You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

3613 lines
115KB

  1. /*
  2. * copyright (c) 2001 Fabrice Bellard
  3. *
  4. * This file is part of FFmpeg.
  5. *
  6. * FFmpeg is free software; you can redistribute it and/or
  7. * modify it under the terms of the GNU Lesser General Public
  8. * License as published by the Free Software Foundation; either
  9. * version 2.1 of the License, or (at your option) any later version.
  10. *
  11. * FFmpeg is distributed in the hope that it will be useful,
  12. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  14. * Lesser General Public License for more details.
  15. *
  16. * You should have received a copy of the GNU Lesser General Public
  17. * License along with FFmpeg; if not, write to the Free Software
  18. * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  19. */
  20. #ifndef AVCODEC_AVCODEC_H
  21. #define AVCODEC_AVCODEC_H
  22. /**
  23. * @file libavcodec/avcodec.h
  24. * external API header
  25. */
  26. #include <errno.h>
  27. #include "libavutil/avutil.h"
  28. #define LIBAVCODEC_VERSION_MAJOR 52
  29. #define LIBAVCODEC_VERSION_MINOR 25
  30. #define LIBAVCODEC_VERSION_MICRO 0
  31. #define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
  32. LIBAVCODEC_VERSION_MINOR, \
  33. LIBAVCODEC_VERSION_MICRO)
  34. #define LIBAVCODEC_VERSION AV_VERSION(LIBAVCODEC_VERSION_MAJOR, \
  35. LIBAVCODEC_VERSION_MINOR, \
  36. LIBAVCODEC_VERSION_MICRO)
  37. #define LIBAVCODEC_BUILD LIBAVCODEC_VERSION_INT
  38. #define LIBAVCODEC_IDENT "Lavc" AV_STRINGIFY(LIBAVCODEC_VERSION)
  39. #define AV_NOPTS_VALUE INT64_C(0x8000000000000000)
  40. #define AV_TIME_BASE 1000000
  41. #define AV_TIME_BASE_Q (AVRational){1, AV_TIME_BASE}
  42. /**
  43. * Identifies the syntax and semantics of the bitstream.
  44. * The principle is roughly:
  45. * Two decoders with the same ID can decode the same streams.
  46. * Two encoders with the same ID can encode compatible streams.
  47. * There may be slight deviations from the principle due to implementation
  48. * details.
  49. *
  50. * If you add a codec ID to this list, add it so that
  51. * 1. no value of a existing codec ID changes (that would break ABI),
  52. * 2. it is as close as possible to similar codecs.
  53. */
  54. enum CodecID {
  55. CODEC_ID_NONE,
  56. /* video codecs */
  57. CODEC_ID_MPEG1VIDEO,
  58. CODEC_ID_MPEG2VIDEO, ///< preferred ID for MPEG-1/2 video decoding
  59. CODEC_ID_MPEG2VIDEO_XVMC,
  60. CODEC_ID_H261,
  61. CODEC_ID_H263,
  62. CODEC_ID_RV10,
  63. CODEC_ID_RV20,
  64. CODEC_ID_MJPEG,
  65. CODEC_ID_MJPEGB,
  66. CODEC_ID_LJPEG,
  67. CODEC_ID_SP5X,
  68. CODEC_ID_JPEGLS,
  69. CODEC_ID_MPEG4,
  70. CODEC_ID_RAWVIDEO,
  71. CODEC_ID_MSMPEG4V1,
  72. CODEC_ID_MSMPEG4V2,
  73. CODEC_ID_MSMPEG4V3,
  74. CODEC_ID_WMV1,
  75. CODEC_ID_WMV2,
  76. CODEC_ID_H263P,
  77. CODEC_ID_H263I,
  78. CODEC_ID_FLV1,
  79. CODEC_ID_SVQ1,
  80. CODEC_ID_SVQ3,
  81. CODEC_ID_DVVIDEO,
  82. CODEC_ID_HUFFYUV,
  83. CODEC_ID_CYUV,
  84. CODEC_ID_H264,
  85. CODEC_ID_INDEO3,
  86. CODEC_ID_VP3,
  87. CODEC_ID_THEORA,
  88. CODEC_ID_ASV1,
  89. CODEC_ID_ASV2,
  90. CODEC_ID_FFV1,
  91. CODEC_ID_4XM,
  92. CODEC_ID_VCR1,
  93. CODEC_ID_CLJR,
  94. CODEC_ID_MDEC,
  95. CODEC_ID_ROQ,
  96. CODEC_ID_INTERPLAY_VIDEO,
  97. CODEC_ID_XAN_WC3,
  98. CODEC_ID_XAN_WC4,
  99. CODEC_ID_RPZA,
  100. CODEC_ID_CINEPAK,
  101. CODEC_ID_WS_VQA,
  102. CODEC_ID_MSRLE,
  103. CODEC_ID_MSVIDEO1,
  104. CODEC_ID_IDCIN,
  105. CODEC_ID_8BPS,
  106. CODEC_ID_SMC,
  107. CODEC_ID_FLIC,
  108. CODEC_ID_TRUEMOTION1,
  109. CODEC_ID_VMDVIDEO,
  110. CODEC_ID_MSZH,
  111. CODEC_ID_ZLIB,
  112. CODEC_ID_QTRLE,
  113. CODEC_ID_SNOW,
  114. CODEC_ID_TSCC,
  115. CODEC_ID_ULTI,
  116. CODEC_ID_QDRAW,
  117. CODEC_ID_VIXL,
  118. CODEC_ID_QPEG,
  119. CODEC_ID_XVID,
  120. CODEC_ID_PNG,
  121. CODEC_ID_PPM,
  122. CODEC_ID_PBM,
  123. CODEC_ID_PGM,
  124. CODEC_ID_PGMYUV,
  125. CODEC_ID_PAM,
  126. CODEC_ID_FFVHUFF,
  127. CODEC_ID_RV30,
  128. CODEC_ID_RV40,
  129. CODEC_ID_VC1,
  130. CODEC_ID_WMV3,
  131. CODEC_ID_LOCO,
  132. CODEC_ID_WNV1,
  133. CODEC_ID_AASC,
  134. CODEC_ID_INDEO2,
  135. CODEC_ID_FRAPS,
  136. CODEC_ID_TRUEMOTION2,
  137. CODEC_ID_BMP,
  138. CODEC_ID_CSCD,
  139. CODEC_ID_MMVIDEO,
  140. CODEC_ID_ZMBV,
  141. CODEC_ID_AVS,
  142. CODEC_ID_SMACKVIDEO,
  143. CODEC_ID_NUV,
  144. CODEC_ID_KMVC,
  145. CODEC_ID_FLASHSV,
  146. CODEC_ID_CAVS,
  147. CODEC_ID_JPEG2000,
  148. CODEC_ID_VMNC,
  149. CODEC_ID_VP5,
  150. CODEC_ID_VP6,
  151. CODEC_ID_VP6F,
  152. CODEC_ID_TARGA,
  153. CODEC_ID_DSICINVIDEO,
  154. CODEC_ID_TIERTEXSEQVIDEO,
  155. CODEC_ID_TIFF,
  156. CODEC_ID_GIF,
  157. CODEC_ID_FFH264,
  158. CODEC_ID_DXA,
  159. CODEC_ID_DNXHD,
  160. CODEC_ID_THP,
  161. CODEC_ID_SGI,
  162. CODEC_ID_C93,
  163. CODEC_ID_BETHSOFTVID,
  164. CODEC_ID_PTX,
  165. CODEC_ID_TXD,
  166. CODEC_ID_VP6A,
  167. CODEC_ID_AMV,
  168. CODEC_ID_VB,
  169. CODEC_ID_PCX,
  170. CODEC_ID_SUNRAST,
  171. CODEC_ID_INDEO4,
  172. CODEC_ID_INDEO5,
  173. CODEC_ID_MIMIC,
  174. CODEC_ID_RL2,
  175. CODEC_ID_8SVX_EXP,
  176. CODEC_ID_8SVX_FIB,
  177. CODEC_ID_ESCAPE124,
  178. CODEC_ID_DIRAC,
  179. CODEC_ID_BFI,
  180. CODEC_ID_CMV,
  181. CODEC_ID_MOTIONPIXELS,
  182. CODEC_ID_TGV,
  183. CODEC_ID_TGQ,
  184. CODEC_ID_TQI,
  185. CODEC_ID_AURA,
  186. CODEC_ID_AURA2,
  187. /* various PCM "codecs" */
  188. CODEC_ID_PCM_S16LE= 0x10000,
  189. CODEC_ID_PCM_S16BE,
  190. CODEC_ID_PCM_U16LE,
  191. CODEC_ID_PCM_U16BE,
  192. CODEC_ID_PCM_S8,
  193. CODEC_ID_PCM_U8,
  194. CODEC_ID_PCM_MULAW,
  195. CODEC_ID_PCM_ALAW,
  196. CODEC_ID_PCM_S32LE,
  197. CODEC_ID_PCM_S32BE,
  198. CODEC_ID_PCM_U32LE,
  199. CODEC_ID_PCM_U32BE,
  200. CODEC_ID_PCM_S24LE,
  201. CODEC_ID_PCM_S24BE,
  202. CODEC_ID_PCM_U24LE,
  203. CODEC_ID_PCM_U24BE,
  204. CODEC_ID_PCM_S24DAUD,
  205. CODEC_ID_PCM_ZORK,
  206. CODEC_ID_PCM_S16LE_PLANAR,
  207. CODEC_ID_PCM_DVD,
  208. CODEC_ID_PCM_F32BE,
  209. CODEC_ID_PCM_F32LE,
  210. CODEC_ID_PCM_F64BE,
  211. CODEC_ID_PCM_F64LE,
  212. /* various ADPCM codecs */
  213. CODEC_ID_ADPCM_IMA_QT= 0x11000,
  214. CODEC_ID_ADPCM_IMA_WAV,
  215. CODEC_ID_ADPCM_IMA_DK3,
  216. CODEC_ID_ADPCM_IMA_DK4,
  217. CODEC_ID_ADPCM_IMA_WS,
  218. CODEC_ID_ADPCM_IMA_SMJPEG,
  219. CODEC_ID_ADPCM_MS,
  220. CODEC_ID_ADPCM_4XM,
  221. CODEC_ID_ADPCM_XA,
  222. CODEC_ID_ADPCM_ADX,
  223. CODEC_ID_ADPCM_EA,
  224. CODEC_ID_ADPCM_G726,
  225. CODEC_ID_ADPCM_CT,
  226. CODEC_ID_ADPCM_SWF,
  227. CODEC_ID_ADPCM_YAMAHA,
  228. CODEC_ID_ADPCM_SBPRO_4,
  229. CODEC_ID_ADPCM_SBPRO_3,
  230. CODEC_ID_ADPCM_SBPRO_2,
  231. CODEC_ID_ADPCM_THP,
  232. CODEC_ID_ADPCM_IMA_AMV,
  233. CODEC_ID_ADPCM_EA_R1,
  234. CODEC_ID_ADPCM_EA_R3,
