE_VERY_STRICT 2 ///< Strictly conform to a older more strict version of the spec or reference software.
1244 #define FF_COMPLIANCE_STRICT 1 ///< Strictly conform to all the things in the spec no matter what consequences.
1245 #define FF_COMPLIANCE_NORMAL 0
1246 #define FF_COMPLIANCE_INOFFICIAL -1 ///< Allow inofficial extensions.
1247 #define FF_COMPLIANCE_EXPERIMENTAL -2 ///< Allow nonstandardized experimental things.
1248
1249 /**
1250 * qscale offset between IP and B-frames
1251 * - encoding: Set by user.
1252 * - decoding: unused
1253 */
1254 float b_quant_offset;
1255
1256 /**
1257 * Error recognization; higher values will detect more errors but may
1258 * misdetect some more or less valid parts as errors.
1259 * - encoding: unused
1260 * - decoding: Set by user.
1261 */
1262 int error_recognition;
1263 #define FF_ER_CAREFUL 1
1264 #define FF_ER_COMPLIANT 2
1265 #define FF_ER_AGGRESSIVE 3
1266 #define FF_ER_VERY_AGGRESSIVE 4
1267
1268 /**
1269 * Called at the beginning of each frame to get a buffer for it.
1270 * If pic.reference is set then the frame will be read later by libavcodec.
1271 * avcodec_align_dimensions() should be used to find the required width and
1272 * height, as they normally need to be rounded up to the next multiple of 16.
1273 * - encoding: unused
1274 * - decoding: Set by libavcodec., user can override.
1275 */
1276 int (*get_buffer)(struct AVCodecContext *c, AVFrame *pic);
1277
1278 /**
1279 * Called to release buffers which were allocated with get_buffer.
1280 * A released buffer can be reused in get_buffer().
1281 * pic.data[*] must be set to NULL.
1282 * - encoding: unused
1283 * - decoding: Set by libavcodec., user can override.
1284 */
1285 void (*release_buffer)(struct AVCodecContext *c, AVFrame *pic);
1286
1287 /**
1288 * Size of the frame reordering buffer in the decoder.
1289 * For MPEG-2 it is 1 IPB or 0 low delay IP.
1290 * - encoding: Set by libavcodec.
1291 * - decoding: Set by libavcodec.
1292 */
1293 int has_b_frames;
1294
1295 /**
1296 * number of bytes per packet if constant and known or 0
1297 * Used by some WAV based audio codecs.
1298 */
1299 int block_align;
1300
1301 int parse_only; /* - decoding only: If true, only parsing is done
1302 (function avcodec_parse_frame()). The frame
1303 data is returned. Only MPEG codecs support this now. */
1304
1305 /**
1306 * 0-> h263 quant 1-> mpeg quant
1307 * - encoding: Set by user.
1308 * - decoding: unused
1309 */
1310 int mpeg_quant;
1311
1312 /**
1313 * pass1 encoding statistics output buffer
1314 * - encoding: Set by libavcodec.
1315 * - decoding: unused
1316 */
1317 char *stats_out;
1318
1319 /**
1320 * pass2 encoding statistics input buffer
1321 * Concatenated stuff from stats_out of pass1 should be placed here.
1322 * - encoding: Allocated/set/freed by user.
1323 * - decoding: unused
1324 */
1325 char *stats_in;
1326
1327 /**
1328 * ratecontrol qmin qmax limiting method
1329 * 0-> clipping, 1-> use a nice continous function to limit qscale wthin qmin/qmax.
1330 * - encoding: Set by user.
1331 * - decoding: unused
1332 */
1333 float rc_qsquish;
1334
1335 float rc_qmod_amp;
1336 int rc_qmod_freq;
1337
1338 /**
1339 * ratecontrol override, see RcOverride
1340 * - encoding: Allocated/set/freed by user.
1341 * - decoding: unused
1342 */
1343 RcOverride *rc_override;
1344 int rc_override_count;
1345
1346 /**
1347 * rate control equation
1348 * - encoding: Set by user
1349 * - decoding: unused
1350 */
1351 const char *rc_eq;
1352
1353 /**
1354 * maximum bitrate
1355 * - encoding: Set by user.
1356 * - decoding: unused
1357 */
1358 int rc_max_rate;
1359
1360 /**
1361 * minimum bitrate
1362 * - encoding: Set by user.
1363 * - decoding: unused
1364 */
1365 int rc_min_rate;
1366
1367 /**
1368 * decoder bitstream buffer size
1369 * - encoding: Set by user.
1370 * - decoding: unused
1371 */
1372 int rc_buffer_size;
1373 float rc_buffer_aggressivity;
1374
1375 /**
1376 * qscale factor between P and I-frames
1377 * If > 0 then the last p frame quantizer will be used (q= lastp_q*factor+offset).
1378 * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
1379 * - encoding: Set by user.
1380 * - decoding: unused
1381 */
1382 float i_quant_factor;
1383
1384 /**
1385 * qscale offset between P and I-frames
1386 * - encoding: Set by user.
1387 * - decoding: unused
1388 */
1389 float i_quant_offset;
1390
1391 /**
1392 * initial complexity for pass1 ratecontrol
1393 * - encoding: Set by user.
1394 * - decoding: unused
1395 */
1396 float rc_initial_cplx;
1397
1398 /**
1399 * DCT algorithm, see FF_DCT_* below
1400 * - encoding: Set by user.
1401 * - decoding: unused
1402 */
1403 int dct_algo;
1404 #define FF_DCT_AUTO 0
1405 #define FF_DCT_FASTINT 1
1406 #define FF_DCT_INT 2
1407 #define FF_DCT_MMX 3
1408 #define FF_DCT_MLIB 4
1409 #define FF_DCT_ALTIVEC 5
1410 #define FF_DCT_FAAN 6
1411
1412 /**
1413 * luminance masking (0-> disabled)
1414 * - encoding: Set by user.
1415 * - decoding: unused
1416 */
1417 float lumi_masking;
1418
1419 /**
1420 * temporary complexity masking (0-> disabled)
1421 * - encoding: Set by user.
1422 * - decoding: unused
1423 */
1424 float temporal_cplx_masking;
1425
1426 /**
1427 * spatial complexity masking (0-> disabled)
1428 * - encoding: Set by user.
1429 * - decoding: unused
1430 */
1431 float spatial_cplx_masking;
1432
1433 /**
1434 * p block masking (0-> disabled)
1435 * - encoding: Set by user.
1436 * - decoding: unused
1437 */
1438 float p_masking;
1439
1440 /**
1441 * darkness masking (0-> disabled)
1442 * - encoding: Set by user.
1443 * - decoding: unused
1444 */
1445 float dark_masking;
1446
1447 /**
1448 * IDCT algorithm, see FF_IDCT_* below.
1449 * - encoding: Set by user.
1450 * - decoding: Set by user.
1451 */
1452 int idct_algo;
1453 #define FF_IDCT_AUTO 0
1454 #define FF_IDCT_INT 1
1455 #define FF_IDCT_SIMPLE 2
1456 #define FF_IDCT_SIMPLEMMX 3
1457 #define FF_IDCT_LIBMPEG2MMX 4
1458 #define FF_IDCT_PS2 5
1459 #define FF_IDCT_MLIB 6
1460 #define FF_IDCT_ARM 7
1461 #define FF_IDCT_ALTIVEC 8
1462 #define FF_IDCT_SH4 9
1463 #define FF_IDCT_SIMPLEARM 10
1464 #define FF_IDCT_H264 11
1465 #define FF_IDCT_VP3 12
1466 #define FF_IDCT_IPP 13
1467 #define FF_IDCT_XVIDMMX 14
1468 #define FF_IDCT_CAVS 15
1469 #define FF_IDCT_SIMPLEARMV5TE 16
1470 #define FF_IDCT_SIMPLEARMV6 17
1471 #define FF_IDCT_SIMPLEVIS 18
1472 #define FF_IDCT_WMV2 19
1473 #define FF_IDCT_FAAN 20
1474 #define FF_IDCT_EA 21
1475 #define FF_IDCT_SIMPLENEON 22
1476 #define FF_IDCT_SIMPLEALPHA 23
1477
1478 /**
1479 * slice count
1480 * - encoding: Set by libavcodec.
1481 * - decoding: Set by user (or 0).
1482 */
1483 int slice_count;
1484 /**
1485 * slice offsets in the frame in bytes
1486 * - encoding: Set/allocated by libavcodec.
1487 * - decoding: Set/allocated by user (or NULL).
1488 */
1489 int *slice_offset;
1490
1491 /**
1492 * error concealment flags
1493 * - encoding: unused
1494 * - decoding: Set by user.
1495 */
1496 int error_concealment;
1497 #define FF_EC_GUESS_MVS 1
1498 #define FF_EC_DEBLOCK 2
1499
1500 /**
1501 * dsp_mask could be add used to disable unwanted CPU features
1502 * CPU features (i.e. MMX, SSE. ...)
1503 *
1504 * With the FORCE flag you may instead enable given CPU features.
1505 * (Dangerous: Usable in case of misdetection, improper usage however will
1506 * result into program crash.)
1507 */
1508 unsigned dsp_mask;
1509 #define FF_MM_FORCE 0x80000000 /* Force usage of selected flags (OR) */
1510 /* lower 16 bits - CPU features */
1511 #define FF_MM_MMX 0x0001 ///< standard MMX
1512 #define FF_MM_3DNOW 0x0004 ///< AMD 3DNOW
1513 #if LIBAVCODEC_VERSION_MAJOR < 53
1514 #define FF_MM_MMXEXT 0x0002 ///< SSE integer functions or AMD MMX ext
1515 #endif
1516 #define FF_MM_MMX2 0x0002 ///< SSE integer functions or AMD MMX ext
1517 #define FF_MM_SSE 0x0008 ///< SSE functions
1518 #define FF_MM_SSE2 0x0010 ///< PIV SSE2 functions
1519 #define FF_MM_3DNOWEXT 0x0020 ///< AMD 3DNowExt
1520 #define FF_MM_SSE3 0x0040 ///< Prescott SSE3 functions
1521 #define FF_MM_SSSE3 0x0080 ///< Conroe SSSE3 functions
1522 #define FF_MM_SSE4 0x0100 ///< Penryn SSE4.1 functions
1523 #define FF_MM_SSE42 0x0200 ///< Nehalem SSE4.2 functions
1524 #define FF_MM_IWMMXT 0x0100 ///< XScale IWMMXT
1525 #define FF_MM_ALTIVEC 0x0001 ///< standard AltiVec
1526
1527 /**
1528 * bits per sample/pixel from the demuxer (needed for huffyuv).
1529 * - encoding: Set by libavcodec.
1530 * - decoding: Set by user.
1531 */
1532 int bits_per_coded_sample;
1533
1534 /**
1535 * prediction method (needed for huffyuv)
1536 * - encoding: Set by user.
1537 * - decoding: unused
1538 */
1539 int prediction_method;
1540 #define FF_PRED_LEFT 0
1541 #define FF_PRED_PLANE 1
1542 #define FF_PRED_MEDIAN 2
1543
1544 /**
1545 * sample aspect ratio (0 if unknown)
1546 * That is the width of a pixel divided by the height of the pixel.
