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 14
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,
195 /* various PCM "codecs" */
196 CODEC_ID_PCM_S16LE= 0x10000,
212 CODEC_ID_PCM_S24DAUD,
214 CODEC_ID_PCM_S16LE_PLANAR,
221 /* various ADPCM codecs */
222 CODEC_ID_ADPCM_IMA_QT= 0x11000,
223 CODEC_ID_ADPCM_IMA_WAV,
224 CODEC_ID_ADPCM_IMA_DK3,
225 CODEC_ID_ADPCM_IMA_DK4,
226 CODEC_ID_ADPCM_IMA_WS,
227 CODEC_ID_ADPCM_IMA_SMJPEG,
236 CODEC_ID_ADPCM_YAMAHA,
237 CODEC_ID_ADPCM_SBPRO_4,
238 CODEC_ID_ADPCM_SBPRO_3,
239 CODEC_ID_ADPCM_SBPRO_2,
241 CODEC_ID_ADPCM_IMA_AMV,
242 CODEC_ID_ADPCM_EA_R1,
243 CODEC_ID_ADPCM_EA_R3,
244 CODEC_ID_ADPCM_EA_R2,
245 CODEC_ID_ADPCM_IMA_EA_SEAD,
246 CODEC_ID_ADPCM_IMA_EA_EACS,
247 CODEC_ID_ADPCM_EA_XAS,
248 CODEC_ID_ADPCM_EA_MAXIS_XA,
249 CODEC_ID_ADPCM_IMA_ISS,
252 CODEC_ID_AMR_NB= 0x12000,
255 /* RealAudio codecs*/
256 CODEC_ID_RA_144= 0x13000,
259 /* various DPCM codecs */
260 CODEC_ID_ROQ_DPCM= 0x14000,
261 CODEC_ID_INTERPLAY_DPCM,
266 CODEC_ID_MP2= 0x15000,
267 CODEC_ID_MP3, ///< preferred ID for decoding MPEG audio layer 1, 2 or 3
285 CODEC_ID_WESTWOOD_SND1,
286 CODEC_ID_GSM, ///< as in Berlin toast format
294 CODEC_ID_DSICINAUDIO,
298 CODEC_ID_GSM_MS, /* as found in WAV */
307 CODEC_ID_WMALOSSLESS,
313 /* subtitle codecs */
314 CODEC_ID_DVD_SUBTITLE= 0x17000,
315 CODEC_ID_DVB_SUBTITLE,
316 CODEC_ID_TEXT, ///< raw UTF-8 text
321 /* other specific kind of codecs (generally used for attachments) */
322 CODEC_ID_TTF= 0x18000,
324 CODEC_ID_PROBE= 0x19000, ///< codec_id is not known (like CODEC_ID_NONE) but lavf should attempt to identify it
326 CODEC_ID_MPEG2TS= 0x20000, /**< _FAKE_ codec to indicate a raw MPEG-2 TS
327 * stream (only used by libavformat) */
331 CODEC_TYPE_UNKNOWN = -1,
336 CODEC_TYPE_ATTACHMENT,
341 * all in native-endian format
344 SAMPLE_FMT_NONE = -1,
345 SAMPLE_FMT_U8, ///< unsigned 8 bits
346 SAMPLE_FMT_S16, ///< signed 16 bits
347 SAMPLE_FMT_S32, ///< signed 32 bits
348 SAMPLE_FMT_FLT, ///< float
349 SAMPLE_FMT_DBL, ///< double
350 SAMPLE_FMT_NB ///< Number of sample formats. DO NOT USE if dynamically linking to libavcodec
353 /* Audio channel masks */
354 #define CH_FRONT_LEFT 0x00000001
355 #define CH_FRONT_RIGHT 0x00000002
356 #define CH_FRONT_CENTER 0x00000004
357 #define CH_LOW_FREQUENCY 0x00000008
358 #define CH_BACK_LEFT 0x00000010
359 #define CH_BACK_RIGHT 0x00000020
360 #define CH_FRONT_LEFT_OF_CENTER 0x00000040
361 #define CH_FRONT_RIGHT_OF_CENTER 0x00000080
362 #define CH_BACK_CENTER 0x00000100
363 #define CH_SIDE_LEFT 0x00000200
364 #define CH_SIDE_RIGHT 0x00000400
365 #define CH_TOP_CENTER 0x00000800
366 #define CH_TOP_FRONT_LEFT 0x00001000
367 #define CH_TOP_FRONT_CENTER 0x00002000
368 #define CH_TOP_FRONT_RIGHT 0x00004000
369 #define CH_TOP_BACK_LEFT 0x00008000
370 #define CH_TOP_BACK_CENTER 0x00010000
371 #define CH_TOP_BACK_RIGHT 0x00020000
372 #define CH_STEREO_LEFT 0x20000000 ///< Stereo downmix.
373 #define CH_STEREO_RIGHT 0x40000000 ///< See CH_STEREO_LEFT.
375 /* Audio channel convenience macros */
376 #define CH_LAYOUT_MONO (CH_FRONT_CENTER)
377 #define CH_LAYOUT_STEREO (CH_FRONT_LEFT|CH_FRONT_RIGHT)
378 #define CH_LAYOUT_SURROUND (CH_LAYOUT_STEREO|CH_FRONT_CENTER)
379 #define CH_LAYOUT_QUAD (CH_LAYOUT_STEREO|CH_BACK_LEFT|CH_BACK_RIGHT)
380 #define CH_LAYOUT_5POINT0 (CH_LAYOUT_SURROUND|CH_SIDE_LEFT|CH_SIDE_RIGHT)
381 #define CH_LAYOUT_5POINT1 (CH_LAYOUT_5POINT0|CH_LOW_FREQUENCY)
382 #define CH_LAYOUT_7POINT1 (CH_LAYOUT_5POINT1|CH_BACK_LEFT|CH_BACK_RIGHT)
383 #define CH_LAYOUT_7POINT1_WIDE (CH_LAYOUT_SURROUND|CH_LOW_FREQUENCY|\
384 CH_BACK_LEFT|CH_BACK_RIGHT|\
385 CH_FRONT_LEFT_OF_CENTER|CH_FRONT_RIGHT_OF_CENTER)
386 #define CH_LAYOUT_STEREO_DOWNMIX (CH_STEREO_LEFT|CH_STEREO_RIGHT)
389 #define AVCODEC_MAX_AUDIO_FRAME_SIZE 192000 // 1 second of 48khz 32bit audio
392 * Required number of additionally allocated bytes at the end of the input bitstream for decoding.
393 * This is mainly needed because some optimized bitstream readers read
394 * 32 or 64 bit at once and could read over the end.<br>
395 * Note: If the first 23 bits of the additional bytes are not 0, then damaged
396 * MPEG bitstreams could cause overread and segfault.
398 #define FF_INPUT_BUFFER_PADDING_SIZE 8
401 * minimum encoding buffer size
402 * Used to avoid some checks during header writing.
404 #define FF_MIN_BUFFER_SIZE 16384
408 * motion estimation type.
411 ME_ZERO = 1, ///< no search, that is use 0,0 vector whenever one is needed
415 ME_EPZS, ///< enhanced predictive zonal search
416 ME_X1, ///< reserved for experiments
417 ME_HEX, ///< hexagon based search
418 ME_UMH, ///< uneven multi-hexagon search
419 ME_ITER, ///< iterative search
420 ME_TESA, ///< transformed exhaustive search algorithm
424 /* We leave some space between them for extensions (drop some
425 * keyframes for intra-only or drop just some bidir frames). */
426 AVDISCARD_NONE =-16, ///< discard nothing
427 AVDISCARD_DEFAULT= 0, ///< discard useless packets like 0 size packets in avi
428 AVDISCARD_NONREF = 8, ///< discard all non reference
429 AVDISCARD_BIDIR = 16, ///< discard all bidirectional frames
430 AVDISCARD_NONKEY = 32, ///< discard all frames except keyframes
431 AVDISCARD_ALL = 48, ///< discard all
434 typedef struct RcOverride{
437 int qscale; // If this is 0 then quality_factor will be used instead.
438 float quality_factor;
441 #define FF_MAX_B_FRAMES 16
444 These flags can be passed in AVCodecContext.flags before initialization.
445 Note: Not everything is supported yet.
448 #define CODEC_FLAG_QSCALE 0x0002 ///< Use fixed qscale.
449 #define CODEC_FLAG_4MV 0x0004 ///< 4 MV per MB allowed / advanced prediction for H.263.
450 #define CODEC_FLAG_QPEL 0x0010 ///< Use qpel MC.
451 #define CODEC_FLAG_GMC 0x0020 ///< Use GMC.
452 #define CODEC_FLAG_MV0 0x0040 ///< Always try a MB with MV=<0,0>.
453 #define CODEC_FLAG_PART 0x0080 ///< Use data partitioning.
455 * The parent program guarantees that the input for B-frames containing
456 * streams is not written to for at least s->max_b_frames+1 frames, if
457 * this is not set the input will be copied.
459 #define CODEC_FLAG_INPUT_PRESERVED 0x0100
460 #define CODEC_FLAG_PASS1 0x0200 ///< Use internal 2pass ratecontrol in first pass mode.
461 #define CODEC_FLAG_PASS2 0x0400 ///< Use internal 2pass ratecontrol in second pass mode.
462 #define CODEC_FLAG_EXTERN_HUFF 0x1000 ///< Use external Huffman table (for MJPEG).
463 #define CODEC_FLAG_GRAY 0x2000 ///< Only decode/encode grayscale.
464 #define CODEC_FLAG_EMU_EDGE 0x4000 ///< Don't draw edges.
465 #define CODEC_FLAG_PSNR 0x8000 ///< error[?] variables will be set during encoding.
466 #define CODEC_FLAG_TRUNCATED 0x00010000 /** Input bitstream might be truncated at a random
467 location instead of only at frame boundaries. */
468 #define CODEC_FLAG_NORMALIZE_AQP 0x00020000 ///< Normalize adaptive quantization.
469 #define CODEC_FLAG_INTERLACED_DCT 0x00040000 ///< Use interlaced DCT.
470 #define CODEC_FLAG_LOW_DELAY 0x00080000 ///< Force low delay.
471 #define CODEC_FLAG_ALT_SCAN 0x00100000 ///< Use alternate scan.
472 #define CODEC_FLAG_GLOBAL_HEADER 0x00400000 ///< Place global headers in extradata instead of every keyframe.
473 #define CODEC_FLAG_BITEXACT 0x00800000 ///< Use only bitexact stuff (except (I)DCT).
474 /* Fx : Flag for h263+ extra options */
475 #define CODEC_FLAG_AC_PRED 0x01000000 ///< H.263 advanced intra coding / MPEG-4 AC prediction
476 #define CODEC_FLAG_H263P_UMV 0x02000000 ///< unlimited motion vector
477 #define CODEC_FLAG_CBP_RD 0x04000000 ///< Use rate distortion optimization for cbp.
478 #define CODEC_FLAG_QP_RD 0x08000000 ///< Use rate distortion optimization for qp selectioon.
479 #define CODEC_FLAG_H263P_AIV 0x00000008 ///< H.263 alternative inter VLC
480 #define CODEC_FLAG_OBMC 0x00000001 ///< OBMC
481 #define CODEC_FLAG_LOOP_FILTER 0x00000800 ///< loop filter
482 #define CODEC_FLAG_H263P_SLICE_STRUCT 0x10000000
483 #define CODEC_FLAG_INTERLACED_ME 0x20000000 ///< interlaced motion estimation
484 #define CODEC_FLAG_SVCD_SCAN_OFFSET 0x40000000 ///< Will reserve space for SVCD scan offset user data.