  235. CODEC_ID_ADPCM_EA_R2,
  236. CODEC_ID_ADPCM_IMA_EA_SEAD,
  237. CODEC_ID_ADPCM_IMA_EA_EACS,
  238. CODEC_ID_ADPCM_EA_XAS,
  239. CODEC_ID_ADPCM_EA_MAXIS_XA,
  240. CODEC_ID_ADPCM_IMA_ISS,
  241. /* AMR */
  242. CODEC_ID_AMR_NB= 0x12000,
  243. CODEC_ID_AMR_WB,
  244. /* RealAudio codecs*/
  245. CODEC_ID_RA_144= 0x13000,
  246. CODEC_ID_RA_288,
  247. /* various DPCM codecs */
  248. CODEC_ID_ROQ_DPCM= 0x14000,
  249. CODEC_ID_INTERPLAY_DPCM,
  250. CODEC_ID_XAN_DPCM,
  251. CODEC_ID_SOL_DPCM,
  252. /* audio codecs */
  253. CODEC_ID_MP2= 0x15000,
  254. CODEC_ID_MP3, ///< preferred ID for decoding MPEG audio layer 1, 2 or 3
  255. CODEC_ID_AAC,
  256. CODEC_ID_AC3,
  257. CODEC_ID_DTS,
  258. CODEC_ID_VORBIS,
  259. CODEC_ID_DVAUDIO,
  260. CODEC_ID_WMAV1,
  261. CODEC_ID_WMAV2,
  262. CODEC_ID_MACE3,
  263. CODEC_ID_MACE6,
  264. CODEC_ID_VMDAUDIO,
  265. CODEC_ID_SONIC,
  266. CODEC_ID_SONIC_LS,
  267. CODEC_ID_FLAC,
  268. CODEC_ID_MP3ADU,
  269. CODEC_ID_MP3ON4,
  270. CODEC_ID_SHORTEN,
  271. CODEC_ID_ALAC,
  272. CODEC_ID_WESTWOOD_SND1,
  273. CODEC_ID_GSM, ///< as in Berlin toast format
  274. CODEC_ID_QDM2,
  275. CODEC_ID_COOK,
  276. CODEC_ID_TRUESPEECH,
  277. CODEC_ID_TTA,
  278. CODEC_ID_SMACKAUDIO,
  279. CODEC_ID_QCELP,
  280. CODEC_ID_WAVPACK,
  281. CODEC_ID_DSICINAUDIO,
  282. CODEC_ID_IMC,
  283. CODEC_ID_MUSEPACK7,
  284. CODEC_ID_MLP,
  285. CODEC_ID_GSM_MS, /* as found in WAV */
  286. CODEC_ID_ATRAC3,
  287. CODEC_ID_VOXWARE,
  288. CODEC_ID_APE,
  289. CODEC_ID_NELLYMOSER,
  290. CODEC_ID_MUSEPACK8,
  291. CODEC_ID_SPEEX,
  292. CODEC_ID_WMAVOICE,
  293. CODEC_ID_WMAPRO,
  294. CODEC_ID_WMALOSSLESS,
  295. CODEC_ID_ATRAC3P,
  296. CODEC_ID_EAC3,
  297. CODEC_ID_SIPR,
  298. CODEC_ID_MP1,
  299. CODEC_ID_TWINVQ,
  300. CODEC_ID_TRUEHD,
  301. /* subtitle codecs */
  302. CODEC_ID_DVD_SUBTITLE= 0x17000,
  303. CODEC_ID_DVB_SUBTITLE,
  304. CODEC_ID_TEXT, ///< raw UTF-8 text
  305. CODEC_ID_XSUB,
  306. CODEC_ID_SSA,
  307. CODEC_ID_MOV_TEXT,
  308. /* other specific kind of codecs (generally used for attachments) */
  309. CODEC_ID_TTF= 0x18000,
  310. CODEC_ID_PROBE= 0x19000, ///< codec_id is not known (like CODEC_ID_NONE) but lavf should attempt to identify it
  311. CODEC_ID_MPEG2TS= 0x20000, /**< _FAKE_ codec to indicate a raw MPEG-2 TS
  312. * stream (only used by libavformat) */
  313. };
  314. enum CodecType {
  315. CODEC_TYPE_UNKNOWN = -1,
  316. CODEC_TYPE_VIDEO,
  317. CODEC_TYPE_AUDIO,
  318. CODEC_TYPE_DATA,
  319. CODEC_TYPE_SUBTITLE,
  320. CODEC_TYPE_ATTACHMENT,
  321. CODEC_TYPE_NB
  322. };
  323. /**
  324. * all in native-endian format
  325. */
  326. enum SampleFormat {
  327. SAMPLE_FMT_NONE = -1,
  328. SAMPLE_FMT_U8, ///< unsigned 8 bits
  329. SAMPLE_FMT_S16, ///< signed 16 bits
  330. SAMPLE_FMT_S32, ///< signed 32 bits
  331. SAMPLE_FMT_FLT, ///< float
  332. SAMPLE_FMT_DBL, ///< double
  333. SAMPLE_FMT_NB ///< Number of sample formats. DO NOT USE if dynamically linking to libavcodec
  334. };
  335. /* Audio channel masks */
  336. #define CH_FRONT_LEFT 0x00000001
  337. #define CH_FRONT_RIGHT 0x00000002
  338. #define CH_FRONT_CENTER 0x00000004
  339. #define CH_LOW_FREQUENCY 0x00000008
  340. #define CH_BACK_LEFT 0x00000010
  341. #define CH_BACK_RIGHT 0x00000020
  342. #define CH_FRONT_LEFT_OF_CENTER 0x00000040
  343. #define CH_FRONT_RIGHT_OF_CENTER 0x00000080
  344. #define CH_BACK_CENTER 0x00000100
  345. #define CH_SIDE_LEFT 0x00000200
  346. #define CH_SIDE_RIGHT 0x00000400
  347. #define CH_TOP_CENTER 0x00000800
  348. #define CH_TOP_FRONT_LEFT 0x00001000
  349. #define CH_TOP_FRONT_CENTER 0x00002000
  350. #define CH_TOP_FRONT_RIGHT 0x00004000
  351. #define CH_TOP_BACK_LEFT 0x00008000
  352. #define CH_TOP_BACK_CENTER 0x00010000
  353. #define CH_TOP_BACK_RIGHT 0x00020000
  354. #define CH_STEREO_LEFT 0x20000000 ///< Stereo downmix.
  355. #define CH_STEREO_RIGHT 0x40000000 ///< See CH_STEREO_LEFT.
  356. /* Audio channel convenience macros */
  357. #define CH_LAYOUT_MONO (CH_FRONT_CENTER)
  358. #define CH_LAYOUT_STEREO (CH_FRONT_LEFT|CH_FRONT_RIGHT)
  359. #define CH_LAYOUT_SURROUND (CH_LAYOUT_STEREO|CH_FRONT_CENTER)
  360. #define CH_LAYOUT_QUAD (CH_LAYOUT_STEREO|CH_BACK_LEFT|CH_BACK_RIGHT)
  361. #define CH_LAYOUT_5POINT0 (CH_LAYOUT_SURROUND|CH_SIDE_LEFT|CH_SIDE_RIGHT)
  362. #define CH_LAYOUT_5POINT1 (CH_LAYOUT_5POINT0|CH_LOW_FREQUENCY)
  363. #define CH_LAYOUT_5POINT0_BACK (CH_LAYOUT_SURROUND|CH_BACK_LEFT|CH_BACK_RIGHT)
  364. #define CH_LAYOUT_5POINT1_BACK (CH_LAYOUT_5POINT0_BACK|CH_LOW_FREQUENCY)
  365. #define CH_LAYOUT_7POINT1 (CH_LAYOUT_5POINT1|CH_BACK_LEFT|CH_BACK_RIGHT)
  366. #define CH_LAYOUT_7POINT1_WIDE (CH_LAYOUT_5POINT1_BACK|\
  367. CH_FRONT_LEFT_OF_CENTER|CH_FRONT_RIGHT_OF_CENTER)
  368. #define CH_LAYOUT_STEREO_DOWNMIX (CH_STEREO_LEFT|CH_STEREO_RIGHT)
  369. /* in bytes */
  370. #define AVCODEC_MAX_AUDIO_FRAME_SIZE 192000 // 1 second of 48khz 32bit audio
  371. /**
  372. * Required number of additionally allocated bytes at the end of the input bitstream for decoding.
  373. * This is mainly needed because some optimized bitstream readers read
  374. * 32 or 64 bit at once and could read over the end.<br>
  375. * Note: If the first 23 bits of the additional bytes are not 0, then damaged
  376. * MPEG bitstreams could cause overread and segfault.
  377. */
  378. #define FF_INPUT_BUFFER_PADDING_SIZE 8
  379. /**
  380. * minimum encoding buffer size
  381. * Used to avoid some checks during header writing.
  382. */
  383. #define FF_MIN_BUFFER_SIZE 16384
  384. /**
  385. * motion estimation type.
  386. */
  387. enum Motion_Est_ID {
  388. ME_ZERO = 1, ///< no search, that is use 0,0 vector whenever one is needed
  389. ME_FULL,
  390. ME_LOG,
  391. ME_PHODS,
  392. ME_EPZS, ///< enhanced predictive zonal search
  393. ME_X1, ///< reserved for experiments
  394. ME_HEX, ///< hexagon based search
  395. ME_UMH, ///< uneven multi-hexagon search
  396. ME_ITER, ///< iterative search
  397. ME_TESA, ///< transformed exhaustive search algorithm
  398. };
  399. enum AVDiscard{
  400. /* We leave some space between them for extensions (drop some
  401. * keyframes for intra-only or drop just some bidir frames). */
  402. AVDISCARD_NONE =-16, ///< discard nothing
  403. AVDISCARD_DEFAULT= 0, ///< discard useless packets like 0 size packets in avi
  404. AVDISCARD_NONREF = 8, ///< discard all non reference
  405. AVDISCARD_BIDIR = 16, ///< discard all bidirectional frames
  406. AVDISCARD_NONKEY = 32, ///< discard all frames except keyframes
  407. AVDISCARD_ALL = 48, ///< discard all
  408. };
  409. typedef struct RcOverride{
  410. int start_frame;
  411. int end_frame;
  412. int qscale; // If this is 0 then quality_factor will be used instead.