1547 * Numerator and denominator must be relatively prime and smaller than 256 for some video standards.
1548 * - encoding: Set by user.
1549 * - decoding: Set by libavcodec.
1550 */
1551 AVRational sample_aspect_ratio;
1552
1553 /**
1554 * the picture in the bitstream
1555 * - encoding: Set by libavcodec.
1556 * - decoding: Set by libavcodec.
1557 */
1558 AVFrame *coded_frame;
1559
1560 /**
1561 * debug
1562 * - encoding: Set by user.
1563 * - decoding: Set by user.
1564 */
1565 int debug;
1566 #define FF_DEBUG_PICT_INFO 1
1567 #define FF_DEBUG_RC 2
1568 #define FF_DEBUG_BITSTREAM 4
1569 #define FF_DEBUG_MB_TYPE 8
1570 #define FF_DEBUG_QP 16
1571 #define FF_DEBUG_MV 32
1572 #define FF_DEBUG_DCT_COEFF 0x00000040
1573 #define FF_DEBUG_SKIP 0x00000080
1574 #define FF_DEBUG_STARTCODE 0x00000100
1575 #define FF_DEBUG_PTS 0x00000200
1576 #define FF_DEBUG_ER 0x00000400
1577 #define FF_DEBUG_MMCO 0x00000800
1578 #define FF_DEBUG_BUGS 0x00001000
1579 #define FF_DEBUG_VIS_QP 0x00002000
1580 #define FF_DEBUG_VIS_MB_TYPE 0x00004000
1581 #define FF_DEBUG_BUFFERS 0x00008000
1582
1583 /**
1584 * debug
1585 * - encoding: Set by user.
1586 * - decoding: Set by user.
1587 */
1588 int debug_mv;
1589 #define FF_DEBUG_VIS_MV_P_FOR 0x00000001 //visualize forward predicted MVs of P frames
1590 #define FF_DEBUG_VIS_MV_B_FOR 0x00000002 //visualize forward predicted MVs of B frames
1591 #define FF_DEBUG_VIS_MV_B_BACK 0x00000004 //visualize backward predicted MVs of B frames
1592
1593 /**
1594 * error
1595 * - encoding: Set by libavcodec if flags&CODEC_FLAG_PSNR.
1596 * - decoding: unused
1597 */
1598 uint64_t error[4];
1599
1600 /**
1601 * minimum MB quantizer
1602 * - encoding: unused
1603 * - decoding: unused
1604 */
1605 int mb_qmin;
1606
1607 /**
1608 * maximum MB quantizer
1609 * - encoding: unused
1610 * - decoding: unused
1611 */
1612 int mb_qmax;
1613
1614 /**
1615 * motion estimation comparison function
1616 * - encoding: Set by user.
1617 * - decoding: unused
1618 */
1619 int me_cmp;
1620 /**
1621 * subpixel motion estimation comparison function
1622 * - encoding: Set by user.
1623 * - decoding: unused
1624 */
1625 int me_sub_cmp;
1626 /**
1627 * macroblock comparison function (not supported yet)
1628 * - encoding: Set by user.
1629 * - decoding: unused
1630 */
1631 int mb_cmp;
1632 /**
1633 * interlaced DCT comparison function
1634 * - encoding: Set by user.
1635 * - decoding: unused
1636 */
1637 int ildct_cmp;
1638 #define FF_CMP_SAD 0
1639 #define FF_CMP_SSE 1
1640 #define FF_CMP_SATD 2
1641 #define FF_CMP_DCT 3
1642 #define FF_CMP_PSNR 4
1643 #define FF_CMP_BIT 5
1644 #define FF_CMP_RD 6
1645 #define FF_CMP_ZERO 7
1646 #define FF_CMP_VSAD 8
1647 #define FF_CMP_VSSE 9
1648 #define FF_CMP_NSSE 10
1649 #define FF_CMP_W53 11
1650 #define FF_CMP_W97 12
1651 #define FF_CMP_DCTMAX 13
1652 #define FF_CMP_DCT264 14
1653 #define FF_CMP_CHROMA 256
1654
1655 /**
1656 * ME diamond size & shape
1657 * - encoding: Set by user.
1658 * - decoding: unused
1659 */
1660 int dia_size;
1661
1662 /**
1663 * amount of previous MV predictors (2a+1 x 2a+1 square)
1664 * - encoding: Set by user.
1665 * - decoding: unused
1666 */
1667 int last_predictor_count;
1668
1669 /**
1670 * prepass for motion estimation
1671 * - encoding: Set by user.
1672 * - decoding: unused
1673 */
1674 int pre_me;
1675
1676 /**
1677 * motion estimation prepass comparison function
1678 * - encoding: Set by user.
1679 * - decoding: unused
1680 */
1681 int me_pre_cmp;
1682
1683 /**
1684 * ME prepass diamond size & shape
1685 * - encoding: Set by user.
1686 * - decoding: unused
1687 */
1688 int pre_dia_size;
1689
1690 /**
1691 * subpel ME quality
1692 * - encoding: Set by user.
1693 * - decoding: unused
1694 */
1695 int me_subpel_quality;
1696
1697 /**
1698 * callback to negotiate the pixelFormat
1699 * @param fmt is the list of formats which are supported by the codec,
1700 * it is terminated by -1 as 0 is a valid format, the formats are ordered by quality.
1701 * The first is always the native one.
1702 * @return the chosen format
1703 * - encoding: unused
1704 * - decoding: Set by user, if not set the native format will be chosen.
1705 */
1706 enum PixelFormat (*get_format)(struct AVCodecContext *s, const enum PixelFormat * fmt);
1707
1708 /**
1709 * DTG active format information (additional aspect ratio
1710 * information only used in DVB MPEG-2 transport streams)
1711 * 0 if not set.
1712 *
1713 * - encoding: unused
1714 * - decoding: Set by decoder.
1715 */
1716 int dtg_active_format;
1717 #define FF_DTG_AFD_SAME 8
1718 #define FF_DTG_AFD_4_3 9
1719 #define FF_DTG_AFD_16_9 10
1720 #define FF_DTG_AFD_14_9 11
1721 #define FF_DTG_AFD_4_3_SP_14_9 13
1722 #define FF_DTG_AFD_16_9_SP_14_9 14
1723 #define FF_DTG_AFD_SP_4_3 15
1724
1725 /**
1726 * maximum motion estimation search range in subpel units
1727 * If 0 then no limit.
1728 *
1729 * - encoding: Set by user.
1730 * - decoding: unused
1731 */
1732 int me_range;
1733
1734 /**
1735 * intra quantizer bias
1736 * - encoding: Set by user.
1737 * - decoding: unused
1738 */
1739 int intra_quant_bias;
1740 #define FF_DEFAULT_QUANT_BIAS 999999
1741
1742 /**
1743 * inter quantizer bias
1744 * - encoding: Set by user.
1745 * - decoding: unused
1746 */
1747 int inter_quant_bias;
1748
1749 /**
1750 * color table ID
1751 * - encoding: unused
1752 * - decoding: Which clrtable should be used for 8bit RGB images.
1753 * Tables have to be stored somewhere. FIXME
1754 */
1755 int color_table_id;
1756
1757 /**
1758 * internal_buffer count
1759 * Don't touch, used by libavcodec default_get_buffer().
1760 */
1761 int internal_buffer_count;
1762
1763 /**
1764 * internal_buffers
1765 * Don't touch, used by libavcodec default_get_buffer().
1766 */
1767 void *internal_buffer;
1768
1769 #define FF_LAMBDA_SHIFT 7
1770 #define FF_LAMBDA_SCALE (1<<FF_LAMBDA_SHIFT)
1771 #define FF_QP2LAMBDA 118 ///< factor to convert from H.263 QP to lambda
1772 #define FF_LAMBDA_MAX (256*128-1)
1773
1774 #define FF_QUALITY_SCALE FF_LAMBDA_SCALE //FIXME maybe remove
1775 /**
1776 * Global quality for codecs which cannot change it per frame.
1777 * This should be proportional to MPEG-1/2/4 qscale.
1778 * - encoding: Set by user.
1779 * - decoding: unused
1780 */
1781 int global_quality;
1782
1783 #define FF_CODER_TYPE_VLC 0
1784 #define FF_CODER_TYPE_AC 1
1785 #define FF_CODER_TYPE_RAW 2
1786 #define FF_CODER_TYPE_RLE 3
1787 #define FF_CODER_TYPE_DEFLATE 4
1788 /**
1789 * coder type
1790 * - encoding: Set by user.
1791 * - decoding: unused
1792 */
1793 int coder_type;
1794
1795 /**
1796 * context model
1797 * - encoding: Set by user.
1798 * - decoding: unused
1799 */
1800 int context_model;
1801 #if 0
1802 /**
1803 *
1804 * - encoding: unused
1805 * - decoding: Set by user.
1806 */
1807 uint8_t * (*realloc)(struct AVCodecContext *s, uint8_t *buf, int buf_size);
1808 #endif
1809
1810 /**
1811 * slice flags
1812 * - encoding: unused
1813 * - decoding: Set by user.
1814 */
1815 int slice_flags;
1816 #define SLICE_FLAG_CODED_ORDER 0x0001 ///< draw_horiz_band() is called in coded order instead of display
1817 #define SLICE_FLAG_ALLOW_FIELD 0x0002 ///< allow draw_horiz_band() with field slices (MPEG2 field pics)
1818 #define SLICE_FLAG_ALLOW_PLANE 0x0004 ///< allow draw_horiz_band() with 1 component at a time (SVQ1)
1819
1820 /**
1821 * XVideo Motion Acceleration
1822 * - encoding: forbidden
1823 * - decoding: set by decoder
1824 */
1825 int xvmc_acceleration;
1826
1827 /**
1828 * macroblock decision mode
1829 * - encoding: Set by user.
1830 * - decoding: unused
1831 */
1832 int mb_decision;
1833 #define FF_MB_DECISION_SIMPLE 0 ///< uses mb_cmp
1834 #define FF_MB_DECISION_BITS 1 ///< chooses the one which needs the fewest bits
1835 #define FF_MB_DECISION_RD 2 ///< rate distortion
1836
1837 /**
1838 * custom intra quantization matrix
1839 * - encoding: Set by user, can be NULL.
1840 * - decoding: Set by libavcodec.
1841 */
1842 uint16_t *intra_matrix;
1843
1844 /**
1845 * custom inter quantization matrix
1846 * - encoding: Set by user, can be NULL.
1847 * - decoding: Set by libavcodec.
1848 */
1849 uint16_t *inter_matrix;
1850
1851 /**
1852 * fourcc from the AVI stream header (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
1853 * This is used to work around some encoder bugs.
1854 * - encoding: unused
1855 * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
1856 */
1857 unsigned int stream_codec_tag;
1858
1859 /**
1860 * scene change detection threshold
1861 * 0 is default, larger means fewer detected scene changes.
1862 * - encoding: Set by user.
1863 * - decoding: unused
1864 */
1865 int scenechange_threshold;
1866
1867 /**
1868 * minimum Lagrange multipler
1869 * - encoding: Set by user.
1870 * - decoding: unused
1871 */
1872 int lmin;
1873
1874 /**
1875 * maximum Lagrange multipler
1876 * - encoding: Set by user.