485 #define CODEC_FLAG_CLOSED_GOP 0x80000000
486 #define CODEC_FLAG2_FAST 0x00000001 ///< Allow non spec compliant speedup tricks.
487 #define CODEC_FLAG2_STRICT_GOP 0x00000002 ///< Strictly enforce GOP size.
488 #define CODEC_FLAG2_NO_OUTPUT 0x00000004 ///< Skip bitstream encoding.
489 #define CODEC_FLAG2_LOCAL_HEADER 0x00000008 ///< Place global headers at every keyframe instead of in extradata.
490 #define CODEC_FLAG2_BPYRAMID 0x00000010 ///< H.264 allow B-frames to be used as references.
491 #define CODEC_FLAG2_WPRED 0x00000020 ///< H.264 weighted biprediction for B-frames
492 #define CODEC_FLAG2_MIXED_REFS 0x00000040 ///< H.264 one reference per partition, as opposed to one reference per macroblock
493 #define CODEC_FLAG2_8X8DCT 0x00000080 ///< H.264 high profile 8x8 transform
494 #define CODEC_FLAG2_FASTPSKIP 0x00000100 ///< H.264 fast pskip
495 #define CODEC_FLAG2_AUD 0x00000200 ///< H.264 access unit delimiters
496 #define CODEC_FLAG2_BRDO 0x00000400 ///< B-frame rate-distortion optimization
497 #define CODEC_FLAG2_INTRA_VLC 0x00000800 ///< Use MPEG-2 intra VLC table.
498 #define CODEC_FLAG2_MEMC_ONLY 0x00001000 ///< Only do ME/MC (I frames -> ref, P frame -> ME+MC).
499 #define CODEC_FLAG2_DROP_FRAME_TIMECODE 0x00002000 ///< timecode is in drop frame format.
500 #define CODEC_FLAG2_SKIP_RD 0x00004000 ///< RD optimal MB level residual skipping
501 #define CODEC_FLAG2_CHUNKS 0x00008000 ///< Input bitstream might be truncated at a packet boundaries instead of only at frame boundaries.
502 #define CODEC_FLAG2_NON_LINEAR_QUANT 0x00010000 ///< Use MPEG-2 nonlinear quantizer.
503 #define CODEC_FLAG2_BIT_RESERVOIR 0x00020000 ///< Use a bit reservoir when encoding if possible
505 /* Unsupported options :
506 * Syntax Arithmetic coding (SAC)
507 * Reference Picture Selection
508 * Independent Segment Decoding */
510 /* codec capabilities */
512 #define CODEC_CAP_DRAW_HORIZ_BAND 0x0001 ///< Decoder can use draw_horiz_band callback.
514 * Codec uses get_buffer() for allocating buffers.
515 * direct rendering method 1
517 #define CODEC_CAP_DR1 0x0002
518 /* If 'parse_only' field is true, then avcodec_parse_frame() can be used. */
519 #define CODEC_CAP_PARSE_ONLY 0x0004
520 #define CODEC_CAP_TRUNCATED 0x0008
521 /* Codec can export data for HW decoding (XvMC). */
522 #define CODEC_CAP_HWACCEL 0x0010
524 * Codec has a nonzero delay and needs to be fed with NULL at the end to get the delayed data.
525 * If this is not set, the codec is guaranteed to never be fed with NULL data.
527 #define CODEC_CAP_DELAY 0x0020
529 * Codec can be fed a final frame with a smaller size.
530 * This can be used to prevent truncation of the last audio samples.
532 #define CODEC_CAP_SMALL_LAST_FRAME 0x0040
534 * Codec can export data for HW decoding (VDPAU).
536 #define CODEC_CAP_HWACCEL_VDPAU 0x0080
538 //The following defines may change, don't expect compatibility if you use them.
539 #define MB_TYPE_INTRA4x4 0x0001
540 #define MB_TYPE_INTRA16x16 0x0002 //FIXME H.264-specific
541 #define MB_TYPE_INTRA_PCM 0x0004 //FIXME H.264-specific
542 #define MB_TYPE_16x16 0x0008
543 #define MB_TYPE_16x8 0x0010
544 #define MB_TYPE_8x16 0x0020
545 #define MB_TYPE_8x8 0x0040
546 #define MB_TYPE_INTERLACED 0x0080
547 #define MB_TYPE_DIRECT2 0x0100 //FIXME
548 #define MB_TYPE_ACPRED 0x0200
549 #define MB_TYPE_GMC 0x0400
550 #define MB_TYPE_SKIP 0x0800
551 #define MB_TYPE_P0L0 0x1000
552 #define MB_TYPE_P1L0 0x2000
553 #define MB_TYPE_P0L1 0x4000
554 #define MB_TYPE_P1L1 0x8000
555 #define MB_TYPE_L0 (MB_TYPE_P0L0 | MB_TYPE_P1L0)
556 #define MB_TYPE_L1 (MB_TYPE_P0L1 | MB_TYPE_P1L1)
557 #define MB_TYPE_L0L1 (MB_TYPE_L0 | MB_TYPE_L1)
558 #define MB_TYPE_QUANT 0x00010000
559 #define MB_TYPE_CBP 0x00020000
560 //Note bits 24-31 are reserved for codec specific use (h264 ref0, mpeg1 0mv, ...)
564 * This specifies the area which should be displayed.
565 * Note there may be multiple such areas for one frame.
567 typedef struct AVPanScan{
570 * - encoding: Set by user.
571 * - decoding: Set by libavcodec.
576 * width and height in 1/16 pel
577 * - encoding: Set by user.
578 * - decoding: Set by libavcodec.
584 * position of the top left corner in 1/16 pel for up to 3 fields/frames
585 * - encoding: Set by user.
586 * - decoding: Set by libavcodec.
588 int16_t position[3][2];
591 #define FF_COMMON_FRAME \
593 * pointer to the picture planes.\
594 * This might be different from the first allocated byte\
601 * pointer to the first allocated byte of the picture. Can be used in get_buffer/release_buffer.\
602 * This isn't used by libavcodec unless the default get/release_buffer() is used.\
608 * 1 -> keyframe, 0-> not\
609 * - encoding: Set by libavcodec.\
610 * - decoding: Set by libavcodec.\
615 * Picture type of the frame, see ?_TYPE below.\
616 * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
617 * - decoding: Set by libavcodec.\
622 * presentation timestamp in time_base units (time when frame should be shown to user)\
623 * If AV_NOPTS_VALUE then frame_rate = 1/time_base will be assumed.\
624 * - encoding: MUST be set by user.\
625 * - decoding: Set by libavcodec.\
630 * picture number in bitstream order\
631 * - encoding: set by\
632 * - decoding: Set by libavcodec.\
634 int coded_picture_number;\
636 * picture number in display order\
637 * - encoding: set by\
638 * - decoding: Set by libavcodec.\
640 int display_picture_number;\
643 * quality (between 1 (good) and FF_LAMBDA_MAX (bad)) \
644 * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
645 * - decoding: Set by libavcodec.\
650 * buffer age (1->was last buffer and dint change, 2->..., ...).\
651 * Set to INT_MAX if the buffer has not been used yet.\
652 * - encoding: unused\
653 * - decoding: MUST be set by get_buffer().\
658 * is this picture used as reference\
659 * The values for this are the same as the MpegEncContext.picture_structure\
660 * variable, that is 1->top field, 2->bottom field, 3->frame/both fields.\
661 * - encoding: unused\
662 * - decoding: Set by libavcodec. (before get_buffer() call)).\
668 * - encoding: unused\
669 * - decoding: Set by libavcodec.\
671 int8_t *qscale_table;\
674 * - encoding: unused\
675 * - decoding: Set by libavcodec.\
680 * mbskip_table[mb]>=1 if MB didn't change\
681 * stride= mb_width = (width+15)>>4\
682 * - encoding: unused\
683 * - decoding: Set by libavcodec.\
685 uint8_t *mbskip_table;\
688 * motion vector table\
691 * int mv_sample_log2= 4 - motion_subsample_log2;\
692 * int mb_width= (width+15)>>4;\
693 * int mv_stride= (mb_width << mv_sample_log2) + 1;\
694 * motion_val[direction][x + y*mv_stride][0->mv_x, 1->mv_y];\
696 * - encoding: Set by user.\
697 * - decoding: Set by libavcodec.\
699 int16_t (*motion_val[2])[2];\
702 * macroblock type table\
703 * mb_type_base + mb_width + 2\
704 * - encoding: Set by user.\
705 * - decoding: Set by libavcodec.\
710 * log2 of the size of the block which a single vector in motion_val represents: \
711 * (4->16x16, 3->8x8, 2-> 4x4, 1-> 2x2)\
712 * - encoding: unused\
713 * - decoding: Set by libavcodec.\
715 uint8_t motion_subsample_log2;\
718 * for some private data of the user\
719 * - encoding: unused\
720 * - decoding: Set by user.\
726 * - encoding: Set by libavcodec. if flags&CODEC_FLAG_PSNR.\
727 * - decoding: unused\
732 * type of the buffer (to keep track of who has to deallocate data[*])\
733 * - encoding: Set by the one who allocates it.\
734 * - decoding: Set by the one who allocates it.\
735 * Note: User allocated (direct rendering) & internal buffers cannot coexist currently.\
740 * When decoding, this signals how much the picture must be delayed.\
741 * extra_delay = repeat_pict / (2*fps)\
742 * - encoding: unused\
743 * - decoding: Set by libavcodec.\
753 * The content of the picture is interlaced.\
754 * - encoding: Set by user.\
755 * - decoding: Set by libavcodec. (default 0)\
757 int interlaced_frame;\
760 * If the content is interlaced, is top field displayed first.\
761 * - encoding: Set by user.\
762 * - decoding: Set by libavcodec.\
764 int top_field_first;\
768 * - encoding: Set by user.\
769 * - decoding: Set by libavcodec.\
771 AVPanScan *pan_scan;\
774 * Tell user application that palette has changed from previous frame.\
775 * - encoding: ??? (no palette-enabled encoder yet)\
776 * - decoding: Set by libavcodec. (default 0).\
778 int palette_has_changed;\
781 * codec suggestion on buffer type if != 0\
782 * - encoding: unused\
783 * - decoding: Set by libavcodec. (before get_buffer() call)).\
789 * - encoding: unused\
790 * - decoding: Set by libavcodec.\
795 * motion referece frame index\
796 * - encoding: Set by user.\
797 * - decoding: Set by libavcodec.\
799 int8_t *ref_index[2];\
802 * reordered opaque 64bit number (generally a PTS) from AVCodecContext.reordered_opaque\
803 * output in AVFrame.reordered_opaque\
804 * - encoding: unused\
805 * - decoding: Read by user.\
807 int64_t reordered_opaque;\
810 #define FF_QSCALE_TYPE_MPEG1 0
811 #define FF_QSCALE_TYPE_MPEG2 1
812 #define FF_QSCALE_TYPE_H264 2
814 #define FF_BUFFER_TYPE_INTERNAL 1
815 #define FF_BUFFER_TYPE_USER 2 ///< direct rendering buffers (image is (de)allocated by user)
816 #define FF_BUFFER_TYPE_SHARED 4 ///< Buffer from somewhere else; don't deallocate image (data/base), all other tables are not shared.
817 #define FF_BUFFER_TYPE_COPY 8 ///< Just a (modified) copy of some other buffer, don't deallocate anything.
820 #define FF_I_TYPE 1 ///< Intra
821 #define FF_P_TYPE 2 ///< Predicted
822 #define FF_B_TYPE 3 ///< Bi-dir predicted
823 #define FF_S_TYPE 4 ///< S(GMC)-VOP MPEG4
824 #define FF_SI_TYPE 5 ///< Switching Intra
825 #define FF_SP_TYPE 6 ///< Switching Predicted
828 #define FF_BUFFER_HINTS_VALID 0x01 // Buffer hints value is meaningful (if 0 ignore).
