2 * copyright (c) 2001 Fabrice Bellard
4 * This file is part of FFmpeg.
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21 #ifndef AVCODEC_AVCODEC_H
22 #define AVCODEC_AVCODEC_H
25 * @file libavcodec/avcodec.h
30 #include "libavutil/avutil.h"
32 #define LIBAVCODEC_VERSION_MAJOR 52
33 #define LIBAVCODEC_VERSION_MINOR 26
34 #define LIBAVCODEC_VERSION_MICRO 0
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
44 #define LIBAVCODEC_IDENT "Lavc" AV_STRINGIFY(LIBAVCODEC_VERSION)
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}
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
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.
67 CODEC_ID_MPEG2VIDEO, ///< preferred ID for MPEG-1/2 video decoding
68 CODEC_ID_MPEG2VIDEO_XVMC,
105 CODEC_ID_INTERPLAY_VIDEO,
117 CODEC_ID_TRUEMOTION1,
145 CODEC_ID_TRUEMOTION2,
162 CODEC_ID_DSICINVIDEO,
163 CODEC_ID_TIERTEXSEQVIDEO,
172 CODEC_ID_BETHSOFTVID,
190 CODEC_ID_MOTIONPIXELS,
197 /* various PCM "codecs" */
198 CODEC_ID_PCM_S16LE= 0x10000,
214 CODEC_ID_PCM_S24DAUD,
216 CODEC_ID_PCM_S16LE_PLANAR,
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,
238 CODEC_ID_ADPCM_YAMAHA,
239 CODEC_ID_ADPCM_SBPRO_4,
240 CODEC_ID_ADPCM_SBPRO_3,
241 CODEC_ID_ADPCM_SBPRO_2,
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,
254 CODEC_ID_AMR_NB= 0x12000,
257 /* RealAudio codecs*/
258 CODEC_ID_RA_144= 0x13000,
261 /* various DPCM codecs */
262 CODEC_ID_ROQ_DPCM= 0x14000,
263 CODEC_ID_INTERPLAY_DPCM,
268 CODEC_ID_MP2= 0x15000,
269 CODEC_ID_MP3, ///< preferred ID for decoding MPEG audio layer 1, 2 or 3
287 CODEC_ID_WESTWOOD_SND1,
288 CODEC_ID_GSM, ///< as in Berlin toast format
296 CODEC_ID_DSICINAUDIO,
300 CODEC_ID_GSM_MS, /* as found in WAV */
309 CODEC_ID_WMALOSSLESS,
318 /* subtitle codecs */
319 CODEC_ID_DVD_SUBTITLE= 0x17000,
320 CODEC_ID_DVB_SUBTITLE,
321 CODEC_ID_TEXT, ///< raw UTF-8 text
326 /* other specific kind of codecs (generally used for attachments) */
327 CODEC_ID_TTF= 0x18000,
329 CODEC_ID_PROBE= 0x19000, ///< codec_id is not known (like CODEC_ID_NONE) but lavf should attempt to identify it
331 CODEC_ID_MPEG2TS= 0x20000, /**< _FAKE_ codec to indicate a raw MPEG-2 TS
332 * stream (only used by libavformat) */
336 CODEC_TYPE_UNKNOWN = -1,
341 CODEC_TYPE_ATTACHMENT,
346 * all in native-endian format
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
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.
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_SURROUND (CH_LAYOUT_STEREO|CH_FRONT_CENTER)
384 #define CH_LAYOUT_QUAD (CH_LAYOUT_STEREO|CH_BACK_LEFT|CH_BACK_RIGHT)
385 #define CH_LAYOUT_5POINT0 (CH_LAYOUT_SURROUND|CH_SIDE_LEFT|CH_SIDE_RIGHT)
386 #define CH_LAYOUT_5POINT1 (CH_LAYOUT_5POINT0|CH_LOW_FREQUENCY)
387 #define CH_LAYOUT_5POINT0_BACK (CH_LAYOUT_SURROUND|CH_BACK_LEFT|CH_BACK_RIGHT)
388 #define CH_LAYOUT_5POINT1_BACK (CH_LAYOUT_5POINT0_BACK|CH_LOW_FREQUENCY)
389 #define CH_LAYOUT_7POINT1 (CH_LAYOUT_5POINT1|CH_BACK_LEFT|CH_BACK_RIGHT)
390 #define CH_LAYOUT_7POINT1_WIDE (CH_LAYOUT_5POINT1_BACK|\
391 CH_FRONT_LEFT_OF_CENTER|CH_FRONT_RIGHT_OF_CENTER)
392 #define CH_LAYOUT_STEREO_DOWNMIX (CH_STEREO_LEFT|CH_STEREO_RIGHT)
395 #define AVCODEC_MAX_AUDIO_FRAME_SIZE 192000 // 1 second of 48khz 32bit audio
398 * Required number of additionally allocated bytes at the end of the input bitstream for decoding.
399 * This is mainly needed because some optimized bitstream readers read
400 * 32 or 64 bit at once and could read over the end.<br>
401 * Note: If the first 23 bits of the additional bytes are not 0, then damaged
402 * MPEG bitstreams could cause overread and segfault.
404 #define FF_INPUT_BUFFER_PADDING_SIZE 8
407 * minimum encoding buffer size
408 * Used to avoid some checks during header writing.
410 #define FF_MIN_BUFFER_SIZE 16384
414 * motion estimation type.
417 ME_ZERO = 1, ///< no search, that is use 0,0 vector whenever one is needed
421 ME_EPZS, ///< enhanced predictive zonal search
422 ME_X1, ///< reserved for experiments
423 ME_HEX, ///< hexagon based search
424 ME_UMH, ///< uneven multi-hexagon search
425 ME_ITER, ///< iterative search
426 ME_TESA, ///< transformed exhaustive search algorithm
430 /* We leave some space between them for extensions (drop some
431 * keyframes for intra-only or drop just some bidir frames). */
432 AVDISCARD_NONE =-16, ///< discard nothing
433 AVDISCARD_DEFAULT= 0, ///< discard useless packets like 0 size packets in avi
434 AVDISCARD_NONREF = 8, ///< discard all non reference
435 AVDISCARD_BIDIR = 16, ///< discard all bidirectional frames
436 AVDISCARD_NONKEY = 32, ///< discard all frames except keyframes
437 AVDISCARD_ALL = 48, ///< discard all
440 typedef struct RcOverride{
443 int qscale; // If this is 0 then quality_factor will be used instead.
444 float quality_factor;
447 #define FF_MAX_B_FRAMES 16
450 These flags can be passed in AVCodecContext.flags before initialization.
451 Note: Not everything is supported yet.
454 #define CODEC_FLAG_QSCALE 0x0002 ///< Use fixed qscale.
455 #define CODEC_FLAG_4MV 0x0004 ///< 4 MV per MB allowed / advanced prediction for H.263.
456 #define CODEC_FLAG_QPEL 0x0010 ///< Use qpel MC.
457 #define CODEC_FLAG_GMC 0x0020 ///< Use GMC.
458 #define CODEC_FLAG_MV0 0x0040 ///< Always try a MB with MV=<0,0>.
459 #define CODEC_FLAG_PART 0x0080 ///< Use data partitioning.
461 * The parent program guarantees that the input for B-frames containing
462 * streams is not written to for at least s->max_b_frames+1 frames, if
463 * this is not set the input will be copied.
465 #define CODEC_FLAG_INPUT_PRESERVED 0x0100
466 #define CODEC_FLAG_PASS1 0x0200 ///< Use internal 2pass ratecontrol in first pass mode.
467 #define CODEC_FLAG_PASS2 0x0400 ///< Use internal 2pass ratecontrol in second pass mode.
468 #define CODEC_FLAG_EXTERN_HUFF 0x1000 ///< Use external Huffman table (for MJPEG).
469 #define CODEC_FLAG_GRAY 0x2000 ///< Only decode/encode grayscale.
470 #define CODEC_FLAG_EMU_EDGE 0x4000 ///< Don't draw edges.
471 #define CODEC_FLAG_PSNR 0x8000 ///< error[?] variables will be set during encoding.
472 #define CODEC_FLAG_TRUNCATED 0x00010000 /** Input bitstream might be truncated at a random
473 location instead of only at frame boundaries. */
474 #define CODEC_FLAG_NORMALIZE_AQP 0x00020000 ///< Normalize adaptive quantization.
475 #define CODEC_FLAG_INTERLACED_DCT 0x00040000 ///< Use interlaced DCT.
476 #define CODEC_FLAG_LOW_DELAY 0x00080000 ///< Force low delay.
477 #define CODEC_FLAG_ALT_SCAN 0x00100000 ///< Use alternate scan.
478 #define CODEC_FLAG_GLOBAL_HEADER 0x00400000 ///< Place global headers in extradata instead of every keyframe.
479 #define CODEC_FLAG_BITEXACT 0x00800000 ///< Use only bitexact stuff (except (I)DCT).
480 /* Fx : Flag for h263+ extra options */
481 #define CODEC_FLAG_AC_PRED 0x01000000 ///< H.263 advanced intra coding / MPEG-4 AC prediction
482 #define CODEC_FLAG_H263P_UMV 0x02000000 ///< unlimited motion vector
483 #define CODEC_FLAG_CBP_RD 0x04000000 ///< Use rate distortion optimization for cbp.
484 #define CODEC_FLAG_QP_RD 0x08000000 ///< Use rate distortion optimization for qp selectioon.
485 #define CODEC_FLAG_H263P_AIV 0x00000008 ///< H.263 alternative inter VLC
486 #define CODEC_FLAG_OBMC 0x00000001 ///< OBMC
487 #define CODEC_FLAG_LOOP_FILTER 0x00000800 ///< loop filter
488 #define CODEC_FLAG_H263P_SLICE_STRUCT 0x10000000
489 #define CODEC_FLAG_INTERLACED_ME 0x20000000 ///< interlaced motion estimation
490 #define CODEC_FLAG_SVCD_SCAN_OFFSET 0x40000000 ///< Will reserve space for SVCD scan offset user data.
491 #define CODEC_FLAG_CLOSED_GOP 0x80000000
492 #define CODEC_FLAG2_FAST 0x00000001 ///< Allow non spec compliant speedup tricks.
493 #define CODEC_FLAG2_STRICT_GOP 0x00000002 ///< Strictly enforce GOP size.
494 #define CODEC_FLAG2_NO_OUTPUT 0x00000004 ///< Skip bitstream encoding.
495 #define CODEC_FLAG2_LOCAL_HEADER 0x00000008 ///< Place global headers at every keyframe instead of in extradata.
496 #define CODEC_FLAG2_BPYRAMID 0x00000010 ///< H.264 allow B-frames to be used as references.
497 #define CODEC_FLAG2_WPRED 0x00000020 ///< H.264 weighted biprediction for B-frames
498 #define CODEC_FLAG2_MIXED_REFS 0x00000040 ///< H.264 one reference per partition, as opposed to one reference per macroblock
499 #define CODEC_FLAG2_8X8DCT 0x00000080 ///< H.264 high profile 8x8 transform
500 #define CODEC_FLAG2_FASTPSKIP 0x00000100 ///< H.264 fast pskip
501 #define CODEC_FLAG2_AUD 0x00000200 ///< H.264 access unit delimiters
502 #define CODEC_FLAG2_BRDO 0x00000400 ///< B-frame rate-distortion optimization
503 #define CODEC_FLAG2_INTRA_VLC 0x00000800 ///< Use MPEG-2 intra VLC table.
504 #define CODEC_FLAG2_MEMC_ONLY 0x00001000 ///< Only do ME/MC (I frames -> ref, P frame -> ME+MC).
505 #define CODEC_FLAG2_DROP_FRAME_TIMECODE 0x00002000 ///< timecode is in drop frame format.
506 #define CODEC_FLAG2_SKIP_RD 0x00004000 ///< RD optimal MB level residual skipping
507 #define CODEC_FLAG2_CHUNKS 0x00008000 ///< Input bitstream might be truncated at a packet boundaries instead of only at frame boundaries.
508 #define CODEC_FLAG2_NON_LINEAR_QUANT 0x00010000 ///< Use MPEG-2 nonlinear quantizer.
509 #define CODEC_FLAG2_BIT_RESERVOIR 0x00020000 ///< Use a bit reservoir when encoding if possible
511 /* Unsupported options :
512 * Syntax Arithmetic coding (SAC)
513 * Reference Picture Selection
514 * Independent Segment Decoding */
516 /* codec capabilities */
518 #define CODEC_CAP_DRAW_HORIZ_BAND 0x0001 ///< Decoder can use draw_horiz_band callback.
520 * Codec uses get_buffer() for allocating buffers.
521 * direct rendering method 1
523 #define CODEC_CAP_DR1 0x0002
524 /* If 'parse_only' field is true, then avcodec_parse_frame() can be used. */
525 #define CODEC_CAP_PARSE_ONLY 0x0004
526 #define CODEC_CAP_TRUNCATED 0x0008
527 /* Codec can export data for HW decoding (XvMC). */
528 #define CODEC_CAP_HWACCEL 0x0010
530 * Codec has a nonzero delay and needs to be fed with NULL at the end to get the delayed data.
531 * If this is not set, the codec is guaranteed to never be fed with NULL data.
533 #define CODEC_CAP_DELAY 0x0020
535 * Codec can be fed a final frame with a smaller size.
536 * This can be used to prevent truncation of the last audio samples.
538 #define CODEC_CAP_SMALL_LAST_FRAME 0x0040
540 * Codec can export data for HW decoding (VDPAU).
542 #define CODEC_CAP_HWACCEL_VDPAU 0x0080
544 //The following defines may change, don't expect compatibility if you use them.