1877 * - decoding: unused
1878 */
1879 int lmax;
1880
1881 /**
1882 * palette control structure
1883 * - encoding: ??? (no palette-enabled encoder yet)
1884 * - decoding: Set by user.
1885 */
1886 struct AVPaletteControl *palctrl;
1887
1888 /**
1889 * noise reduction strength
1890 * - encoding: Set by user.
1891 * - decoding: unused
1892 */
1893 int noise_reduction;
1894
1895 /**
1896 * Called at the beginning of a frame to get cr buffer for it.
1897 * Buffer type (size, hints) must be the same. libavcodec won't check it.
1898 * libavcodec will pass previous buffer in pic, function should return
1899 * same buffer or new buffer with old frame "painted" into it.
1900 * If pic.data[0] == NULL must behave like get_buffer().
1901 * - encoding: unused
1902 * - decoding: Set by libavcodec., user can override
1903 */
1904 int (*reget_buffer)(struct AVCodecContext *c, AVFrame *pic);
1905
1906 /**
1907 * Number of bits which should be loaded into the rc buffer before decoding starts.
1908 * - encoding: Set by user.
1909 * - decoding: unused
1910 */
1911 int rc_initial_buffer_occupancy;
1912
1913 /**
1914 *
1915 * - encoding: Set by user.
1916 * - decoding: unused
1917 */
1918 int inter_threshold;
1919
1920 /**
1921 * CODEC_FLAG2_*
1922 * - encoding: Set by user.
1923 * - decoding: Set by user.
1924 */
1925 int flags2;
1926
1927 /**
1928 * Simulates errors in the bitstream to test error concealment.
1929 * - encoding: Set by user.
1930 * - decoding: unused
1931 */
1932 int error_rate;
1933
1934 /**
1935 * MP3 antialias algorithm, see FF_AA_* below.
1936 * - encoding: unused
1937 * - decoding: Set by user.
1938 */
1939 int antialias_algo;
1940 #define FF_AA_AUTO 0
1941 #define FF_AA_FASTINT 1 //not implemented yet
1942 #define FF_AA_INT 2
1943 #define FF_AA_FLOAT 3
1944 /**
1945 * quantizer noise shaping
1946 * - encoding: Set by user.
1947 * - decoding: unused
1948 */
1949 int quantizer_noise_shaping;
1950
1951 /**
1952 * thread count
1953 * is used to decide how many independent tasks should be passed to execute()
1954 * - encoding: Set by user.
1955 * - decoding: Set by user.
1956 */
1957 int thread_count;
1958
1959 /**
1960 * The codec may call this to execute several independent things.
1961 * It will return only after finishing all tasks.
1962 * The user may replace this with some multithreaded implementation,
1963 * the default implementation will execute the parts serially.
1964 * @param count the number of things to execute
1965 * - encoding: Set by libavcodec, user can override.
1966 * - decoding: Set by libavcodec, user can override.
1967 */
1968 int (*execute)(struct AVCodecContext *c, int (*func)(struct AVCodecContext *c2, void *arg), void *arg2, int *ret, int count, int size);
1969
1970 /**
1971 * thread opaque
1972 * Can be used by execute() to store some per AVCodecContext stuff.
1973 * - encoding: set by execute()
1974 * - decoding: set by execute()
1975 */
1976 void *thread_opaque;
1977
1978 /**
1979 * Motion estimation threshold below which no motion estimation is
1980 * performed, but instead the user specified motion vectors are used.
1981 *
1982 * - encoding: Set by user.
1983 * - decoding: unused
1984 */
1985 int me_threshold;
1986
1987 /**
1988 * Macroblock threshold below which the user specified macroblock types will be used.
1989 * - encoding: Set by user.
1990 * - decoding: unused
1991 */
1992 int mb_threshold;
1993
1994 /**
1995 * precision of the intra DC coefficient - 8
1996 * - encoding: Set by user.
1997 * - decoding: unused
1998 */
1999 int intra_dc_precision;
2000
2001 /**
2002 * noise vs. sse weight for the nsse comparsion function
2003 * - encoding: Set by user.
2004 * - decoding: unused
2005 */
2006 int nsse_weight;
2007
2008 /**
2009 * Number of macroblock rows at the top which are skipped.
2010 * - encoding: unused
2011 * - decoding: Set by user.
2012 */
2013 int skip_top;
2014
2015 /**
2016 * Number of macroblock rows at the bottom which are skipped.
2017 * - encoding: unused
2018 * - decoding: Set by user.
2019 */
2020 int skip_bottom;
2021
2022 /**
2023 * profile
2024 * - encoding: Set by user.
2025 * - decoding: Set by libavcodec.
2026 */
2027 int profile;
2028 #define FF_PROFILE_UNKNOWN -99
2029 #define FF_PROFILE_AAC_MAIN 0
2030 #define FF_PROFILE_AAC_LOW 1
2031 #define FF_PROFILE_AAC_SSR 2
2032 #define FF_PROFILE_AAC_LTP 3
2033
2034 /**
2035 * level
2036 * - encoding: Set by user.
2037 * - decoding: Set by libavcodec.
2038 */
2039 int level;
2040 #define FF_LEVEL_UNKNOWN -99
2041
2042 /**
2043 * low resolution decoding, 1-> 1/2 size, 2->1/4 size
2044 * - encoding: unused
2045 * - decoding: Set by user.
2046 */
2047 int lowres;
2048
2049 /**
2050 * Bitstream width / height, may be different from width/height if lowres
2051 * or other things are used.
2052 * - encoding: unused
2053 * - decoding: Set by user before init if known. Codec should override / dynamically change if needed.
2054 */
2055 int coded_width, coded_height;
2056
2057 /**
2058 * frame skip threshold
2059 * - encoding: Set by user.
2060 * - decoding: unused
2061 */
2062 int frame_skip_threshold;
2063
2064 /**
2065 * frame skip factor
2066 * - encoding: Set by user.
2067 * - decoding: unused
2068 */
2069 int frame_skip_factor;
2070
2071 /**
2072 * frame skip exponent
2073 * - encoding: Set by user.
2074 * - decoding: unused
2075 */
2076 int frame_skip_exp;
2077
2078 /**
2079 * frame skip comparison function
2080 * - encoding: Set by user.
2081 * - decoding: unused
2082 */
2083 int frame_skip_cmp;
2084
2085 /**
2086 * Border processing masking, raises the quantizer for mbs on the borders
2087 * of the picture.
2088 * - encoding: Set by user.
2089 * - decoding: unused
2090 */
2091 float border_masking;
2092
2093 /**
2094 * minimum MB lagrange multipler
2095 * - encoding: Set by user.
2096 * - decoding: unused
2097 */
2098 int mb_lmin;
2099
2100 /**
2101 * maximum MB lagrange multipler
2102 * - encoding: Set by user.
2103 * - decoding: unused
2104 */
2105 int mb_lmax;
2106
2107 /**
2108 *
2109 * - encoding: Set by user.
2110 * - decoding: unused
2111 */
2112 int me_penalty_compensation;
2113
2114 /**
2115 *
2116 * - encoding: unused
2117 * - decoding: Set by user.
2118 */
2119 enum AVDiscard skip_loop_filter;
2120
2121 /**
2122 *
2123 * - encoding: unused
2124 * - decoding: Set by user.
2125 */
2126 enum AVDiscard skip_idct;
2127
2128 /**
2129 *
2130 * - encoding: unused
2131 * - decoding: Set by user.
2132 */
2133 enum AVDiscard skip_frame;
2134
2135 /**
2136 *
2137 * - encoding: Set by user.
2138 * - decoding: unused
2139 */
2140 int bidir_refine;
2141
2142 /**
2143 *
2144 * - encoding: Set by user.
2145 * - decoding: unused
2146 */
2147 int brd_scale;
2148
2149 /**
2150 * constant rate factor - quality-based VBR - values ~correspond to qps
2151 * - encoding: Set by user.
2152 * - decoding: unused
2153 */
2154 float crf;
2155
2156 /**
2157 * constant quantization parameter rate control method
2158 * - encoding: Set by user.
2159 * - decoding: unused
2160 */
2161 int cqp;
2162
2163 /**
2164 * minimum GOP size
2165 * - encoding: Set by user.
2166 * - decoding: unused
2167 */
2168 int keyint_min;
2169
2170 /**
2171 * number of reference frames
2172 * - encoding: Set by user.
2173 * - decoding: Set by lavc.
2174 */
2175 int refs;
2176
2177 /**
2178 * chroma qp offset from luma
2179 * - encoding: Set by user.
2180 * - decoding: unused
2181 */
2182 int chromaoffset;
2183
2184 /**
2185 * Influences how often B-frames are used.
2186 * - encoding: Set by user.
2187 * - decoding: unused
2188 */
2189 int bframebias;
2190
2191 /**
2192 * trellis RD quantization
2193 * - encoding: Set by user.
2194 * - decoding: unused
2195 */
2196 int trellis;
2197
2198 /**
2199 * Reduce fluctuations in qp (before curve compression).
2200 * - encoding: Set by user.
2201 * - decoding: unused
2202 */
2203 float complexityblur;
2204
2205 /**
2206 * in-loop deblocking filter alphac0 parameter
2207 * alpha is in the range -6...6
2208 * - encoding: Set by user.
2209 * - decoding: unused
2210 */
2211 int deblockalpha;
2212
2213 /**
2214 * in-loop deblocking filter beta parameter
2215 * beta is in the range -6...6
2216 * - encoding: Set by user.
2217 * - decoding: unused
2218 */
2219 int deblockbeta;
2220
2221 /**
2222 * macroblock subpartition sizes to consider - p8x8, p4x4, b8x8, i8x8, i4x4
2223 * - encoding: Set by user.
2224 * - decoding: unused
2225 */
2226 int partitions;
2227 #define X264_PART_I4X4 0x001 /* Analyze i4x4 */
2228 #define X264_PART_I8X8 0x002 /* Analyze i8x8 (requires 8x8 transform) */
2229 #define X264_PART_P8X8 0x010 /* Analyze p16x8, p8x16 and p8x8 */
2230 #define X264_PART_P4X4 0x020 /* Analyze p8x4, p4x8, p4x4 */
2231 #define X264_PART_B8X8 0x100 /* Analyze b16x8, b8x16 and b8x8 */
2232
2233 /**
2234 * direct MV prediction mode - 0 (none), 1 (spatial), 2 (temporal), 3 (auto)
2235 * - encoding: Set by user.
2236 * - decoding: unused
2237 */
2238 int directpred;
2239
2240 /**
2241 * Audio cutoff bandwidth (0 means "automatic")
2242 * - encoding: Set by user.
2243 * - decoding: unused
2244 */
2245 int cutoff;
2246
2247 /**
2248 * Multiplied by qscale for each frame and added to scene_change_score.
2249 * - encoding: Set by user.
2250 * - decoding: unused
2251 */
2252 int scenechange_factor;
2253
2254 /**
2255 *
2256 * Note: Value depends upon the compare function used for fullpel ME.
2257 * - encoding: Set by user.
2258 * - decoding: unused
2259 */
2260 int mv0_threshold;
2261
2262 /**
2263 * Adjusts sensitivity of b_frame_strategy 1.