829 #define FF_BUFFER_HINTS_READABLE 0x02 // Codec will read from buffer.
830 #define FF_BUFFER_HINTS_PRESERVE 0x04 // User must not alter buffer content.
831 #define FF_BUFFER_HINTS_REUSABLE 0x08 // Codec will reuse the buffer (update).
835 * New fields can be added to the end of FF_COMMON_FRAME with minor version
837 * Removal, reordering and changes to existing fields require a major
838 * version bump. No fields should be added into AVFrame before or after
840 * sizeof(AVFrame) must not be used outside libav*.
842 typedef struct AVFrame {
847 * main external API structure.
848 * New fields can be added to the end with minor version bumps.
849 * Removal, reordering and changes to existing fields require a major
851 * sizeof(AVCodecContext) must not be used outside libav*.
853 typedef struct AVCodecContext {
855 * information on struct for av_log
856 * - set by avcodec_alloc_context
858 const AVClass *av_class;
860 * the average bitrate
861 * - encoding: Set by user; unused for constant quantizer encoding.
862 * - decoding: Set by libavcodec. 0 or some bitrate if this info is available in the stream.
867 * number of bits the bitstream is allowed to diverge from the reference.
868 * the reference can be CBR (for CBR pass1) or VBR (for pass2)
869 * - encoding: Set by user; unused for constant quantizer encoding.
872 int bit_rate_tolerance;
876 * - encoding: Set by user.
877 * - decoding: Set by user.
882 * Some codecs need additional format info. It is stored here.
883 * If any muxer uses this then ALL demuxers/parsers AND encoders for the
884 * specific codec MUST set it correctly otherwise stream copy breaks.
885 * In general use of this field by muxers is not recommanded.
886 * - encoding: Set by libavcodec.
887 * - decoding: Set by libavcodec. (FIXME: Is this OK?)
892 * Motion estimation algorithm used for video coding.
893 * 1 (zero), 2 (full), 3 (log), 4 (phods), 5 (epzs), 6 (x1), 7 (hex),
894 * 8 (umh), 9 (iter), 10 (tesa) [7, 8, 10 are x264 specific, 9 is snow specific]
895 * - encoding: MUST be set by user.
901 * some codecs need / can use extradata like Huffman tables.
902 * mjpeg: Huffman tables
903 * rv10: additional flags
904 * mpeg4: global headers (they can be in the bitstream or here)
905 * The allocated memory should be FF_INPUT_BUFFER_PADDING_SIZE bytes larger
906 * than extradata_size to avoid prolems if it is read with the bitstream reader.
907 * The bytewise contents of extradata must not depend on the architecture or CPU endianness.
908 * - encoding: Set/allocated/freed by libavcodec.
909 * - decoding: Set/allocated/freed by user.
915 * This is the fundamental unit of time (in seconds) in terms
916 * of which frame timestamps are represented. For fixed-fps content,
917 * timebase should be 1/framerate and timestamp increments should be
919 * - encoding: MUST be set by user.
920 * - decoding: Set by libavcodec.
922 AVRational time_base;
926 * picture width / height.
927 * - encoding: MUST be set by user.
928 * - decoding: Set by libavcodec.
929 * Note: For compatibility it is possible to set this instead of
930 * coded_width/height before decoding.
934 #define FF_ASPECT_EXTENDED 15
937 * the number of pictures in a group of pictures, or 0 for intra_only
938 * - encoding: Set by user.
944 * Pixel format, see PIX_FMT_xxx.
945 * - encoding: Set by user.
946 * - decoding: Set by libavcodec.
948 enum PixelFormat pix_fmt;
951 * Frame rate emulation. If not zero, the lower layer (i.e. format handler)
952 * has to read frames at native frame rate.
953 * - encoding: Set by user.
959 * If non NULL, 'draw_horiz_band' is called by the libavcodec
960 * decoder to draw a horizontal band. It improves cache usage. Not
961 * all codecs can do that. You must check the codec capabilities
964 * - decoding: Set by user.
965 * @param height the height of the slice
966 * @param y the y position of the slice
967 * @param type 1->top field, 2->bottom field, 3->frame
968 * @param offset offset into the AVFrame.data from which the slice should be read
970 void (*draw_horiz_band)(struct AVCodecContext *s,
971 const AVFrame *src, int offset[4],
972 int y, int type, int height);
975 int sample_rate; ///< samples per second
976 int channels; ///< number of audio channels
979 * audio sample format
980 * - encoding: Set by user.
981 * - decoding: Set by libavcodec.
983 enum SampleFormat sample_fmt; ///< sample format, currently unused
985 /* The following data should not be initialized. */
987 * Samples per packet, initialized when calling 'init'.
990 int frame_number; ///< audio or video frame number
991 int real_pict_num; ///< Returns the real picture number of previous encoded frame.
994 * Number of frames the decoded output will be delayed relative to
996 * - encoding: Set by libavcodec.
1001 /* - encoding parameters */
1002 float qcompress; ///< amount of qscale change between easy & hard scenes (0.0-1.0)
1003 float qblur; ///< amount of qscale smoothing over time (0.0-1.0)
1007 * - encoding: Set by user.
1008 * - decoding: unused
1014 * - encoding: Set by user.
1015 * - decoding: unused
1020 * maximum quantizer difference between frames
1021 * - encoding: Set by user.
1022 * - decoding: unused
1027 * maximum number of B-frames between non-B-frames
1028 * Note: The output will be delayed by max_b_frames+1 relative to the input.
1029 * - encoding: Set by user.
1030 * - decoding: unused
1035 * qscale factor between IP and B-frames
1036 * If > 0 then the last P-frame quantizer will be used (q= lastp_q*factor+offset).
1037 * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
1038 * - encoding: Set by user.
1039 * - decoding: unused
1041 float b_quant_factor;
1043 /** obsolete FIXME remove */
1045 #define FF_RC_STRATEGY_XVID 1
1047 int b_frame_strategy;
1051 * - encoding: unused
1052 * - decoding: Set by user. 1-> Skip B-frames, 2-> Skip IDCT/dequant too, 5-> Skip everything except header
1053 * @deprecated Deprecated in favor of skip_idct and skip_frame.
1057 struct AVCodec *codec;
1061 int rtp_payload_size; /* The size of the RTP payload: the coder will */
1062 /* do its best to deliver a chunk with size */
1063 /* below rtp_payload_size, the chunk will start */
1064 /* with a start code on some codecs like H.263. */
1065 /* This doesn't take account of any particular */
1066 /* headers inside the transmitted RTP payload. */
1069 /* The RTP callback: This function is called */
1070 /* every time the encoder has a packet to send. */
1071 /* It depends on the encoder if the data starts */
1072 /* with a Start Code (it should). H.263 does. */
1073 /* mb_nb contains the number of macroblocks */
1074 /* encoded in the RTP payload. */
1075 void (*rtp_callback)(struct AVCodecContext *avctx, void *data, int size, int mb_nb);
1077 /* statistics, used for 2-pass encoding */
1088 * number of bits used for the previously encoded frame
1089 * - encoding: Set by libavcodec.
1090 * - decoding: unused
1095 * Private data of the user, can be used to carry app specific stuff.
1096 * - encoding: Set by user.
1097 * - decoding: Set by user.
1101 char codec_name[32];
1102 enum CodecType codec_type; /* see CODEC_TYPE_xxx */
1103 enum CodecID codec_id; /* see CODEC_ID_xxx */
1106 * fourcc (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
1107 * This is used to work around some encoder bugs.
1108 * A demuxer should set this to what is stored in the field used to identify the codec.
1109 * If there are multiple such fields in a container then the demuxer should choose the one
1110 * which maximizes the information about the used codec.
1111 * If the codec tag field in a container is larger then 32 bits then the demuxer should
1112 * remap the longer ID to 32 bits with a table or other structure. Alternatively a new
1113 * extra_codec_tag + size could be added but for this a clear advantage must be demonstrated
1115 * - encoding: Set by user, if not then the default based on codec_id will be used.
1116 * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
1118 unsigned int codec_tag;
1121 * Work around bugs in encoders which sometimes cannot be detected automatically.
1122 * - encoding: Set by user
1123 * - decoding: Set by user
1125 int workaround_bugs;
1126 #define FF_BUG_AUTODETECT 1 ///< autodetection
1127 #define FF_BUG_OLD_MSMPEG4 2
1128 #define FF_BUG_XVID_ILACE 4
1129 #define FF_BUG_UMP4 8
1130 #define FF_BUG_NO_PADDING 16
1131 #define FF_BUG_AMV 32
1132 #define FF_BUG_AC_VLC 0 ///< Will be removed, libavcodec can now handle these non-compliant files by default.
1133 #define FF_BUG_QPEL_CHROMA 64
1134 #define FF_BUG_STD_QPEL 128
1135 #define FF_BUG_QPEL_CHROMA2 256
1136 #define FF_BUG_DIRECT_BLOCKSIZE 512
1137 #define FF_BUG_EDGE 1024
1138 #define FF_BUG_HPEL_CHROMA 2048
1139 #define FF_BUG_DC_CLIP 4096
1140 #define FF_BUG_MS 8192 ///< Work around various bugs in Microsoft's broken decoders.
1141 //#define FF_BUG_FAKE_SCALABILITY 16 //Autodetection should work 100%.
1144 * luma single coefficient elimination threshold
1145 * - encoding: Set by user.
1146 * - decoding: unused
1148 int luma_elim_threshold;
1151 * chroma single coeff elimination threshold
1152 * - encoding: Set by user.
1153 * - decoding: unused
1155 int chroma_elim_threshold;
1158 * strictly follow the standard (MPEG4, ...).
1159 * - encoding: Set by user.
1160 * - decoding: Set by user.
1161 * Setting this to STRICT or higher means the encoder and decoder will
1162 * generally do stupid things. While setting it to inofficial or lower
1163 * will mean the encoder might use things that are not supported by all
1164 * spec compliant decoders. Decoders make no difference between normal,
1165 * inofficial and experimental, that is they always try to decode things
1166 * when they can unless they are explicitly asked to behave stupid
1167 * (=strictly conform to the specs)
1169 int strict_std_compliance;
1170 #define FF_COMPLIANCE_VERY_STRICT 2 ///< Strictly conform to a older more strict version of the spec or reference software.
1171 #define FF_COMPLIANCE_STRICT 1 ///< Strictly conform to all the things in the spec no matter what consequences.
1172 #define FF_COMPLIANCE_NORMAL 0
1173 #define FF_COMPLIANCE_INOFFICIAL -1 ///< Allow inofficial extensions.
1174 #define FF_COMPLIANCE_EXPERIMENTAL -2 ///< Allow nonstandardized experimental things.
1177 * qscale offset between IP and B-frames
1178 * - encoding: Set by user.
1179 * - decoding: unused
1181 float b_quant_offset;
1184 * Error recognization; higher values will detect more errors but may
1185 * misdetect some more or less valid parts as errors.
1186 * - encoding: unused
1187 * - decoding: Set by user.
1189 int error_recognition;
1190 #define FF_ER_CAREFUL 1
1191 #define FF_ER_COMPLIANT 2
1192 #define FF_ER_AGGRESSIVE 3
1193 #define FF_ER_VERY_AGGRESSIVE 4
1196 * Called at the beginning of each frame to get a buffer for it.
1197 * If pic.reference is set then the frame will be read later by libavcodec.
1198 * avcodec_align_dimensions() should be used to find the required width and
1199 * height, as they normally need to be rounded up to the next multiple of 16.