545 #define MB_TYPE_INTRA4x4 0x0001
546 #define MB_TYPE_INTRA16x16 0x0002 //FIXME H.264-specific
547 #define MB_TYPE_INTRA_PCM 0x0004 //FIXME H.264-specific
548 #define MB_TYPE_16x16 0x0008
549 #define MB_TYPE_16x8 0x0010
550 #define MB_TYPE_8x16 0x0020
551 #define MB_TYPE_8x8 0x0040
552 #define MB_TYPE_INTERLACED 0x0080
553 #define MB_TYPE_DIRECT2 0x0100 //FIXME
554 #define MB_TYPE_ACPRED 0x0200
555 #define MB_TYPE_GMC 0x0400
556 #define MB_TYPE_SKIP 0x0800
557 #define MB_TYPE_P0L0 0x1000
558 #define MB_TYPE_P1L0 0x2000
559 #define MB_TYPE_P0L1 0x4000
560 #define MB_TYPE_P1L1 0x8000
561 #define MB_TYPE_L0 (MB_TYPE_P0L0 | MB_TYPE_P1L0)
562 #define MB_TYPE_L1 (MB_TYPE_P0L1 | MB_TYPE_P1L1)
563 #define MB_TYPE_L0L1 (MB_TYPE_L0 | MB_TYPE_L1)
564 #define MB_TYPE_QUANT 0x00010000
565 #define MB_TYPE_CBP 0x00020000
566 //Note bits 24-31 are reserved for codec specific use (h264 ref0, mpeg1 0mv, ...)
570 * This specifies the area which should be displayed.
571 * Note there may be multiple such areas for one frame.
573 typedef struct AVPanScan{
576 * - encoding: Set by user.
577 * - decoding: Set by libavcodec.
582 * width and height in 1/16 pel
583 * - encoding: Set by user.
584 * - decoding: Set by libavcodec.
590 * position of the top left corner in 1/16 pel for up to 3 fields/frames
591 * - encoding: Set by user.
592 * - decoding: Set by libavcodec.
594 int16_t position[3][2];
597 #define FF_COMMON_FRAME \
599 * pointer to the picture planes.\
600 * This might be different from the first allocated byte\
607 * pointer to the first allocated byte of the picture. Can be used in get_buffer/release_buffer.\
608 * This isn't used by libavcodec unless the default get/release_buffer() is used.\
614 * 1 -> keyframe, 0-> not\
615 * - encoding: Set by libavcodec.\
616 * - decoding: Set by libavcodec.\
621 * Picture type of the frame, see ?_TYPE below.\
622 * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
623 * - decoding: Set by libavcodec.\
628 * presentation timestamp in time_base units (time when frame should be shown to user)\
629 * If AV_NOPTS_VALUE then frame_rate = 1/time_base will be assumed.\
630 * - encoding: MUST be set by user.\
631 * - decoding: Set by libavcodec.\
636 * picture number in bitstream order\
637 * - encoding: set by\
638 * - decoding: Set by libavcodec.\
640 int coded_picture_number;\
642 * picture number in display order\
643 * - encoding: set by\
644 * - decoding: Set by libavcodec.\
646 int display_picture_number;\
649 * quality (between 1 (good) and FF_LAMBDA_MAX (bad)) \
650 * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
651 * - decoding: Set by libavcodec.\
656 * buffer age (1->was last buffer and dint change, 2->..., ...).\
657 * Set to INT_MAX if the buffer has not been used yet.\
658 * - encoding: unused\
659 * - decoding: MUST be set by get_buffer().\
664 * is this picture used as reference\
665 * The values for this are the same as the MpegEncContext.picture_structure\
666 * variable, that is 1->top field, 2->bottom field, 3->frame/both fields.\
667 * Set to 4 for delayed, non-reference frames.\
668 * - encoding: unused\
669 * - decoding: Set by libavcodec. (before get_buffer() call)).\
675 * - encoding: unused\
676 * - decoding: Set by libavcodec.\
678 int8_t *qscale_table;\
681 * - encoding: unused\
682 * - decoding: Set by libavcodec.\
687 * mbskip_table[mb]>=1 if MB didn't change\
688 * stride= mb_width = (width+15)>>4\
689 * - encoding: unused\
690 * - decoding: Set by libavcodec.\
692 uint8_t *mbskip_table;\
695 * motion vector table\
698 * int mv_sample_log2= 4 - motion_subsample_log2;\
699 * int mb_width= (width+15)>>4;\
700 * int mv_stride= (mb_width << mv_sample_log2) + 1;\
701 * motion_val[direction][x + y*mv_stride][0->mv_x, 1->mv_y];\
703 * - encoding: Set by user.\
704 * - decoding: Set by libavcodec.\
706 int16_t (*motion_val[2])[2];\
709 * macroblock type table\
710 * mb_type_base + mb_width + 2\
711 * - encoding: Set by user.\
712 * - decoding: Set by libavcodec.\
717 * log2 of the size of the block which a single vector in motion_val represents: \
718 * (4->16x16, 3->8x8, 2-> 4x4, 1-> 2x2)\
719 * - encoding: unused\
720 * - decoding: Set by libavcodec.\
722 uint8_t motion_subsample_log2;\
725 * for some private data of the user\
726 * - encoding: unused\
727 * - decoding: Set by user.\
733 * - encoding: Set by libavcodec. if flags&CODEC_FLAG_PSNR.\
734 * - decoding: unused\
739 * type of the buffer (to keep track of who has to deallocate data[*])\
740 * - encoding: Set by the one who allocates it.\
741 * - decoding: Set by the one who allocates it.\
742 * Note: User allocated (direct rendering) & internal buffers cannot coexist currently.\
747 * When decoding, this signals how much the picture must be delayed.\
748 * extra_delay = repeat_pict / (2*fps)\
749 * - encoding: unused\
750 * - decoding: Set by libavcodec.\
760 * The content of the picture is interlaced.\
761 * - encoding: Set by user.\
762 * - decoding: Set by libavcodec. (default 0)\
764 int interlaced_frame;\
767 * If the content is interlaced, is top field displayed first.\
768 * - encoding: Set by user.\
769 * - decoding: Set by libavcodec.\
771 int top_field_first;\
775 * - encoding: Set by user.\
776 * - decoding: Set by libavcodec.\
778 AVPanScan *pan_scan;\
781 * Tell user application that palette has changed from previous frame.\
782 * - encoding: ??? (no palette-enabled encoder yet)\
783 * - decoding: Set by libavcodec. (default 0).\
785 int palette_has_changed;\
788 * codec suggestion on buffer type if != 0\
789 * - encoding: unused\
790 * - decoding: Set by libavcodec. (before get_buffer() call)).\
796 * - encoding: unused\
797 * - decoding: Set by libavcodec.\
802 * motion referece frame index\
803 * - encoding: Set by user.\
804 * - decoding: Set by libavcodec.\
806 int8_t *ref_index[2];\
809 * reordered opaque 64bit number (generally a PTS) from AVCodecContext.reordered_opaque\
810 * output in AVFrame.reordered_opaque\
811 * - encoding: unused\
812 * - decoding: Read by user.\
814 int64_t reordered_opaque;\
817 * hardware accelerator private data (FFmpeg allocated)\
818 * - encoding: unused\
819 * - decoding: Set by libavcodec\
821 void *hwaccel_picture_private;\
824 #define FF_QSCALE_TYPE_MPEG1 0
825 #define FF_QSCALE_TYPE_MPEG2 1
826 #define FF_QSCALE_TYPE_H264 2
828 #define FF_BUFFER_TYPE_INTERNAL 1
829 #define FF_BUFFER_TYPE_USER 2 ///< direct rendering buffers (image is (de)allocated by user)
830 #define FF_BUFFER_TYPE_SHARED 4 ///< Buffer from somewhere else; don't deallocate image (data/base), all other tables are not shared.
831 #define FF_BUFFER_TYPE_COPY 8 ///< Just a (modified) copy of some other buffer, don't deallocate anything.
834 #define FF_I_TYPE 1 ///< Intra
835 #define FF_P_TYPE 2 ///< Predicted
836 #define FF_B_TYPE 3 ///< Bi-dir predicted
837 #define FF_S_TYPE 4 ///< S(GMC)-VOP MPEG4
838 #define FF_SI_TYPE 5 ///< Switching Intra
839 #define FF_SP_TYPE 6 ///< Switching Predicted
842 #define FF_BUFFER_HINTS_VALID 0x01 // Buffer hints value is meaningful (if 0 ignore).
843 #define FF_BUFFER_HINTS_READABLE 0x02 // Codec will read from buffer.
844 #define FF_BUFFER_HINTS_PRESERVE 0x04 // User must not alter buffer content.
845 #define FF_BUFFER_HINTS_REUSABLE 0x08 // Codec will reuse the buffer (update).
847 typedef struct AVPacket {
849 * Presentation timestamp in AVStream->time_base units; the time at which
850 * the decompressed packet will be presented to the user.
851 * Can be AV_NOPTS_VALUE if it is not stored in the file.
852 * pts MUST be larger or equal to dts as presentation cannot happen before
853 * decompression, unless one wants to view hex dumps. Some formats misuse
854 * the terms dts and pts/cts to mean something different. Such timestamps
855 * must be converted to true pts/dts before they are stored in AVPacket.
859 * Decompression timestamp in AVStream->time_base units; the time at which
860 * the packet is decompressed.
861 * Can be AV_NOPTS_VALUE if it is not stored in the file.
869 * Duration of this packet in AVStream->time_base units, 0 if unknown.
870 * Equals next_pts - this_pts in presentation order.
873 void (*destruct)(struct AVPacket *);
875 int64_t pos; ///< byte position in stream, -1 if unknown
878 * Time difference in AVStream->time_base units from the pts of this
879 * packet to the point at which the output from the decoder has converged
880 * independent from the availability of previous frames. That is, the
881 * frames are virtually identical no matter if decoding started from
882 * the very first frame or from this keyframe.
883 * Is AV_NOPTS_VALUE if unknown.
884 * This field is not the display duration of the current packet.
886 * The purpose of this field is to allow seeking in streams that have no
887 * keyframes in the conventional sense. It corresponds to the
888 * recovery point SEI in H.264 and match_time_delta in NUT. It is also
889 * essential for some types of subtitle streams to ensure that all
890 * subtitles are correctly displayed after seeking.
892 int64_t convergence_duration;
894 #define PKT_FLAG_KEY 0x0001
898 * New fields can be added to the end of FF_COMMON_FRAME with minor version
900 * Removal, reordering and changes to existing fields require a major
901 * version bump. No fields should be added into AVFrame before or after
903 * sizeof(AVFrame) must not be used outside libav*.
905 typedef struct AVFrame {
910 * main external API structure.
911 * New fields can be added to the end with minor version bumps.
912 * Removal, reordering and changes to existing fields require a major
914 * sizeof(AVCodecContext) must not be used outside libav*.
916 typedef struct AVCodecContext {
918 * information on struct for av_log
919 * - set by avcodec_alloc_context
921 const AVClass *av_class;
923 * the average bitrate
924 * - encoding: Set by user; unused for constant quantizer encoding.
925 * - decoding: Set by libavcodec. 0 or some bitrate if this info is available in the stream.
930 * number of bits the bitstream is allowed to diverge from the reference.
931 * the reference can be CBR (for CBR pass1) or VBR (for pass2)
932 * - encoding: Set by user; unused for constant quantizer encoding.
935 int bit_rate_tolerance;
939 * - encoding: Set by user.
940 * - decoding: Set by user.
945 * Some codecs need additional format info. It is stored here.
946 * If any muxer uses this then ALL demuxers/parsers AND encoders for the
947 * specific codec MUST set it correctly otherwise stream copy breaks.
948 * In general use of this field by muxers is not recommanded.
949 * - encoding: Set by libavcodec.
950 * - decoding: Set by libavcodec. (FIXME: Is this OK?)
955 * Motion estimation algorithm used for video coding.
956 * 1 (zero), 2 (full), 3 (log), 4 (phods), 5 (epzs), 6 (x1), 7 (hex),
957 * 8 (umh), 9 (iter), 10 (tesa) [7, 8, 10 are x264 specific, 9 is snow specific]
958 * - encoding: MUST be set by user.
964 * some codecs need / can use extradata like Huffman tables.
965 * mjpeg: Huffman tables
966 * rv10: additional flags
967 * mpeg4: global headers (they can be in the bitstream or here)
968 * The allocated memory should be FF_INPUT_BUFFER_PADDING_SIZE bytes larger
969 * than extradata_size to avoid prolems if it is read with the bitstream reader.
970 * The bytewise contents of extradata must not depend on the architecture or CPU endianness.
971 * - encoding: Set/allocated/freed by libavcodec.
972 * - decoding: Set/allocated/freed by user.
978 * This is the fundamental unit of time (in seconds) in terms
979 * of which frame timestamps are represented. For fixed-fps content,
980 * timebase should be 1/framerate and timestamp increments should be
982 * - encoding: MUST be set by user.
983 * - decoding: Set by libavcodec.
985 AVRational time_base;
989 * picture width / height.
990 * - encoding: MUST be set by user.
991 * - decoding: Set by libavcodec.
992 * Note: For compatibility it is possible to set this instead of
993 * coded_width/height before decoding.
997 #define FF_ASPECT_EXTENDED 15
1000 * the number of pictures in a group of pictures, or 0 for intra_only
1001 * - encoding: Set by user.
1002 * - decoding: unused
1007 * Pixel format, see PIX_FMT_xxx.
1008 * - encoding: Set by user.
1009 * - decoding: Set by libavcodec.