  413. float quality_factor;
  414. } RcOverride;
  415. #define FF_MAX_B_FRAMES 16
  416. /* encoding support
  417. These flags can be passed in AVCodecContext.flags before initialization.
  418. Note: Not everything is supported yet.
  419. */
  420. #define CODEC_FLAG_QSCALE 0x0002 ///< Use fixed qscale.
  421. #define CODEC_FLAG_4MV 0x0004 ///< 4 MV per MB allowed / advanced prediction for H.263.
  422. #define CODEC_FLAG_QPEL 0x0010 ///< Use qpel MC.
  423. #define CODEC_FLAG_GMC 0x0020 ///< Use GMC.
  424. #define CODEC_FLAG_MV0 0x0040 ///< Always try a MB with MV=<0,0>.
  425. #define CODEC_FLAG_PART 0x0080 ///< Use data partitioning.
  426. /**
  427. * The parent program guarantees that the input for B-frames containing
  428. * streams is not written to for at least s->max_b_frames+1 frames, if
  429. * this is not set the input will be copied.
  430. */
  431. #define CODEC_FLAG_INPUT_PRESERVED 0x0100
  432. #define CODEC_FLAG_PASS1 0x0200 ///< Use internal 2pass ratecontrol in first pass mode.
  433. #define CODEC_FLAG_PASS2 0x0400 ///< Use internal 2pass ratecontrol in second pass mode.
  434. #define CODEC_FLAG_EXTERN_HUFF 0x1000 ///< Use external Huffman table (for MJPEG).
  435. #define CODEC_FLAG_GRAY 0x2000 ///< Only decode/encode grayscale.
  436. #define CODEC_FLAG_EMU_EDGE 0x4000 ///< Don't draw edges.
  437. #define CODEC_FLAG_PSNR 0x8000 ///< error[?] variables will be set during encoding.
  438. #define CODEC_FLAG_TRUNCATED 0x00010000 /** Input bitstream might be truncated at a random
  439. location instead of only at frame boundaries. */
  440. #define CODEC_FLAG_NORMALIZE_AQP 0x00020000 ///< Normalize adaptive quantization.
  441. #define CODEC_FLAG_INTERLACED_DCT 0x00040000 ///< Use interlaced DCT.
  442. #define CODEC_FLAG_LOW_DELAY 0x00080000 ///< Force low delay.
  443. #define CODEC_FLAG_ALT_SCAN 0x00100000 ///< Use alternate scan.
  444. #define CODEC_FLAG_GLOBAL_HEADER 0x00400000 ///< Place global headers in extradata instead of every keyframe.
  445. #define CODEC_FLAG_BITEXACT 0x00800000 ///< Use only bitexact stuff (except (I)DCT).
  446. /* Fx : Flag for h263+ extra options */
  447. #define CODEC_FLAG_AC_PRED 0x01000000 ///< H.263 advanced intra coding / MPEG-4 AC prediction
  448. #define CODEC_FLAG_H263P_UMV 0x02000000 ///< unlimited motion vector
  449. #define CODEC_FLAG_CBP_RD 0x04000000 ///< Use rate distortion optimization for cbp.
  450. #define CODEC_FLAG_QP_RD 0x08000000 ///< Use rate distortion optimization for qp selectioon.
  451. #define CODEC_FLAG_H263P_AIV 0x00000008 ///< H.263 alternative inter VLC
  452. #define CODEC_FLAG_OBMC 0x00000001 ///< OBMC
  453. #define CODEC_FLAG_LOOP_FILTER 0x00000800 ///< loop filter
  454. #define CODEC_FLAG_H263P_SLICE_STRUCT 0x10000000
  455. #define CODEC_FLAG_INTERLACED_ME 0x20000000 ///< interlaced motion estimation
  456. #define CODEC_FLAG_SVCD_SCAN_OFFSET 0x40000000 ///< Will reserve space for SVCD scan offset user data.
  457. #define CODEC_FLAG_CLOSED_GOP 0x80000000
  458. #define CODEC_FLAG2_FAST 0x00000001 ///< Allow non spec compliant speedup tricks.
  459. #define CODEC_FLAG2_STRICT_GOP 0x00000002 ///< Strictly enforce GOP size.
  460. #define CODEC_FLAG2_NO_OUTPUT 0x00000004 ///< Skip bitstream encoding.
  461. #define CODEC_FLAG2_LOCAL_HEADER 0x00000008 ///< Place global headers at every keyframe instead of in extradata.
  462. #define CODEC_FLAG2_BPYRAMID 0x00000010 ///< H.264 allow B-frames to be used as references.
  463. #define CODEC_FLAG2_WPRED 0x00000020 ///< H.264 weighted biprediction for B-frames
  464. #define CODEC_FLAG2_MIXED_REFS 0x00000040 ///< H.264 one reference per partition, as opposed to one reference per macroblock
  465. #define CODEC_FLAG2_8X8DCT 0x00000080 ///< H.264 high profile 8x8 transform
  466. #define CODEC_FLAG2_FASTPSKIP 0x00000100 ///< H.264 fast pskip
  467. #define CODEC_FLAG2_AUD 0x00000200 ///< H.264 access unit delimiters
  468. #define CODEC_FLAG2_BRDO 0x00000400 ///< B-frame rate-distortion optimization
  469. #define CODEC_FLAG2_INTRA_VLC 0x00000800 ///< Use MPEG-2 intra VLC table.
  470. #define CODEC_FLAG2_MEMC_ONLY 0x00001000 ///< Only do ME/MC (I frames -> ref, P frame -> ME+MC).
  471. #define CODEC_FLAG2_DROP_FRAME_TIMECODE 0x00002000 ///< timecode is in drop frame format.
  472. #define CODEC_FLAG2_SKIP_RD 0x00004000 ///< RD optimal MB level residual skipping
  473. #define CODEC_FLAG2_CHUNKS 0x00008000 ///< Input bitstream might be truncated at a packet boundaries instead of only at frame boundaries.
  474. #define CODEC_FLAG2_NON_LINEAR_QUANT 0x00010000 ///< Use MPEG-2 nonlinear quantizer.
  475. #define CODEC_FLAG2_BIT_RESERVOIR 0x00020000 ///< Use a bit reservoir when encoding if possible
  476. /* Unsupported options :
  477. * Syntax Arithmetic coding (SAC)
  478. * Reference Picture Selection
  479. * Independent Segment Decoding */
  480. /* /Fx */
  481. /* codec capabilities */
  482. #define CODEC_CAP_DRAW_HORIZ_BAND 0x0001 ///< Decoder can use draw_horiz_band callback.
  483. /**
  484. * Codec uses get_buffer() for allocating buffers.
  485. * direct rendering method 1
  486. */
  487. #define CODEC_CAP_DR1 0x0002
  488. /* If 'parse_only' field is true, then avcodec_parse_frame() can be used. */
  489. #define CODEC_CAP_PARSE_ONLY 0x0004
  490. #define CODEC_CAP_TRUNCATED 0x0008
  491. /* Codec can export data for HW decoding (XvMC). */
  492. #define CODEC_CAP_HWACCEL 0x0010
  493. /**
  494. * Codec has a nonzero delay and needs to be fed with NULL at the end to get the delayed data.
  495. * If this is not set, the codec is guaranteed to never be fed with NULL data.
  496. */
  497. #define CODEC_CAP_DELAY 0x0020
  498. /**
  499. * Codec can be fed a final frame with a smaller size.
  500. * This can be used to prevent truncation of the last audio samples.
  501. */
  502. #define CODEC_CAP_SMALL_LAST_FRAME 0x0040
  503. /**
  504. * Codec can export data for HW decoding (VDPAU).
  505. */
  506. #define CODEC_CAP_HWACCEL_VDPAU 0x0080
  507. //The following defines may change, don't expect compatibility if you use them.
  508. #define MB_TYPE_INTRA4x4 0x0001
  509. #define MB_TYPE_INTRA16x16 0x0002 //FIXME H.264-specific
  510. #define MB_TYPE_INTRA_PCM 0x0004 //FIXME H.264-specific
  511. #define MB_TYPE_16x16 0x0008
  512. #define MB_TYPE_16x8 0x0010
  513. #define MB_TYPE_8x16 0x0020
  514. #define MB_TYPE_8x8 0x0040
  515. #define MB_TYPE_INTERLACED 0x0080
  516. #define MB_TYPE_DIRECT2 0x0100 //FIXME
  517. #define MB_TYPE_ACPRED 0x0200
  518. #define MB_TYPE_GMC 0x0400
  519. #define MB_TYPE_SKIP 0x0800
  520. #define MB_TYPE_P0L0 0x1000
  521. #define MB_TYPE_P1L0 0x2000
  522. #define MB_TYPE_P0L1 0x4000
  523. #define MB_TYPE_P1L1 0x8000
  524. #define MB_TYPE_L0 (MB_TYPE_P0L0 | MB_TYPE_P1L0)
  525. #define MB_TYPE_L1 (MB_TYPE_P0L1 | MB_TYPE_P1L1)
  526. #define MB_TYPE_L0L1 (MB_TYPE_L0 | MB_TYPE_L1)
  527. #define MB_TYPE_QUANT 0x00010000
  528. #define MB_TYPE_CBP 0x00020000
  529. //Note bits 24-31 are reserved for codec specific use (h264 ref0, mpeg1 0mv, ...)
  530. /**
  531. * Pan Scan area.
  532. * This specifies the area which should be displayed.
  533. * Note there may be multiple such areas for one frame.
  534. */
  535. typedef struct AVPanScan{
  536. /**
  537. * id
  538. * - encoding: Set by user.
  539. * - decoding: Set by libavcodec.
  540. */
  541. int id;
  542. /**
  543. * width and height in 1/16 pel
  544. * - encoding: Set by user.
  545. * - decoding: Set by libavcodec.
  546. */
  547. int width;
  548. int height;
  549. /**
  550. * position of the top left corner in 1/16 pel for up to 3 fields/frames
  551. * - encoding: Set by user.
  552. * - decoding: Set by libavcodec.