1200 * - encoding: unused
1201 * - decoding: Set by libavcodec., user can override.
1203 int (*get_buffer)(struct AVCodecContext *c, AVFrame *pic);
1206 * Called to release buffers which were allocated with get_buffer.
1207 * A released buffer can be reused in get_buffer().
1208 * pic.data[*] must be set to NULL.
1209 * - encoding: unused
1210 * - decoding: Set by libavcodec., user can override.
1212 void (*release_buffer)(struct AVCodecContext *c, AVFrame *pic);
1215 * If 1 the stream has a 1 frame delay during decoding.
1216 * - encoding: Set by libavcodec.
1217 * - decoding: Set by libavcodec.
1222 * number of bytes per packet if constant and known or 0
1223 * Used by some WAV based audio codecs.
1227 int parse_only; /* - decoding only: If true, only parsing is done
1228 (function avcodec_parse_frame()). The frame
1229 data is returned. Only MPEG codecs support this now. */
1232 * 0-> h263 quant 1-> mpeg quant
1233 * - encoding: Set by user.
1234 * - decoding: unused
1239 * pass1 encoding statistics output buffer
1240 * - encoding: Set by libavcodec.
1241 * - decoding: unused
1246 * pass2 encoding statistics input buffer
1247 * Concatenated stuff from stats_out of pass1 should be placed here.
1248 * - encoding: Allocated/set/freed by user.
1249 * - decoding: unused
1254 * ratecontrol qmin qmax limiting method
1255 * 0-> clipping, 1-> use a nice continous function to limit qscale wthin qmin/qmax.
1256 * - encoding: Set by user.
1257 * - decoding: unused
1265 * ratecontrol override, see RcOverride
1266 * - encoding: Allocated/set/freed by user.
1267 * - decoding: unused
1269 RcOverride *rc_override;
1270 int rc_override_count;
1273 * rate control equation
1274 * - encoding: Set by user
1275 * - decoding: unused
1281 * - encoding: Set by user.
1282 * - decoding: unused
1288 * - encoding: Set by user.
1289 * - decoding: unused
1294 * decoder bitstream buffer size
1295 * - encoding: Set by user.
1296 * - decoding: unused
1299 float rc_buffer_aggressivity;
1302 * qscale factor between P and I-frames
1303 * If > 0 then the last p frame quantizer will be used (q= lastp_q*factor+offset).
1304 * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
1305 * - encoding: Set by user.
1306 * - decoding: unused
1308 float i_quant_factor;
1311 * qscale offset between P and I-frames
1312 * - encoding: Set by user.
1313 * - decoding: unused
1315 float i_quant_offset;
1318 * initial complexity for pass1 ratecontrol
1319 * - encoding: Set by user.
1320 * - decoding: unused
1322 float rc_initial_cplx;
1325 * DCT algorithm, see FF_DCT_* below
1326 * - encoding: Set by user.
1327 * - decoding: unused
1330 #define FF_DCT_AUTO 0
1331 #define FF_DCT_FASTINT 1
1332 #define FF_DCT_INT 2
1333 #define FF_DCT_MMX 3
1334 #define FF_DCT_MLIB 4
1335 #define FF_DCT_ALTIVEC 5
1336 #define FF_DCT_FAAN 6
1339 * luminance masking (0-> disabled)
1340 * - encoding: Set by user.
1341 * - decoding: unused
1346 * temporary complexity masking (0-> disabled)
1347 * - encoding: Set by user.
1348 * - decoding: unused
1350 float temporal_cplx_masking;
1353 * spatial complexity masking (0-> disabled)
1354 * - encoding: Set by user.
1355 * - decoding: unused
1357 float spatial_cplx_masking;
1360 * p block masking (0-> disabled)
1361 * - encoding: Set by user.
1362 * - decoding: unused
1367 * darkness masking (0-> disabled)
1368 * - encoding: Set by user.
1369 * - decoding: unused
1374 * IDCT algorithm, see FF_IDCT_* below.
1375 * - encoding: Set by user.
1376 * - decoding: Set by user.
1379 #define FF_IDCT_AUTO 0
1380 #define FF_IDCT_INT 1
1381 #define FF_IDCT_SIMPLE 2
1382 #define FF_IDCT_SIMPLEMMX 3
1383 #define FF_IDCT_LIBMPEG2MMX 4
1384 #define FF_IDCT_PS2 5
1385 #define FF_IDCT_MLIB 6
1386 #define FF_IDCT_ARM 7
1387 #define FF_IDCT_ALTIVEC 8
1388 #define FF_IDCT_SH4 9
1389 #define FF_IDCT_SIMPLEARM 10
1390 #define FF_IDCT_H264 11
1391 #define FF_IDCT_VP3 12
1392 #define FF_IDCT_IPP 13
1393 #define FF_IDCT_XVIDMMX 14
1394 #define FF_IDCT_CAVS 15
1395 #define FF_IDCT_SIMPLEARMV5TE 16
1396 #define FF_IDCT_SIMPLEARMV6 17
1397 #define FF_IDCT_SIMPLEVIS 18
1398 #define FF_IDCT_WMV2 19
1399 #define FF_IDCT_FAAN 20
1400 #define FF_IDCT_EA 21
1401 #define FF_IDCT_SIMPLENEON 22
1402 #define FF_IDCT_SIMPLEALPHA 23
1406 * - encoding: Set by libavcodec.
1407 * - decoding: Set by user (or 0).
1411 * slice offsets in the frame in bytes
1412 * - encoding: Set/allocated by libavcodec.
1413 * - decoding: Set/allocated by user (or NULL).
1418 * error concealment flags
1419 * - encoding: unused
1420 * - decoding: Set by user.
1422 int error_concealment;
1423 #define FF_EC_GUESS_MVS 1
1424 #define FF_EC_DEBLOCK 2
1427 * dsp_mask could be add used to disable unwanted CPU features
1428 * CPU features (i.e. MMX, SSE. ...)
1430 * With the FORCE flag you may instead enable given CPU features.
1431 * (Dangerous: Usable in case of misdetection, improper usage however will
1432 * result into program crash.)
1435 #define FF_MM_FORCE 0x80000000 /* Force usage of selected flags (OR) */
1436 /* lower 16 bits - CPU features */
1437 #define FF_MM_MMX 0x0001 ///< standard MMX
1438 #define FF_MM_3DNOW 0x0004 ///< AMD 3DNOW
1439 #define FF_MM_MMXEXT 0x0002 ///< SSE integer functions or AMD MMX ext
1440 #define FF_MM_SSE 0x0008 ///< SSE functions
1441 #define FF_MM_SSE2 0x0010 ///< PIV SSE2 functions
1442 #define FF_MM_3DNOWEXT 0x0020 ///< AMD 3DNowExt
1443 #define FF_MM_SSE3 0x0040 ///< Prescott SSE3 functions
1444 #define FF_MM_SSSE3 0x0080 ///< Conroe SSSE3 functions
1445 #define FF_MM_IWMMXT 0x0100 ///< XScale IWMMXT
1446 #define FF_MM_ALTIVEC 0x0001 ///< standard AltiVec
1449 * bits per sample/pixel from the demuxer (needed for huffyuv).
1450 * - encoding: Set by libavcodec.
1451 * - decoding: Set by user.
1453 int bits_per_coded_sample;
1456 * prediction method (needed for huffyuv)
1457 * - encoding: Set by user.
1458 * - decoding: unused
1460 int prediction_method;
1461 #define FF_PRED_LEFT 0
1462 #define FF_PRED_PLANE 1
1463 #define FF_PRED_MEDIAN 2
1466 * sample aspect ratio (0 if unknown)
1467 * That is the width of a pixel divided by the height of the pixel.
1468 * Numerator and denominator must be relatively prime and smaller than 256 for some video standards.
1469 * - encoding: Set by user.
1470 * - decoding: Set by libavcodec.
1472 AVRational sample_aspect_ratio;
1475 * the picture in the bitstream
1476 * - encoding: Set by libavcodec.
1477 * - decoding: Set by libavcodec.
1479 AVFrame *coded_frame;
1483 * - encoding: Set by user.
1484 * - decoding: Set by user.
1487 #define FF_DEBUG_PICT_INFO 1
1488 #define FF_DEBUG_RC 2
1489 #define FF_DEBUG_BITSTREAM 4
1490 #define FF_DEBUG_MB_TYPE 8
1491 #define FF_DEBUG_QP 16
1492 #define FF_DEBUG_MV 32
1493 #define FF_DEBUG_DCT_COEFF 0x00000040
1494 #define FF_DEBUG_SKIP 0x00000080
1495 #define FF_DEBUG_STARTCODE 0x00000100
1496 #define FF_DEBUG_PTS 0x00000200
1497 #define FF_DEBUG_ER 0x00000400
1498 #define FF_DEBUG_MMCO 0x00000800
1499 #define FF_DEBUG_BUGS 0x00001000
1500 #define FF_DEBUG_VIS_QP 0x00002000
1501 #define FF_DEBUG_VIS_MB_TYPE 0x00004000
1502 #define FF_DEBUG_BUFFERS 0x00008000
1506 * - encoding: Set by user.
1507 * - decoding: Set by user.
1510 #define FF_DEBUG_VIS_MV_P_FOR 0x00000001 //visualize forward predicted MVs of P frames
1511 #define FF_DEBUG_VIS_MV_B_FOR 0x00000002 //visualize forward predicted MVs of B frames
1512 #define FF_DEBUG_VIS_MV_B_BACK 0x00000004 //visualize backward predicted MVs of B frames
1516 * - encoding: Set by libavcodec if flags&CODEC_FLAG_PSNR.
1517 * - decoding: unused
1522 * minimum MB quantizer
1523 * - encoding: unused
1524 * - decoding: unused
1529 * maximum MB quantizer
1530 * - encoding: unused
1531 * - decoding: unused
1536 * motion estimation comparison function
1537 * - encoding: Set by user.
1538 * - decoding: unused
1542 * subpixel motion estimation comparison function
1543 * - encoding: Set by user.
1544 * - decoding: unused
1548 * macroblock comparison function (not supported yet)
1549 * - encoding: Set by user.
1550 * - decoding: unused
1554 * interlaced DCT comparison function
1555 * - encoding: Set by user.
1556 * - decoding: unused
1559 #define FF_CMP_SAD 0
1560 #define FF_CMP_SSE 1
1561 #define FF_CMP_SATD 2
1562 #define FF_CMP_DCT 3
1563 #define FF_CMP_PSNR 4
1564 #define FF_CMP_BIT 5
1566 #define FF_CMP_ZERO 7
1567 #define FF_CMP_VSAD 8
1568 #define FF_CMP_VSSE 9
1569 #define FF_CMP_NSSE 10
1570 #define FF_CMP_W53 11
1571 #define FF_CMP_W97 12
1572 #define FF_CMP_DCTMAX 13
1573 #define FF_CMP_DCT264 14
1574 #define FF_CMP_CHROMA 256
1577 * ME diamond size & shape
1578 * - encoding: Set by user.
1579 * - decoding: unused
1584 * amount of previous MV predictors (2a+1 x 2a+1 square)
1585 * - encoding: Set by user.
1586 * - decoding: unused
1588 int last_predictor_count;
1591 * prepass for motion estimation
1592 * - encoding: Set by user.
1593 * - decoding: unused
1598 * motion estimation prepass comparison function
1599 * - encoding: Set by user.
1600 * - decoding: unused
1605 * ME prepass diamond size & shape
1606 * - encoding: Set by user.
1607 * - decoding: unused
1613 * - encoding: Set by user.
1614 * - decoding: unused
1616 int me_subpel_quality;
1619 * callback to negotiate the pixelFormat
1620 * @param fmt is the list of formats which are supported by the codec,
1621 * it is terminated by -1 as 0 is a valid format, the formats are ordered by quality.