1011 enum PixelFormat pix_fmt;
1014 * Frame rate emulation. If not zero, the lower layer (i.e. format handler)
1015 * has to read frames at native frame rate.
1016 * - encoding: Set by user.
1017 * - decoding: unused
1022 * If non NULL, 'draw_horiz_band' is called by the libavcodec
1023 * decoder to draw a horizontal band. It improves cache usage. Not
1024 * all codecs can do that. You must check the codec capabilities
1026 * The function is also used by hardware acceleration APIs.
1027 * It is called at least once during frame decoding to pass
1028 * the data needed for hardware render.
1029 * In that mode instead of pixel data, AVFrame points to
1030 * a structure specific to the acceleration API. The application
1031 * reads the structure and can change some fields to indicate progress
1033 * - encoding: unused
1034 * - decoding: Set by user.
1035 * @param height the height of the slice
1036 * @param y the y position of the slice
1037 * @param type 1->top field, 2->bottom field, 3->frame
1038 * @param offset offset into the AVFrame.data from which the slice should be read
1040 void (*draw_horiz_band)(struct AVCodecContext *s,
1041 const AVFrame *src, int offset[4],
1042 int y, int type, int height);
1045 int sample_rate; ///< samples per second
1046 int channels; ///< number of audio channels
1049 * audio sample format
1050 * - encoding: Set by user.
1051 * - decoding: Set by libavcodec.
1053 enum SampleFormat sample_fmt; ///< sample format, currently unused
1055 /* The following data should not be initialized. */
1057 * Samples per packet, initialized when calling 'init'.
1060 int frame_number; ///< audio or video frame number
1061 int real_pict_num; ///< Returns the real picture number of previous encoded frame.
1064 * Number of frames the decoded output will be delayed relative to
1065 * the encoded input.
1066 * - encoding: Set by libavcodec.
1067 * - decoding: unused
1071 /* - encoding parameters */
1072 float qcompress; ///< amount of qscale change between easy & hard scenes (0.0-1.0)
1073 float qblur; ///< amount of qscale smoothing over time (0.0-1.0)
1077 * - encoding: Set by user.
1078 * - decoding: unused
1084 * - encoding: Set by user.
1085 * - decoding: unused
1090 * maximum quantizer difference between frames
1091 * - encoding: Set by user.
1092 * - decoding: unused
1097 * maximum number of B-frames between non-B-frames
1098 * Note: The output will be delayed by max_b_frames+1 relative to the input.
1099 * - encoding: Set by user.
1100 * - decoding: unused
1105 * qscale factor between IP and B-frames
1106 * If > 0 then the last P-frame quantizer will be used (q= lastp_q*factor+offset).
1107 * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
1108 * - encoding: Set by user.
1109 * - decoding: unused
1111 float b_quant_factor;
1113 /** obsolete FIXME remove */
1115 #define FF_RC_STRATEGY_XVID 1
1117 int b_frame_strategy;
1121 * - encoding: unused
1122 * - decoding: Set by user. 1-> Skip B-frames, 2-> Skip IDCT/dequant too, 5-> Skip everything except header
1123 * @deprecated Deprecated in favor of skip_idct and skip_frame.
1127 struct AVCodec *codec;
1131 int rtp_payload_size; /* The size of the RTP payload: the coder will */
1132 /* do its best to deliver a chunk with size */
1133 /* below rtp_payload_size, the chunk will start */
1134 /* with a start code on some codecs like H.263. */
1135 /* This doesn't take account of any particular */
1136 /* headers inside the transmitted RTP payload. */
1139 /* The RTP callback: This function is called */
1140 /* every time the encoder has a packet to send. */
1141 /* It depends on the encoder if the data starts */
1142 /* with a Start Code (it should). H.263 does. */
1143 /* mb_nb contains the number of macroblocks */
1144 /* encoded in the RTP payload. */
1145 void (*rtp_callback)(struct AVCodecContext *avctx, void *data, int size, int mb_nb);
1147 /* statistics, used for 2-pass encoding */
1158 * number of bits used for the previously encoded frame
1159 * - encoding: Set by libavcodec.
1160 * - decoding: unused
1165 * Private data of the user, can be used to carry app specific stuff.
1166 * - encoding: Set by user.
1167 * - decoding: Set by user.
1171 char codec_name[32];
1172 enum CodecType codec_type; /* see CODEC_TYPE_xxx */
1173 enum CodecID codec_id; /* see CODEC_ID_xxx */
1176 * fourcc (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
1177 * This is used to work around some encoder bugs.
1178 * A demuxer should set this to what is stored in the field used to identify the codec.
1179 * If there are multiple such fields in a container then the demuxer should choose the one
1180 * which maximizes the information about the used codec.
1181 * If the codec tag field in a container is larger then 32 bits then the demuxer should
1182 * remap the longer ID to 32 bits with a table or other structure. Alternatively a new
1183 * extra_codec_tag + size could be added but for this a clear advantage must be demonstrated
1185 * - encoding: Set by user, if not then the default based on codec_id will be used.
1186 * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
1188 unsigned int codec_tag;
1191 * Work around bugs in encoders which sometimes cannot be detected automatically.
1192 * - encoding: Set by user
1193 * - decoding: Set by user
1195 int workaround_bugs;
1196 #define FF_BUG_AUTODETECT 1 ///< autodetection
1197 #define FF_BUG_OLD_MSMPEG4 2
1198 #define FF_BUG_XVID_ILACE 4
1199 #define FF_BUG_UMP4 8
1200 #define FF_BUG_NO_PADDING 16
1201 #define FF_BUG_AMV 32
1202 #define FF_BUG_AC_VLC 0 ///< Will be removed, libavcodec can now handle these non-compliant files by default.
1203 #define FF_BUG_QPEL_CHROMA 64
1204 #define FF_BUG_STD_QPEL 128
1205 #define FF_BUG_QPEL_CHROMA2 256
1206 #define FF_BUG_DIRECT_BLOCKSIZE 512
1207 #define FF_BUG_EDGE 1024
1208 #define FF_BUG_HPEL_CHROMA 2048
1209 #define FF_BUG_DC_CLIP 4096
1210 #define FF_BUG_MS 8192 ///< Work around various bugs in Microsoft's broken decoders.
1211 //#define FF_BUG_FAKE_SCALABILITY 16 //Autodetection should work 100%.
1214 * luma single coefficient elimination threshold
1215 * - encoding: Set by user.
1216 * - decoding: unused
1218 int luma_elim_threshold;
1221 * chroma single coeff elimination threshold
1222 * - encoding: Set by user.
1223 * - decoding: unused
1225 int chroma_elim_threshold;
1228 * strictly follow the standard (MPEG4, ...).
1229 * - encoding: Set by user.
1230 * - decoding: Set by user.
1231 * Setting this to STRICT or higher means the encoder and decoder will
1232 * generally do stupid things. While setting it to inofficial or lower
1233 * will mean the encoder might use things that are not supported by all
1234 * spec compliant decoders. Decoders make no difference between normal,
1235 * inofficial and experimental, that is they always try to decode things
1236 * when they can unless they are explicitly asked to behave stupid
1237 * (=strictly conform to the specs)
1239 int strict_std_compliance;
1240 #define FF_COMPLIANCE_VERY_STRICT 2 ///< Strictly conform to a older more strict version of the spec or reference software.
1241 #define FF_COMPLIANCE_STRICT 1 ///< Strictly conform to all the things in the spec no matter what consequences.
1242 #define FF_COMPLIANCE_NORMAL 0
1243 #define FF_COMPLIANCE_INOFFICIAL -1 ///< Allow inofficial extensions.
1244 #define FF_COMPLIANCE_EXPERIMENTAL -2 ///< Allow nonstandardized experimental things.
1247 * qscale offset between IP and B-frames
1248 * - encoding: Set by user.
1249 * - decoding: unused
1251 float b_quant_offset;
1254 * Error recognization; higher values will detect more errors but may
1255 * misdetect some more or less valid parts as errors.
1256 * - encoding: unused
1257 * - decoding: Set by user.
1259 int error_recognition;
1260 #define FF_ER_CAREFUL 1
1261 #define FF_ER_COMPLIANT 2
1262 #define FF_ER_AGGRESSIVE 3
1263 #define FF_ER_VERY_AGGRESSIVE 4
1266 * Called at the beginning of each frame to get a buffer for it.
1267 * If pic.reference is set then the frame will be read later by libavcodec.
1268 * avcodec_align_dimensions() should be used to find the required width and
1269 * height, as they normally need to be rounded up to the next multiple of 16.
1270 * - encoding: unused
1271 * - decoding: Set by libavcodec., user can override.
1273 int (*get_buffer)(struct AVCodecContext *c, AVFrame *pic);
1276 * Called to release buffers which were allocated with get_buffer.
1277 * A released buffer can be reused in get_buffer().
1278 * pic.data[*] must be set to NULL.
1279 * - encoding: unused
1280 * - decoding: Set by libavcodec., user can override.
1282 void (*release_buffer)(struct AVCodecContext *c, AVFrame *pic);
1285 * Size of the frame reordering buffer in the decoder.
1286 * For MPEG-2 it is 1 IPB or 0 low delay IP.
1287 * - encoding: Set by libavcodec.
1288 * - decoding: Set by libavcodec.
1293 * number of bytes per packet if constant and known or 0
1294 * Used by some WAV based audio codecs.
1298 int parse_only; /* - decoding only: If true, only parsing is done
1299 (function avcodec_parse_frame()). The frame
1300 data is returned. Only MPEG codecs support this now. */
1303 * 0-> h263 quant 1-> mpeg quant
1304 * - encoding: Set by user.
1305 * - decoding: unused
1310 * pass1 encoding statistics output buffer
1311 * - encoding: Set by libavcodec.
1312 * - decoding: unused
1317 * pass2 encoding statistics input buffer
1318 * Concatenated stuff from stats_out of pass1 should be placed here.
1319 * - encoding: Allocated/set/freed by user.
1320 * - decoding: unused
1325 * ratecontrol qmin qmax limiting method
1326 * 0-> clipping, 1-> use a nice continous function to limit qscale wthin qmin/qmax.
1327 * - encoding: Set by user.
1328 * - decoding: unused
1336 * ratecontrol override, see RcOverride
1337 * - encoding: Allocated/set/freed by user.
1338 * - decoding: unused
1340 RcOverride *rc_override;
1341 int rc_override_count;
1344 * rate control equation
1345 * - encoding: Set by user
1346 * - decoding: unused
1352 * - encoding: Set by user.
1353 * - decoding: unused
1359 * - encoding: Set by user.
1360 * - decoding: unused
1365 * decoder bitstream buffer size
1366 * - encoding: Set by user.
1367 * - decoding: unused
1370 float rc_buffer_aggressivity;
1373 * qscale factor between P and I-frames
1374 * If > 0 then the last p frame quantizer will be used (q= lastp_q*factor+offset).
1375 * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
1376 * - encoding: Set by user.
1377 * - decoding: unused
1379 float i_quant_factor;
1382 * qscale offset between P and I-frames
1383 * - encoding: Set by user.
1384 * - decoding: unused
1386 float i_quant_offset;
1389 * initial complexity for pass1 ratecontrol
1390 * - encoding: Set by user.
1391 * - decoding: unused
1393 float rc_initial_cplx;
1396 * DCT algorithm, see FF_DCT_* below
1397 * - encoding: Set by user.
1398 * - decoding: unused
1401 #define FF_DCT_AUTO 0
1402 #define FF_DCT_FASTINT 1
1403 #define FF_DCT_INT 2
1404 #define FF_DCT_MMX 3
1405 #define FF_DCT_MLIB 4
1406 #define FF_DCT_ALTIVEC 5
1407 #define FF_DCT_FAAN 6
1410 * luminance masking (0-> disabled)
1411 * - encoding: Set by user.
1412 * - decoding: unused
1417 * temporary complexity masking (0-> disabled)
1418 * - encoding: Set by user.
1419 * - decoding: unused
1421 float temporal_cplx_masking;
1424 * spatial complexity masking (0-> disabled)
1425 * - encoding: Set by user.
1426 * - decoding: unused
1428 float spatial_cplx_masking;
1431 * p block masking (0-> disabled)
1432 * - encoding: Set by user.
1433 * - decoding: unused
1438 * darkness masking (0-> disabled)
1439 * - encoding: Set by user.
1440 * - decoding: unused
1445 * IDCT algorithm, see FF_IDCT_* below.
1446 * - encoding: Set by user.
1447 * - decoding: Set by user.
1450 #define FF_IDCT_AUTO 0
1451 #define FF_IDCT_INT 1
1452 #define FF_IDCT_SIMPLE 2
1453 #define FF_IDCT_SIMPLEMMX 3
1454 #define FF_IDCT_LIBMPEG2MMX 4
1455 #define FF_IDCT_PS2 5
1456 #define FF_IDCT_MLIB 6
1457 #define FF_IDCT_ARM 7
1458 #define FF_IDCT_ALTIVEC 8
1459 #define FF_IDCT_SH4 9
1460 #define FF_IDCT_SIMPLEARM 10
1461 #define FF_IDCT_H264 11
1462 #define FF_IDCT_VP3 12
1463 #define FF_IDCT_IPP 13
1464 #define FF_IDCT_XVIDMMX 14
1465 #define FF_IDCT_CAVS 15
1466 #define FF_IDCT_SIMPLEARMV5TE 16
1467 #define FF_IDCT_SIMPLEARMV6 17
1468 #define FF_IDCT_SIMPLEVIS 18
1469 #define FF_IDCT_WMV2 19
1470 #define FF_IDCT_FAAN 20
1471 #define FF_IDCT_EA 21
1472 #define FF_IDCT_SIMPLENEON 22
1473 #define FF_IDCT_SIMPLEALPHA 23
1477 * - encoding: Set by libavcodec.