2264 * - encoding: Set by user.
2265 * - decoding: unused
2266 */
2267 int b_sensitivity;
2268
2269 /**
2270 * - encoding: Set by user.
2271 * - decoding: unused
2272 */
2273 int compression_level;
2274 #define FF_COMPRESSION_DEFAULT -1
2275
2276 /**
2277 * Sets whether to use LPC mode - used by FLAC encoder.
2278 * - encoding: Set by user.
2279 * - decoding: unused
2280 */
2281 int use_lpc;
2282
2283 /**
2284 * LPC coefficient precision - used by FLAC encoder
2285 * - encoding: Set by user.
2286 * - decoding: unused
2287 */
2288 int lpc_coeff_precision;
2289
2290 /**
2291 * - encoding: Set by user.
2292 * - decoding: unused
2293 */
2294 int min_prediction_order;
2295
2296 /**
2297 * - encoding: Set by user.
2298 * - decoding: unused
2299 */
2300 int max_prediction_order;
2301
2302 /**
2303 * search method for selecting prediction order
2304 * - encoding: Set by user.
2305 * - decoding: unused
2306 */
2307 int prediction_order_method;
2308
2309 /**
2310 * - encoding: Set by user.
2311 * - decoding: unused
2312 */
2313 int min_partition_order;
2314
2315 /**
2316 * - encoding: Set by user.
2317 * - decoding: unused
2318 */
2319 int max_partition_order;
2320
2321 /**
2322 * GOP timecode frame start number, in non drop frame format
2323 * - encoding: Set by user.
2324 * - decoding: unused
2325 */
2326 int64_t timecode_frame_start;
2327
2328 #if LIBAVCODEC_VERSION_MAJOR < 53
2329 /**
2330 * Decoder should decode to this many channels if it can (0 for default)
2331 * - encoding: unused
2332 * - decoding: Set by user.
2333 * @deprecated Deprecated in favor of request_channel_layout.
2334 */
2335 int request_channels;
2336 #endif
2337
2338 /**
2339 * Percentage of dynamic range compression to be applied by the decoder.
2340 * The default value is 1.0, corresponding to full compression.
2341 * - encoding: unused
2342 * - decoding: Set by user.
2343 */
2344 float drc_scale;
2345
2346 /**
2347 * opaque 64bit number (generally a PTS) that will be reordered and
2348 * output in AVFrame.reordered_opaque
2349 * - encoding: unused
2350 * - decoding: Set by user.
2351 */
2352 int64_t reordered_opaque;
2353
2354 /**
2355 * Bits per sample/pixel of internal libavcodec pixel/sample format.
2356 * This field is applicable only when sample_fmt is SAMPLE_FMT_S32.
2357 * - encoding: set by user.
2358 * - decoding: set by libavcodec.
2359 */
2360 int bits_per_raw_sample;
2361
2362 /**
2363 * Audio channel layout.
2364 * - encoding: set by user.
2365 * - decoding: set by libavcodec.
2366 */
2367 int64_t channel_layout;
2368
2369 /**
2370 * Request decoder to use this channel layout if it can (0 for default)
2371 * - encoding: unused
2372 * - decoding: Set by user.
2373 */
2374 int64_t request_channel_layout;
2375
2376 /**
2377 * Ratecontrol attempt to use, at maximum, <value> of what can be used without an underflow.
2378 * - encoding: Set by user.
2379 * - decoding: unused.
2380 */
2381 float rc_max_available_vbv_use;
2382
2383 /**
2384 * Ratecontrol attempt to use, at least, <value> times the amount needed to prevent a vbv overflow.
2385 * - encoding: Set by user.
2386 * - decoding: unused.
2387 */
2388 float rc_min_vbv_overflow_use;
2389
2390 /**
2391 * Hardware accelerator in use
2392 * - encoding: unused.
2393 * - decoding: Set by libavcodec
2394 */
2395 struct AVHWAccel *hwaccel;
2396
2397 /**
2398 * For some codecs, the time base is closer to the field rate than the frame rate.
2399 * Most notably, H.264 and MPEG-2 specify time_base as half of frame duration
2400 * if no telecine is used ...
2401 *
2402 * Set to time_base ticks per frame. Default 1, e.g., H.264/MPEG-2 set it to 2.
2403 */
2404 int ticks_per_frame;
2405
2406 /**
2407 * Hardware accelerator context.
2408 * For some hardware accelerators, a global context needs to be
2409 * provided by the user. In that case, this holds display-dependent
2410 * data FFmpeg cannot instantiate itself. Please refer to the
2411 * FFmpeg HW accelerator documentation to know how to fill this
2412 * is. e.g. for VA API, this is a struct vaapi_context.
2413 * - encoding: unused
2414 * - decoding: Set by user
2415 */
2416 void *hwaccel_context;
2417 } AVCodecContext;
2418
2419 /**
2420 * AVCodec.
2421 */
2422 typedef struct AVCodec {
2423 /**
2424 * Name of the codec implementation.
2425 * The name is globally unique among encoders and among decoders (but an
2426 * encoder and a decoder can share the same name).
2427 * This is the primary way to find a codec from the user perspective.
2428 */
2429 const char *name;
2430 enum CodecType type;
2431 enum CodecID id;
2432 int priv_data_size;
2433 int (*init)(AVCodecContext *);
2434 int (*encode)(AVCodecContext *, uint8_t *buf, int buf_size, void *data);
2435 int (*close)(AVCodecContext *);
2436 int (*decode)(AVCodecContext *, void *outdata, int *outdata_size, AVPacket *avpkt);
2437 /**
2438 * Codec capabilities.
2439 * see CODEC_CAP_*
2440 */
2441 int capabilities;
2442 struct AVCodec *next;
2443 /**
2444 * Flush buffers.
2445 * Will be called when seeking
2446 */
2447 void (*flush)(AVCodecContext *);
2448 const AVRational *supported_framerates; ///< array of supported framerates, or NULL if any, array is terminated by {0,0}
2449 const enum PixelFormat *pix_fmts; ///< array of supported pixel formats, or NULL if unknown, array is terminated by -1
2450 /**
2451 * Descriptive name for the codec, meant to be more human readable than \p name.
2452 * You \e should use the NULL_IF_CONFIG_SMALL() macro to define it.
2453 */
2454 const char *long_name;
2455 const int *supported_samplerates; ///< array of supported audio samplerates, or NULL if unknown, array is terminated by 0
2456 const enum SampleFormat *sample_fmts; ///< array of supported sample formats, or NULL if unknown, array is terminated by -1
2457 const int64_t *channel_layouts; ///< array of support channel layouts, or NULL if unknown. array is terminated by 0
2458 } AVCodec;
2459
2460 /**
2461 * AVHWAccel.
2462 */
2463 typedef struct AVHWAccel {
2464 /**
2465 * Name of the hardware accelerated codec.
2466 * The name is globally unique among encoders and among decoders (but an
2467 * encoder and a decoder can share the same name).
2468 */
2469 const char *name;
2470
2471 /**
2472 * Type of codec implemented by the hardware accelerator.
2473 *
2474 * See CODEC_TYPE_xxx
2475 */
2476 enum CodecType type;
2477
2478 /**
2479 * Codec implemented by the hardware accelerator.
2480 *
2481 * See CODEC_ID_xxx
2482 */
2483 enum CodecID id;
2484
2485 /**
2486 * Supported pixel format.
2487 *
2488 * Only hardware accelerated formats are supported here.
2489 */
2490 enum PixelFormat pix_fmt;
2491
2492 /**
2493 * Hardware accelerated codec capabilities.
2494 * see FF_HWACCEL_CODEC_CAP_*
2495 */
2496 int capabilities;
2497
2498 struct AVHWAccel *next;
2499
2500 /**
2501 * Called at the beginning of each frame or field picture.
2502 *
2503 * Meaningful frame information (codec specific) is guaranteed to
2504 * be parsed at this point. This function is mandatory.
2505 *
2506 * Note that \p buf can be NULL along with \p buf_size set to 0.
2507 * Otherwise, this means the whole frame is available at this point.
2508 *
2509 * @param avctx the codec context
2510 * @param buf the frame data buffer base
2511 * @param buf_size the size of the frame in bytes
2512 * @return zero if successful, a negative value otherwise
2513 */
2514 int (*start_frame)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
2515
2516 /**
2517 * Callback for each slice.
2518 *
2519 * Meaningful slice information (codec specific) is guaranteed to
2520 * be parsed at this point. This function is mandatory.
2521 *
2522 * @param avctx the codec context
2523 * @param buf the slice data buffer base
2524 * @param buf_size the size of the slice in bytes
2525 * @return zero if successful, a negative value otherwise
2526 */
2527 int (*decode_slice)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
2528
2529 /**
2530 * Called at the end of each frame or field picture.
2531 *
2532 * The whole picture is parsed at this point and can now be sent
2533 * to the hardware accelerator. This function is mandatory.
2534 *
2535 * @param avctx the codec context
2536 * @return zero if successful, a negative value otherwise
2537 */
2538 int (*end_frame)(AVCodecContext *avctx);
2539
2540 /**
2541 * Size of HW accelerator private data.
2542 *
2543 * Private data is allocated with av_mallocz() before
2544 * AVCodecContext.get_buffer() and deallocated after
2545 * AVCodecContext.release_buffer().
2546 */
2547 int priv_data_size;
2548 } AVHWAccel;
2549
2550 /**
2551 * four components are given, that's all.
2552 * the last component is alpha
2553 */
2554 typedef struct AVPicture {
2555 uint8_t *data[4];
2556 int linesize[4]; ///< number of bytes per line
2557 } AVPicture;
2558
2559 #if LIBAVCODEC_VERSION_MAJOR < 53
2560 /**
2561 * AVPaletteControl
2562 * This structure defines a method for communicating palette changes
2563 * between and demuxer and a decoder.
2564 *
2565 * @deprecated Use AVPacket to send palette changes instead.
2566 * This is totally broken.
2567 */
2568 #define AVPALETTE_SIZE 1024
2569 #define AVPALETTE_COUNT 256
2570 typedef struct AVPaletteControl {
2571
2572 /* Demuxer sets this to 1 to indicate the palette has changed;
2573 * decoder resets to 0. */
2574 int palette_changed;
2575
2576 /* 4-byte ARGB palette entries, stored in native byte order; note that
2577 * the individual palette components should be on a 8-bit scale; if
2578 * the palette data comes from an IBM VGA native format, the component
2579 * data is probably 6 bits in size and needs to be scaled. */
2580 unsigned int palette[AVPALETTE_COUNT];
2581
2582 } AVPaletteControl attribute_deprecated;
2583 #endif
2584
2585 enum AVSubtitleType {
2586 SUBTITLE_NONE,
2587
2588 SUBTITLE_BITMAP, ///< A bitmap, pict will be set
2589
2590 /**
2591 * Plain text, the text field must be set by the decoder and is
2592 * authoritative. ass and pict fields may contain approximations.
2593 */
2594 SUBTITLE_TEXT,
2595
2596 /**
2597 * Formatted text, the ass field must be set by the decoder and is
2598 * authoritative. pict and text fields may contain approximations.