1622 * The first is always the native one.
1623 * @return the chosen format
1624 * - encoding: unused
1625 * - decoding: Set by user, if not set the native format will be chosen.
1627 enum PixelFormat (*get_format)(struct AVCodecContext *s, const enum PixelFormat * fmt);
1630 * DTG active format information (additional aspect ratio
1631 * information only used in DVB MPEG-2 transport streams)
1634 * - encoding: unused
1635 * - decoding: Set by decoder.
1637 int dtg_active_format;
1638 #define FF_DTG_AFD_SAME 8
1639 #define FF_DTG_AFD_4_3 9
1640 #define FF_DTG_AFD_16_9 10
1641 #define FF_DTG_AFD_14_9 11
1642 #define FF_DTG_AFD_4_3_SP_14_9 13
1643 #define FF_DTG_AFD_16_9_SP_14_9 14
1644 #define FF_DTG_AFD_SP_4_3 15
1647 * maximum motion estimation search range in subpel units
1648 * If 0 then no limit.
1650 * - encoding: Set by user.
1651 * - decoding: unused
1656 * intra quantizer bias
1657 * - encoding: Set by user.
1658 * - decoding: unused
1660 int intra_quant_bias;
1661 #define FF_DEFAULT_QUANT_BIAS 999999
1664 * inter quantizer bias
1665 * - encoding: Set by user.
1666 * - decoding: unused
1668 int inter_quant_bias;
1672 * - encoding: unused
1673 * - decoding: Which clrtable should be used for 8bit RGB images.
1674 * Tables have to be stored somewhere. FIXME
1679 * internal_buffer count
1680 * Don't touch, used by libavcodec default_get_buffer().
1682 int internal_buffer_count;
1686 * Don't touch, used by libavcodec default_get_buffer().
1688 void *internal_buffer;
1690 #define FF_LAMBDA_SHIFT 7
1691 #define FF_LAMBDA_SCALE (1<<FF_LAMBDA_SHIFT)
1692 #define FF_QP2LAMBDA 118 ///< factor to convert from H.263 QP to lambda
1693 #define FF_LAMBDA_MAX (256*128-1)
1695 #define FF_QUALITY_SCALE FF_LAMBDA_SCALE //FIXME maybe remove
1697 * Global quality for codecs which cannot change it per frame.
1698 * This should be proportional to MPEG-1/2/4 qscale.
1699 * - encoding: Set by user.
1700 * - decoding: unused
1704 #define FF_CODER_TYPE_VLC 0
1705 #define FF_CODER_TYPE_AC 1
1706 #define FF_CODER_TYPE_RAW 2
1707 #define FF_CODER_TYPE_RLE 3
1708 #define FF_CODER_TYPE_DEFLATE 4
1711 * - encoding: Set by user.
1712 * - decoding: unused
1718 * - encoding: Set by user.
1719 * - decoding: unused
1725 * - encoding: unused
1726 * - decoding: Set by user.
1728 uint8_t * (*realloc)(struct AVCodecContext *s, uint8_t *buf, int buf_size);
1733 * - encoding: unused
1734 * - decoding: Set by user.
1737 #define SLICE_FLAG_CODED_ORDER 0x0001 ///< draw_horiz_band() is called in coded order instead of display
1738 #define SLICE_FLAG_ALLOW_FIELD 0x0002 ///< allow draw_horiz_band() with field slices (MPEG2 field pics)
1739 #define SLICE_FLAG_ALLOW_PLANE 0x0004 ///< allow draw_horiz_band() with 1 component at a time (SVQ1)
1742 * XVideo Motion Acceleration
1743 * - encoding: forbidden
1744 * - decoding: set by decoder
1746 int xvmc_acceleration;
1749 * macroblock decision mode
1750 * - encoding: Set by user.
1751 * - decoding: unused
1754 #define FF_MB_DECISION_SIMPLE 0 ///< uses mb_cmp
1755 #define FF_MB_DECISION_BITS 1 ///< chooses the one which needs the fewest bits
1756 #define FF_MB_DECISION_RD 2 ///< rate distortion
1759 * custom intra quantization matrix
1760 * - encoding: Set by user, can be NULL.
1761 * - decoding: Set by libavcodec.
1763 uint16_t *intra_matrix;
1766 * custom inter quantization matrix
1767 * - encoding: Set by user, can be NULL.
1768 * - decoding: Set by libavcodec.
1770 uint16_t *inter_matrix;
1773 * fourcc from the AVI stream header (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
1774 * This is used to work around some encoder bugs.
1775 * - encoding: unused
1776 * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
1778 unsigned int stream_codec_tag;
1781 * scene change detection threshold
1782 * 0 is default, larger means fewer detected scene changes.
1783 * - encoding: Set by user.
1784 * - decoding: unused
1786 int scenechange_threshold;
1789 * minimum Lagrange multipler
1790 * - encoding: Set by user.
1791 * - decoding: unused
1796 * maximum Lagrange multipler
1797 * - encoding: Set by user.
1798 * - decoding: unused
1803 * palette control structure
1804 * - encoding: ??? (no palette-enabled encoder yet)
1805 * - decoding: Set by user.
1807 struct AVPaletteControl *palctrl;
1810 * noise reduction strength
1811 * - encoding: Set by user.
1812 * - decoding: unused
1814 int noise_reduction;
1817 * Called at the beginning of a frame to get cr buffer for it.
1818 * Buffer type (size, hints) must be the same. libavcodec won't check it.
1819 * libavcodec will pass previous buffer in pic, function should return
1820 * same buffer or new buffer with old frame "painted" into it.
1821 * If pic.data[0] == NULL must behave like get_buffer().
1822 * - encoding: unused
1823 * - decoding: Set by libavcodec., user can override
1825 int (*reget_buffer)(struct AVCodecContext *c, AVFrame *pic);
1828 * Number of bits which should be loaded into the rc buffer before decoding starts.
1829 * - encoding: Set by user.
1830 * - decoding: unused
1832 int rc_initial_buffer_occupancy;
1836 * - encoding: Set by user.
1837 * - decoding: unused
1839 int inter_threshold;
1843 * - encoding: Set by user.
1844 * - decoding: Set by user.
1849 * Simulates errors in the bitstream to test error concealment.
1850 * - encoding: Set by user.
1851 * - decoding: unused
1856 * MP3 antialias algorithm, see FF_AA_* below.
1857 * - encoding: unused
1858 * - decoding: Set by user.
1861 #define FF_AA_AUTO 0
1862 #define FF_AA_FASTINT 1 //not implemented yet
1864 #define FF_AA_FLOAT 3
1866 * quantizer noise shaping
1867 * - encoding: Set by user.
1868 * - decoding: unused
1870 int quantizer_noise_shaping;
1874 * is used to decide how many independent tasks should be passed to execute()
1875 * - encoding: Set by user.
1876 * - decoding: Set by user.
1881 * The codec may call this to execute several independent things.
1882 * It will return only after finishing all tasks.
1883 * The user may replace this with some multithreaded implementation,
1884 * the default implementation will execute the parts serially.
1885 * @param count the number of things to execute
1886 * - encoding: Set by libavcodec, user can override.
1887 * - decoding: Set by libavcodec, user can override.
1889 int (*execute)(struct AVCodecContext *c, int (*func)(struct AVCodecContext *c2, void *arg), void *arg2, int *ret, int count, int size);
1893 * Can be used by execute() to store some per AVCodecContext stuff.
1894 * - encoding: set by execute()
1895 * - decoding: set by execute()
1897 void *thread_opaque;
1900 * Motion estimation threshold below which no motion estimation is
1901 * performed, but instead the user specified motion vectors are used.
1903 * - encoding: Set by user.
1904 * - decoding: unused
1909 * Macroblock threshold below which the user specified macroblock types will be used.
1910 * - encoding: Set by user.
1911 * - decoding: unused
1916 * precision of the intra DC coefficient - 8
1917 * - encoding: Set by user.
1918 * - decoding: unused
1920 int intra_dc_precision;
1923 * noise vs. sse weight for the nsse comparsion function
1924 * - encoding: Set by user.
1925 * - decoding: unused
1930 * Number of macroblock rows at the top which are skipped.
1931 * - encoding: unused
1932 * - decoding: Set by user.
1937 * Number of macroblock rows at the bottom which are skipped.
1938 * - encoding: unused
1939 * - decoding: Set by user.
1945 * - encoding: Set by user.
1946 * - decoding: Set by libavcodec.
1949 #define FF_PROFILE_UNKNOWN -99
1950 #define FF_PROFILE_AAC_MAIN 0
1951 #define FF_PROFILE_AAC_LOW 1
1952 #define FF_PROFILE_AAC_SSR 2
1953 #define FF_PROFILE_AAC_LTP 3
1957 * - encoding: Set by user.
1958 * - decoding: Set by libavcodec.
1961 #define FF_LEVEL_UNKNOWN -99
1964 * low resolution decoding, 1-> 1/2 size, 2->1/4 size
1965 * - encoding: unused
1966 * - decoding: Set by user.
1971 * Bitstream width / height, may be different from width/height if lowres
1972 * or other things are used.
1973 * - encoding: unused
1974 * - decoding: Set by user before init if known. Codec should override / dynamically change if needed.
1976 int coded_width, coded_height;
1979 * frame skip threshold
1980 * - encoding: Set by user.
1981 * - decoding: unused
1983 int frame_skip_threshold;
1987 * - encoding: Set by user.
1988 * - decoding: unused
1990 int frame_skip_factor;
1993 * frame skip exponent
1994 * - encoding: Set by user.
1995 * - decoding: unused
2000 * frame skip comparison function
2001 * - encoding: Set by user.
2002 * - decoding: unused
2007 * Border processing masking, raises the quantizer for mbs on the borders
2009 * - encoding: Set by user.
2010 * - decoding: unused
2012 float border_masking;
2015 * minimum MB lagrange multipler
2016 * - encoding: Set by user.
2017 * - decoding: unused
2022 * maximum MB lagrange multipler
2023 * - encoding: Set by user.
2024 * - decoding: unused
2030 * - encoding: Set by user.
2031 * - decoding: unused
2033 int me_penalty_compensation;
2037 * - encoding: unused
2038 * - decoding: Set by user.
2040 enum AVDiscard skip_loop_filter;
2044 * - encoding: unused
2045 * - decoding: Set by user.
2047 enum AVDiscard skip_idct;
2051 * - encoding: unused
2052 * - decoding: Set by user.
2054 enum AVDiscard skip_frame;
2058 * - encoding: Set by user.
2059 * - decoding: unused
2065 * - encoding: Set by user.
2066 * - decoding: unused
2071 * constant rate factor - quality-based VBR - values ~correspond to qps
2072 * - encoding: Set by user.
2073 * - decoding: unused
2078 * constant quantization parameter rate control method
2079 * - encoding: Set by user.
2080 * - decoding: unused
2086 * - encoding: Set by user.
2087 * - decoding: unused
2092 * number of reference frames
2093 * - encoding: Set by user.
2094 * - decoding: Set by lavc.
2099 * chroma qp offset from luma
2100 * - encoding: Set by user.
2101 * - decoding: unused
2106 * Influences how often B-frames are used.
2107 * - encoding: Set by user.
2108 * - decoding: unused
2113 * trellis RD quantization
2114 * - encoding: Set by user.
2115 * - decoding: unused
2120 * Reduce fluctuations in qp (before curve compression).
2121 * - encoding: Set by user.
2122 * - decoding: unused
2124 float complexityblur;
2127 * in-loop deblocking filter alphac0 parameter
2128 * alpha is in the range -6...6
2129 * - encoding: Set by user.