1478 * - decoding: Set by user (or 0).
1482 * slice offsets in the frame in bytes
1483 * - encoding: Set/allocated by libavcodec.
1484 * - decoding: Set/allocated by user (or NULL).
1489 * error concealment flags
1490 * - encoding: unused
1491 * - decoding: Set by user.
1493 int error_concealment;
1494 #define FF_EC_GUESS_MVS 1
1495 #define FF_EC_DEBLOCK 2
1498 * dsp_mask could be add used to disable unwanted CPU features
1499 * CPU features (i.e. MMX, SSE. ...)
1501 * With the FORCE flag you may instead enable given CPU features.
1502 * (Dangerous: Usable in case of misdetection, improper usage however will
1503 * result into program crash.)
1506 #define FF_MM_FORCE 0x80000000 /* Force usage of selected flags (OR) */
1507 /* lower 16 bits - CPU features */
1508 #define FF_MM_MMX 0x0001 ///< standard MMX
1509 #define FF_MM_3DNOW 0x0004 ///< AMD 3DNOW
1510 #if LIBAVCODEC_VERSION_MAJOR < 53
1511 #define FF_MM_MMXEXT 0x0002 ///< SSE integer functions or AMD MMX ext
1513 #define FF_MM_MMX2 0x0002 ///< SSE integer functions or AMD MMX ext
1514 #define FF_MM_SSE 0x0008 ///< SSE functions
1515 #define FF_MM_SSE2 0x0010 ///< PIV SSE2 functions
1516 #define FF_MM_3DNOWEXT 0x0020 ///< AMD 3DNowExt
1517 #define FF_MM_SSE3 0x0040 ///< Prescott SSE3 functions
1518 #define FF_MM_SSSE3 0x0080 ///< Conroe SSSE3 functions
1519 #define FF_MM_SSE4 0x0100 ///< Penryn SSE4.1 functions
1520 #define FF_MM_SSE42 0x0200 ///< Nehalem SSE4.2 functions
1521 #define FF_MM_IWMMXT 0x0100 ///< XScale IWMMXT
1522 #define FF_MM_ALTIVEC 0x0001 ///< standard AltiVec
1525 * bits per sample/pixel from the demuxer (needed for huffyuv).
1526 * - encoding: Set by libavcodec.
1527 * - decoding: Set by user.
1529 int bits_per_coded_sample;
1532 * prediction method (needed for huffyuv)
1533 * - encoding: Set by user.
1534 * - decoding: unused
1536 int prediction_method;
1537 #define FF_PRED_LEFT 0
1538 #define FF_PRED_PLANE 1
1539 #define FF_PRED_MEDIAN 2
1542 * sample aspect ratio (0 if unknown)
1543 * That is the width of a pixel divided by the height of the pixel.
1544 * Numerator and denominator must be relatively prime and smaller than 256 for some video standards.
1545 * - encoding: Set by user.
1546 * - decoding: Set by libavcodec.
1548 AVRational sample_aspect_ratio;
1551 * the picture in the bitstream
1552 * - encoding: Set by libavcodec.
1553 * - decoding: Set by libavcodec.
1555 AVFrame *coded_frame;
1559 * - encoding: Set by user.
1560 * - decoding: Set by user.
1563 #define FF_DEBUG_PICT_INFO 1
1564 #define FF_DEBUG_RC 2
1565 #define FF_DEBUG_BITSTREAM 4
1566 #define FF_DEBUG_MB_TYPE 8
1567 #define FF_DEBUG_QP 16
1568 #define FF_DEBUG_MV 32
1569 #define FF_DEBUG_DCT_COEFF 0x00000040
1570 #define FF_DEBUG_SKIP 0x00000080
1571 #define FF_DEBUG_STARTCODE 0x00000100
1572 #define FF_DEBUG_PTS 0x00000200
1573 #define FF_DEBUG_ER 0x00000400
1574 #define FF_DEBUG_MMCO 0x00000800
1575 #define FF_DEBUG_BUGS 0x00001000
1576 #define FF_DEBUG_VIS_QP 0x00002000
1577 #define FF_DEBUG_VIS_MB_TYPE 0x00004000
1578 #define FF_DEBUG_BUFFERS 0x00008000
1582 * - encoding: Set by user.
1583 * - decoding: Set by user.
1586 #define FF_DEBUG_VIS_MV_P_FOR 0x00000001 //visualize forward predicted MVs of P frames
1587 #define FF_DEBUG_VIS_MV_B_FOR 0x00000002 //visualize forward predicted MVs of B frames
1588 #define FF_DEBUG_VIS_MV_B_BACK 0x00000004 //visualize backward predicted MVs of B frames
1592 * - encoding: Set by libavcodec if flags&CODEC_FLAG_PSNR.
1593 * - decoding: unused
1598 * minimum MB quantizer
1599 * - encoding: unused
1600 * - decoding: unused
1605 * maximum MB quantizer
1606 * - encoding: unused
1607 * - decoding: unused
1612 * motion estimation comparison function
1613 * - encoding: Set by user.
1614 * - decoding: unused
1618 * subpixel motion estimation comparison function
1619 * - encoding: Set by user.
1620 * - decoding: unused
1624 * macroblock comparison function (not supported yet)
1625 * - encoding: Set by user.
1626 * - decoding: unused
1630 * interlaced DCT comparison function
1631 * - encoding: Set by user.
1632 * - decoding: unused
1635 #define FF_CMP_SAD 0
1636 #define FF_CMP_SSE 1
1637 #define FF_CMP_SATD 2
1638 #define FF_CMP_DCT 3
1639 #define FF_CMP_PSNR 4
1640 #define FF_CMP_BIT 5
1642 #define FF_CMP_ZERO 7
1643 #define FF_CMP_VSAD 8
1644 #define FF_CMP_VSSE 9
1645 #define FF_CMP_NSSE 10
1646 #define FF_CMP_W53 11
1647 #define FF_CMP_W97 12
1648 #define FF_CMP_DCTMAX 13
1649 #define FF_CMP_DCT264 14
1650 #define FF_CMP_CHROMA 256
1653 * ME diamond size & shape
1654 * - encoding: Set by user.
1655 * - decoding: unused
1660 * amount of previous MV predictors (2a+1 x 2a+1 square)
1661 * - encoding: Set by user.
1662 * - decoding: unused
1664 int last_predictor_count;
1667 * prepass for motion estimation
1668 * - encoding: Set by user.
1669 * - decoding: unused
1674 * motion estimation prepass comparison function
1675 * - encoding: Set by user.
1676 * - decoding: unused
1681 * ME prepass diamond size & shape
1682 * - encoding: Set by user.
1683 * - decoding: unused
1689 * - encoding: Set by user.
1690 * - decoding: unused
1692 int me_subpel_quality;
1695 * callback to negotiate the pixelFormat
1696 * @param fmt is the list of formats which are supported by the codec,
1697 * it is terminated by -1 as 0 is a valid format, the formats are ordered by quality.
1698 * The first is always the native one.
1699 * @return the chosen format
1700 * - encoding: unused
1701 * - decoding: Set by user, if not set the native format will be chosen.
1703 enum PixelFormat (*get_format)(struct AVCodecContext *s, const enum PixelFormat * fmt);
1706 * DTG active format information (additional aspect ratio
1707 * information only used in DVB MPEG-2 transport streams)
1710 * - encoding: unused
1711 * - decoding: Set by decoder.
1713 int dtg_active_format;
1714 #define FF_DTG_AFD_SAME 8
1715 #define FF_DTG_AFD_4_3 9
1716 #define FF_DTG_AFD_16_9 10
1717 #define FF_DTG_AFD_14_9 11
1718 #define FF_DTG_AFD_4_3_SP_14_9 13
1719 #define FF_DTG_AFD_16_9_SP_14_9 14
1720 #define FF_DTG_AFD_SP_4_3 15
1723 * maximum motion estimation search range in subpel units
1724 * If 0 then no limit.
1726 * - encoding: Set by user.
1727 * - decoding: unused
1732 * intra quantizer bias
1733 * - encoding: Set by user.
1734 * - decoding: unused
1736 int intra_quant_bias;
1737 #define FF_DEFAULT_QUANT_BIAS 999999
1740 * inter quantizer bias
1741 * - encoding: Set by user.
1742 * - decoding: unused
1744 int inter_quant_bias;
1748 * - encoding: unused
1749 * - decoding: Which clrtable should be used for 8bit RGB images.
1750 * Tables have to be stored somewhere. FIXME
1755 * internal_buffer count
1756 * Don't touch, used by libavcodec default_get_buffer().
1758 int internal_buffer_count;
1762 * Don't touch, used by libavcodec default_get_buffer().
1764 void *internal_buffer;
1766 #define FF_LAMBDA_SHIFT 7
1767 #define FF_LAMBDA_SCALE (1<<FF_LAMBDA_SHIFT)
1768 #define FF_QP2LAMBDA 118 ///< factor to convert from H.263 QP to lambda
1769 #define FF_LAMBDA_MAX (256*128-1)
1771 #define FF_QUALITY_SCALE FF_LAMBDA_SCALE //FIXME maybe remove
1773 * Global quality for codecs which cannot change it per frame.
1774 * This should be proportional to MPEG-1/2/4 qscale.
1775 * - encoding: Set by user.
1776 * - decoding: unused
1780 #define FF_CODER_TYPE_VLC 0
1781 #define FF_CODER_TYPE_AC 1
1782 #define FF_CODER_TYPE_RAW 2
1783 #define FF_CODER_TYPE_RLE 3
1784 #define FF_CODER_TYPE_DEFLATE 4
1787 * - encoding: Set by user.
1788 * - decoding: unused
1794 * - encoding: Set by user.
1795 * - decoding: unused
1801 * - encoding: unused
1802 * - decoding: Set by user.
1804 uint8_t * (*realloc)(struct AVCodecContext *s, uint8_t *buf, int buf_size);
1809 * - encoding: unused
1810 * - decoding: Set by user.
1813 #define SLICE_FLAG_CODED_ORDER 0x0001 ///< draw_horiz_band() is called in coded order instead of display
1814 #define SLICE_FLAG_ALLOW_FIELD 0x0002 ///< allow draw_horiz_band() with field slices (MPEG2 field pics)
1815 #define SLICE_FLAG_ALLOW_PLANE 0x0004 ///< allow draw_horiz_band() with 1 component at a time (SVQ1)
1818 * XVideo Motion Acceleration
1819 * - encoding: forbidden
1820 * - decoding: set by decoder
1822 int xvmc_acceleration;
1825 * macroblock decision mode
1826 * - encoding: Set by user.
1827 * - decoding: unused
1830 #define FF_MB_DECISION_SIMPLE 0 ///< uses mb_cmp
1831 #define FF_MB_DECISION_BITS 1 ///< chooses the one which needs the fewest bits
1832 #define FF_MB_DECISION_RD 2 ///< rate distortion
1835 * custom intra quantization matrix
1836 * - encoding: Set by user, can be NULL.
1837 * - decoding: Set by libavcodec.
1839 uint16_t *intra_matrix;
1842 * custom inter quantization matrix
1843 * - encoding: Set by user, can be NULL.
1844 * - decoding: Set by libavcodec.
1846 uint16_t *inter_matrix;
1849 * fourcc from the AVI stream header (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
1850 * This is used to work around some encoder bugs.
1851 * - encoding: unused
1852 * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
1854 unsigned int stream_codec_tag;
1857 * scene change detection threshold
1858 * 0 is default, larger means fewer detected scene changes.
1859 * - encoding: Set by user.
1860 * - decoding: unused
1862 int scenechange_threshold;
1865 * minimum Lagrange multipler
1866 * - encoding: Set by user.
1867 * - decoding: unused
1872 * maximum Lagrange multipler
1873 * - encoding: Set by user.
1874 * - decoding: unused
1879 * palette control structure
1880 * - encoding: ??? (no palette-enabled encoder yet)
1881 * - decoding: Set by user.
1883 struct AVPaletteControl *palctrl;
1886 * noise reduction strength
1887 * - encoding: Set by user.
1888 * - decoding: unused
1890 int noise_reduction;
1893 * Called at the beginning of a frame to get cr buffer for it.
1894 * Buffer type (size, hints) must be the same. libavcodec won't check it.
1895 * libavcodec will pass previous buffer in pic, function should return
1896 * same buffer or new buffer with old frame "painted" into it.
1897 * If pic.data[0] == NULL must behave like get_buffer().
1898 * - encoding: unused
1899 * - decoding: Set by libavcodec., user can override
1901 int (*reget_buffer)(struct AVCodecContext *c, AVFrame *pic);
1904 * Number of bits which should be loaded into the rc buffer before decoding starts.
1905 * - encoding: Set by user.
1906 * - decoding: unused
1908 int rc_initial_buffer_occupancy;
1912 * - encoding: Set by user.
1913 * - decoding: unused
1915 int inter_threshold;
1919 * - encoding: Set by user.
1920 * - decoding: Set by user.
1925 * Simulates errors in the bitstream to test error concealment.
1926 * - encoding: Set by user.
1927 * - decoding: unused
1932 * MP3 antialias algorithm, see FF_AA_* below.
1933 * - encoding: unused
1934 * - decoding: Set by user.
1937 #define FF_AA_AUTO 0
1938 #define FF_AA_FASTINT 1 //not implemented yet
1940 #define FF_AA_FLOAT 3
1942 * quantizer noise shaping
1943 * - encoding: Set by user.