2599 */
2600 SUBTITLE_ASS,
2601 };
2602
2603 typedef struct AVSubtitleRect {
2604 int x; ///< top left corner of pict, undefined when pict is not set
2605 int y; ///< top left corner of pict, undefined when pict is not set
2606 int w; ///< width of pict, undefined when pict is not set
2607 int h; ///< height of pict, undefined when pict is not set
2608 int nb_colors; ///< number of colors in pict, undefined when pict is not set
2609
2610 /**
2611 * data+linesize for the bitmap of this subtitle.
2612 * can be set for text/ass as well once they where rendered
2613 */
2614 AVPicture pict;
2615 enum AVSubtitleType type;
2616
2617 char *text; ///< 0 terminated plain UTF-8 text
2618
2619 /**
2620 * 0 terminated ASS/SSA compatible event line.
2621 * The pressentation of this is unaffected by the other values in this
2622 * struct.
2623 */
2624 char *ass;
2625 } AVSubtitleRect;
2626
2627 typedef struct AVSubtitle {
2628 uint16_t format; /* 0 = graphics */
2629 uint32_t start_display_time; /* relative to packet pts, in ms */
2630 uint32_t end_display_time; /* relative to packet pts, in ms */
2631 unsigned num_rects;
2632 AVSubtitleRect **rects;
2633 } AVSubtitle;
2634
2635 /* packet functions */
2636
2637 /**
2638 * @deprecated use NULL instead
2639 */
2640 attribute_deprecated void av_destruct_packet_nofree(AVPacket *pkt);
2641
2642 /**
2643 * Default packet destructor.
2644 */
2645 void av_destruct_packet(AVPacket *pkt);
2646
2647 /**
2648 * Initialize optional fields of a packet with default values.
2649 *
2650 * @param pkt packet
2651 */
2652 void av_init_packet(AVPacket *pkt);
2653
2654 /**
2655 * Allocate the payload of a packet and initialize its fields with
2656 * default values.
2657 *
2658 * @param pkt packet
2659 * @param size wanted payload size
2660 * @return 0 if OK, AVERROR_xxx otherwise
2661 */
2662 int av_new_packet(AVPacket *pkt, int size);
2663
2664 /**
2665 * Reduce packet size, correctly zeroing padding
2666 *
2667 * @param pkt packet
2668 * @param size new size
2669 */
2670 void av_shrink_packet(AVPacket *pkt, int size);
2671
2672 /**
2673 * @warning This is a hack - the packet memory allocation stuff is broken. The
2674 * packet is allocated if it was not really allocated.
2675 */
2676 int av_dup_packet(AVPacket *pkt);
2677
2678 /**
2679 * Free a packet.
2680 *
2681 * @param pkt packet to free
2682 */
2683 static inline void av_free_packet(AVPacket *pkt)
2684 {
2685 if (pkt) {
2686 if (pkt->destruct) pkt->destruct(pkt);
2687 pkt->data = NULL; pkt->size = 0;
2688 }
2689 }
2690
2691 /* resample.c */
2692
2693 struct ReSampleContext;
2694 struct AVResampleContext;
2695
2696 typedef struct ReSampleContext ReSampleContext;
2697
2698 #if LIBAVCODEC_VERSION_MAJOR < 53
2699 /**
2700 * @deprecated Use av_audio_resample_init() instead.
2701 */
2702 attribute_deprecated ReSampleContext *audio_resample_init(int output_channels, int input_channels,
2703 int output_rate, int input_rate);
2704 #endif
2705 /**
2706 * Initializes audio resampling context
2707 *
2708 * @param output_channels number of output channels
2709 * @param input_channels number of input channels
2710 * @param output_rate output sample rate
2711 * @param input_rate input sample rate
2712 * @param sample_fmt_out requested output sample format
2713 * @param sample_fmt_in input sample format
2714 * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
2715 * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
2716 * @param linear If 1 then the used FIR filter will be linearly interpolated
2717 between the 2 closest, if 0 the closest will be used
2718 * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
2719 * @return allocated ReSampleContext, NULL if error occured
2720 */
2721 ReSampleContext *av_audio_resample_init(int output_channels, int input_channels,
2722 int output_rate, int input_rate,
2723 enum SampleFormat sample_fmt_out,
2724 enum SampleFormat sample_fmt_in,
2725 int filter_length, int log2_phase_count,
2726 int linear, double cutoff);
2727
2728 int audio_resample(ReSampleContext *s, short *output, short *input, int nb_samples);
2729 void audio_resample_close(ReSampleContext *s);
2730
2731
2732 /**
2733 * Initializes an audio resampler.
2734 * Note, if either rate is not an integer then simply scale both rates up so they are.
2735 * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
2736 * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
2737 * @param linear If 1 then the used FIR filter will be linearly interpolated
2738 between the 2 closest, if 0 the closest will be used
2739 * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
2740 */
2741 struct AVResampleContext *av_resample_init(int out_rate, int in_rate, int filter_length, int log2_phase_count, int linear, double cutoff);
2742
2743 /**
2744 * resamples.
2745 * @param src an array of unconsumed samples
2746 * @param consumed the number of samples of src which have been consumed are returned here
2747 * @param src_size the number of unconsumed samples available
2748 * @param dst_size the amount of space in samples available in dst
2749 * @param update_ctx If this is 0 then the context will not be modified, that way several channels can be resampled with the same context.
2750 * @return the number of samples written in dst or -1 if an error occurred
2751 */
2752 int av_resample(struct AVResampleContext *c, short *dst, short *src, int *consumed, int src_size, int dst_size, int update_ctx);
2753
2754
2755 /**
2756 * Compensates samplerate/timestamp drift. The compensation is done by changing
2757 * the resampler parameters, so no audible clicks or similar distortions occur
2758 * @param compensation_distance distance in output samples over which the compensation should be performed
2759 * @param sample_delta number of output samples which should be output less
2760 *
2761 * example: av_resample_compensate(c, 10, 500)
2762 * here instead of 510 samples only 500 samples would be output
2763 *
2764 * note, due to rounding the actual compensation might be slightly different,
2765 * especially if the compensation_distance is large and the in_rate used during init is small
2766 */
2767 void av_resample_compensate(struct AVResampleContext *c, int sample_delta, int compensation_distance);
2768 void av_resample_close(struct AVResampleContext *c);
2769
2770 /**
2771 * Allocate memory for a picture. Call avpicture_free to free it.
2772 *
2773 * @param picture the picture to be filled in
2774 * @param pix_fmt the format of the picture
2775 * @param width the width of the picture
2776 * @param height the height of the picture
2777 * @return zero if successful, a negative value if not
2778 */
2779 int avpicture_alloc(AVPicture *picture, enum PixelFormat pix_fmt, int width, int height);
2780
2781 /**
2782 * Free a picture previously allocated by avpicture_alloc().
2783 *
2784 * @param picture the AVPicture to be freed
2785 */
2786 void avpicture_free(AVPicture *picture);
2787
2788 /**
2789 * Fill in the AVPicture fields.
2790 * The fields of the given AVPicture are filled in by using the 'ptr' address
2791 * which points to the image data buffer. Depending on the specified picture
2792 * format, one or multiple image data pointers and line sizes will be set.
2793 * If a planar format is specified, several pointers will be set pointing to
2794 * the different picture planes and the line sizes of the different planes
2795 * will be stored in the lines_sizes array.
2796 * Call with ptr == NULL to get the required size for the ptr buffer.
2797 *
2798 * @param picture AVPicture whose fields are to be filled in
2799 * @param ptr Buffer which will contain or contains the actual image data
2800 * @param pix_fmt The format in which the picture data is stored.
2801 * @param width the width of the image in pixels
2802 * @param height the height of the image in pixels
2803 * @return size of the image data in bytes
2804 */
2805 int avpicture_fill(AVPicture *picture, uint8_t *ptr,
2806 int pix_fmt, int width, int height);
2807 int avpicture_layout(const AVPicture* src, enum PixelFormat pix_fmt, int width, int height,
2808 unsigned char *dest, int dest_size);
2809
2810 /**
2811 * Calculate the size in bytes that a picture of the given width and height
2812 * would occupy if stored in the given picture format.
2813 * Note that this returns the size of a compact representation as generated
2814 * by avpicture_layout, which can be smaller than the size required for e.g.
2815 * avpicture_fill.
2816 *
2817 * @param pix_fmt the given picture format
2818 * @param width the width of the image
2819 * @param height the height of the image
2820 * @return Image data size in bytes or -1 on error (e.g. too large dimensions).
2821 */
2822 int avpicture_get_size(enum PixelFormat pix_fmt, int width, int height);
2823 void avcodec_get_chroma_sub_sample(enum PixelFormat pix_fmt, int *h_shift, int *v_shift);
2824 const char *avcodec_get_pix_fmt_name(enum PixelFormat pix_fmt);
2825 void avcodec_set_dimensions(AVCodecContext *s, int width, int height);
2826
2827 /**
2828 * Returns the pixel format corresponding to the name \p name.
2829 *
2830 * If there is no pixel format with name \p name, then looks for a
2831 * pixel format with the name corresponding to the native endian
2832 * format of \p name.
2833 * For example in a little-endian system, first looks for "gray16",
2834 * then for "gray16le".
2835 *
2836 * Finally if no pixel format has been found, returns \c PIX_FMT_NONE.
2837 */
2838 enum PixelFormat avcodec_get_pix_fmt(const char* name);
2839 unsigned int avcodec_pix_fmt_to_codec_tag(enum PixelFormat p);
2840
2841 #define FF_LOSS_RESOLUTION 0x0001 /**< loss due to resolution change */
2842 #define FF_LOSS_DEPTH 0x0002 /**< loss due to color depth change */
2843 #define FF_LOSS_COLORSPACE 0x0004 /**< loss due to color space conversion */
2844 #define FF_LOSS_ALPHA 0x0008 /**< loss of alpha bits */
2845 #define FF_LOSS_COLORQUANT 0x0010 /**< loss due to color quantization */
2846 #define FF_LOSS_CHROMA 0x0020 /**< loss of chroma (e.g. RGB to gray conversion) */
2847
2848 /**
2849 * Computes what kind of losses will occur when converting from one specific
2850 * pixel format to another.
2851 * When converting from one pixel format to another, information loss may occur.
2852 * For example, when converting from RGB24 to GRAY, the color information will
2853 * be lost. Similarly, other losses occur when converting from some formats to
2854 * other formats. These losses can involve loss of chroma, but also loss of
2855 * resolution, loss of color depth, loss due to the color space conversion, loss
2856 * of the alpha bits or loss due to color quantization.
2857 * avcodec_get_fix_fmt_loss() informs you about the various types of losses
2858 * which will occur when converting from one pixel format to another.
2859 *
2860 * @param[in] dst_pix_fmt destination pixel format
2861 * @param[in] src_pix_fmt source pixel format
2862 * @param[in] has_alpha Whether the source pixel format alpha channel is used.
2863 * @return Combination of flags informing you what kind of losses will occur.
2864 */
2865 int avcodec_get_pix_fmt_loss(enum PixelFormat dst_pix_fmt, enum PixelFormat src_pix_fmt,
2866 int has_alpha);
2867
2868 /**
2869 * Finds the best pixel format to convert to given a certain source pixel
2870 * format. When converting from one pixel format to another, information loss
2871 * may occur. For example, when converting from RGB24 to GRAY, the color
2872 * information will be lost. Similarly, other losses occur when converting from
2873 * some formats to other formats. avcodec_find_best_pix_fmt() searches which of
2874 * the given pixel formats should be used to suffer the least amount of loss.