  553. */
  554. int16_t position[3][2];
  555. }AVPanScan;
  556. #define FF_COMMON_FRAME \
  557. /**\
  558. * pointer to the picture planes.\
  559. * This might be different from the first allocated byte\
  560. * - encoding: \
  561. * - decoding: \
  562. */\
  563. uint8_t *data[4];\
  564. int linesize[4];\
  565. /**\
  566. * pointer to the first allocated byte of the picture. Can be used in get_buffer/release_buffer.\
  567. * This isn't used by libavcodec unless the default get/release_buffer() is used.\
  568. * - encoding: \
  569. * - decoding: \
  570. */\
  571. uint8_t *base[4];\
  572. /**\
  573. * 1 -> keyframe, 0-> not\
  574. * - encoding: Set by libavcodec.\
  575. * - decoding: Set by libavcodec.\
  576. */\
  577. int key_frame;\
  578. \
  579. /**\
  580. * Picture type of the frame, see ?_TYPE below.\
  581. * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
  582. * - decoding: Set by libavcodec.\
  583. */\
  584. int pict_type;\
  585. \
  586. /**\
  587. * presentation timestamp in time_base units (time when frame should be shown to user)\
  588. * If AV_NOPTS_VALUE then frame_rate = 1/time_base will be assumed.\
  589. * - encoding: MUST be set by user.\
  590. * - decoding: Set by libavcodec.\
  591. */\
  592. int64_t pts;\
  593. \
  594. /**\
  595. * picture number in bitstream order\
  596. * - encoding: set by\
  597. * - decoding: Set by libavcodec.\
  598. */\
  599. int coded_picture_number;\
  600. /**\
  601. * picture number in display order\
  602. * - encoding: set by\
  603. * - decoding: Set by libavcodec.\
  604. */\
  605. int display_picture_number;\
  606. \
  607. /**\
  608. * quality (between 1 (good) and FF_LAMBDA_MAX (bad)) \
  609. * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
  610. * - decoding: Set by libavcodec.\
  611. */\
  612. int quality; \
  613. \
  614. /**\
  615. * buffer age (1->was last buffer and dint change, 2->..., ...).\
  616. * Set to INT_MAX if the buffer has not been used yet.\
  617. * - encoding: unused\
  618. * - decoding: MUST be set by get_buffer().\
  619. */\
  620. int age;\
  621. \
  622. /**\
  623. * is this picture used as reference\
  624. * The values for this are the same as the MpegEncContext.picture_structure\
  625. * variable, that is 1->top field, 2->bottom field, 3->frame/both fields.\
  626. * Set to 4 for delayed, non-reference frames.\
  627. * - encoding: unused\
  628. * - decoding: Set by libavcodec. (before get_buffer() call)).\
  629. */\
  630. int reference;\
  631. \
  632. /**\
  633. * QP table\
  634. * - encoding: unused\
  635. * - decoding: Set by libavcodec.\
  636. */\
  637. int8_t *qscale_table;\
  638. /**\
  639. * QP store stride\
  640. * - encoding: unused\
  641. * - decoding: Set by libavcodec.\
  642. */\
  643. int qstride;\
  644. \
  645. /**\
  646. * mbskip_table[mb]>=1 if MB didn't change\
  647. * stride= mb_width = (width+15)>>4\
  648. * - encoding: unused\
  649. * - decoding: Set by libavcodec.\
  650. */\
  651. uint8_t *mbskip_table;\
  652. \
  653. /**\
  654. * motion vector table\
  655. * @code\
  656. * example:\
  657. * int mv_sample_log2= 4 - motion_subsample_log2;\
  658. * int mb_width= (width+15)>>4;\
  659. * int mv_stride= (mb_width << mv_sample_log2) + 1;\
  660. * motion_val[direction][x + y*mv_stride][0->mv_x, 1->mv_y];\
  661. * @endcode\
  662. * - encoding: Set by user.\
  663. * - decoding: Set by libavcodec.\
  664. */\
  665. int16_t (*motion_val[2])[2];\
  666. \
  667. /**\
  668. * macroblock type table\
  669. * mb_type_base + mb_width + 2\
  670. * - encoding: Set by user.\
  671. * - decoding: Set by libavcodec.\
  672. */\
  673. uint32_t *mb_type;\
  674. \
  675. /**\
  676. * log2 of the size of the block which a single vector in motion_val represents: \
  677. * (4->16x16, 3->8x8, 2-> 4x4, 1-> 2x2)\
  678. * - encoding: unused\
  679. * - decoding: Set by libavcodec.\
  680. */\
  681. uint8_t motion_subsample_log2;\
  682. \
  683. /**\
  684. * for some private data of the user\
  685. * - encoding: unused\
  686. * - decoding: Set by user.\
  687. */\
  688. void *opaque;\
  689. \
  690. /**\
  691. * error\
  692. * - encoding: Set by libavcodec. if flags&CODEC_FLAG_PSNR.\
  693. * - decoding: unused\
  694. */\
  695. uint64_t error[4];\
  696. \
  697. /**\
  698. * type of the buffer (to keep track of who has to deallocate data[*])\
  699. * - encoding: Set by the one who allocates it.\
  700. * - decoding: Set by the one who allocates it.\
  701. * Note: User allocated (direct rendering) & internal buffers cannot coexist currently.\
  702. */\
  703. int type;\
  704. \
  705. /**\
  706. * When decoding, this signals how much the picture must be delayed.\
  707. * extra_delay = repeat_pict / (2*fps)\
  708. * - encoding: unused\
  709. * - decoding: Set by libavcodec.\
  710. */\
  711. int repeat_pict;\
  712. \
  713. /**\
  714. * \
  715. */\
  716. int qscale_type;\
  717. \
  718. /**\
  719. * The content of the picture is interlaced.\
  720. * - encoding: Set by user.\
  721. * - decoding: Set by libavcodec. (default 0)\
  722. */\
  723. int interlaced_frame;\
  724. \
  725. /**\
  726. * If the content is interlaced, is top field displayed first.\
  727. * - encoding: Set by user.\
  728. * - decoding: Set by libavcodec.\
  729. */\
  730. int top_field_first;\
  731. \
  732. /**\
  733. * Pan scan.\
  734. * - encoding: Set by user.\
  735. * - decoding: Set by libavcodec.\
  736. */\
  737. AVPanScan *pan_scan;\
  738. \
  739. /**\
  740. * Tell user application that palette has changed from previous frame.\
  741. * - encoding: ??? (no palette-enabled encoder yet)\
  742. * - decoding: Set by libavcodec. (default 0).\
  743. */\
  744. int palette_has_changed;\
  745. \
  746. /**\
  747. * codec suggestion on buffer type if != 0\
  748. * - encoding: unused\
  749. * - decoding: Set by libavcodec. (before get_buffer() call)).\
  750. */\
  751. int buffer_hints;\
  752. \
  753. /**\
  754. * DCT coefficients\
  755. * - encoding: unused\
  756. * - decoding: Set by libavcodec.\
  757. */\
  758. short *dct_coeff;\
  759. \
  760. /**\
  761. * motion referece frame index\
  762. * - encoding: Set by user.\
  763. * - decoding: Set by libavcodec.\
  764. */\
  765. int8_t *ref_index[2];\
  766. \
  767. /**\
  768. * reordered opaque 64bit number (generally a PTS) from AVCodecContext.reordered_opaque\
  769. * output in AVFrame.reordered_opaque\
  770. * - encoding: unused\
  771. * - decoding: Read by user.\
  772. */\
  773. int64_t reordered_opaque;\
  774. \
  775. /**\
  776. * hardware accelerator private data (FFmpeg allocated)\
  777. * - encoding: unused\
  778. * - decoding: Set by libavcodec\
  779. */\
  780. void *hwaccel_picture_private;\
  781. #define FF_QSCALE_TYPE_MPEG1 0
  782. #define FF_QSCALE_TYPE_MPEG2 1
  783. #define FF_QSCALE_TYPE_H264 2
  784. #define FF_BUFFER_TYPE_INTERNAL 1
  785. #define FF_BUFFER_TYPE_USER 2 ///< direct rendering buffers (image is (de)allocated by user)
  786. #define FF_BUFFER_TYPE_SHARED 4 ///< Buffer from somewhere else; don't deallocate image (data/base), all other tables are not shared.
  787. #define FF_BUFFER_TYPE_COPY 8 ///< Just a (modified) copy of some other buffer, don't deallocate anything.
  788. #define FF_I_TYPE 1 ///< Intra
  789. #define FF_P_TYPE 2 ///< Predicted
  790. #define FF_B_TYPE 3 ///< Bi-dir predicted
  791. #define FF_S_TYPE 4 ///< S(GMC)-VOP MPEG4
  792. #define FF_SI_TYPE 5 ///< Switching Intra
  793. #define FF_SP_TYPE 6 ///< Switching Predicted
  794. #define FF_BI_TYPE 7
  795. #define FF_BUFFER_HINTS_VALID 0x01 // Buffer hints value is meaningful (if 0 ignore).
  796. #define FF_BUFFER_HINTS_READABLE 0x02 // Codec will read from buffer.
  797. #define FF_BUFFER_HINTS_PRESERVE 0x04 // User must not alter buffer content.
  798. #define FF_BUFFER_HINTS_REUSABLE 0x08 // Codec will reuse the buffer (update).
  799. typedef struct AVPacket {
  800. /**
  801. * Presentation timestamp in AVStream->time_base units; the time at which
  802. * the decompressed packet will be presented to the user.
  803. * Can be AV_NOPTS_VALUE if it is not stored in the file.
  804. * pts MUST be larger or equal to dts as presentation cannot happen before
  805. * decompression, unless one wants to view hex dumps. Some formats misuse
  806. * the terms dts and pts/cts to mean something different. Such timestamps
  807. * must be converted to true pts/dts before they are stored in AVPacket.
  808. */
  809. int64_t pts;
  810. /**
  811. * Decompression timestamp in AVStream->time_base units; the time at which
  812. * the packet is decompressed.
  813. * Can be AV_NOPTS_VALUE if it is not stored in the file.
  814. */
  815. int64_t dts;
  816. uint8_t *data;
  817. int size;
  818. int stream_index;
  819. int flags;
  820. /**
  821. * Duration of this packet in AVStream->time_base units, 0 if unknown.
  822. * Equals next_pts - this_pts in presentation order.
  823. */
  824. int duration;
  825. void (*destruct)(struct AVPacket *);
  826. void *priv;
  827. int64_t pos; ///< byte position in stream, -1 if unknown
  828. /**
  829. * Time difference in AVStream->time_base units from the pts of this
  830. * packet to the point at which the output from the decoder has converged
  831. * independent from the availability of previous frames. That is, the
  832. * frames are virtually identical no matter if decoding started from
  833. * the very first frame or from this keyframe.
  834. * Is AV_NOPTS_VALUE if unknown.
  835. * This field is not the display duration of the current packet.