1944 * - decoding: unused
1946 int quantizer_noise_shaping;
1950 * is used to decide how many independent tasks should be passed to execute()
1951 * - encoding: Set by user.
1952 * - decoding: Set by user.
1957 * The codec may call this to execute several independent things.
1958 * It will return only after finishing all tasks.
1959 * The user may replace this with some multithreaded implementation,
1960 * the default implementation will execute the parts serially.
1961 * @param count the number of things to execute
1962 * - encoding: Set by libavcodec, user can override.
1963 * - decoding: Set by libavcodec, user can override.
1965 int (*execute)(struct AVCodecContext *c, int (*func)(struct AVCodecContext *c2, void *arg), void *arg2, int *ret, int count, int size);
1969 * Can be used by execute() to store some per AVCodecContext stuff.
1970 * - encoding: set by execute()
1971 * - decoding: set by execute()
1973 void *thread_opaque;
1976 * Motion estimation threshold below which no motion estimation is
1977 * performed, but instead the user specified motion vectors are used.
1979 * - encoding: Set by user.
1980 * - decoding: unused
1985 * Macroblock threshold below which the user specified macroblock types will be used.
1986 * - encoding: Set by user.
1987 * - decoding: unused
1992 * precision of the intra DC coefficient - 8
1993 * - encoding: Set by user.
1994 * - decoding: unused
1996 int intra_dc_precision;
1999 * noise vs. sse weight for the nsse comparsion function
2000 * - encoding: Set by user.
2001 * - decoding: unused
2006 * Number of macroblock rows at the top which are skipped.
2007 * - encoding: unused
2008 * - decoding: Set by user.
2013 * Number of macroblock rows at the bottom which are skipped.
2014 * - encoding: unused
2015 * - decoding: Set by user.
2021 * - encoding: Set by user.
2022 * - decoding: Set by libavcodec.
2025 #define FF_PROFILE_UNKNOWN -99
2026 #define FF_PROFILE_AAC_MAIN 0
2027 #define FF_PROFILE_AAC_LOW 1
2028 #define FF_PROFILE_AAC_SSR 2
2029 #define FF_PROFILE_AAC_LTP 3
2033 * - encoding: Set by user.
2034 * - decoding: Set by libavcodec.
2037 #define FF_LEVEL_UNKNOWN -99
2040 * low resolution decoding, 1-> 1/2 size, 2->1/4 size
2041 * - encoding: unused
2042 * - decoding: Set by user.
2047 * Bitstream width / height, may be different from width/height if lowres
2048 * or other things are used.
2049 * - encoding: unused
2050 * - decoding: Set by user before init if known. Codec should override / dynamically change if needed.
2052 int coded_width, coded_height;
2055 * frame skip threshold
2056 * - encoding: Set by user.
2057 * - decoding: unused
2059 int frame_skip_threshold;
2063 * - encoding: Set by user.
2064 * - decoding: unused
2066 int frame_skip_factor;
2069 * frame skip exponent
2070 * - encoding: Set by user.
2071 * - decoding: unused
2076 * frame skip comparison function
2077 * - encoding: Set by user.
2078 * - decoding: unused
2083 * Border processing masking, raises the quantizer for mbs on the borders
2085 * - encoding: Set by user.
2086 * - decoding: unused
2088 float border_masking;
2091 * minimum MB lagrange multipler
2092 * - encoding: Set by user.
2093 * - decoding: unused
2098 * maximum MB lagrange multipler
2099 * - encoding: Set by user.
2100 * - decoding: unused
2106 * - encoding: Set by user.
2107 * - decoding: unused
2109 int me_penalty_compensation;
2113 * - encoding: unused
2114 * - decoding: Set by user.
2116 enum AVDiscard skip_loop_filter;
2120 * - encoding: unused
2121 * - decoding: Set by user.
2123 enum AVDiscard skip_idct;
2127 * - encoding: unused
2128 * - decoding: Set by user.
2130 enum AVDiscard skip_frame;
2134 * - encoding: Set by user.
2135 * - decoding: unused
2141 * - encoding: Set by user.
2142 * - decoding: unused
2147 * constant rate factor - quality-based VBR - values ~correspond to qps
2148 * - encoding: Set by user.
2149 * - decoding: unused
2154 * constant quantization parameter rate control method
2155 * - encoding: Set by user.
2156 * - decoding: unused
2162 * - encoding: Set by user.
2163 * - decoding: unused
2168 * number of reference frames
2169 * - encoding: Set by user.
2170 * - decoding: Set by lavc.
2175 * chroma qp offset from luma
2176 * - encoding: Set by user.
2177 * - decoding: unused
2182 * Influences how often B-frames are used.
2183 * - encoding: Set by user.
2184 * - decoding: unused
2189 * trellis RD quantization
2190 * - encoding: Set by user.
2191 * - decoding: unused
2196 * Reduce fluctuations in qp (before curve compression).
2197 * - encoding: Set by user.
2198 * - decoding: unused
2200 float complexityblur;
2203 * in-loop deblocking filter alphac0 parameter
2204 * alpha is in the range -6...6
2205 * - encoding: Set by user.
2206 * - decoding: unused
2211 * in-loop deblocking filter beta parameter
2212 * beta is in the range -6...6
2213 * - encoding: Set by user.
2214 * - decoding: unused
2219 * macroblock subpartition sizes to consider - p8x8, p4x4, b8x8, i8x8, i4x4
2220 * - encoding: Set by user.
2221 * - decoding: unused
2224 #define X264_PART_I4X4 0x001 /* Analyze i4x4 */
2225 #define X264_PART_I8X8 0x002 /* Analyze i8x8 (requires 8x8 transform) */
2226 #define X264_PART_P8X8 0x010 /* Analyze p16x8, p8x16 and p8x8 */
2227 #define X264_PART_P4X4 0x020 /* Analyze p8x4, p4x8, p4x4 */
2228 #define X264_PART_B8X8 0x100 /* Analyze b16x8, b8x16 and b8x8 */
2231 * direct MV prediction mode - 0 (none), 1 (spatial), 2 (temporal), 3 (auto)
2232 * - encoding: Set by user.
2233 * - decoding: unused
2238 * Audio cutoff bandwidth (0 means "automatic")
2239 * - encoding: Set by user.
2240 * - decoding: unused
2245 * Multiplied by qscale for each frame and added to scene_change_score.
2246 * - encoding: Set by user.
2247 * - decoding: unused
2249 int scenechange_factor;
2253 * Note: Value depends upon the compare function used for fullpel ME.
2254 * - encoding: Set by user.
2255 * - decoding: unused
2260 * Adjusts sensitivity of b_frame_strategy 1.
2261 * - encoding: Set by user.
2262 * - decoding: unused
2267 * - encoding: Set by user.
2268 * - decoding: unused
2270 int compression_level;
2271 #define FF_COMPRESSION_DEFAULT -1
2274 * Sets whether to use LPC mode - used by FLAC encoder.
2275 * - encoding: Set by user.
2276 * - decoding: unused
2281 * LPC coefficient precision - used by FLAC encoder
2282 * - encoding: Set by user.
2283 * - decoding: unused
2285 int lpc_coeff_precision;
2288 * - encoding: Set by user.
2289 * - decoding: unused
2291 int min_prediction_order;
2294 * - encoding: Set by user.
2295 * - decoding: unused
2297 int max_prediction_order;
2300 * search method for selecting prediction order
2301 * - encoding: Set by user.
2302 * - decoding: unused
2304 int prediction_order_method;
2307 * - encoding: Set by user.
2308 * - decoding: unused
2310 int min_partition_order;
2313 * - encoding: Set by user.
2314 * - decoding: unused
2316 int max_partition_order;
2319 * GOP timecode frame start number, in non drop frame format
2320 * - encoding: Set by user.
2321 * - decoding: unused
2323 int64_t timecode_frame_start;
2325 #if LIBAVCODEC_VERSION_MAJOR < 53
2327 * Decoder should decode to this many channels if it can (0 for default)
2328 * - encoding: unused
2329 * - decoding: Set by user.
2330 * @deprecated Deprecated in favor of request_channel_layout.
2332 int request_channels;
2336 * Percentage of dynamic range compression to be applied by the decoder.
2337 * The default value is 1.0, corresponding to full compression.
2338 * - encoding: unused
2339 * - decoding: Set by user.
2344 * opaque 64bit number (generally a PTS) that will be reordered and
2345 * output in AVFrame.reordered_opaque
2346 * - encoding: unused
2347 * - decoding: Set by user.
2349 int64_t reordered_opaque;
2352 * Bits per sample/pixel of internal libavcodec pixel/sample format.
2353 * This field is applicable only when sample_fmt is SAMPLE_FMT_S32.
2354 * - encoding: set by user.
2355 * - decoding: set by libavcodec.
2357 int bits_per_raw_sample;
2360 * Audio channel layout.
2361 * - encoding: set by user.
2362 * - decoding: set by libavcodec.
2364 int64_t channel_layout;
2367 * Request decoder to use this channel layout if it can (0 for default)
2368 * - encoding: unused
2369 * - decoding: Set by user.
2371 int64_t request_channel_layout;
2374 * Ratecontrol attempt to use, at maximum, <value> of what can be used without an underflow.
2375 * - encoding: Set by user.
2376 * - decoding: unused.
2378 float rc_max_available_vbv_use;
2381 * Ratecontrol attempt to use, at least, <value> times the amount needed to prevent a vbv overflow.
2382 * - encoding: Set by user.
2383 * - decoding: unused.
2385 float rc_min_vbv_overflow_use;
2388 * Hardware accelerator in use
2389 * - encoding: unused.
2390 * - decoding: Set by libavcodec
2392 struct AVHWAccel *hwaccel;
2395 * For some codecs, the time base is closer to the field rate than the frame rate.
2396 * Most notably, H.264 and MPEG-2 specify time_base as half of frame duration
2397 * if no telecine is used ...
2399 * Set to time_base ticks per frame. Default 1, e.g., H.264/MPEG-2 set it to 2.
2401 int ticks_per_frame;
2404 * Hardware accelerator context.
2405 * For some hardware accelerators, a global context needs to be
2406 * provided by the user. In that case, this holds display-dependent
2407 * data FFmpeg cannot instantiate itself. Please refer to the
2408 * FFmpeg HW accelerator documentation to know how to fill this
2409 * is. e.g. for VA API, this is a struct vaapi_context.
2410 * - encoding: unused
2411 * - decoding: Set by user
2413 void *hwaccel_context;
2419 typedef struct AVCodec {
2421 * Name of the codec implementation.
2422 * The name is globally unique among encoders and among decoders (but an
2423 * encoder and a decoder can share the same name).
2424 * This is the primary way to find a codec from the user perspective.
2427 enum CodecType type;
2430 int (*init)(AVCodecContext *);
2431 int (*encode)(AVCodecContext *, uint8_t *buf, int buf_size, void *data);
2432 int (*close)(AVCodecContext *);
2433 int (*decode)(AVCodecContext *, void *outdata, int *outdata_size, AVPacket *avpkt);
2435 * Codec capabilities.
2439 struct AVCodec *next;
2442 * Will be called when seeking
2444 void (*flush)(AVCodecContext *);
2445 const AVRational *supported_framerates; ///< array of supported framerates, or NULL if any, array is terminated by {0,0}
2446 const enum PixelFormat *pix_fmts; ///< array of supported pixel formats, or NULL if unknown, array is terminated by -1
2448 * Descriptive name for the codec, meant to be more human readable than \p name.
2449 * You \e should use the NULL_IF_CONFIG_SMALL() macro to define it.
2451 const char *long_name;
2452 const int *supported_samplerates; ///< array of supported audio samplerates, or NULL if unknown, array is terminated by 0
2453 const enum SampleFormat *sample_fmts; ///< array of supported sample formats, or NULL if unknown, array is terminated by -1
2454 const int64_t *channel_layouts; ///< array of support channel layouts, or NULL if unknown. array is terminated by 0
2460 typedef struct AVHWAccel {
2462 * Name of the hardware accelerated codec.
2463 * The name is globally unique among encoders and among decoders (but an
2464 * encoder and a decoder can share the same name).
2469 * Type of codec implemented by the hardware accelerator.
2471 * See CODEC_TYPE_xxx
2473 enum CodecType type;
2476 * Codec implemented by the hardware accelerator.
2483 * Supported pixel format.
2485 * Only hardware accelerated formats are supported here.
2487 enum PixelFormat pix_fmt;
2490 * Hardware accelerated codec capabilities.
2491 * see FF_HWACCEL_CODEC_CAP_*
2495 struct AVHWAccel *next;
2498 * Called at the beginning of each frame or field picture.
2500 * Meaningful frame information (codec specific) is guaranteed to
2501 * be parsed at this point. This function is mandatory.
2503 * Note that \p buf can be NULL along with \p buf_size set to 0.
2504 * Otherwise, this means the whole frame is available at this point.
2506 * @param avctx the codec context
2507 * @param buf the frame data buffer base
2508 * @param buf_size the size of the frame in bytes
2509 * @return zero if successful, a negative value otherwise
2511 int (*start_frame)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
2514 * Callback for each slice.
2516 * Meaningful slice information (codec specific) is guaranteed to
2517 * be parsed at this point. This function is mandatory.
2519 * @param avctx the codec context
2520 * @param buf the slice data buffer base
2521 * @param buf_size the size of the slice in bytes
2522 * @return zero if successful, a negative value otherwise
2524 int (*decode_slice)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
2527 * Called at the end of each frame or field picture.
2529 * The whole picture is parsed at this point and can now be sent
2530 * to the hardware accelerator. This function is mandatory.