2875 * The pixel formats from which it chooses one, are determined by the
2876 * \p pix_fmt_mask parameter.
2877 *
2878 * @code
2879 * src_pix_fmt = PIX_FMT_YUV420P;
2880 * pix_fmt_mask = (1 << PIX_FMT_YUV422P) || (1 << PIX_FMT_RGB24);
2881 * dst_pix_fmt = avcodec_find_best_pix_fmt(pix_fmt_mask, src_pix_fmt, alpha, &loss);
2882 * @endcode
2883 *
2884 * @param[in] pix_fmt_mask bitmask determining which pixel format to choose from
2885 * @param[in] src_pix_fmt source pixel format
2886 * @param[in] has_alpha Whether the source pixel format alpha channel is used.
2887 * @param[out] loss_ptr Combination of flags informing you what kind of losses will occur.
2888 * @return The best pixel format to convert to or -1 if none was found.
2889 */
2890 enum PixelFormat avcodec_find_best_pix_fmt(int64_t pix_fmt_mask, enum PixelFormat src_pix_fmt,
2891 int has_alpha, int *loss_ptr);
2892
2893
2894 /**
2895 * Print in buf the string corresponding to the pixel format with
2896 * number pix_fmt, or an header if pix_fmt is negative.
2897 *
2898 * @param[in] buf the buffer where to write the string
2899 * @param[in] buf_size the size of buf
2900 * @param[in] pix_fmt the number of the pixel format to print the corresponding info string, or
2901 * a negative value to print the corresponding header.
2902 * Meaningful values for obtaining a pixel format info vary from 0 to PIX_FMT_NB -1.
2903 */
2904 void avcodec_pix_fmt_string (char *buf, int buf_size, enum PixelFormat pix_fmt);
2905
2906 #define FF_ALPHA_TRANSP 0x0001 /* image has some totally transparent pixels */
2907 #define FF_ALPHA_SEMI_TRANSP 0x0002 /* image has some transparent pixels */
2908
2909 /**
2910 * Tell if an image really has transparent alpha values.
2911 * @return ored mask of FF_ALPHA_xxx constants
2912 */
2913 int img_get_alpha_info(const AVPicture *src,
2914 enum PixelFormat pix_fmt, int width, int height);
2915
2916 /* deinterlace a picture */
2917 /* deinterlace - if not supported return -1 */
2918 int avpicture_deinterlace(AVPicture *dst, const AVPicture *src,
2919 enum PixelFormat pix_fmt, int width, int height);
2920
2921 /* external high level API */
2922
2923 /**
2924 * If c is NULL, returns the first registered codec,
2925 * if c is non-NULL, returns the next registered codec after c,
2926 * or NULL if c is the last one.
2927 */
2928 AVCodec *av_codec_next(AVCodec *c);
2929
2930 /**
2931 * Returns the LIBAVCODEC_VERSION_INT constant.
2932 */
2933 unsigned avcodec_version(void);
2934
2935 /**
2936 * Initializes libavcodec.
2937 *
2938 * @warning This function \e must be called before any other libavcodec
2939 * function.
2940 */
2941 void avcodec_init(void);
2942
2943 #if LIBAVCODEC_VERSION_MAJOR < 53
2944 /**
2945 * @deprecated Deprecated in favor of avcodec_register().
2946 */
2947 attribute_deprecated void register_avcodec(AVCodec *codec);
2948 #endif
2949
2950 /**
2951 * Register the codec \p codec and initialize libavcodec.
2952 *
2953 * @see avcodec_init()
2954 */
2955 void avcodec_register(AVCodec *codec);
2956
2957 /**
2958 * Finds a registered encoder with a matching codec ID.
2959 *
2960 * @param id CodecID of the requested encoder
2961 * @return An encoder if one was found, NULL otherwise.
2962 */
2963 AVCodec *avcodec_find_encoder(enum CodecID id);
2964
2965 /**
2966 * Finds a registered encoder with the specified name.
2967 *
2968 * @param name name of the requested encoder
2969 * @return An encoder if one was found, NULL otherwise.
2970 */
2971 AVCodec *avcodec_find_encoder_by_name(const char *name);
2972
2973 /**
2974 * Finds a registered decoder with a matching codec ID.
2975 *
2976 * @param id CodecID of the requested decoder
2977 * @return A decoder if one was found, NULL otherwise.
2978 */
2979 AVCodec *avcodec_find_decoder(enum CodecID id);
2980
2981 /**
2982 * Finds a registered decoder with the specified name.
2983 *
2984 * @param name name of the requested decoder
2985 * @return A decoder if one was found, NULL otherwise.
2986 */
2987 AVCodec *avcodec_find_decoder_by_name(const char *name);
2988 void avcodec_string(char *buf, int buf_size, AVCodecContext *enc, int encode);
2989
2990 /**
2991 * Sets the fields of the given AVCodecContext to default values.
2992 *
2993 * @param s The AVCodecContext of which the fields should be set to default values.
2994 */
2995 void avcodec_get_context_defaults(AVCodecContext *s);
2996
2997 /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
2998 * we WILL change its arguments and name a few times! */
2999 void avcodec_get_context_defaults2(AVCodecContext *s, enum CodecType);
3000
3001 /**
3002 * Allocates an AVCodecContext and sets its fields to default values. The
3003 * resulting struct can be deallocated by simply calling av_free().
3004 *
3005 * @return An AVCodecContext filled with default values or NULL on failure.
3006 * @see avcodec_get_context_defaults
3007 */
3008 AVCodecContext *avcodec_alloc_context(void);
3009
3010 /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
3011 * we WILL change its arguments and name a few times! */
3012 AVCodecContext *avcodec_alloc_context2(enum CodecType);
3013
3014 /**
3015 * Sets the fields of the given AVFrame to default values.
3016 *
3017 * @param pic The AVFrame of which the fields should be set to default values.
3018 */
3019 void avcodec_get_frame_defaults(AVFrame *pic);
3020
3021 /**
3022 * Allocates an AVFrame and sets its fields to default values. The resulting
3023 * struct can be deallocated by simply calling av_free().
3024 *
3025 * @return An AVFrame filled with default values or NULL on failure.
3026 * @see avcodec_get_frame_defaults
3027 */
3028 AVFrame *avcodec_alloc_frame(void);
3029
3030 int avcodec_default_get_buffer(AVCodecContext *s, AVFrame *pic);
3031 void avcodec_default_release_buffer(AVCodecContext *s, AVFrame *pic);
3032 int avcodec_default_reget_buffer(AVCodecContext *s, AVFrame *pic);
3033 void avcodec_align_dimensions(AVCodecContext *s, int *width, int *height);
3034
3035 /**
3036 * Checks if the given dimension of a picture is valid, meaning that all
3037 * bytes of the picture can be addressed with a signed int.
3038 *
3039 * @param[in] w Width of the picture.
3040 * @param[in] h Height of the picture.
3041 * @return Zero if valid, a negative value if invalid.
3042 */
3043 int avcodec_check_dimensions(void *av_log_ctx, unsigned int w, unsigned int h);
3044 enum PixelFormat avcodec_default_get_format(struct AVCodecContext *s, const enum PixelFormat * fmt);
3045
3046 int avcodec_thread_init(AVCodecContext *s, int thread_count);
3047 void avcodec_thread_free(AVCodecContext *s);
3048 int avcodec_thread_execute(AVCodecContext *s, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
3049 int avcodec_default_execute(AVCodecContext *c, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
3050 //FIXME func typedef
3051
3052 /**
3053 * Initializes the AVCodecContext to use the given AVCodec. Prior to using this
3054 * function the context has to be allocated.
3055 *
3056 * The functions avcodec_find_decoder_by_name(), avcodec_find_encoder_by_name(),
3057 * avcodec_find_decoder() and avcodec_find_encoder() provide an easy way for
3058 * retrieving a codec.
3059 *
3060 * @warning This function is not thread safe!
3061 *
3062 * @code
3063 * avcodec_register_all();
3064 * codec = avcodec_find_decoder(CODEC_ID_H264);
3065 * if (!codec)
3066 * exit(1);
3067 *
3068 * context = avcodec_alloc_context();
3069 *
3070 * if (avcodec_open(context, codec) < 0)
3071 * exit(1);
3072 * @endcode
3073 *
3074 * @param avctx The context which will be set up to use the given codec.
3075 * @param codec The codec to use within the context.
3076 * @return zero on success, a negative value on error
3077 * @see avcodec_alloc_context, avcodec_find_decoder, avcodec_find_encoder
3078 */
3079 int avcodec_open(AVCodecContext *avctx, AVCodec *codec);
3080
3081 #if LIBAVCODEC_VERSION_MAJOR < 53
3082 /**
3083 * Decodes an audio frame from \p buf into \p samples.
3084 * Wrapper function which calls avcodec_decode_audio3.
3085 *
3086 * @deprecated Use avcodec_decode_audio3 instead.
3087 * @param avctx the codec context
3088 * @param[out] samples the output buffer
3089 * @param[in,out] frame_size_ptr the output buffer size in bytes
3090 * @param[in] buf the input buffer
3091 * @param[in] buf_size the input buffer size in bytes
3092 * @return On error a negative value is returned, otherwise the number of bytes
3093 * used or zero if no frame could be decompressed.
3094 */
3095 attribute_deprecated int avcodec_decode_audio2(AVCodecContext *avctx, int16_t *samples,
3096 int *frame_size_ptr,
3097 const uint8_t *buf, int buf_size);
3098 #endif
3099
3100 /**
3101 * Decodes an audio frame from \p avpkt->data into \p samples.
3102 * The avcodec_decode_audio3() function decodes an audio frame from the input
3103 * buffer \p avpkt->data of size \p avpkt->size. To decode it, it makes use of the
3104 * audio codec which was coupled with \p avctx using avcodec_open(). The
3105 * resulting decoded frame is stored in output buffer \p samples. If no frame
3106 * could be decompressed, \p frame_size_ptr is zero. Otherwise, it is the
3107 * decompressed frame size in \e bytes.
3108 *
3109 * @warning You \e must set \p frame_size_ptr to the allocated size of the
3110 * output buffer before calling avcodec_decode_audio3().
3111 *
3112 * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
3113 * the actual read bytes because some optimized bitstream readers read 32 or 64
3114 * bits at once and could read over the end.
3115 *
3116 * @warning The end of the input buffer \p avpkt->data should be set to 0 to ensure that
3117 * no overreading happens for damaged MPEG streams.
3118 *
3119 * @note You might have to align the input buffer \p avpkt->data and output buffer \p
3120 * samples. The alignment requirements depend on the CPU: On some CPUs it isn't
3121 * necessary at all, on others it won't work at all if not aligned and on others
3122 * it will work but it will have an impact on performance. In practice, the
3123 * bitstream should have 4 byte alignment at minimum and all sample data should
3124 * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
3125 * the linesize is not a multiple of 16 then there's no sense in aligning the
3126 * start of the buffer to 16.