2130 * - decoding: unused
2135 * in-loop deblocking filter beta parameter
2136 * beta is in the range -6...6
2137 * - encoding: Set by user.
2138 * - decoding: unused
2143 * macroblock subpartition sizes to consider - p8x8, p4x4, b8x8, i8x8, i4x4
2144 * - encoding: Set by user.
2145 * - decoding: unused
2148 #define X264_PART_I4X4 0x001 /* Analyze i4x4 */
2149 #define X264_PART_I8X8 0x002 /* Analyze i8x8 (requires 8x8 transform) */
2150 #define X264_PART_P8X8 0x010 /* Analyze p16x8, p8x16 and p8x8 */
2151 #define X264_PART_P4X4 0x020 /* Analyze p8x4, p4x8, p4x4 */
2152 #define X264_PART_B8X8 0x100 /* Analyze b16x8, b8x16 and b8x8 */
2155 * direct MV prediction mode - 0 (none), 1 (spatial), 2 (temporal), 3 (auto)
2156 * - encoding: Set by user.
2157 * - decoding: unused
2162 * Audio cutoff bandwidth (0 means "automatic")
2163 * - encoding: Set by user.
2164 * - decoding: unused
2169 * Multiplied by qscale for each frame and added to scene_change_score.
2170 * - encoding: Set by user.
2171 * - decoding: unused
2173 int scenechange_factor;
2177 * Note: Value depends upon the compare function used for fullpel ME.
2178 * - encoding: Set by user.
2179 * - decoding: unused
2184 * Adjusts sensitivity of b_frame_strategy 1.
2185 * - encoding: Set by user.
2186 * - decoding: unused
2191 * - encoding: Set by user.
2192 * - decoding: unused
2194 int compression_level;
2195 #define FF_COMPRESSION_DEFAULT -1
2198 * Sets whether to use LPC mode - used by FLAC encoder.
2199 * - encoding: Set by user.
2200 * - decoding: unused
2205 * LPC coefficient precision - used by FLAC encoder
2206 * - encoding: Set by user.
2207 * - decoding: unused
2209 int lpc_coeff_precision;
2212 * - encoding: Set by user.
2213 * - decoding: unused
2215 int min_prediction_order;
2218 * - encoding: Set by user.
2219 * - decoding: unused
2221 int max_prediction_order;
2224 * search method for selecting prediction order
2225 * - encoding: Set by user.
2226 * - decoding: unused
2228 int prediction_order_method;
2231 * - encoding: Set by user.
2232 * - decoding: unused
2234 int min_partition_order;
2237 * - encoding: Set by user.
2238 * - decoding: unused
2240 int max_partition_order;
2243 * GOP timecode frame start number, in non drop frame format
2244 * - encoding: Set by user.
2245 * - decoding: unused
2247 int64_t timecode_frame_start;
2249 #if LIBAVCODEC_VERSION_MAJOR < 53
2251 * Decoder should decode to this many channels if it can (0 for default)
2252 * - encoding: unused
2253 * - decoding: Set by user.
2254 * @deprecated Deprecated in favor of request_channel_layout.
2256 int request_channels;
2260 * Percentage of dynamic range compression to be applied by the decoder.
2261 * The default value is 1.0, corresponding to full compression.
2262 * - encoding: unused
2263 * - decoding: Set by user.
2268 * opaque 64bit number (generally a PTS) that will be reordered and
2269 * output in AVFrame.reordered_opaque
2270 * - encoding: unused
2271 * - decoding: Set by user.
2273 int64_t reordered_opaque;
2276 * Bits per sample/pixel of internal libavcodec pixel/sample format.
2277 * This field is applicable only when sample_fmt is SAMPLE_FMT_S32.
2278 * - encoding: set by user.
2279 * - decoding: set by libavcodec.
2281 int bits_per_raw_sample;
2284 * Audio channel layout.
2285 * - encoding: set by user.
2286 * - decoding: set by libavcodec.
2288 int64_t channel_layout;
2291 * Request decoder to use this channel layout if it can (0 for default)
2292 * - encoding: unused
2293 * - decoding: Set by user.
2295 int64_t request_channel_layout;
2298 * Ratecontrol attempt to use, at maximum, <value> of what can be used without an underflow.
2299 * - encoding: Set by user.
2300 * - decoding: unused.
2302 float rc_max_available_vbv_use;
2305 * Ratecontrol attempt to use, at least, <value> times the amount needed to prevent a vbv overflow.
2306 * - encoding: Set by user.
2307 * - decoding: unused.
2309 float rc_min_vbv_overflow_use;
2315 typedef struct AVCodec {
2317 * Name of the codec implementation.
2318 * The name is globally unique among encoders and among decoders (but an
2319 * encoder and a decoder can share the same name).
2320 * This is the primary way to find a codec from the user perspective.
2323 enum CodecType type;
2326 int (*init)(AVCodecContext *);
2327 int (*encode)(AVCodecContext *, uint8_t *buf, int buf_size, void *data);
2328 int (*close)(AVCodecContext *);
2329 int (*decode)(AVCodecContext *, void *outdata, int *outdata_size,
2330 const uint8_t *buf, int buf_size);
2332 * Codec capabilities.
2336 struct AVCodec *next;
2339 * Will be called when seeking
2341 void (*flush)(AVCodecContext *);
2342 const AVRational *supported_framerates; ///< array of supported framerates, or NULL if any, array is terminated by {0,0}
2343 const enum PixelFormat *pix_fmts; ///< array of supported pixel formats, or NULL if unknown, array is terminated by -1
2345 * Descriptive name for the codec, meant to be more human readable than \p name.
2346 * You \e should use the NULL_IF_CONFIG_SMALL() macro to define it.
2348 const char *long_name;
2349 const int *supported_samplerates; ///< array of supported audio samplerates, or NULL if unknown, array is terminated by 0
2350 const enum SampleFormat *sample_fmts; ///< array of supported sample formats, or NULL if unknown, array is terminated by -1
2351 const int64_t *channel_layouts; ///< array of support channel layouts, or NULL if unknown. array is terminated by 0
2355 * four components are given, that's all.
2356 * the last component is alpha
2358 typedef struct AVPicture {
2360 int linesize[4]; ///< number of bytes per line
2365 * This structure defines a method for communicating palette changes
2366 * between and demuxer and a decoder.
2368 * @deprecated Use AVPacket to send palette changes instead.
2369 * This is totally broken.
2371 #define AVPALETTE_SIZE 1024
2372 #define AVPALETTE_COUNT 256
2373 typedef struct AVPaletteControl {
2375 /* Demuxer sets this to 1 to indicate the palette has changed;
2376 * decoder resets to 0. */
2377 int palette_changed;
2379 /* 4-byte ARGB palette entries, stored in native byte order; note that
2380 * the individual palette components should be on a 8-bit scale; if
2381 * the palette data comes from an IBM VGA native format, the component
2382 * data is probably 6 bits in size and needs to be scaled. */
2383 unsigned int palette[AVPALETTE_COUNT];
2385 } AVPaletteControl attribute_deprecated;
2387 enum AVSubtitleType {
2390 SUBTITLE_BITMAP, ///< A bitmap, pict will be set
2393 * Plain text, the text field must be set by the decoder and is
2394 * authoritative. ass and pict fields may contain approximations.
2399 * Formatted text, the ass field must be set by the decoder and is
2400 * authoritative. pict and text fields may contain approximations.
2405 typedef struct AVSubtitleRect {
2406 int x; ///< top left corner of pict, undefined when pict is not set
2407 int y; ///< top left corner of pict, undefined when pict is not set
2408 int w; ///< width of pict, undefined when pict is not set
2409 int h; ///< height of pict, undefined when pict is not set
2410 int nb_colors; ///< number of colors in pict, undefined when pict is not set
2413 * data+linesize for the bitmap of this subtitle.
2414 * can be set for text/ass as well once they where rendered
2417 enum AVSubtitleType type;
2419 char *text; ///< 0 terminated plain UTF-8 text
2422 * 0 terminated ASS/SSA compatible event line.
2423 * The pressentation of this is unaffected by the other values in this
2429 typedef struct AVSubtitle {
2430 uint16_t format; /* 0 = graphics */
2431 uint32_t start_display_time; /* relative to packet pts, in ms */
2432 uint32_t end_display_time; /* relative to packet pts, in ms */
2434 AVSubtitleRect **rects;
2440 struct ReSampleContext;
2441 struct AVResampleContext;
2443 typedef struct ReSampleContext ReSampleContext;
2445 ReSampleContext *audio_resample_init(int output_channels, int input_channels,
2446 int output_rate, int input_rate);
2447 int audio_resample(ReSampleContext *s, short *output, short *input, int nb_samples);
2448 void audio_resample_close(ReSampleContext *s);
2450 struct AVResampleContext *av_resample_init(int out_rate, int in_rate, int filter_length, int log2_phase_count, int linear, double cutoff);
2451 int av_resample(struct AVResampleContext *c, short *dst, short *src, int *consumed, int src_size, int dst_size, int update_ctx);
2452 void av_resample_compensate(struct AVResampleContext *c, int sample_delta, int compensation_distance);
2453 void av_resample_close(struct AVResampleContext *c);
2456 * Allocate memory for a picture. Call avpicture_free to free it.
2458 * @param picture the picture to be filled in
2459 * @param pix_fmt the format of the picture
2460 * @param width the width of the picture
2461 * @param height the height of the picture
2462 * @return zero if successful, a negative value if not
2464 int avpicture_alloc(AVPicture *picture, int pix_fmt, int width, int height);
2467 * Free a picture previously allocated by avpicture_alloc().
2469 * @param picture the AVPicture to be freed
2471 void avpicture_free(AVPicture *picture);
2474 * Fill in the AVPicture fields.
2475 * The fields of the given AVPicture are filled in by using the 'ptr' address
2476 * which points to the image data buffer. Depending on the specified picture
2477 * format, one or multiple image data pointers and line sizes will be set.
2478 * If a planar format is specified, several pointers will be set pointing to
2479 * the different picture planes and the line sizes of the different planes
2480 * will be stored in the lines_sizes array.
2482 * @param picture AVPicture whose fields are to be filled in
2483 * @param ptr Buffer which will contain or contains the actual image data
2484 * @param pix_fmt The format in which the picture data is stored.
2485 * @param width the width of the image in pixels
2486 * @param height the height of the image in pixels
2487 * @return size of the image data in bytes
2489 int avpicture_fill(AVPicture *picture, uint8_t *ptr,
2490 int pix_fmt, int width, int height);
2491 int avpicture_layout(const AVPicture* src, int pix_fmt, int width, int height,
2492 unsigned char *dest, int dest_size);
2495 * Calculate the size in bytes that a picture of the given width and height
2496 * would occupy if stored in the given picture format.
2498 * @param pix_fmt the given picture format
2499 * @param width the width of the image
2500 * @param height the height of the image
2501 * @return Image data size in bytes
2503 int avpicture_get_size(int pix_fmt, int width, int height);
2504 void avcodec_get_chroma_sub_sample(int pix_fmt, int *h_shift, int *v_shift);
2505 const char *avcodec_get_pix_fmt_name(int pix_fmt);
2506 void avcodec_set_dimensions(AVCodecContext *s, int width, int height);
2507 enum PixelFormat avcodec_get_pix_fmt(const char* name);
2508 unsigned int avcodec_pix_fmt_to_codec_tag(enum PixelFormat p);
2510 #define FF_LOSS_RESOLUTION 0x0001 /**< loss due to resolution change */
2511 #define FF_LOSS_DEPTH 0x0002 /**< loss due to color depth change */
2512 #define FF_LOSS_COLORSPACE 0x0004 /**< loss due to color space conversion */
2513 #define FF_LOSS_ALPHA 0x0008 /**< loss of alpha bits */
2514 #define FF_LOSS_COLORQUANT 0x0010 /**< loss due to color quantization */
2515 #define FF_LOSS_CHROMA 0x0020 /**< loss of chroma (e.g. RGB to gray conversion) */
2518 * Computes what kind of losses will occur when converting from one specific
2519 * pixel format to another.