2532 * @param avctx the codec context
2533 * @return zero if successful, a negative value otherwise
2535 int (*end_frame)(AVCodecContext *avctx);
2538 * Size of HW accelerator private data.
2540 * Private data is allocated with av_mallocz() before
2541 * AVCodecContext.get_buffer() and deallocated after
2542 * AVCodecContext.release_buffer().
2548 * four components are given, that's all.
2549 * the last component is alpha
2551 typedef struct AVPicture {
2553 int linesize[4]; ///< number of bytes per line
2556 #if LIBAVCODEC_VERSION_MAJOR < 53
2559 * This structure defines a method for communicating palette changes
2560 * between and demuxer and a decoder.
2562 * @deprecated Use AVPacket to send palette changes instead.
2563 * This is totally broken.
2565 #define AVPALETTE_SIZE 1024
2566 #define AVPALETTE_COUNT 256
2567 typedef struct AVPaletteControl {
2569 /* Demuxer sets this to 1 to indicate the palette has changed;
2570 * decoder resets to 0. */
2571 int palette_changed;
2573 /* 4-byte ARGB palette entries, stored in native byte order; note that
2574 * the individual palette components should be on a 8-bit scale; if
2575 * the palette data comes from an IBM VGA native format, the component
2576 * data is probably 6 bits in size and needs to be scaled. */
2577 unsigned int palette[AVPALETTE_COUNT];
2579 } AVPaletteControl attribute_deprecated;
2582 enum AVSubtitleType {
2585 SUBTITLE_BITMAP, ///< A bitmap, pict will be set
2588 * Plain text, the text field must be set by the decoder and is
2589 * authoritative. ass and pict fields may contain approximations.
2594 * Formatted text, the ass field must be set by the decoder and is
2595 * authoritative. pict and text fields may contain approximations.
2600 typedef struct AVSubtitleRect {
2601 int x; ///< top left corner of pict, undefined when pict is not set
2602 int y; ///< top left corner of pict, undefined when pict is not set
2603 int w; ///< width of pict, undefined when pict is not set
2604 int h; ///< height of pict, undefined when pict is not set
2605 int nb_colors; ///< number of colors in pict, undefined when pict is not set
2608 * data+linesize for the bitmap of this subtitle.
2609 * can be set for text/ass as well once they where rendered
2612 enum AVSubtitleType type;
2614 char *text; ///< 0 terminated plain UTF-8 text
2617 * 0 terminated ASS/SSA compatible event line.
2618 * The pressentation of this is unaffected by the other values in this
2624 typedef struct AVSubtitle {
2625 uint16_t format; /* 0 = graphics */
2626 uint32_t start_display_time; /* relative to packet pts, in ms */
2627 uint32_t end_display_time; /* relative to packet pts, in ms */
2629 AVSubtitleRect **rects;
2632 /* packet functions */
2635 * @deprecated use NULL instead
2637 attribute_deprecated void av_destruct_packet_nofree(AVPacket *pkt);
2640 * Default packet destructor.
2642 void av_destruct_packet(AVPacket *pkt);
2645 * Initialize optional fields of a packet with default values.
2649 void av_init_packet(AVPacket *pkt);
2652 * Allocate the payload of a packet and initialize its fields with
2656 * @param size wanted payload size
2657 * @return 0 if OK, AVERROR_xxx otherwise
2659 int av_new_packet(AVPacket *pkt, int size);
2662 * Reduce packet size, correctly zeroing padding
2665 * @param size new size
2667 void av_shrink_packet(AVPacket *pkt, int size);
2670 * @warning This is a hack - the packet memory allocation stuff is broken. The
2671 * packet is allocated if it was not really allocated.
2673 int av_dup_packet(AVPacket *pkt);
2678 * @param pkt packet to free
2680 static inline void av_free_packet(AVPacket *pkt)
2683 if (pkt->destruct) pkt->destruct(pkt);
2684 pkt->data = NULL; pkt->size = 0;
2690 struct ReSampleContext;
2691 struct AVResampleContext;
2693 typedef struct ReSampleContext ReSampleContext;
2695 #if LIBAVCODEC_VERSION_MAJOR < 53
2697 * @deprecated Use av_audio_resample_init() instead.
2699 attribute_deprecated ReSampleContext *audio_resample_init(int output_channels, int input_channels,
2700 int output_rate, int input_rate);
2703 * Initializes audio resampling context
2705 * @param output_channels number of output channels
2706 * @param input_channels number of input channels
2707 * @param output_rate output sample rate
2708 * @param input_rate input sample rate
2709 * @param sample_fmt_out requested output sample format
2710 * @param sample_fmt_in input sample format
2711 * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
2712 * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
2713 * @param linear If 1 then the used FIR filter will be linearly interpolated
2714 between the 2 closest, if 0 the closest will be used
2715 * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
2716 * @return allocated ReSampleContext, NULL if error occured
2718 ReSampleContext *av_audio_resample_init(int output_channels, int input_channels,
2719 int output_rate, int input_rate,
2720 enum SampleFormat sample_fmt_out,
2721 enum SampleFormat sample_fmt_in,
2722 int filter_length, int log2_phase_count,
2723 int linear, double cutoff);
2725 int audio_resample(ReSampleContext *s, short *output, short *input, int nb_samples);
2726 void audio_resample_close(ReSampleContext *s);
2730 * Initializes an audio resampler.
2731 * Note, if either rate is not an integer then simply scale both rates up so they are.
2732 * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
2733 * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
2734 * @param linear If 1 then the used FIR filter will be linearly interpolated
2735 between the 2 closest, if 0 the closest will be used
2736 * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
2738 struct AVResampleContext *av_resample_init(int out_rate, int in_rate, int filter_length, int log2_phase_count, int linear, double cutoff);
2742 * @param src an array of unconsumed samples
2743 * @param consumed the number of samples of src which have been consumed are returned here
2744 * @param src_size the number of unconsumed samples available
2745 * @param dst_size the amount of space in samples available in dst
2746 * @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.
2747 * @return the number of samples written in dst or -1 if an error occurred
2749 int av_resample(struct AVResampleContext *c, short *dst, short *src, int *consumed, int src_size, int dst_size, int update_ctx);
2753 * Compensates samplerate/timestamp drift. The compensation is done by changing
2754 * the resampler parameters, so no audible clicks or similar distortions occur
2755 * @param compensation_distance distance in output samples over which the compensation should be performed
2756 * @param sample_delta number of output samples which should be output less
2758 * example: av_resample_compensate(c, 10, 500)
2759 * here instead of 510 samples only 500 samples would be output
2761 * note, due to rounding the actual compensation might be slightly different,
2762 * especially if the compensation_distance is large and the in_rate used during init is small
2764 void av_resample_compensate(struct AVResampleContext *c, int sample_delta, int compensation_distance);
2765 void av_resample_close(struct AVResampleContext *c);
2768 * Allocate memory for a picture. Call avpicture_free to free it.
2770 * @param picture the picture to be filled in
2771 * @param pix_fmt the format of the picture
2772 * @param width the width of the picture
2773 * @param height the height of the picture
2774 * @return zero if successful, a negative value if not
2776 int avpicture_alloc(AVPicture *picture, enum PixelFormat pix_fmt, int width, int height);
2779 * Free a picture previously allocated by avpicture_alloc().
2781 * @param picture the AVPicture to be freed
2783 void avpicture_free(AVPicture *picture);
2786 * Fill in the AVPicture fields.
2787 * The fields of the given AVPicture are filled in by using the 'ptr' address
2788 * which points to the image data buffer. Depending on the specified picture
2789 * format, one or multiple image data pointers and line sizes will be set.
2790 * If a planar format is specified, several pointers will be set pointing to
2791 * the different picture planes and the line sizes of the different planes
2792 * will be stored in the lines_sizes array.
2793 * Call with ptr == NULL to get the required size for the ptr buffer.
2795 * @param picture AVPicture whose fields are to be filled in
2796 * @param ptr Buffer which will contain or contains the actual image data
2797 * @param pix_fmt The format in which the picture data is stored.
2798 * @param width the width of the image in pixels
2799 * @param height the height of the image in pixels
2800 * @return size of the image data in bytes
2802 int avpicture_fill(AVPicture *picture, uint8_t *ptr,
2803 int pix_fmt, int width, int height);
2804 int avpicture_layout(const AVPicture* src, enum PixelFormat pix_fmt, int width, int height,
2805 unsigned char *dest, int dest_size);
2808 * Calculate the size in bytes that a picture of the given width and height
2809 * would occupy if stored in the given picture format.
2810 * Note that this returns the size of a compact representation as generated
2811 * by avpicture_layout, which can be smaller than the size required for e.g.
2814 * @param pix_fmt the given picture format
2815 * @param width the width of the image
2816 * @param height the height of the image
2817 * @return Image data size in bytes or -1 on error (e.g. too large dimensions).
2819 int avpicture_get_size(enum PixelFormat pix_fmt, int width, int height);
2820 void avcodec_get_chroma_sub_sample(enum PixelFormat pix_fmt, int *h_shift, int *v_shift);
2821 const char *avcodec_get_pix_fmt_name(enum PixelFormat pix_fmt);
2822 void avcodec_set_dimensions(AVCodecContext *s, int width, int height);
2825 * Returns the pixel format corresponding to the name \p name.
2827 * If there is no pixel format with name \p name, then looks for a
2828 * pixel format with the name corresponding to the native endian
2829 * format of \p name.
2830 * For example in a little-endian system, first looks for "gray16",
2831 * then for "gray16le".
2833 * Finally if no pixel format has been found, returns \c PIX_FMT_NONE.
2835 enum PixelFormat avcodec_get_pix_fmt(const char* name);
2836 unsigned int avcodec_pix_fmt_to_codec_tag(enum PixelFormat p);
2838 #define FF_LOSS_RESOLUTION 0x0001 /**< loss due to resolution change */
2839 #define FF_LOSS_DEPTH 0x0002 /**< loss due to color depth change */
2840 #define FF_LOSS_COLORSPACE 0x0004 /**< loss due to color space conversion */
2841 #define FF_LOSS_ALPHA 0x0008 /**< loss of alpha bits */
2842 #define FF_LOSS_COLORQUANT 0x0010 /**< loss due to color quantization */
2843 #define FF_LOSS_CHROMA 0x0020 /**< loss of chroma (e.g. RGB to gray conversion) */
2846 * Computes what kind of losses will occur when converting from one specific
2847 * pixel format to another.
2848 * When converting from one pixel format to another, information loss may occur.
2849 * For example, when converting from RGB24 to GRAY, the color information will
2850 * be lost. Similarly, other losses occur when converting from some formats to
2851 * other formats. These losses can involve loss of chroma, but also loss of
2852 * resolution, loss of color depth, loss due to the color space conversion, loss
2853 * of the alpha bits or loss due to color quantization.
2854 * avcodec_get_fix_fmt_loss() informs you about the various types of losses
2855 * which will occur when converting from one pixel format to another.
2857 * @param[in] dst_pix_fmt destination pixel format
2858 * @param[in] src_pix_fmt source pixel format
2859 * @param[in] has_alpha Whether the source pixel format alpha channel is used.
2860 * @return Combination of flags informing you what kind of losses will occur.
2862 int avcodec_get_pix_fmt_loss(enum PixelFormat dst_pix_fmt, enum PixelFormat src_pix_fmt,
2866 * Finds the best pixel format to convert to given a certain source pixel
2867 * format. When converting from one pixel format to another, information loss
2868 * may occur. For example, when converting from RGB24 to GRAY, the color
2869 * information will be lost. Similarly, other losses occur when converting from
2870 * some formats to other formats. avcodec_find_best_pix_fmt() searches which of
2871 * the given pixel formats should be used to suffer the least amount of loss.
2872 * The pixel formats from which it chooses one, are determined by the
2873 * \p pix_fmt_mask parameter.
2876 * src_pix_fmt = PIX_FMT_YUV420P;
2877 * pix_fmt_mask = (1 << PIX_FMT_YUV422P) || (1 << PIX_FMT_RGB24);
2878 * dst_pix_fmt = avcodec_find_best_pix_fmt(pix_fmt_mask, src_pix_fmt, alpha, &loss);
2881 * @param[in] pix_fmt_mask bitmask determining which pixel format to choose from
2882 * @param[in] src_pix_fmt source pixel format
2883 * @param[in] has_alpha Whether the source pixel format alpha channel is used.
2884 * @param[out] loss_ptr Combination of flags informing you what kind of losses will occur.
2885 * @return The best pixel format to convert to or -1 if none was found.
2887 enum PixelFormat avcodec_find_best_pix_fmt(int64_t pix_fmt_mask, enum PixelFormat src_pix_fmt,
2888 int has_alpha, int *loss_ptr);
2892 * Print in buf the string corresponding to the pixel format with
2893 * number pix_fmt, or an header if pix_fmt is negative.
2895 * @param[in] buf the buffer where to write the string
2896 * @param[in] buf_size the size of buf
2897 * @param[in] pix_fmt the number of the pixel format to print the corresponding info string, or
2898 * a negative value to print the corresponding header.
2899 * Meaningful values for obtaining a pixel format info vary from 0 to PIX_FMT_NB -1.
2901 void avcodec_pix_fmt_string (char *buf, int buf_size, enum PixelFormat pix_fmt);
2903 #define FF_ALPHA_TRANSP 0x0001 /* image has some totally transparent pixels */
2904 #define FF_ALPHA_SEMI_TRANSP 0x0002 /* image has some transparent pixels */
2907 * Tell if an image really has transparent alpha values.