3127 *
3128 * @param avctx the codec context
3129 * @param[out] samples the output buffer
3130 * @param[in,out] frame_size_ptr the output buffer size in bytes
3131 * @param[in] avpkt The input AVPacket containing the input buffer.
3132 * @return On error a negative value is returned, otherwise the number of bytes
3133 * used or zero if no frame could be decompressed.
3134 */
3135 int avcodec_decode_audio3(AVCodecContext *avctx, int16_t *samples,
3136 int *frame_size_ptr,
3137 AVPacket *avpkt);
3138
3139 #if LIBAVCODEC_VERSION_MAJOR < 53
3140 /**
3141 * Decodes a video frame from \p buf into \p picture.
3142 * Wrapper function which calls avcodec_decode_video2.
3143 *
3144 * @deprecated Use avcodec_decode_video2 instead.
3145 * @param avctx the codec context
3146 * @param[out] picture The AVFrame in which the decoded video frame will be stored.
3147 * @param[in] buf the input buffer
3148 * @param[in] buf_size the size of the input buffer in bytes
3149 * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
3150 * @return On error a negative value is returned, otherwise the number of bytes
3151 * used or zero if no frame could be decompressed.
3152 */
3153 attribute_deprecated int avcodec_decode_video(AVCodecContext *avctx, AVFrame *picture,
3154 int *got_picture_ptr,
3155 const uint8_t *buf, int buf_size);
3156 #endif
3157
3158 /**
3159 * Decodes a video frame from \p avpkt->data into \p picture.
3160 * The avcodec_decode_video2() function decodes a video frame from the input
3161 * buffer \p avpkt->data of size \p avpkt->size. To decode it, it makes use of the
3162 * video codec which was coupled with \p avctx using avcodec_open(). The
3163 * resulting decoded frame is stored in \p picture.
3164 *
3165 * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
3166 * the actual read bytes because some optimized bitstream readers read 32 or 64
3167 * bits at once and could read over the end.
3168 *
3169 * @warning The end of the input buffer \p buf should be set to 0 to ensure that
3170 * no overreading happens for damaged MPEG streams.
3171 *
3172 * @note You might have to align the input buffer \p avpkt->data and output buffer \p
3173 * samples. The alignment requirements depend on the CPU: on some CPUs it isn't
3174 * necessary at all, on others it won't work at all if not aligned and on others
3175 * it will work but it will have an impact on performance. In practice, the
3176 * bitstream should have 4 byte alignment at minimum and all sample data should
3177 * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
3178 * the linesize is not a multiple of 16 then there's no sense in aligning the
3179 * start of the buffer to 16.
3180 *
3181 * @note Some codecs have a delay between input and output, these need to be
3182 * feeded with avpkt->data=NULL, avpkt->size=0 at the end to return the remaining frames.
3183 *
3184 * @param avctx the codec context
3185 * @param[out] picture The AVFrame in which the decoded video frame will be stored.
3186 * @param[in] avpkt The input AVpacket containing the input buffer.
3187 * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
3188 * @return On error a negative value is returned, otherwise the number of bytes
3189 * used or zero if no frame could be decompressed.
3190 */
3191 int avcodec_decode_video2(AVCodecContext *avctx, AVFrame *picture,
3192 int *got_picture_ptr,
3193 AVPacket *avpkt);
3194
3195 #if LIBAVCODEC_VERSION_MAJOR < 53
3196 /* Decode a subtitle message. Return -1 if error, otherwise return the
3197 * number of bytes used. If no subtitle could be decompressed,
3198 * got_sub_ptr is zero. Otherwise, the subtitle is stored in *sub. */
3199 attribute_deprecated int avcodec_decode_subtitle(AVCodecContext *avctx, AVSubtitle *sub,
3200 int *got_sub_ptr,
3201 const uint8_t *buf, int buf_size);
3202 #endif
3203
3204 /**
3205 * Decodes a subtitle message.
3206 * Returns -1 if error, otherwise returns the number of bytes used.
3207 * If no subtitle could be decompressed, \p got_sub_ptr is zero.
3208 * Otherwise, the subtitle is stored in \p *sub.
3209 *
3210 * @param avctx the codec context
3211 * @param[out] sub The AVSubtitle in which the decoded subtitle will be stored.
3212 * @param[in,out] got_sub_ptr Zero if no subtitle could be decompressed, otherwise, it is nonzero.
3213 * @param[in] avpkt The input AVPacket containing the input buffer.
3214 */
3215 int avcodec_decode_subtitle2(AVCodecContext *avctx, AVSubtitle *sub,
3216 int *got_sub_ptr,
3217 AVPacket *avpkt);
3218 int avcodec_parse_frame(AVCodecContext *avctx, uint8_t **pdata,
3219 int *data_size_ptr,
3220 uint8_t *buf, int buf_size);
3221
3222 /**
3223 * Encodes an audio frame from \p samples into \p buf.
3224 * The avcodec_encode_audio() function encodes an audio frame from the input
3225 * buffer \p samples. To encode it, it makes use of the audio codec which was
3226 * coupled with \p avctx using avcodec_open(). The resulting encoded frame is
3227 * stored in output buffer \p buf.
3228 *
3229 * @note The output buffer should be at least \c FF_MIN_BUFFER_SIZE bytes large.
3230 * However, for PCM audio the user will know how much space is needed
3231 * because it depends on the value passed in \p buf_size as described
3232 * below. In that case a lower value can be used.
3233 *
3234 * @param avctx the codec context
3235 * @param[out] buf the output buffer
3236 * @param[in] buf_size the output buffer size
3237 * @param[in] samples the input buffer containing the samples
3238 * The number of samples read from this buffer is frame_size*channels,
3239 * both of which are defined in \p avctx.
3240 * For PCM audio the number of samples read from \p samples is equal to
3241 * \p buf_size * input_sample_size / output_sample_size.
3242 * @return On error a negative value is returned, on success zero or the number
3243 * of bytes used to encode the data read from the input buffer.
3244 */
3245 int avcodec_encode_audio(AVCodecContext *avctx, uint8_t *buf, int buf_size,
3246 const short *samples);
3247
3248 /**
3249 * Encodes a video frame from \p pict into \p buf.
3250 * The avcodec_encode_video() function encodes a video frame from the input
3251 * \p pict. To encode it, it makes use of the video codec which was coupled with
3252 * \p avctx using avcodec_open(). The resulting encoded bytes representing the
3253 * frame are stored in the output buffer \p buf. The input picture should be
3254 * stored using a specific format, namely \c avctx.pix_fmt.
3255 *
3256 * @param avctx the codec context
3257 * @param[out] buf the output buffer for the bitstream of encoded frame
3258 * @param[in] buf_size the size of the output buffer in bytes
3259 * @param[in] pict the input picture to encode
3260 * @return On error a negative value is returned, on success zero or the number
3261 * of bytes used from the output buffer.
3262 */
3263 int avcodec_encode_video(AVCodecContext *avctx, uint8_t *buf, int buf_size,
3264 const AVFrame *pict);
3265 int avcodec_encode_subtitle(AVCodecContext *avctx, uint8_t *buf, int buf_size,
3266 const AVSubtitle *sub);
3267
3268 int avcodec_close(AVCodecContext *avctx);
3269
3270 /**
3271 * Register all the codecs, parsers and bitstream filters which were enabled at
3272 * configuration time. If you do not call this function you can select exactly
3273 * which formats you want to support, by using the individual registration
3274 * functions.
3275 *
3276 * @see avcodec_register
3277 * @see av_register_codec_parser
3278 * @see av_register_bitstream_filter
3279 */
3280 void avcodec_register_all(void);
3281
3282 /**
3283 * Flush buffers, should be called when seeking or when switching to a different stream.
3284 */
3285 void avcodec_flush_buffers(AVCodecContext *avctx);
3286
3287 void avcodec_default_free_buffers(AVCodecContext *s);
3288
3289 /* misc useful functions */
3290
3291 /**
3292 * Returns a single letter to describe the given picture type \p pict_type.
3293 *
3294 * @param[in] pict_type the picture type
3295 * @return A single character representing the picture type.
3296 */
3297 char av_get_pict_type_char(int pict_type);
3298
3299 /**
3300 * Returns codec bits per sample.
3301 *
3302 * @param[in] codec_id the codec
3303 * @return Number of bits per sample or zero if unknown for the given codec.
3304 */
3305 int av_get_bits_per_sample(enum CodecID codec_id);
3306
3307 /**
3308 * Returns sample format bits per sample.
3309 *
3310 * @param[in] sample_fmt the sample format
3311 * @return Number of bits per sample or zero if unknown for the given sample format.
3312 */
3313 int av_get_bits_per_sample_format(enum SampleFormat sample_fmt);
3314
3315 /* frame parsing */
3316 typedef struct AVCodecParserContext {
3317 void *priv_data;
3318 struct AVCodecParser *parser;
3319 int64_t frame_offset; /* offset of the current frame */
3320 int64_t cur_offset; /* current offset
3321 (incremented by each av_parser_parse()) */
3322 int64_t next_frame_offset; /* offset of the next frame */
3323 /* video info */
3324 int pict_type; /* XXX: Put it back in AVCodecContext. */
3325 /**
3326 * This field is used for proper frame duration computation in lavf.
3327 * It signals, how much longer the frame duration of the current frame
3328 * is compared to normal frame duration.
3329 *
3330 * frame_duration = (1 + repeat_pict) * time_base
3331 *
3332 * It is used by codecs like H.264 to display telecined material.
3333 */
3334 int repeat_pict; /* XXX: Put it back in AVCodecContext. */
3335 int64_t pts; /* pts of the current frame */
3336 int64_t dts; /* dts of the current frame */
3337
3338 /* private data */
3339 int64_t last_pts;
3340 int64_t last_dts;
3341 int fetch_timestamp;
3342
3343 #define AV_PARSER_PTS_NB 4
3344 int cur_frame_start_index;
3345 int64_t cur_frame_offset[AV_PARSER_PTS_NB];
3346 int64_t cur_frame_pts[AV_PARSER_PTS_NB];
3347 int64_t cur_frame_dts[AV_PARSER_PTS_NB];
3348
3349 int flags;
3350 #define PARSER_FLAG_COMPLETE_FRAMES 0x0001
3351
3352 int64_t offset; ///< byte offset from starting packet start
3353 int64_t cur_frame_end[AV_PARSER_PTS_NB];
3354
3355 /*!
3356 * Set by parser to 1 for key frames and 0 for non-key frames.
3357 * It is initialized to -1, so if the parser doesn't set this flag,
3358 * old-style fallback using FF_I_TYPE picture type as key frames
3359 * will be used.
3360 */
3361 int key_frame;
3362
3363 /**
3364 * Time difference in stream time base units from the pts of this
3365 * packet to the point at which the output from the decoder has converged
3366 * independent from the availability of previous frames. That is, the
3367 * frames are virtually identical no matter if decoding started from
3368 * the very first frame or from this keyframe.
3369 * Is AV_NOPTS_VALUE if unknown.
3370 * This field is not the display duration of the current frame.
3371 *
3372 * The purpose of this field is to allow seeking in streams that have no
3373 * keyframes in the conventional sense. It corresponds to the
3374 * recovery point SEI in H.264 and match_time_delta in NUT. It is also
3375 * essential for some types of subtitle streams to ensure that all
3376 * subtitles are correctly displayed after seeking.