2520 * When converting from one pixel format to another, information loss may occur.
2521 * For example, when converting from RGB24 to GRAY, the color information will
2522 * be lost. Similarly, other losses occur when converting from some formats to
2523 * other formats. These losses can involve loss of chroma, but also loss of
2524 * resolution, loss of color depth, loss due to the color space conversion, loss
2525 * of the alpha bits or loss due to color quantization.
2526 * avcodec_get_fix_fmt_loss() informs you about the various types of losses
2527 * which will occur when converting from one pixel format to another.
2529 * @param[in] dst_pix_fmt destination pixel format
2530 * @param[in] src_pix_fmt source pixel format
2531 * @param[in] has_alpha Whether the source pixel format alpha channel is used.
2532 * @return Combination of flags informing you what kind of losses will occur.
2534 int avcodec_get_pix_fmt_loss(int dst_pix_fmt, int src_pix_fmt,
2538 * Finds the best pixel format to convert to given a certain source pixel
2539 * format. When converting from one pixel format to another, information loss
2540 * may occur. For example, when converting from RGB24 to GRAY, the color
2541 * information will be lost. Similarly, other losses occur when converting from
2542 * some formats to other formats. avcodec_find_best_pix_fmt() searches which of
2543 * the given pixel formats should be used to suffer the least amount of loss.
2544 * The pixel formats from which it chooses one, are determined by the
2545 * \p pix_fmt_mask parameter.
2548 * src_pix_fmt = PIX_FMT_YUV420P;
2549 * pix_fmt_mask = (1 << PIX_FMT_YUV422P) || (1 << PIX_FMT_RGB24);
2550 * dst_pix_fmt = avcodec_find_best_pix_fmt(pix_fmt_mask, src_pix_fmt, alpha, &loss);
2553 * @param[in] pix_fmt_mask bitmask determining which pixel format to choose from
2554 * @param[in] src_pix_fmt source pixel format
2555 * @param[in] has_alpha Whether the source pixel format alpha channel is used.
2556 * @param[out] loss_ptr Combination of flags informing you what kind of losses will occur.
2557 * @return The best pixel format to convert to or -1 if none was found.
2559 int avcodec_find_best_pix_fmt(int64_t pix_fmt_mask, int src_pix_fmt,
2560 int has_alpha, int *loss_ptr);
2564 * Print in buf the string corresponding to the pixel format with
2565 * number pix_fmt, or an header if pix_fmt is negative.
2567 * @param[in] buf the buffer where to write the string
2568 * @param[in] buf_size the size of buf
2569 * @param[in] pix_fmt the number of the pixel format to print the corresponding info string, or
2570 * a negative value to print the corresponding header.
2571 * Meaningful values for obtaining a pixel format info vary from 0 to PIX_FMT_NB -1.
2573 void avcodec_pix_fmt_string (char *buf, int buf_size, int pix_fmt);
2575 #define FF_ALPHA_TRANSP 0x0001 /* image has some totally transparent pixels */
2576 #define FF_ALPHA_SEMI_TRANSP 0x0002 /* image has some transparent pixels */
2579 * Tell if an image really has transparent alpha values.
2580 * @return ored mask of FF_ALPHA_xxx constants
2582 int img_get_alpha_info(const AVPicture *src,
2583 int pix_fmt, int width, int height);
2585 /* deinterlace a picture */
2586 /* deinterlace - if not supported return -1 */
2587 int avpicture_deinterlace(AVPicture *dst, const AVPicture *src,
2588 int pix_fmt, int width, int height);
2590 /* external high level API */
2592 AVCodec *av_codec_next(AVCodec *c);
2595 * Returns the LIBAVCODEC_VERSION_INT constant.
2597 unsigned avcodec_version(void);
2600 * Initializes libavcodec.
2602 * @warning This function \e must be called before any other libavcodec
2605 void avcodec_init(void);
2608 * @deprecated Deprecated in favor of avcodec_register().
2610 attribute_deprecated void register_avcodec(AVCodec *codec);
2613 * Register the codec \p codec and initialize libavcodec.
2615 * @see avcodec_init()
2617 void avcodec_register(AVCodec *codec);
2620 * Finds a registered encoder with a matching codec ID.
2622 * @param id CodecID of the requested encoder
2623 * @return An encoder if one was found, NULL otherwise.
2625 AVCodec *avcodec_find_encoder(enum CodecID id);
2628 * Finds a registered encoder with the specified name.
2630 * @param name name of the requested encoder
2631 * @return An encoder if one was found, NULL otherwise.
2633 AVCodec *avcodec_find_encoder_by_name(const char *name);
2636 * Finds a registered decoder with a matching codec ID.
2638 * @param id CodecID of the requested decoder
2639 * @return A decoder if one was found, NULL otherwise.
2641 AVCodec *avcodec_find_decoder(enum CodecID id);
2644 * Finds a registered decoder with the specified name.
2646 * @param name name of the requested decoder
2647 * @return A decoder if one was found, NULL otherwise.
2649 AVCodec *avcodec_find_decoder_by_name(const char *name);
2650 void avcodec_string(char *buf, int buf_size, AVCodecContext *enc, int encode);
2653 * Sets the fields of the given AVCodecContext to default values.
2655 * @param s The AVCodecContext of which the fields should be set to default values.
2657 void avcodec_get_context_defaults(AVCodecContext *s);
2659 /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
2660 * we WILL change its arguments and name a few times! */
2661 void avcodec_get_context_defaults2(AVCodecContext *s, enum CodecType);
2664 * Allocates an AVCodecContext and sets its fields to default values. The
2665 * resulting struct can be deallocated by simply calling av_free().
2667 * @return An AVCodecContext filled with default values or NULL on failure.
2668 * @see avcodec_get_context_defaults
2670 AVCodecContext *avcodec_alloc_context(void);
2672 /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
2673 * we WILL change its arguments and name a few times! */
2674 AVCodecContext *avcodec_alloc_context2(enum CodecType);
2677 * Sets the fields of the given AVFrame to default values.
2679 * @param pic The AVFrame of which the fields should be set to default values.
2681 void avcodec_get_frame_defaults(AVFrame *pic);
2684 * Allocates an AVFrame and sets its fields to default values. The resulting
2685 * struct can be deallocated by simply calling av_free().
2687 * @return An AVFrame filled with default values or NULL on failure.
2688 * @see avcodec_get_frame_defaults
2690 AVFrame *avcodec_alloc_frame(void);
2692 int avcodec_default_get_buffer(AVCodecContext *s, AVFrame *pic);
2693 void avcodec_default_release_buffer(AVCodecContext *s, AVFrame *pic);
2694 int avcodec_default_reget_buffer(AVCodecContext *s, AVFrame *pic);
2695 void avcodec_align_dimensions(AVCodecContext *s, int *width, int *height);
2698 * Checks if the given dimension of a picture is valid, meaning that all
2699 * bytes of the picture can be addressed with a signed int.
2701 * @param[in] w Width of the picture.
2702 * @param[in] h Height of the picture.
2703 * @return Zero if valid, a negative value if invalid.
2705 int avcodec_check_dimensions(void *av_log_ctx, unsigned int w, unsigned int h);
2706 enum PixelFormat avcodec_default_get_format(struct AVCodecContext *s, const enum PixelFormat * fmt);
2708 int avcodec_thread_init(AVCodecContext *s, int thread_count);
2709 void avcodec_thread_free(AVCodecContext *s);
2710 int avcodec_thread_execute(AVCodecContext *s, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
2711 int avcodec_default_execute(AVCodecContext *c, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
2712 //FIXME func typedef
2715 * Initializes the AVCodecContext to use the given AVCodec. Prior to using this
2716 * function the context has to be allocated.
2718 * The functions avcodec_find_decoder_by_name(), avcodec_find_encoder_by_name(),
2719 * avcodec_find_decoder() and avcodec_find_encoder() provide an easy way for
2720 * retrieving a codec.
2722 * @warning This function is not thread safe!
2725 * avcodec_register_all();
2726 * codec = avcodec_find_decoder(CODEC_ID_H264);
2730 * context = avcodec_alloc_context();
2732 * if (avcodec_open(context, codec) < 0)
2736 * @param avctx The context which will be set up to use the given codec.
2737 * @param codec The codec to use within the context.
2738 * @return zero on success, a negative value on error
2739 * @see avcodec_alloc_context, avcodec_find_decoder, avcodec_find_encoder
2741 int avcodec_open(AVCodecContext *avctx, AVCodec *codec);
2744 * Decodes an audio frame from \p buf into \p samples.
2745 * The avcodec_decode_audio2() function decodes an audio frame from the input
2746 * buffer \p buf of size \p buf_size. To decode it, it makes use of the
2747 * audio codec which was coupled with \p avctx using avcodec_open(). The
2748 * resulting decoded frame is stored in output buffer \p samples. If no frame
2749 * could be decompressed, \p frame_size_ptr is zero. Otherwise, it is the
2750 * decompressed frame size in \e bytes.
2752 * @warning You \e must set \p frame_size_ptr to the allocated size of the
2753 * output buffer before calling avcodec_decode_audio2().
2755 * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
2756 * the actual read bytes because some optimized bitstream readers read 32 or 64
2757 * bits at once and could read over the end.
2759 * @warning The end of the input buffer \p buf should be set to 0 to ensure that
2760 * no overreading happens for damaged MPEG streams.
2762 * @note You might have to align the input buffer \p buf and output buffer \p
2763 * samples. The alignment requirements depend on the CPU: On some CPUs it isn't
2764 * necessary at all, on others it won't work at all if not aligned and on others
2765 * it will work but it will have an impact on performance. In practice, the
2766 * bitstream should have 4 byte alignment at minimum and all sample data should
2767 * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
2768 * the linesize is not a multiple of 16 then there's no sense in aligning the
2769 * start of the buffer to 16.
2771 * @param avctx the codec context
2772 * @param[out] samples the output buffer
2773 * @param[in,out] frame_size_ptr the output buffer size in bytes
2774 * @param[in] buf the input buffer
2775 * @param[in] buf_size the input buffer size in bytes
2776 * @return On error a negative value is returned, otherwise the number of bytes
2777 * used or zero if no frame could be decompressed.
2779 int avcodec_decode_audio2(AVCodecContext *avctx, int16_t *samples,
2780 int *frame_size_ptr,
2781 const uint8_t *buf, int buf_size);
2784 * Decodes a video frame from \p buf into \p picture.
2785 * The avcodec_decode_video() function decodes a video frame from the input
2786 * buffer \p buf of size \p buf_size. To decode it, it makes use of the
2787 * video codec which was coupled with \p avctx using avcodec_open(). The
2788 * resulting decoded frame is stored in \p picture.
2790 * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
2791 * the actual read bytes because some optimized bitstream readers read 32 or 64
2792 * bits at once and could read over the end.
2794 * @warning The end of the input buffer \p buf should be set to 0 to ensure that
2795 * no overreading happens for damaged MPEG streams.
2797 * @note You might have to align the input buffer \p buf and output buffer \p
2798 * samples. The alignment requirements depend on the CPU: on some CPUs it isn't
2799 * necessary at all, on others it won't work at all if not aligned and on others
2800 * it will work but it will have an impact on performance. In practice, the
2801 * bitstream should have 4 byte alignment at minimum and all sample data should
2802 * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
2803 * the linesize is not a multiple of 16 then there's no sense in aligning the
2804 * start of the buffer to 16.