2908 * @return ored mask of FF_ALPHA_xxx constants
2910 int img_get_alpha_info(const AVPicture *src,
2911 enum PixelFormat pix_fmt, int width, int height);
2913 /* deinterlace a picture */
2914 /* deinterlace - if not supported return -1 */
2915 int avpicture_deinterlace(AVPicture *dst, const AVPicture *src,
2916 enum PixelFormat pix_fmt, int width, int height);
2918 /* external high level API */
2921 * If c is NULL, returns the first registered codec,
2922 * if c is non-NULL, returns the next registered codec after c,
2923 * or NULL if c is the last one.
2925 AVCodec *av_codec_next(AVCodec *c);
2928 * Returns the LIBAVCODEC_VERSION_INT constant.
2930 unsigned avcodec_version(void);
2933 * Initializes libavcodec.
2935 * @warning This function \e must be called before any other libavcodec
2938 void avcodec_init(void);
2940 #if LIBAVCODEC_VERSION_MAJOR < 53
2942 * @deprecated Deprecated in favor of avcodec_register().
2944 attribute_deprecated void register_avcodec(AVCodec *codec);
2948 * Register the codec \p codec and initialize libavcodec.
2950 * @see avcodec_init()
2952 void avcodec_register(AVCodec *codec);
2955 * Finds a registered encoder with a matching codec ID.
2957 * @param id CodecID of the requested encoder
2958 * @return An encoder if one was found, NULL otherwise.
2960 AVCodec *avcodec_find_encoder(enum CodecID id);
2963 * Finds a registered encoder with the specified name.
2965 * @param name name of the requested encoder
2966 * @return An encoder if one was found, NULL otherwise.
2968 AVCodec *avcodec_find_encoder_by_name(const char *name);
2971 * Finds a registered decoder with a matching codec ID.
2973 * @param id CodecID of the requested decoder
2974 * @return A decoder if one was found, NULL otherwise.
2976 AVCodec *avcodec_find_decoder(enum CodecID id);
2979 * Finds a registered decoder with the specified name.
2981 * @param name name of the requested decoder
2982 * @return A decoder if one was found, NULL otherwise.
2984 AVCodec *avcodec_find_decoder_by_name(const char *name);
2985 void avcodec_string(char *buf, int buf_size, AVCodecContext *enc, int encode);
2988 * Sets the fields of the given AVCodecContext to default values.
2990 * @param s The AVCodecContext of which the fields should be set to default values.
2992 void avcodec_get_context_defaults(AVCodecContext *s);
2994 /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
2995 * we WILL change its arguments and name a few times! */
2996 void avcodec_get_context_defaults2(AVCodecContext *s, enum CodecType);
2999 * Allocates an AVCodecContext and sets its fields to default values. The
3000 * resulting struct can be deallocated by simply calling av_free().
3002 * @return An AVCodecContext filled with default values or NULL on failure.
3003 * @see avcodec_get_context_defaults
3005 AVCodecContext *avcodec_alloc_context(void);
3007 /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
3008 * we WILL change its arguments and name a few times! */
3009 AVCodecContext *avcodec_alloc_context2(enum CodecType);
3012 * Sets the fields of the given AVFrame to default values.
3014 * @param pic The AVFrame of which the fields should be set to default values.
3016 void avcodec_get_frame_defaults(AVFrame *pic);
3019 * Allocates an AVFrame and sets its fields to default values. The resulting
3020 * struct can be deallocated by simply calling av_free().
3022 * @return An AVFrame filled with default values or NULL on failure.
3023 * @see avcodec_get_frame_defaults
3025 AVFrame *avcodec_alloc_frame(void);
3027 int avcodec_default_get_buffer(AVCodecContext *s, AVFrame *pic);
3028 void avcodec_default_release_buffer(AVCodecContext *s, AVFrame *pic);
3029 int avcodec_default_reget_buffer(AVCodecContext *s, AVFrame *pic);
3030 void avcodec_align_dimensions(AVCodecContext *s, int *width, int *height);
3033 * Checks if the given dimension of a picture is valid, meaning that all
3034 * bytes of the picture can be addressed with a signed int.
3036 * @param[in] w Width of the picture.
3037 * @param[in] h Height of the picture.
3038 * @return Zero if valid, a negative value if invalid.
3040 int avcodec_check_dimensions(void *av_log_ctx, unsigned int w, unsigned int h);
3041 enum PixelFormat avcodec_default_get_format(struct AVCodecContext *s, const enum PixelFormat * fmt);
3043 int avcodec_thread_init(AVCodecContext *s, int thread_count);
3044 void avcodec_thread_free(AVCodecContext *s);
3045 int avcodec_thread_execute(AVCodecContext *s, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
3046 int avcodec_default_execute(AVCodecContext *c, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
3047 //FIXME func typedef
3050 * Initializes the AVCodecContext to use the given AVCodec. Prior to using this
3051 * function the context has to be allocated.
3053 * The functions avcodec_find_decoder_by_name(), avcodec_find_encoder_by_name(),
3054 * avcodec_find_decoder() and avcodec_find_encoder() provide an easy way for
3055 * retrieving a codec.
3057 * @warning This function is not thread safe!
3060 * avcodec_register_all();
3061 * codec = avcodec_find_decoder(CODEC_ID_H264);
3065 * context = avcodec_alloc_context();
3067 * if (avcodec_open(context, codec) < 0)
3071 * @param avctx The context which will be set up to use the given codec.
3072 * @param codec The codec to use within the context.
3073 * @return zero on success, a negative value on error
3074 * @see avcodec_alloc_context, avcodec_find_decoder, avcodec_find_encoder
3076 int avcodec_open(AVCodecContext *avctx, AVCodec *codec);
3078 #if LIBAVCODEC_VERSION_MAJOR < 53
3080 * Decodes an audio frame from \p buf into \p samples.
3081 * Wrapper function which calls avcodec_decode_audio3.
3083 * @deprecated Use avcodec_decode_audio3 instead.
3084 * @param avctx the codec context
3085 * @param[out] samples the output buffer
3086 * @param[in,out] frame_size_ptr the output buffer size in bytes
3087 * @param[in] buf the input buffer
3088 * @param[in] buf_size the input buffer size in bytes
3089 * @return On error a negative value is returned, otherwise the number of bytes
3090 * used or zero if no frame could be decompressed.
3092 attribute_deprecated int avcodec_decode_audio2(AVCodecContext *avctx, int16_t *samples,
3093 int *frame_size_ptr,
3094 const uint8_t *buf, int buf_size);
3098 * Decodes an audio frame from \p avpkt->data into \p samples.
3099 * The avcodec_decode_audio3() function decodes an audio frame from the input
3100 * buffer \p avpkt->data of size \p avpkt->size. To decode it, it makes use of the
3101 * audio codec which was coupled with \p avctx using avcodec_open(). The
3102 * resulting decoded frame is stored in output buffer \p samples. If no frame
3103 * could be decompressed, \p frame_size_ptr is zero. Otherwise, it is the
3104 * decompressed frame size in \e bytes.
3106 * @warning You \e must set \p frame_size_ptr to the allocated size of the
3107 * output buffer before calling avcodec_decode_audio3().
3109 * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
3110 * the actual read bytes because some optimized bitstream readers read 32 or 64
3111 * bits at once and could read over the end.
3113 * @warning The end of the input buffer \p avpkt->data should be set to 0 to ensure that
3114 * no overreading happens for damaged MPEG streams.
3116 * @note You might have to align the input buffer \p avpkt->data and output buffer \p
3117 * samples. The alignment requirements depend on the CPU: On some CPUs it isn't
3118 * necessary at all, on others it won't work at all if not aligned and on others
3119 * it will work but it will have an impact on performance. In practice, the
3120 * bitstream should have 4 byte alignment at minimum and all sample data should
3121 * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
3122 * the linesize is not a multiple of 16 then there's no sense in aligning the
3123 * start of the buffer to 16.
3125 * @param avctx the codec context
3126 * @param[out] samples the output buffer
3127 * @param[in,out] frame_size_ptr the output buffer size in bytes
3128 * @param[in] avpkt The input AVPacket containing the input buffer.
3129 * @return On error a negative value is returned, otherwise the number of bytes
3130 * used or zero if no frame could be decompressed.
3132 int avcodec_decode_audio3(AVCodecContext *avctx, int16_t *samples,
3133 int *frame_size_ptr,
3136 #if LIBAVCODEC_VERSION_MAJOR < 53
3138 * Decodes a video frame from \p buf into \p picture.
3139 * Wrapper function which calls avcodec_decode_video2.
3141 * @deprecated Use avcodec_decode_video2 instead.
3142 * @param avctx the codec context
3143 * @param[out] picture The AVFrame in which the decoded video frame will be stored.
3144 * @param[in] buf the input buffer
3145 * @param[in] buf_size the size of the input buffer in bytes
3146 * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
3147 * @return On error a negative value is returned, otherwise the number of bytes
3148 * used or zero if no frame could be decompressed.
3150 attribute_deprecated int avcodec_decode_video(AVCodecContext *avctx, AVFrame *picture,
3151 int *got_picture_ptr,
3152 const uint8_t *buf, int buf_size);
3156 * Decodes a video frame from \p avpkt->data into \p picture.
3157 * The avcodec_decode_video2() function decodes a video frame from the input
3158 * buffer \p avpkt->data of size \p avpkt->size. To decode it, it makes use of the
3159 * video codec which was coupled with \p avctx using avcodec_open(). The
3160 * resulting decoded frame is stored in \p picture.
3162 * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
3163 * the actual read bytes because some optimized bitstream readers read 32 or 64
3164 * bits at once and could read over the end.
3166 * @warning The end of the input buffer \p buf should be set to 0 to ensure that
3167 * no overreading happens for damaged MPEG streams.
3169 * @note You might have to align the input buffer \p avpkt->data and output buffer \p
3170 * samples. The alignment requirements depend on the CPU: on some CPUs it isn't
3171 * necessary at all, on others it won't work at all if not aligned and on others
3172 * it will work but it will have an impact on performance. In practice, the
3173 * bitstream should have 4 byte alignment at minimum and all sample data should
3174 * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
3175 * the linesize is not a multiple of 16 then there's no sense in aligning the
3176 * start of the buffer to 16.
3178 * @note Some codecs have a delay between input and output, these need to be
3179 * feeded with avpkt->data=NULL, avpkt->size=0 at the end to return the remaining frames.
3181 * @param avctx the codec context
3182 * @param[out] picture The AVFrame in which the decoded video frame will be stored.
3183 * @param[in] avpkt The input AVpacket containing the input buffer.
3184 * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
3185 * @return On error a negative value is returned, otherwise the number of bytes
3186 * used or zero if no frame could be decompressed.
3188 int avcodec_decode_video2(AVCodecContext *avctx, AVFrame *picture,
3189 int *got_picture_ptr,
3192 #if LIBAVCODEC_VERSION_MAJOR < 53
3193 /* Decode a subtitle message. Return -1 if error, otherwise return the
3194 * number of bytes used. If no subtitle could be decompressed,
3195 * got_sub_ptr is zero. Otherwise, the subtitle is stored in *sub. */
3196 attribute_deprecated int avcodec_decode_subtitle(AVCodecContext *avctx, AVSubtitle *sub,
3198 const uint8_t *buf, int buf_size);
3202 * Decodes a subtitle message.
3203 * Returns -1 if error, otherwise returns the number of bytes used.
3204 * If no subtitle could be decompressed, \p got_sub_ptr is zero.
3205 * Otherwise, the subtitle is stored in \p *sub.
3207 * @param avctx the codec context
3208 * @param[out] sub The AVSubtitle in which the decoded subtitle will be stored.
3209 * @param[in,out] got_sub_ptr Zero if no subtitle could be decompressed, otherwise, it is nonzero.
3210 * @param[in] avpkt The input AVPacket containing the input buffer.
3212 int avcodec_decode_subtitle2(AVCodecContext *avctx, AVSubtitle *sub,
3215 int avcodec_parse_frame(AVCodecContext *avctx, uint8_t **pdata,
3217 uint8_t *buf, int buf_size);
3220 * Encodes an audio frame from \p samples into \p buf.
3221 * The avcodec_encode_audio() function encodes an audio frame from the input
3222 * buffer \p samples. To encode it, it makes use of the audio codec which was
3223 * coupled with \p avctx using avcodec_open(). The resulting encoded frame is
3224 * stored in output buffer \p buf.
3226 * @note The output buffer should be at least \c FF_MIN_BUFFER_SIZE bytes large.
3227 * However, for PCM audio the user will know how much space is needed
3228 * because it depends on the value passed in \p buf_size as described
3229 * below. In that case a lower value can be used.
3231 * @param avctx the codec context
3232 * @param[out] buf the output buffer
3233 * @param[in] buf_size the output buffer size
3234 * @param[in] samples the input buffer containing the samples
3235 * The number of samples read from this buffer is frame_size*channels,
3236 * both of which are defined in \p avctx.
3237 * For PCM audio the number of samples read from \p samples is equal to
3238 * \p buf_size * input_sample_size / output_sample_size.
3239 * @return On error a negative value is returned, on success zero or the number
3240 * of bytes used to encode the data read from the input buffer.
3242 int avcodec_encode_audio(AVCodecContext *avctx, uint8_t *buf, int buf_size,
3243 const short *samples);
3246 * Encodes a video frame from \p pict into \p buf.
3247 * The avcodec_encode_video() function encodes a video frame from the input
3248 * \p pict. To encode it, it makes use of the video codec which was coupled with
3249 * \p avctx using avcodec_open(). The resulting encoded bytes representing the
3250 * frame are stored in the output buffer \p buf. The input picture should be
3251 * stored using a specific format, namely \c avctx.pix_fmt.