  836. *
  837. * The purpose of this field is to allow seeking in streams that have no
  838. * keyframes in the conventional sense. It corresponds to the
  839. * recovery point SEI in H.264 and match_time_delta in NUT. It is also
  840. * essential for some types of subtitle streams to ensure that all
  841. * subtitles are correctly displayed after seeking.
  842. */
  843. int64_t convergence_duration;
  844. } AVPacket;
  845. #define PKT_FLAG_KEY 0x0001
  846. /**
  847. * Audio Video Frame.
  848. * New fields can be added to the end of FF_COMMON_FRAME with minor version
  849. * bumps.
  850. * Removal, reordering and changes to existing fields require a major
  851. * version bump. No fields should be added into AVFrame before or after
  852. * FF_COMMON_FRAME!
  853. * sizeof(AVFrame) must not be used outside libav*.
  854. */
  855. typedef struct AVFrame {
  856. FF_COMMON_FRAME
  857. } AVFrame;
  858. /**
  859. * main external API structure.
  860. * New fields can be added to the end with minor version bumps.
  861. * Removal, reordering and changes to existing fields require a major
  862. * version bump.
  863. * sizeof(AVCodecContext) must not be used outside libav*.
  864. */
  865. typedef struct AVCodecContext {
  866. /**
  867. * information on struct for av_log
  868. * - set by avcodec_alloc_context
  869. */
  870. const AVClass *av_class;
  871. /**
  872. * the average bitrate
  873. * - encoding: Set by user; unused for constant quantizer encoding.
  874. * - decoding: Set by libavcodec. 0 or some bitrate if this info is available in the stream.
  875. */
  876. int bit_rate;
  877. /**
  878. * number of bits the bitstream is allowed to diverge from the reference.
  879. * the reference can be CBR (for CBR pass1) or VBR (for pass2)
  880. * - encoding: Set by user; unused for constant quantizer encoding.
  881. * - decoding: unused
  882. */
  883. int bit_rate_tolerance;
  884. /**
  885. * CODEC_FLAG_*.
  886. * - encoding: Set by user.
  887. * - decoding: Set by user.
  888. */
  889. int flags;
  890. /**
  891. * Some codecs need additional format info. It is stored here.
  892. * If any muxer uses this then ALL demuxers/parsers AND encoders for the
  893. * specific codec MUST set it correctly otherwise stream copy breaks.
  894. * In general use of this field by muxers is not recommanded.
  895. * - encoding: Set by libavcodec.
  896. * - decoding: Set by libavcodec. (FIXME: Is this OK?)
  897. */
  898. int sub_id;
  899. /**
  900. * Motion estimation algorithm used for video coding.
  901. * 1 (zero), 2 (full), 3 (log), 4 (phods), 5 (epzs), 6 (x1), 7 (hex),
  902. * 8 (umh), 9 (iter), 10 (tesa) [7, 8, 10 are x264 specific, 9 is snow specific]
  903. * - encoding: MUST be set by user.
  904. * - decoding: unused
  905. */
  906. int me_method;
  907. /**
  908. * some codecs need / can use extradata like Huffman tables.
  909. * mjpeg: Huffman tables
  910. * rv10: additional flags
  911. * mpeg4: global headers (they can be in the bitstream or here)
  912. * The allocated memory should be FF_INPUT_BUFFER_PADDING_SIZE bytes larger
  913. * than extradata_size to avoid prolems if it is read with the bitstream reader.
  914. * The bytewise contents of extradata must not depend on the architecture or CPU endianness.
  915. * - encoding: Set/allocated/freed by libavcodec.
  916. * - decoding: Set/allocated/freed by user.
  917. */
  918. uint8_t *extradata;
  919. int extradata_size;
  920. /**
  921. * This is the fundamental unit of time (in seconds) in terms
  922. * of which frame timestamps are represented. For fixed-fps content,
  923. * timebase should be 1/framerate and timestamp increments should be
  924. * identically 1.
  925. * - encoding: MUST be set by user.
  926. * - decoding: Set by libavcodec.
  927. */
  928. AVRational time_base;
  929. /* video only */
  930. /**
  931. * picture width / height.
  932. * - encoding: MUST be set by user.
  933. * - decoding: Set by libavcodec.
  934. * Note: For compatibility it is possible to set this instead of
  935. * coded_width/height before decoding.
  936. */
  937. int width, height;
  938. #define FF_ASPECT_EXTENDED 15
  939. /**
  940. * the number of pictures in a group of pictures, or 0 for intra_only
  941. * - encoding: Set by user.
  942. * - decoding: unused
  943. */
  944. int gop_size;
  945. /**
  946. * Pixel format, see PIX_FMT_xxx.
  947. * - encoding: Set by user.
  948. * - decoding: Set by libavcodec.
  949. */
  950. enum PixelFormat pix_fmt;
  951. /**
  952. * Frame rate emulation. If not zero, the lower layer (i.e. format handler)
  953. * has to read frames at native frame rate.
  954. * - encoding: Set by user.
  955. * - decoding: unused
  956. */
  957. int rate_emu;
  958. /**
  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
  962. * beforehand.
  963. * The function is also used by hardware acceleration APIs.
  964. * It is called at least once during frame decoding to pass
  965. * the data needed for hardware render.
  966. * In that mode instead of pixel data, AVFrame points to
  967. * a structure specific to the acceleration API. The application
  968. * reads the structure and can change some fields to indicate progress
  969. * or mark state.
  970. * - encoding: unused
  971. * - decoding: Set by user.
  972. * @param height the height of the slice
  973. * @param y the y position of the slice
  974. * @param type 1->top field, 2->bottom field, 3->frame
  975. * @param offset offset into the AVFrame.data from which the slice should be read
  976. */
  977. void (*draw_horiz_band)(struct AVCodecContext *s,
  978. const AVFrame *src, int offset[4],
  979. int y, int type, int height);
  980. /* audio only */
  981. int sample_rate; ///< samples per second
  982. int channels; ///< number of audio channels
  983. /**
  984. * audio sample format
  985. * - encoding: Set by user.
  986. * - decoding: Set by libavcodec.
  987. */
  988. enum SampleFormat sample_fmt; ///< sample format, currently unused
  989. /* The following data should not be initialized. */
  990. /**
  991. * Samples per packet, initialized when calling 'init'.
  992. */
  993. int frame_size;
  994. int frame_number; ///< audio or video frame number
  995. int real_pict_num; ///< Returns the real picture number of previous encoded frame.
  996. /**
  997. * Number of frames the decoded output will be delayed relative to
  998. * the encoded input.
  999. * - encoding: Set by libavcodec.
  1000. * - decoding: unused
  1001. */
  1002. int delay;
  1003. /* - encoding parameters */
  1004. float qcompress; ///< amount of qscale change between easy & hard scenes (0.0-1.0)
  1005. float qblur; ///< amount of qscale smoothing over time (0.0-1.0)
  1006. /**
  1007. * minimum quantizer
  1008. * - encoding: Set by user.
  1009. * - decoding: unused
  1010. */
  1011. int qmin;
  1012. /**
  1013. * maximum quantizer
  1014. * - encoding: Set by user.
  1015. * - decoding: unused
  1016. */
  1017. int qmax;
  1018. /**
  1019. * maximum quantizer difference between frames
  1020. * - encoding: Set by user.
  1021. * - decoding: unused
  1022. */
  1023. int max_qdiff;
  1024. /**
  1025. * maximum number of B-frames between non-B-frames
  1026. * Note: The output will be delayed by max_b_frames+1 relative to the input.
  1027. * - encoding: Set by user.
  1028. * - decoding: unused
  1029. */
  1030. int max_b_frames;
  1031. /**
  1032. * qscale factor between IP and B-frames
  1033. * If > 0 then the last P-frame quantizer will be used (q= lastp_q*factor+offset).
  1034. * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
  1035. * - encoding: Set by user.
  1036. * - decoding: unused
  1037. */
  1038. float b_quant_factor;
  1039. /** obsolete FIXME remove */
  1040. int rc_strategy;
  1041. #define FF_RC_STRATEGY_XVID 1
  1042. int b_frame_strategy;
  1043. /**
  1044. * hurry up amount
  1045. * - encoding: unused
  1046. * - decoding: Set by user. 1-> Skip B-frames, 2-> Skip IDCT/dequant too, 5-> Skip everything except header
  1047. * @deprecated Deprecated in favor of skip_idct and skip_frame.
  1048. */
  1049. int hurry_up;
  1050. struct AVCodec *codec;
  1051. void *priv_data;
  1052. int rtp_payload_size; /* The size of the RTP payload: the coder will */
  1053. /* do its best to deliver a chunk with size */
  1054. /* below rtp_payload_size, the chunk will start */
  1055. /* with a start code on some codecs like H.263. */
  1056. /* This doesn't take account of any particular */
  1057. /* headers inside the transmitted RTP payload. */
  1058. /* The RTP callback: This function is called */
  1059. /* every time the encoder has a packet to send. */
  1060. /* It depends on the encoder if the data starts */
  1061. /* with a Start Code (it should). H.263 does. */
  1062. /* mb_nb contains the number of macroblocks */
  1063. /* encoded in the RTP payload. */
  1064. void (*rtp_callback)(struct AVCodecContext *avctx, void *data, int size, int mb_nb);
  1065. /* statistics, used for 2-pass encoding */
  1066. int mv_bits;
  1067. int header_bits;
  1068. int i_tex_bits;
  1069. int p_tex_bits;
  1070. int i_count;
  1071. int p_count;
  1072. int skip_count;
  1073. int misc_bits;
  1074. /**
  1075. * number of bits used for the previously encoded frame
  1076. * - encoding: Set by libavcodec.
  1077. * - decoding: unused
  1078. */
  1079. int frame_bits;
  1080. /**
  1081. * Private data of the user, can be used to carry app specific stuff.
  1082. * - encoding: Set by user.
  1083. * - decoding: Set by user.
  1084. */
  1085. void *opaque;
  1086. char codec_name[32];
  1087. enum CodecType codec_type; /* see CODEC_TYPE_xxx */
  1088. enum CodecID codec_id; /* see CODEC_ID_xxx */
  1089. /**
  1090. * fourcc (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
  1091. * This is used to work around some encoder bugs.
  1092. * A demuxer should set this to what is stored in the field used to identify the codec.
  1093. * If there are multiple such fields in a container then the demuxer should choose the one
  1094. * which maximizes the information about the used codec.
  1095. * If the codec tag field in a container is larger then 32 bits then the demuxer should
  1096. * remap the longer ID to 32 bits with a table or other structure. Alternatively a new
  1097. * extra_codec_tag + size could be added but for this a clear advantage must be demonstrated
  1098. * first.
  1099. * - encoding: Set by user, if not then the default based on codec_id will be used.