2806 * @note Some codecs have a delay between input and output, these need to be
2807 * feeded with buf=NULL, buf_size=0 at the end to return the remaining frames.
2809 * @param avctx the codec context
2810 * @param[out] picture The AVFrame in which the decoded video frame will be stored.
2811 * @param[in] buf the input buffer
2812 * @param[in] buf_size the size of the input buffer in bytes
2813 * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
2814 * @return On error a negative value is returned, otherwise the number of bytes
2815 * used or zero if no frame could be decompressed.
2817 int avcodec_decode_video(AVCodecContext *avctx, AVFrame *picture,
2818 int *got_picture_ptr,
2819 const uint8_t *buf, int buf_size);
2821 /* Decode a subtitle message. Return -1 if error, otherwise return the
2822 * number of bytes used. If no subtitle could be decompressed,
2823 * got_sub_ptr is zero. Otherwise, the subtitle is stored in *sub. */
2824 int avcodec_decode_subtitle(AVCodecContext *avctx, AVSubtitle *sub,
2826 const uint8_t *buf, int buf_size);
2827 int avcodec_parse_frame(AVCodecContext *avctx, uint8_t **pdata,
2829 uint8_t *buf, int buf_size);
2832 * Encodes an audio frame from \p samples into \p buf.
2833 * The avcodec_encode_audio() function encodes an audio frame from the input
2834 * buffer \p samples. To encode it, it makes use of the audio codec which was
2835 * coupled with \p avctx using avcodec_open(). The resulting encoded frame is
2836 * stored in output buffer \p buf.
2838 * @note The output buffer should be at least \c FF_MIN_BUFFER_SIZE bytes large.
2840 * @param avctx the codec context
2841 * @param[out] buf the output buffer
2842 * @param[in] buf_size the output buffer size
2843 * @param[in] samples the input buffer containing the samples
2844 * The number of samples read from this buffer is frame_size*channels,
2845 * both of which are defined in \p avctx.
2846 * For PCM audio the number of samples read from \p samples is equal to
2847 * \p buf_size * input_sample_size / output_sample_size.
2848 * @return On error a negative value is returned, on success zero or the number
2849 * of bytes used to encode the data read from the input buffer.
2851 int avcodec_encode_audio(AVCodecContext *avctx, uint8_t *buf, int buf_size,
2852 const short *samples);
2855 * Encodes a video frame from \p pict into \p buf.
2856 * The avcodec_encode_video() function encodes a video frame from the input
2857 * \p pict. To encode it, it makes use of the video codec which was coupled with
2858 * \p avctx using avcodec_open(). The resulting encoded bytes representing the
2859 * frame are stored in the output buffer \p buf. The input picture should be
2860 * stored using a specific format, namely \c avctx.pix_fmt.
2862 * @param avctx the codec context
2863 * @param[out] buf the output buffer for the bitstream of encoded frame
2864 * @param[in] buf_size the size of the output buffer in bytes
2865 * @param[in] pict the input picture to encode
2866 * @return On error a negative value is returned, on success zero or the number
2867 * of bytes used from the output buffer.
2869 int avcodec_encode_video(AVCodecContext *avctx, uint8_t *buf, int buf_size,
2870 const AVFrame *pict);
2871 int avcodec_encode_subtitle(AVCodecContext *avctx, uint8_t *buf, int buf_size,
2872 const AVSubtitle *sub);
2874 int avcodec_close(AVCodecContext *avctx);
2877 * Register all the codecs, parsers and bitstream filters which were enabled at
2878 * configuration time. If you do not call this function you can select exactly
2879 * which formats you want to support, by using the individual registration
2882 * @see avcodec_register
2883 * @see av_register_codec_parser
2884 * @see av_register_bitstream_filter
2886 void avcodec_register_all(void);
2889 * Flush buffers, should be called when seeking or when switching to a different stream.
2891 void avcodec_flush_buffers(AVCodecContext *avctx);
2893 void avcodec_default_free_buffers(AVCodecContext *s);
2895 /* misc useful functions */
2898 * Returns a single letter to describe the given picture type \p pict_type.
2900 * @param[in] pict_type the picture type
2901 * @return A single character representing the picture type.
2903 char av_get_pict_type_char(int pict_type);
2906 * Returns codec bits per sample.
2908 * @param[in] codec_id the codec
2909 * @return Number of bits per sample or zero if unknown for the given codec.
2911 int av_get_bits_per_sample(enum CodecID codec_id);
2914 * Returns sample format bits per sample.
2916 * @param[in] sample_fmt the sample format
2917 * @return Number of bits per sample or zero if unknown for the given sample format.
2919 int av_get_bits_per_sample_format(enum SampleFormat sample_fmt);
2922 typedef struct AVCodecParserContext {
2924 struct AVCodecParser *parser;
2925 int64_t frame_offset; /* offset of the current frame */
2926 int64_t cur_offset; /* current offset
2927 (incremented by each av_parser_parse()) */
2928 int64_t next_frame_offset; /* offset of the next frame */
2930 int pict_type; /* XXX: Put it back in AVCodecContext. */
2931 int repeat_pict; /* XXX: Put it back in AVCodecContext. */
2932 int64_t pts; /* pts of the current frame */
2933 int64_t dts; /* dts of the current frame */
2938 int fetch_timestamp;
2940 #define AV_PARSER_PTS_NB 4
2941 int cur_frame_start_index;
2942 int64_t cur_frame_offset[AV_PARSER_PTS_NB];
2943 int64_t cur_frame_pts[AV_PARSER_PTS_NB];
2944 int64_t cur_frame_dts[AV_PARSER_PTS_NB];
2947 #define PARSER_FLAG_COMPLETE_FRAMES 0x0001
2949 int64_t offset; ///< byte offset from starting packet start
2950 int64_t cur_frame_end[AV_PARSER_PTS_NB];
2951 } AVCodecParserContext;
2953 typedef struct AVCodecParser {
2954 int codec_ids[5]; /* several codec IDs are permitted */
2956 int (*parser_init)(AVCodecParserContext *s);
2957 int (*parser_parse)(AVCodecParserContext *s,
2958 AVCodecContext *avctx,
2959 const uint8_t **poutbuf, int *poutbuf_size,
2960 const uint8_t *buf, int buf_size);
2961 void (*parser_close)(AVCodecParserContext *s);
2962 int (*split)(AVCodecContext *avctx, const uint8_t *buf, int buf_size);
2963 struct AVCodecParser *next;
2966 AVCodecParser *av_parser_next(AVCodecParser *c);
2968 void av_register_codec_parser(AVCodecParser *parser);
2969 AVCodecParserContext *av_parser_init(int codec_id);
2970 int av_parser_parse(AVCodecParserContext *s,
2971 AVCodecContext *avctx,
2972 uint8_t **poutbuf, int *poutbuf_size,
2973 const uint8_t *buf, int buf_size,
2974 int64_t pts, int64_t dts);
2975 int av_parser_change(AVCodecParserContext *s,
2976 AVCodecContext *avctx,
2977 uint8_t **poutbuf, int *poutbuf_size,
2978 const uint8_t *buf, int buf_size, int keyframe);
2979 void av_parser_close(AVCodecParserContext *s);
2982 typedef struct AVBitStreamFilterContext {
2984 struct AVBitStreamFilter *filter;
2985 AVCodecParserContext *parser;
2986 struct AVBitStreamFilterContext *next;
2987 } AVBitStreamFilterContext;
2990 typedef struct AVBitStreamFilter {
2993 int (*filter)(AVBitStreamFilterContext *bsfc,
2994 AVCodecContext *avctx, const char *args,
2995 uint8_t **poutbuf, int *poutbuf_size,
2996 const uint8_t *buf, int buf_size, int keyframe);
2997 void (*close)(AVBitStreamFilterContext *bsfc);
2998 struct AVBitStreamFilter *next;
2999 } AVBitStreamFilter;
3001 void av_register_bitstream_filter(AVBitStreamFilter *bsf);
3002 AVBitStreamFilterContext *av_bitstream_filter_init(const char *name);
3003 int av_bitstream_filter_filter(AVBitStreamFilterContext *bsfc,
3004 AVCodecContext *avctx, const char *args,
3005 uint8_t **poutbuf, int *poutbuf_size,
3006 const uint8_t *buf, int buf_size, int keyframe);
3007 void av_bitstream_filter_close(AVBitStreamFilterContext *bsf);
3009 AVBitStreamFilter *av_bitstream_filter_next(AVBitStreamFilter *f);
3014 * Reallocates the given block if it is not large enough, otherwise it
3019 void *av_fast_realloc(void *ptr, unsigned int *size, unsigned int min_size);
3022 * Copy image 'src' to 'dst'.
3024 void av_picture_copy(AVPicture *dst, const AVPicture *src,
3025 int pix_fmt, int width, int height);
3028 * Crop image top and left side.
3030 int av_picture_crop(AVPicture *dst, const AVPicture *src,
3031 int pix_fmt, int top_band, int left_band);
3036 int av_picture_pad(AVPicture *dst, const AVPicture *src, int height, int width, int pix_fmt,
3037 int padtop, int padbottom, int padleft, int padright, int *color);
3039 unsigned int av_xiphlacing(unsigned char *s, unsigned int v);
3042 * Parses \p str and put in \p width_ptr and \p height_ptr the detected values.
3044 * @return 0 in case of a successful parsing, a negative value otherwise
3045 * @param[in] str the string to parse: it has to be a string in the format
3046 * <width>x<height> or a valid video frame size abbreviation.
3047 * @param[in,out] width_ptr pointer to the variable which will contain the detected
3049 * @param[in,out] height_ptr pointer to the variable which will contain the detected
3050 * frame height value
3052 int av_parse_video_frame_size(int *width_ptr, int *height_ptr, const char *str);
3055 * Parses \p str and put in \p frame_rate the detected values.
3057 * @return 0 in case of a successful parsing, a negative value otherwise
3058 * @param[in] str the string to parse: it has to be a string in the format
3059 * <frame_rate_nom>/<frame_rate_den>, a float number or a valid video rate abbreviation
3060 * @param[in,out] frame_rate pointer to the AVRational which will contain the detected
3063 int av_parse_video_frame_rate(AVRational *frame_rate, const char *str);
3065 /* error handling */
3067 #define AVERROR(e) (-(e)) /**< Returns a negative error code from a POSIX error code, to return from library functions. */
3068 #define AVUNERROR(e) (-(e)) /**< Returns a POSIX error code from a library function error return value. */
3070 /* Some platforms have E* and errno already negated. */
3071 #define AVERROR(e) (e)
3072 #define AVUNERROR(e) (e)
3074 #define AVERROR_UNKNOWN AVERROR(EINVAL) /**< unknown error */
3075 #define AVERROR_IO AVERROR(EIO) /**< I/O error */
3076 #define AVERROR_NUMEXPECTED AVERROR(EDOM) /**< Number syntax expected in filename. */
3077 #define AVERROR_INVALIDDATA AVERROR(EINVAL) /**< invalid data found */
3078 #define AVERROR_NOMEM AVERROR(ENOMEM) /**< not enough memory */
3079 #define AVERROR_NOFMT AVERROR(EILSEQ) /**< unknown format */
3080 #define AVERROR_NOTSUPP AVERROR(ENOSYS) /**< Operation not supported. */
3081 #define AVERROR_NOENT AVERROR(ENOENT) /**< No such file or directory. */
3082 #define AVERROR_PATCHWELCOME -MKTAG('P','A','W','E') /**< Not yet implemented in FFmpeg. Patches welcome. */
3084 #endif /* AVCODEC_AVCODEC_H */