3377 */
3378 int64_t convergence_duration;
3379
3380 // Timestamp generation support:
3381 /**
3382 * Synchronization point for start of timestamp generation.
3383 *
3384 * Set to >0 for sync point, 0 for no sync point and <0 for undefined
3385 * (default).
3386 *
3387 * For example, this corresponds to presence of H.264 buffering period
3388 * SEI message.
3389 */
3390 int dts_sync_point;
3391
3392 /**
3393 * Offset of the current timestamp against last timestamp sync point in
3394 * units of AVCodecContext.time_base.
3395 *
3396 * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
3397 * contain a valid timestamp offset.
3398 *
3399 * Note that the timestamp of sync point has usually a nonzero
3400 * dts_ref_dts_delta, which refers to the previous sync point. Offset of
3401 * the next frame after timestamp sync point will be usually 1.
3402 *
3403 * For example, this corresponds to H.264 cpb_removal_delay.
3404 */
3405 int dts_ref_dts_delta;
3406
3407 /**
3408 * Presentation delay of current frame in units of AVCodecContext.time_base.
3409 *
3410 * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
3411 * contain valid non-negative timestamp delta (presentation time of a frame
3412 * must not lie in the past).
3413 *
3414 * This delay represents the difference between decoding and presentation
3415 * time of the frame.
3416 *
3417 * For example, this corresponds to H.264 dpb_output_delay.
3418 */
3419 int pts_dts_delta;
3420
3421 /**
3422 * Position of the packet in file.
3423 *
3424 * Analogous to cur_frame_pts/dts
3425 */
3426 int64_t cur_frame_pos[AV_PARSER_PTS_NB];
3427
3428 /**
3429 * Byte position of currently parsed frame in stream.
3430 */
3431 int64_t pos;
3432
3433 /**
3434 * Previous frame byte position.
3435 */
3436 int64_t last_pos;
3437 } AVCodecParserContext;
3438
3439 typedef struct AVCodecParser {
3440 int codec_ids[5]; /* several codec IDs are permitted */
3441 int priv_data_size;
3442 int (*parser_init)(AVCodecParserContext *s);
3443 int (*parser_parse)(AVCodecParserContext *s,
3444 AVCodecContext *avctx,
3445 const uint8_t **poutbuf, int *poutbuf_size,
3446 const uint8_t *buf, int buf_size);
3447 void (*parser_close)(AVCodecParserContext *s);
3448 int (*split)(AVCodecContext *avctx, const uint8_t *buf, int buf_size);
3449 struct AVCodecParser *next;
3450 } AVCodecParser;
3451
3452 AVCodecParser *av_parser_next(AVCodecParser *c);
3453
3454 void av_register_codec_parser(AVCodecParser *parser);
3455 AVCodecParserContext *av_parser_init(int codec_id);
3456
3457 #if LIBAVCODEC_VERSION_MAJOR < 53
3458 attribute_deprecated
3459 int av_parser_parse(AVCodecParserContext *s,
3460 AVCodecContext *avctx,
3461 uint8_t **poutbuf, int *poutbuf_size,
3462 const uint8_t *buf, int buf_size,
3463 int64_t pts, int64_t dts);
3464 #endif
3465
3466 /**
3467 * Parse a packet.
3468 *
3469 * @param s parser context.
3470 * @param avctx codec context.
3471 * @param poutbuf set to pointer to parsed buffer or NULL if not yet finished.
3472 * @param poutbuf_size set to size of parsed buffer or zero if not yet finished.
3473 * @param buf input buffer.
3474 * @param buf_size input length, to signal EOF, this should be 0 (so that the last frame can be output).
3475 * @param pts input presentation timestamp.
3476 * @param dts input decoding timestamp.
3477 * @param pos input byte position in stream.
3478 * @return the number of bytes of the input bitstream used.
3479 *
3480 * Example:
3481 * @code
3482 * while(in_len){
3483 * len = av_parser_parse2(myparser, AVCodecContext, &data, &size,
3484 * in_data, in_len,
3485 * pts, dts, pos);
3486 * in_data += len;
3487 * in_len -= len;
3488 *
3489 * if(size)
3490 * decode_frame(data, size);
3491 * }
3492 * @endcode
3493 */
3494 int av_parser_parse2(AVCodecParserContext *s,
3495 AVCodecContext *avctx,
3496 uint8_t **poutbuf, int *poutbuf_size,
3497 const uint8_t *buf, int buf_size,
3498 int64_t pts, int64_t dts,
3499 int64_t pos);
3500
3501 int av_parser_change(AVCodecParserContext *s,
3502 AVCodecContext *avctx,
3503 uint8_t **poutbuf, int *poutbuf_size,
3504 const uint8_t *buf, int buf_size, int keyframe);
3505 void av_parser_close(AVCodecParserContext *s);
3506
3507
3508 typedef struct AVBitStreamFilterContext {
3509 void *priv_data;
3510 struct AVBitStreamFilter *filter;
3511 AVCodecParserContext *parser;
3512 struct AVBitStreamFilterContext *next;
3513 } AVBitStreamFilterContext;
3514
3515
3516 typedef struct AVBitStreamFilter {
3517 const char *name;
3518 int priv_data_size;
3519 int (*filter)(AVBitStreamFilterContext *bsfc,
3520 AVCodecContext *avctx, const char *args,
3521 uint8_t **poutbuf, int *poutbuf_size,
3522 const uint8_t *buf, int buf_size, int keyframe);
3523 void (*close)(AVBitStreamFilterContext *bsfc);
3524 struct AVBitStreamFilter *next;
3525 } AVBitStreamFilter;
3526
3527 void av_register_bitstream_filter(AVBitStreamFilter *bsf);
3528 AVBitStreamFilterContext *av_bitstream_filter_init(const char *name);
3529 int av_bitstream_filter_filter(AVBitStreamFilterContext *bsfc,
3530 AVCodecContext *avctx, const char *args,
3531 uint8_t **poutbuf, int *poutbuf_size,
3532 const uint8_t *buf, int buf_size, int keyframe);
3533 void av_bitstream_filter_close(AVBitStreamFilterContext *bsf);
3534
3535 AVBitStreamFilter *av_bitstream_filter_next(AVBitStreamFilter *f);
3536
3537 /* memory */
3538
3539 /**
3540 * Reallocates the given block if it is not large enough, otherwise it
3541 * does nothing.
3542 *
3543 * @see av_realloc
3544 */
3545 void *av_fast_realloc(void *ptr, unsigned int *size, unsigned int min_size);
3546
3547 /**
3548 * Allocates a buffer, reusing the given one if large enough.
3549 *
3550 * Contrary to av_fast_realloc the current buffer contents might not be
3551 * preserved and on error the old buffer is freed, thus no special
3552 * handling to avoid memleaks is necessary.
3553 *
3554 * @param ptr pointer to pointer to already allocated buffer, overwritten with pointer to new buffer
3555 * @param size size of the buffer *ptr points to
3556 * @param min_size minimum size of *ptr buffer after returning, *ptr will be NULL and
3557 * *size 0 if an error occurred.
3558 */
3559 void av_fast_malloc(void *ptr, unsigned int *size, unsigned int min_size);
3560
3561 /**
3562 * Copy image 'src' to 'dst'.
3563 */
3564 void av_picture_copy(AVPicture *dst, const AVPicture *src,
3565 enum PixelFormat pix_fmt, int width, int height);
3566
3567 /**
3568 * Crop image top and left side.
3569 */
3570 int av_picture_crop(AVPicture *dst, const AVPicture *src,
3571 enum PixelFormat pix_fmt, int top_band, int left_band);
3572
3573 /**
3574 * Pad image.
3575 */
3576 int av_picture_pad(AVPicture *dst, const AVPicture *src, int height, int width, enum PixelFormat pix_fmt,
3577 int padtop, int padbottom, int padleft, int padright, int *color);
3578
3579 unsigned int av_xiphlacing(unsigned char *s, unsigned int v);
3580
3581 /**
3582 * Parses \p str and put in \p width_ptr and \p height_ptr the detected values.
3583 *
3584 * @return 0 in case of a successful parsing, a negative value otherwise
3585 * @param[in] str the string to parse: it has to be a string in the format
3586 * <width>x<height> or a valid video frame size abbreviation.
3587 * @param[in,out] width_ptr pointer to the variable which will contain the detected
3588 * frame width value
3589 * @param[in,out] height_ptr pointer to the variable which will contain the detected
3590 * frame height value
3591 */
3592 int av_parse_video_frame_size(int *width_ptr, int *height_ptr, const char *str);
3593
3594 /**
3595 * Parses \p str and put in \p frame_rate the detected values.
3596 *
3597 * @return 0 in case of a successful parsing, a negative value otherwise
3598 * @param[in] str the string to parse: it has to be a string in the format
3599 * <frame_rate_num>/<frame_rate_den>, a float number or a valid video rate abbreviation
3600 * @param[in,out] frame_rate pointer to the AVRational which will contain the detected
3601 * frame rate
3602 */
3603 int av_parse_video_frame_rate(AVRational *frame_rate, const char *str);
3604
3605 /* error handling */
3606 #if EINVAL > 0
3607 #define AVERROR(e) (-(e)) /**< Returns a negative error code from a POSIX error code, to return from library functions. */
3608 #define AVUNERROR(e) (-(e)) /**< Returns a POSIX error code from a library function error return value. */
3609 #else
3610 /* Some platforms have E* and errno already negated. */
3611 #define AVERROR(e) (e)
3612 #define AVUNERROR(e) (e)
3613 #endif
3614 #define AVERROR_UNKNOWN AVERROR(EINVAL) /**< unknown error */
3615 #define AVERROR_IO AVERROR(EIO) /**< I/O error */
3616 #define AVERROR_NUMEXPECTED AVERROR(EDOM) /**< Number syntax expected in filename. */
3617 #define AVERROR_INVALIDDATA AVERROR(EINVAL) /**< invalid data found */
3618 #define AVERROR_NOMEM AVERROR(ENOMEM) /**< not enough memory */
3619 #define AVERROR_NOFMT AVERROR(EILSEQ) /**< unknown format */
3620 #define AVERROR_NOTSUPP AVERROR(ENOSYS) /**< Operation not supported. */
3621 #define AVERROR_NOENT AVERROR(ENOENT) /**< No such file or directory. */
3622 #define AVERROR_EOF AVERROR(EPIPE) /**< End of file. */
3623 #define AVERROR_PATCHWELCOME -MKTAG('P','A','W','E') /**< Not yet implemented in FFmpeg. Patches welcome. */
3624
3625 /**
3626 * Registers the hardware accelerator \p hwaccel.
3627 */
3628 void av_register_hwaccel(AVHWAccel *hwaccel);
3629
3630 /**
3631 * If hwaccel is NULL, returns the first registered hardware accelerator,
3632 * if hwaccel is non-NULL, returns the next registered hardware accelerator
3633 * after hwaccel, or NULL if hwaccel is the last one.
3634 */
3635 AVHWAccel *av_hwaccel_next(AVHWAccel *hwaccel);
3636
3637 #endif /* AVCODEC_AVCODEC_H */