  1100. * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
  1101. */
  1102. unsigned int codec_tag;
  1103. /**
  1104. * Work around bugs in encoders which sometimes cannot be detected automatically.
  1105. * - encoding: Set by user
  1106. * - decoding: Set by user
  1107. */
  1108. int workaround_bugs;
  1109. #define FF_BUG_AUTODETECT 1 ///< autodetection
  1110. #define FF_BUG_OLD_MSMPEG4 2
  1111. #define FF_BUG_XVID_ILACE 4
  1112. #define FF_BUG_UMP4 8
  1113. #define FF_BUG_NO_PADDING 16
  1114. #define FF_BUG_AMV 32
  1115. #define FF_BUG_AC_VLC 0 ///< Will be removed, libavcodec can now handle these non-compliant files by default.
  1116. #define FF_BUG_QPEL_CHROMA 64
  1117. #define FF_BUG_STD_QPEL 128
  1118. #define FF_BUG_QPEL_CHROMA2 256
  1119. #define FF_BUG_DIRECT_BLOCKSIZE 512
  1120. #define FF_BUG_EDGE 1024
  1121. #define FF_BUG_HPEL_CHROMA 2048
  1122. #define FF_BUG_DC_CLIP 4096
  1123. #define FF_BUG_MS 8192 ///< Work around various bugs in Microsoft's broken decoders.
  1124. //#define FF_BUG_FAKE_SCALABILITY 16 //Autodetection should work 100%.
  1125. /**
  1126. * luma single coefficient elimination threshold
  1127. * - encoding: Set by user.
  1128. * - decoding: unused
  1129. */
  1130. int luma_elim_threshold;
  1131. /**
  1132. * chroma single coeff elimination threshold
  1133. * - encoding: Set by user.
  1134. * - decoding: unused
  1135. */
  1136. int chroma_elim_threshold;
  1137. /**
  1138. * strictly follow the standard (MPEG4, ...).
  1139. * - encoding: Set by user.
  1140. * - decoding: Set by user.
  1141. * Setting this to STRICT or higher means the encoder and decoder will
  1142. * generally do stupid things. While setting it to inofficial or lower
  1143. * will mean the encoder might use things that are not supported by all
  1144. * spec compliant decoders. Decoders make no difference between normal,
  1145. * inofficial and experimental, that is they always try to decode things
  1146. * when they can unless they are explicitly asked to behave stupid
  1147. * (=strictly conform to the specs)
  1148. */
  1149. int strict_std_compliance;
  1150. #define FF_COMPLIANCE_VERY_STRICT 2 ///< Strictly conform to a older more strict version of the spec or reference software.
  1151. #define FF_COMPLIANCE_STRICT 1 ///< Strictly conform to all the things in the spec no matter what consequences.
  1152. #define FF_COMPLIANCE_NORMAL 0
  1153. #define FF_COMPLIANCE_INOFFICIAL -1 ///< Allow inofficial extensions.
  1154. #define FF_COMPLIANCE_EXPERIMENTAL -2 ///< Allow nonstandardized experimental things.
  1155. /**
  1156. * qscale offset between IP and B-frames
  1157. * - encoding: Set by user.
  1158. * - decoding: unused
  1159. */
  1160. float b_quant_offset;
  1161. /**
  1162. * Error recognization; higher values will detect more errors but may
  1163. * misdetect some more or less valid parts as errors.
  1164. * - encoding: unused
  1165. * - decoding: Set by user.
  1166. */
  1167. int error_recognition;
  1168. #define FF_ER_CAREFUL 1
  1169. #define FF_ER_COMPLIANT 2
  1170. #define FF_ER_AGGRESSIVE 3
  1171. #define FF_ER_VERY_AGGRESSIVE 4
  1172. /**
  1173. * Called at the beginning of each frame to get a buffer for it.
  1174. * If pic.reference is set then the frame will be read later by libavcodec.
  1175. * avcodec_align_dimensions() should be used to find the required width and
  1176. * height, as they normally need to be rounded up to the next multiple of 16.
  1177. * - encoding: unused
  1178. * - decoding: Set by libavcodec., user can override.
  1179. */
  1180. int (*get_buffer)(struct AVCodecContext *c, AVFrame *pic);
  1181. /**
  1182. * Called to release buffers which were allocated with get_buffer.
  1183. * A released buffer can be reused in get_buffer().
  1184. * pic.data[*] must be set to NULL.
  1185. * - encoding: unused
  1186. * - decoding: Set by libavcodec., user can override.
  1187. */
  1188. void (*release_buffer)(struct AVCodecContext *c, AVFrame *pic);
  1189. /**
  1190. * Size of the frame reordering buffer in the decoder.
  1191. * For MPEG-2 it is 1 IPB or 0 low delay IP.
  1192. * - encoding: Set by libavcodec.
  1193. * - decoding: Set by libavcodec.
  1194. */
  1195. int has_b_frames;
  1196. /**
  1197. * number of bytes per packet if constant and known or 0
  1198. * Used by some WAV based audio codecs.
  1199. */
  1200. int block_align;
  1201. int parse_only; /* - decoding only: If true, only parsing is done
  1202. (function avcodec_parse_frame()). The frame
  1203. data is returned. Only MPEG codecs support this now. */
  1204. /**
  1205. * 0-> h263 quant 1-> mpeg quant
  1206. * - encoding: Set by user.
  1207. * - decoding: unused
  1208. */
  1209. int mpeg_quant;
  1210. /**
  1211. * pass1 encoding statistics output buffer
  1212. * - encoding: Set by libavcodec.
  1213. * - decoding: unused
  1214. */
  1215. char *stats_out;
  1216. /**
  1217. * pass2 encoding statistics input buffer
  1218. * Concatenated stuff from stats_out of pass1 should be placed here.
  1219. * - encoding: Allocated/set/freed by user.
  1220. * - decoding: unused
  1221. */
  1222. char *stats_in;
  1223. /**
  1224. * ratecontrol qmin qmax limiting method
  1225. * 0-> clipping, 1-> use a nice continous function to limit qscale wthin qmin/qmax.
  1226. * - encoding: Set by user.
  1227. * - decoding: unused
  1228. */
  1229. float rc_qsquish;
  1230. float rc_qmod_amp;
  1231. int rc_qmod_freq;
  1232. /**
  1233. * ratecontrol override, see RcOverride
  1234. * - encoding: Allocated/set/freed by user.
  1235. * - decoding: unused
  1236. */
  1237. RcOverride *rc_override;
  1238. int rc_override_count;
  1239. /**
  1240. * rate control equation
  1241. * - encoding: Set by user
  1242. * - decoding: unused
  1243. */
  1244. const char *rc_eq;
  1245. /**
  1246. * maximum bitrate
  1247. * - encoding: Set by user.
  1248. * - decoding: unused
  1249. */
  1250. int rc_max_rate;
  1251. /**
  1252. * minimum bitrate
  1253. * - encoding: Set by user.
  1254. * - decoding: unused
  1255. */
  1256. int rc_min_rate;
  1257. /**
  1258. * decoder bitstream buffer size
  1259. * - encoding: Set by user.
  1260. * - decoding: unused
  1261. */
  1262. int rc_buffer_size;
  1263. float rc_buffer_aggressivity;
  1264. /**
  1265. * qscale factor between P and I-frames
  1266. * If > 0 then the last p frame quantizer will be used (q= lastp_q*factor+offset).
  1267. * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
  1268. * - encoding: Set by user.
  1269. * - decoding: unused
  1270. */
  1271. float i_quant_factor;
  1272. /**
  1273. * qscale offset between P and I-frames
  1274. * - encoding: Set by user.
  1275. * - decoding: unused
  1276. */
  1277. float i_quant_offset;
  1278. /**
  1279. * initial complexity for pass1 ratecontrol
  1280. * - encoding: Set by user.
  1281. * - decoding: unused
  1282. */
  1283. float rc_initial_cplx;
  1284. /**
  1285. * DCT algorithm, see FF_DCT_* below
  1286. * - encoding: Set by user.
  1287. * - decoding: unused
  1288. */
  1289. int dct_algo;
  1290. #define FF_DCT_AUTO 0
  1291. #define FF_DCT_FASTINT 1
  1292. #define FF_DCT_INT 2
  1293. #define FF_DCT_MMX 3
  1294. #define FF_DCT_MLIB 4
  1295. #define FF_DCT_ALTIVEC 5
  1296. #define FF_DCT_FAAN 6
  1297. /**
  1298. * luminance masking (0-> disabled)
  1299. * - encoding: Set by user.
  1300. * - decoding: unused
  1301. */
  1302. float lumi_masking;
  1303. /**
  1304. * temporary complexity masking (0-> disabled)
  1305. * - encoding: Set by user.
  1306. * - decoding: unused
  1307. */
  1308. float temporal_cplx_masking;
  1309. /**
  1310. * spatial complexity masking (0-> disabled)
  1311. * - encoding: Set by user.
  1312. * - decoding: unused
  1313. */
  1314. float spatial_cplx_masking;
  1315. /**
  1316. * p block masking (0-> disabled)
  1317. * - encoding: Set by user.
  1318. * - decoding: unused
  1319. */
  1320. float p_masking;
  1321. /**
  1322. * darkness masking (0-> disabled)
  1323. * - encoding: Set by user.
  1324. * - decoding: unused
  1325. */
  1326. float dark_masking;
  1327. /**
  1328. * IDCT algorithm, see FF_IDCT_* below.
  1329. * - encoding: Set by user.
  1330. * - decoding: Set by user.
  1331. */
  1332. int idct_algo;
  1333. #define FF_IDCT_AUTO 0
  1334. #define FF_IDCT_INT 1
  1335. #define FF_IDCT_SIMPLE 2
  1336. #define FF_IDCT_SIMPLEMMX 3
  1337. #define FF_IDCT_LIBMPEG2MMX 4
  1338. #define FF_IDCT_PS2 5
  1339. #define FF_IDCT_MLIB 6
  1340. #define FF_IDCT_ARM 7
  1341. #define FF_IDCT_ALTIVEC 8
  1342. #define FF_IDCT_SH4 9
  1343. #define FF_IDCT_SIMPLEARM 10
  1344. #define FF_IDCT_H264 11
  1345. #define FF_IDCT_VP3 12
  1346. #define FF_IDCT_IPP 13
  1347. #define FF_IDCT_XVIDMMX 14
  1348. #define FF_IDCT_CAVS 15
  1349. #define FF_IDCT_SIMPLEARMV5TE 16
  1350. #define FF_IDCT_SIMPLEARMV6 17
  1351. #define FF_IDCT_SIMPLEVIS 18
  1352. #define FF_IDCT_WMV2 19
  1353. #define FF_IDCT_FAAN 20
  1354. #define FF_IDCT_EA 21
  1355. #define FF_IDCT_SIMPLENEON 22
  1356. #define FF_IDCT_SIMPLEALPHA 23
  1357. /**
  1358. * slice count
  1359. * - encoding: Set by libavcodec.