3253 * @param avctx the codec context
3254 * @param[out] buf the output buffer for the bitstream of encoded frame
3255 * @param[in] buf_size the size of the output buffer in bytes
3256 * @param[in] pict the input picture to encode
3257 * @return On error a negative value is returned, on success zero or the number
3258 * of bytes used from the output buffer.
3260 int avcodec_encode_video(AVCodecContext *avctx, uint8_t *buf, int buf_size,
3261 const AVFrame *pict);
3262 int avcodec_encode_subtitle(AVCodecContext *avctx, uint8_t *buf, int buf_size,
3263 const AVSubtitle *sub);
3265 int avcodec_close(AVCodecContext *avctx);
3268 * Register all the codecs, parsers and bitstream filters which were enabled at
3269 * configuration time. If you do not call this function you can select exactly
3270 * which formats you want to support, by using the individual registration
3273 * @see avcodec_register
3274 * @see av_register_codec_parser
3275 * @see av_register_bitstream_filter
3277 void avcodec_register_all(void);
3280 * Flush buffers, should be called when seeking or when switching to a different stream.
3282 void avcodec_flush_buffers(AVCodecContext *avctx);
3284 void avcodec_default_free_buffers(AVCodecContext *s);
3286 /* misc useful functions */
3289 * Returns a single letter to describe the given picture type \p pict_type.
3291 * @param[in] pict_type the picture type
3292 * @return A single character representing the picture type.
3294 char av_get_pict_type_char(int pict_type);
3297 * Returns codec bits per sample.
3299 * @param[in] codec_id the codec
3300 * @return Number of bits per sample or zero if unknown for the given codec.
3302 int av_get_bits_per_sample(enum CodecID codec_id);
3305 * Returns sample format bits per sample.
3307 * @param[in] sample_fmt the sample format
3308 * @return Number of bits per sample or zero if unknown for the given sample format.
3310 int av_get_bits_per_sample_format(enum SampleFormat sample_fmt);
3313 typedef struct AVCodecParserContext {
3315 struct AVCodecParser *parser;
3316 int64_t frame_offset; /* offset of the current frame */
3317 int64_t cur_offset; /* current offset
3318 (incremented by each av_parser_parse()) */
3319 int64_t next_frame_offset; /* offset of the next frame */
3321 int pict_type; /* XXX: Put it back in AVCodecContext. */
3323 * This field is used for proper frame duration computation in lavf.
3324 * It signals, how much longer the frame duration of the current frame
3325 * is compared to normal frame duration.
3327 * frame_duration = (1 + repeat_pict) * time_base
3329 * It is used by codecs like H.264 to display telecined material.
3331 int repeat_pict; /* XXX: Put it back in AVCodecContext. */
3332 int64_t pts; /* pts of the current frame */
3333 int64_t dts; /* dts of the current frame */
3338 int fetch_timestamp;
3340 #define AV_PARSER_PTS_NB 4
3341 int cur_frame_start_index;
3342 int64_t cur_frame_offset[AV_PARSER_PTS_NB];
3343 int64_t cur_frame_pts[AV_PARSER_PTS_NB];
3344 int64_t cur_frame_dts[AV_PARSER_PTS_NB];
3347 #define PARSER_FLAG_COMPLETE_FRAMES 0x0001
3349 int64_t offset; ///< byte offset from starting packet start
3350 int64_t cur_frame_end[AV_PARSER_PTS_NB];
3353 * Set by parser to 1 for key frames and 0 for non-key frames.
3354 * It is initialized to -1, so if the parser doesn't set this flag,
3355 * old-style fallback using FF_I_TYPE picture type as key frames
3361 * Time difference in stream time base units from the pts of this
3362 * packet to the point at which the output from the decoder has converged
3363 * independent from the availability of previous frames. That is, the
3364 * frames are virtually identical no matter if decoding started from
3365 * the very first frame or from this keyframe.
3366 * Is AV_NOPTS_VALUE if unknown.
3367 * This field is not the display duration of the current frame.
3369 * The purpose of this field is to allow seeking in streams that have no
3370 * keyframes in the conventional sense. It corresponds to the
3371 * recovery point SEI in H.264 and match_time_delta in NUT. It is also
3372 * essential for some types of subtitle streams to ensure that all
3373 * subtitles are correctly displayed after seeking.
3375 int64_t convergence_duration;
3377 // Timestamp generation support:
3379 * Synchronization point for start of timestamp generation.
3381 * Set to >0 for sync point, 0 for no sync point and <0 for undefined
3384 * For example, this corresponds to presence of H.264 buffering period
3390 * Offset of the current timestamp against last timestamp sync point in
3391 * units of AVCodecContext.time_base.
3393 * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
3394 * contain a valid timestamp offset.
3396 * Note that the timestamp of sync point has usually a nonzero
3397 * dts_ref_dts_delta, which refers to the previous sync point. Offset of
3398 * the next frame after timestamp sync point will be usually 1.
3400 * For example, this corresponds to H.264 cpb_removal_delay.
3402 int dts_ref_dts_delta;
3405 * Presentation delay of current frame in units of AVCodecContext.time_base.
3407 * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
3408 * contain valid non-negative timestamp delta (presentation time of a frame
3409 * must not lie in the past).
3411 * This delay represents the difference between decoding and presentation
3412 * time of the frame.
3414 * For example, this corresponds to H.264 dpb_output_delay.
3419 * Position of the packet in file.
3421 * Analogous to cur_frame_pts/dts
3423 int64_t cur_frame_pos[AV_PARSER_PTS_NB];
3426 * Byte position of currently parsed frame in stream.
3431 * Previous frame byte position.
3434 } AVCodecParserContext;
3436 typedef struct AVCodecParser {
3437 int codec_ids[5]; /* several codec IDs are permitted */
3439 int (*parser_init)(AVCodecParserContext *s);
3440 int (*parser_parse)(AVCodecParserContext *s,
3441 AVCodecContext *avctx,
3442 const uint8_t **poutbuf, int *poutbuf_size,
3443 const uint8_t *buf, int buf_size);
3444 void (*parser_close)(AVCodecParserContext *s);
3445 int (*split)(AVCodecContext *avctx, const uint8_t *buf, int buf_size);
3446 struct AVCodecParser *next;
3449 AVCodecParser *av_parser_next(AVCodecParser *c);
3451 void av_register_codec_parser(AVCodecParser *parser);
3452 AVCodecParserContext *av_parser_init(int codec_id);
3454 #if LIBAVCODEC_VERSION_MAJOR < 53
3455 attribute_deprecated
3456 int av_parser_parse(AVCodecParserContext *s,
3457 AVCodecContext *avctx,
3458 uint8_t **poutbuf, int *poutbuf_size,
3459 const uint8_t *buf, int buf_size,
3460 int64_t pts, int64_t dts);
3466 * @param s parser context.
3467 * @param avctx codec context.
3468 * @param poutbuf set to pointer to parsed buffer or NULL if not yet finished.
3469 * @param poutbuf_size set to size of parsed buffer or zero if not yet finished.
3470 * @param buf input buffer.
3471 * @param buf_size input length, to signal EOF, this should be 0 (so that the last frame can be output).
3472 * @param pts input presentation timestamp.
3473 * @param dts input decoding timestamp.
3474 * @param pos input byte position in stream.
3475 * @return the number of bytes of the input bitstream used.
3480 * len = av_parser_parse2(myparser, AVCodecContext, &data, &size,
3487 * decode_frame(data, size);
3491 int av_parser_parse2(AVCodecParserContext *s,
3492 AVCodecContext *avctx,
3493 uint8_t **poutbuf, int *poutbuf_size,
3494 const uint8_t *buf, int buf_size,
3495 int64_t pts, int64_t dts,
3498 int av_parser_change(AVCodecParserContext *s,
3499 AVCodecContext *avctx,
3500 uint8_t **poutbuf, int *poutbuf_size,
3501 const uint8_t *buf, int buf_size, int keyframe);
3502 void av_parser_close(AVCodecParserContext *s);
3505 typedef struct AVBitStreamFilterContext {
3507 struct AVBitStreamFilter *filter;
3508 AVCodecParserContext *parser;
3509 struct AVBitStreamFilterContext *next;
3510 } AVBitStreamFilterContext;
3513 typedef struct AVBitStreamFilter {
3516 int (*filter)(AVBitStreamFilterContext *bsfc,
3517 AVCodecContext *avctx, const char *args,
3518 uint8_t **poutbuf, int *poutbuf_size,
3519 const uint8_t *buf, int buf_size, int keyframe);
3520 void (*close)(AVBitStreamFilterContext *bsfc);
3521 struct AVBitStreamFilter *next;
3522 } AVBitStreamFilter;
3524 void av_register_bitstream_filter(AVBitStreamFilter *bsf);
3525 AVBitStreamFilterContext *av_bitstream_filter_init(const char *name);
3526 int av_bitstream_filter_filter(AVBitStreamFilterContext *bsfc,
3527 AVCodecContext *avctx, const char *args,
3528 uint8_t **poutbuf, int *poutbuf_size,
3529 const uint8_t *buf, int buf_size, int keyframe);
3530 void av_bitstream_filter_close(AVBitStreamFilterContext *bsf);
3532 AVBitStreamFilter *av_bitstream_filter_next(AVBitStreamFilter *f);
3537 * Reallocates the given block if it is not large enough, otherwise it
3542 void *av_fast_realloc(void *ptr, unsigned int *size, unsigned int min_size);
3545 * Allocates a buffer, reusing the given one if large enough.
3547 * Contrary to av_fast_realloc the current buffer contents might not be
3548 * preserved and on error the old buffer is freed, thus no special
3549 * handling to avoid memleaks is necessary.
3551 * @param ptr pointer to pointer to already allocated buffer, overwritten with pointer to new buffer
3552 * @param size size of the buffer *ptr points to
3553 * @param min_size minimum size of *ptr buffer after returning, *ptr will be NULL and
3554 * *size 0 if an error occurred.
3556 void av_fast_malloc(void *ptr, unsigned int *size, unsigned int min_size);
3559 * Copy image 'src' to 'dst'.
3561 void av_picture_copy(AVPicture *dst, const AVPicture *src,
3562 enum PixelFormat pix_fmt, int width, int height);
3565 * Crop image top and left side.
3567 int av_picture_crop(AVPicture *dst, const AVPicture *src,
3568 enum PixelFormat pix_fmt, int top_band, int left_band);
3573 int av_picture_pad(AVPicture *dst, const AVPicture *src, int height, int width, enum PixelFormat pix_fmt,
3574 int padtop, int padbottom, int padleft, int padright, int *color);
3576 unsigned int av_xiphlacing(unsigned char *s, unsigned int v);
3579 * Parses \p str and put in \p width_ptr and \p height_ptr the detected values.
3581 * @return 0 in case of a successful parsing, a negative value otherwise
3582 * @param[in] str the string to parse: it has to be a string in the format
3583 * <width>x<height> or a valid video frame size abbreviation.
3584 * @param[in,out] width_ptr pointer to the variable which will contain the detected
3586 * @param[in,out] height_ptr pointer to the variable which will contain the detected
3587 * frame height value
3589 int av_parse_video_frame_size(int *width_ptr, int *height_ptr, const char *str);
3592 * Parses \p str and put in \p frame_rate the detected values.
3594 * @return 0 in case of a successful parsing, a negative value otherwise
3595 * @param[in] str the string to parse: it has to be a string in the format
3596 * <frame_rate_num>/<frame_rate_den>, a float number or a valid video rate abbreviation
3597 * @param[in,out] frame_rate pointer to the AVRational which will contain the detected
3600 int av_parse_video_frame_rate(AVRational *frame_rate, const char *str);
3602 /* error handling */
3604 #define AVERROR(e) (-(e)) /**< Returns a negative error code from a POSIX error code, to return from library functions. */
3605 #define AVUNERROR(e) (-(e)) /**< Returns a POSIX error code from a library function error return value. */
3607 /* Some platforms have E* and errno already negated. */
3608 #define AVERROR(e) (e)
3609 #define AVUNERROR(e) (e)
3611 #define AVERROR_UNKNOWN AVERROR(EINVAL) /**< unknown error */
3612 #define AVERROR_IO AVERROR(EIO) /**< I/O error */
3613 #define AVERROR_NUMEXPECTED AVERROR(EDOM) /**< Number syntax expected in filename. */
3614 #define AVERROR_INVALIDDATA AVERROR(EINVAL) /**< invalid data found */
3615 #define AVERROR_NOMEM AVERROR(ENOMEM) /**< not enough memory */
3616 #define AVERROR_NOFMT AVERROR(EILSEQ) /**< unknown format */
3617 #define AVERROR_NOTSUPP AVERROR(ENOSYS) /**< Operation not supported. */
3618 #define AVERROR_NOENT AVERROR(ENOENT) /**< No such file or directory. */
3619 #define AVERROR_EOF AVERROR(EPIPE) /**< End of file. */
3620 #define AVERROR_PATCHWELCOME -MKTAG('P','A','W','E') /**< Not yet implemented in FFmpeg. Patches welcome. */
3623 * Registers the hardware accelerator \p hwaccel.
3625 void av_register_hwaccel(AVHWAccel *hwaccel);
3628 * If hwaccel is NULL, returns the first registered hardware accelerator,
3629 * if hwaccel is non-NULL, returns the next registered hardware accelerator
3630 * after hwaccel, or NULL if hwaccel is the last one.
3632 AVHWAccel *av_hwaccel_next(AVHWAccel *hwaccel);
3634 #endif /* AVCODEC_AVCODEC_H */