  1360. * - decoding: Set by user (or 0).
  1361. */
  1362. int slice_count;
  1363. /**
  1364. * slice offsets in the frame in bytes
  1365. * - encoding: Set/allocated by libavcodec.
  1366. * - decoding: Set/allocated by user (or NULL).
  1367. */
  1368. int *slice_offset;
  1369. /**
  1370. * error concealment flags
  1371. * - encoding: unused
  1372. * - decoding: Set by user.
  1373. */
  1374. int error_concealment;
  1375. #define FF_EC_GUESS_MVS 1
  1376. #define FF_EC_DEBLOCK 2
  1377. /**
  1378. * dsp_mask could be add used to disable unwanted CPU features
  1379. * CPU features (i.e. MMX, SSE. ...)
  1380. *
  1381. * With the FORCE flag you may instead enable given CPU features.
  1382. * (Dangerous: Usable in case of misdetection, improper usage however will
  1383. * result into program crash.)
  1384. */
  1385. unsigned dsp_mask;
  1386. #define FF_MM_FORCE 0x80000000 /* Force usage of selected flags (OR) */
  1387. /* lower 16 bits - CPU features */
  1388. #define FF_MM_MMX 0x0001 ///< standard MMX
  1389. #define FF_MM_3DNOW 0x0004 ///< AMD 3DNOW
  1390. #if LIBAVCODEC_VERSION_MAJOR < 53
  1391. #define FF_MM_MMXEXT 0x0002 ///< SSE integer functions or AMD MMX ext
  1392. #endif
  1393. #define FF_MM_MMX2 0x0002 ///< SSE integer functions or AMD MMX ext
  1394. #define FF_MM_SSE 0x0008 ///< SSE functions
  1395. #define FF_MM_SSE2 0x0010 ///< PIV SSE2 functions
  1396. #define FF_MM_3DNOWEXT 0x0020 ///< AMD 3DNowExt
  1397. #define FF_MM_SSE3 0x0040 ///< Prescott SSE3 functions
  1398. #define FF_MM_SSSE3 0x0080 ///< Conroe SSSE3 functions
  1399. #define FF_MM_SSE4 0x0100 ///< Penryn SSE4.1 functions
  1400. #define FF_MM_SSE42 0x0200 ///< Nehalem SSE4.2 functions
  1401. #define FF_MM_IWMMXT 0x0100 ///< XScale IWMMXT
  1402. #define FF_MM_ALTIVEC 0x0001 ///< standard AltiVec
  1403. /**
  1404. * bits per sample/pixel from the demuxer (needed for huffyuv).
  1405. * - encoding: Set by libavcodec.
  1406. * - decoding: Set by user.
  1407. */
  1408. int bits_per_coded_sample;
  1409. /**
  1410. * prediction method (needed for huffyuv)
  1411. * - encoding: Set by user.
  1412. * - decoding: unused
  1413. */
  1414. int prediction_method;
  1415. #define FF_PRED_LEFT 0
  1416. #define FF_PRED_PLANE 1
  1417. #define FF_PRED_MEDIAN 2
  1418. /**
  1419. * sample aspect ratio (0 if unknown)
  1420. * That is the width of a pixel divided by the height of the pixel.
  1421. * Numerator and denominator must be relatively prime and smaller than 256 for some video standards.
  1422. * - encoding: Set by user.
  1423. * - decoding: Set by libavcodec.
  1424. */
  1425. AVRational sample_aspect_ratio;
  1426. /**
  1427. * the picture in the bitstream
  1428. * - encoding: Set by libavcodec.
  1429. * - decoding: Set by libavcodec.
  1430. */
  1431. AVFrame *coded_frame;
  1432. /**
  1433. * debug
  1434. * - encoding: Set by user.
  1435. * - decoding: Set by user.
  1436. */
  1437. int debug;
  1438. #define FF_DEBUG_PICT_INFO 1
  1439. #define FF_DEBUG_RC 2
  1440. #define FF_DEBUG_BITSTREAM 4
  1441. #define FF_DEBUG_MB_TYPE 8
  1442. #define FF_DEBUG_QP 16
  1443. #define FF_DEBUG_MV 32
  1444. #define FF_DEBUG_DCT_COEFF 0x00000040
  1445. #define FF_DEBUG_SKIP 0x00000080
  1446. #define FF_DEBUG_STARTCODE 0x00000100
  1447. #define FF_DEBUG_PTS 0x00000200
  1448. #define FF_DEBUG_ER 0x00000400
  1449. #define FF_DEBUG_MMCO 0x00000800
  1450. #define FF_DEBUG_BUGS 0x00001000
  1451. #define FF_DEBUG_VIS_QP 0x00002000
  1452. #define FF_DEBUG_VIS_MB_TYPE 0x00004000
  1453. #define FF_DEBUG_BUFFERS 0x00008000
  1454. /**
  1455. * debug
  1456. * - encoding: Set by user.
  1457. * - decoding: Set by user.
  1458. */
  1459. int debug_mv;
  1460. #define FF_DEBUG_VIS_MV_P_FOR 0x00000001 //visualize forward predicted MVs of P frames
  1461. #define FF_DEBUG_VIS_MV_B_FOR 0x00000002 //visualize forward predicted MVs of B frames
  1462. #define FF_DEBUG_VIS_MV_B_BACK 0x00000004 //visualize backward predicted MVs of B frames
  1463. /**
  1464. * error
  1465. * - encoding: Set by libavcodec if flags&CODEC_FLAG_PSNR.
  1466. * - decoding: unused
  1467. */
  1468. uint64_t error[4];
  1469. /**
  1470. * minimum MB quantizer
  1471. * - encoding: unused
  1472. * - decoding: unused
  1473. */
  1474. int mb_qmin;
  1475. /**
  1476. * maximum MB quantizer
  1477. * - encoding: unused
  1478. * - decoding: unused
  1479. */
  1480. int mb_qmax;
  1481. /**
  1482. * motion estimation comparison function
  1483. * - encoding: Set by user.
  1484. * - decoding: unused
  1485. */
  1486. int me_cmp;
  1487. /**
  1488. * subpixel motion estimation comparison function
  1489. * - encoding: Set by user.
  1490. * - decoding: unused
  1491. */
  1492. int me_sub_cmp;
  1493. /**
  1494. * macroblock comparison function (not supported yet)
  1495. * - encoding: Set by user.
  1496. * - decoding: unused
  1497. */
  1498. int mb_cmp;
  1499. /**
  1500. * interlaced DCT comparison function
  1501. * - encoding: Set by user.
  1502. * - decoding: unused
  1503. */
  1504. int ildct_cmp;
  1505. #define FF_CMP_SAD 0
  1506. #define FF_CMP_SSE 1
  1507. #define FF_CMP_SATD 2
  1508. #define FF_CMP_DCT 3
  1509. #define FF_CMP_PSNR 4
  1510. #define FF_CMP_BIT 5
  1511. #define FF_CMP_RD 6
  1512. #define FF_CMP_ZERO 7
  1513. #define FF_CMP_VSAD 8
  1514. #define FF_CMP_VSSE 9
  1515. #define FF_CMP_NSSE 10
  1516. #define FF_CMP_W53 11
  1517. #define FF_CMP_W97 12
  1518. #define FF_CMP_DCTMAX 13
  1519. #define FF_CMP_DCT264 14
  1520. #define FF_CMP_CHROMA 256
  1521. /**
  1522. * ME diamond size & shape
  1523. * - encoding: Set by user.
  1524. * - decoding: unused
  1525. */
  1526. int dia_size;
  1527. /**
  1528. * amount of previous MV predictors (2a+1 x 2a+1 square)
  1529. * - encoding: Set by user.
  1530. * - decoding: unused
  1531. */
  1532. int last_predictor_count;
  1533. /**
  1534. * prepass for motion estimation
  1535. * - encoding: Set by user.
  1536. * - decoding: unused
  1537. */
  1538. int pre_me;
  1539. /**
  1540. * motion estimation prepass comparison function
  1541. * - encoding: Set by user.
  1542. * - decoding: unused
  1543. */
  1544. int me_pre_cmp;
  1545. /**
  1546. * ME prepass diamond size & shape
  1547. * - encoding: Set by user.
  1548. * - decoding: unused
  1549. */
  1550. int pre_dia_size;
  1551. /**
  1552. * subpel ME quality
  1553. * - encoding: Set by user.
  1554. * - decoding: unused
  1555. */
  1556. int me_subpel_quality;
  1557. /**
  1558. * callback to negotiate the pixelFormat
  1559. * @param fmt is the list of formats which are supported by the codec,
  1560. * it is terminated by -1 as 0 is a valid format, the formats are ordered by quality.
  1561. * The first is always the native one.
  1562. * @return the chosen format
  1563. * - encoding: unused
  1564. * - decoding: Set by user, if not set the native format will be chosen.
  1565. */
  1566. enum PixelFormat (*get_format)(struct AVCodecContext *s, const enum PixelFormat * fmt);
  1567. /**
  1568. * DTG active format information (additional aspect ratio
  1569. * information only used in DVB MPEG-2 transport streams)
  1570. * 0 if not set.
  1571. *
  1572. * - encoding: unused
  1573. * - decoding: Set by decoder.
  1574. */
  1575. int dtg_active_format;
  1576. #define FF_DTG_AFD_SAME 8
  1577. #define FF_DTG_AFD_4_3 9
  1578. #define FF_DTG_AFD_16_9 10
  1579. #define FF_DTG_AFD_14_9 11
  1580. #define FF_DTG_AFD_4_3_SP_14_9 13
  1581. #define FF_DTG_AFD_16_9_SP_14_9 14
  1582. #define FF_DTG_AFD_SP_4_3 15
  1583. /**
  1584. * maximum motion estimation search range in subpel units
  1585. * If 0 then no limit.
  1586. *
  1587. * - encoding: Set by user.
  1588. * - decoding: unused
  1589. */
  1590. int me_range;
  1591. /**
  1592. * intra quantizer bias
  1593. * - encoding: Set by user.
  1594. * - decoding: unused
  1595. */
  1596. int intra_quant_bias;
  1597. #define FF_DEFAULT_QUANT_BIAS 999999
  1598. /**
  1599. * inter quantizer bias
  1600. * - encoding: Set by user.
  1601. * - decoding: unused
  1602. */
  1603. int inter_quant_bias;
  1604. /**
  1605. * color table ID
  1606. * - encoding: unused
  1607. * - decoding: Which clrtable should be used for 8bit RGB images.
  1608. * Tables have to be stored somewhere. FIXME
  1609. */
  1610. int color_table_id;
  1611. /**
  1612. * internal_buffer count
  1613. * Don't touch, used by libavcodec default_get_buffer().
  1614. */
  1615. int internal_buffer_count;
  1616. /**
  1617. * internal_buffers
  1618. * Don't touch, used by libavcodec default_get_buffer().
  1619. */
  1620. void *internal_buffer;
  1621. #define FF_LAMBDA_SHIFT 7
  1622. #define FF_LAMBDA_SCALE (1<<FF_LAMBDA_SHIFT)
  1623. #define FF_QP2LAMBDA 118 ///< factor to convert from H.263 QP to lambda
  1624. #define FF_LAMBDA_MAX (256*128-1)
  1625. #define FF_QUALITY_SCALE FF_LAMBDA_SCALE //FIXME maybe remove
  1626. /**
  1627. * Global quality for codecs which cannot change it per frame.
  1628. * This should be proportional to MPEG-1/2/4 qscale.
  1629. * - encoding: Set by user.
  1630. * - decoding: unused
  1631. */
  1632. int global_quality;
  1633. #define FF_CODER_TYPE_VLC 0
  1634. #define FF_CODER_TYPE_AC 1
  1635. #define FF_CODER_TYPE_RAW 2
  1636. #define FF_CODER_TYPE_RLE 3
  1637. #define FF_CODER_TYPE_DEFLATE 4
  1638. /**
  1639. * coder type
  1640. * - encoding: Set by user.
  1641. * - decoding: unused
  1642. */
  1643. int coder_type;
  1644. /**
  1645. * context model
  1646. * - encoding: Set by user.
  1647. * - decoding: unused
  1648. */
  1649. int context_model;
  1650. #if 0
  1651. /**
  1652. *
  1653. * - encoding: unused
  1654. * - decoding: Set by user.
  1655. */
  1656. uint8_t * (*realloc)(struct AVCodecContext *s, uint8_t *buf, int buf_size);
  1657. #endif
  1658. /**
  1659. * slice flags
  1660. * - encoding: unused
  1661. * - decoding: Set by user.
  1662. */
  1663. int slice_flags;
  1664. #define SLICE_FLAG_CODED_ORDER 0x0001 ///< draw_horiz_band() is called in coded order instead of display
  1665. #define SLICE_FLAG_ALLOW_FIELD 0x0002 ///< allow draw_horiz_band() with field slices (MPEG2 field pics)
  1666. #define SLICE_FLAG_ALLOW_PLANE 0x0004 ///< allow draw_horiz_band() with 1 component at a time (SVQ1)
  1667. /**
  1668. * XVideo Motion Acceleration
  1669. * - encoding: forbidden
  1670. * - decoding: set by decoder
  1671. */
  1672. int xvmc_acceleration;
  1673. /**
  1674. * macroblock decision mode
  1675. * - encoding: Set by user.
  1676. * - decoding: unused
  1677. */
  1678. int mb_decision;
  1679. #define FF_MB_DECISION_SIMPLE 0 ///< uses mb_cmp
  1680. #define FF_MB_DECISION_BITS 1 ///< chooses the one which needs the fewest bits
  1681. #define FF_MB_DECISION_RD 2 ///< rate distortion
  1682. /**
  1683. * custom intra quantization matrix
  1684. * - encoding: Set by user, can be NULL.
  1685. * - decoding: Set by libavcodec.
  1686. */
  1687. uint16_t *intra_matrix;
  1688. /**
  1689. * custom inter quantization matrix
  1690. * - encoding: Set by user, can be NULL.
  1691. * - decoding: Set by libavcodec.
  1692. */
  1693. uint16_t *inter_matrix;
  1694. /**
  1695. * fourcc from the AVI stream header (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
  1696. * This is used to work around some encoder bugs.
  1697. * - encoding: unused
  1698. * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
  1699. */
  1700. unsigned int stream_codec_tag;
  1701. /**
  1702. * scene change detection threshold
  1703. * 0 is default, larger means fewer detected scene changes.
  1704. * - encoding: Set by user.
  1705. * - decoding: unused
  1706. */
  1707. int scenechange_threshold;
  1708. /**
  1709. * minimum Lagrange multipler
  1710. * - encoding: Set by user.
  1711. * - decoding: unused
  1712. */
  1713. int lmin;
  1714. /**
  1715. * maximum Lagrange multipler
  1716. * - encoding: Set by user.
  1717. * - decoding: unused
  1718. */
  1719. int lmax;
  1720. /**
  1721. * palette control structure
  1722. * - encoding: ??? (no palette-enabled encoder yet)
  1723. * - decoding: Set by user.
  1724. */
  1725. struct AVPaletteControl *palctrl;
  1726. /**
  1727. * noise reduction strength
  1728. * - encoding: Set by user.
  1729. * - decoding: unused
  1730. */
  1731. int noise_reduction;
  1732. /**
  1733. * Called at the beginning of a frame to get cr buffer for it.
  1734. * Buffer type (size, hints) must be the same. libavcodec won't check it.
  1735. * libavcodec will pass previous buffer in pic, function should return
  1736. * same buffer or new buffer with old frame "painted" into it.
  1737. * If pic.data[0] == NULL must behave like get_buffer().
  1738. * - encoding: unused
  1739. * - decoding: Set by libavcodec., user can override
  1740. */
  1741. int (*reget_buffer)(struct AVCodecContext *c, AVFrame *pic);
  1742. /**
  1743. * Number of bits which should be loaded into the rc buffer before decoding starts.
  1744. * - encoding: Set by user.
  1745. * - decoding: unused
  1746. */
  1747. int rc_initial_buffer_occupancy;
  1748. /**
  1749. *
  1750. * - encoding: Set by user.
  1751. * - decoding: unused
  1752. */
  1753. int inter_threshold;
  1754. /**
  1755. * CODEC_FLAG2_*
  1756. * - encoding: Set by user.
  1757. * - decoding: Set by user.
  1758. */
  1759. int flags2;
  1760. /**
  1761. * Simulates errors in the bitstream to test error concealment.
  1762. * - encoding: Set by user.
  1763. * - decoding: unused
  1764. */
  1765. int error_rate;
  1766. /**
  1767. * MP3 antialias algorithm, see FF_AA_* below.
  1768. * - encoding: unused
  1769. * - decoding: Set by user.
  1770. */
  1771. int antialias_algo;
  1772. #define FF_AA_AUTO 0
  1773. #define FF_AA_FASTINT 1 //not implemented yet
  1774. #define FF_AA_INT 2
  1775. #define FF_AA_FLOAT 3
  1776. /**
  1777. * quantizer noise shaping
  1778. * - encoding: Set by user.
  1779. * - decoding: unused
  1780. */
  1781. int quantizer_noise_shaping;
  1782. /**
  1783. * thread count
  1784. * is used to decide how many independent tasks should be passed to execute()
  1785. * - encoding: Set by user.
  1786. * - decoding: Set by user.
  1787. */
  1788. int thread_count;
  1789. /**
  1790. * The codec may call this to execute several independent things.
  1791. * It will return only after finishing all tasks.
  1792. * The user may replace this with some multithreaded implementation,
  1793. * the default implementation will execute the parts serially.
  1794. * @param count the number of things to execute
  1795. * - encoding: Set by libavcodec, user can override.
  1796. * - decoding: Set by libavcodec, user can override.
  1797. */
  1798. int (*execute)(struct AVCodecContext *c, int (*func)(struct AVCodecContext *c2, void *arg), void *arg2, int *ret, int count, int size);
  1799. /**
  1800. * thread opaque
  1801. * Can be used by execute() to store some per AVCodecContext stuff.
  1802. * - encoding: set by execute()
  1803. * - decoding: set by execute()
  1804. */
  1805. void *thread_opaque;
  1806. /**
  1807. * Motion estimation threshold below which no motion estimation is
  1808. * performed, but instead the user specified motion vectors are used.
  1809. *
  1810. * - encoding: Set by user.
  1811. * - decoding: unused
  1812. */
  1813. int me_threshold;
  1814. /**
  1815. * Macroblock threshold below which the user specified macroblock types will be used.
  1816. * - encoding: Set by user.
  1817. * - decoding: unused
  1818. */
  1819. int mb_threshold;
  1820. /**
  1821. * precision of the intra DC coefficient - 8
  1822. * - encoding: Set by user.
  1823. * - decoding: unused
  1824. */
  1825. int intra_dc_precision;
  1826. /**
  1827. * noise vs. sse weight for the nsse comparsion function
  1828. * - encoding: Set by user.
  1829. * - decoding: unused
  1830. */
  1831. int nsse_weight;
  1832. /**
  1833. * Number of macroblock rows at the top which are skipped.
  1834. * - encoding: unused
  1835. * - decoding: Set by user.
  1836. */
  1837. int skip_top;
  1838. /**
  1839. * Number of macroblock rows at the bottom which are skipped.
  1840. * - encoding: unused
  1841. * - decoding: Set by user.
  1842. */
  1843. int skip_bottom;
  1844. /**
  1845. * profile
  1846. * - encoding: Set by user.
  1847. * - decoding: Set by libavcodec.
  1848. */
  1849. int profile;
  1850. #define FF_PROFILE_UNKNOWN -99
  1851. #define FF_PROFILE_AAC_MAIN 0
  1852. #define FF_PROFILE_AAC_LOW 1
  1853. #define FF_PROFILE_AAC_SSR 2
  1854. #define FF_PROFILE_AAC_LTP 3
  1855. /**
  1856. * level
  1857. * - encoding: Set by user.
  1858. * - decoding: Set by libavcodec.
  1859. */
  1860. int level;
  1861. #define FF_LEVEL_UNKNOWN -99
  1862. /**
  1863. * low resolution decoding, 1-> 1/2 size, 2->1/4 size
  1864. * - encoding: unused
  1865. * - decoding: Set by user.
  1866. */
  1867. int lowres;
  1868. /**
  1869. * Bitstream width / height, may be different from width/height if lowres
  1870. * or other things are used.
  1871. * - encoding: unused
  1872. * - decoding: Set by user before init if known. Codec should override / dynamically change if needed.
  1873. */
  1874. int coded_width, coded_height;
  1875. /**
  1876. * frame skip threshold
  1877. * - encoding: Set by user.
  1878. * - decoding: unused
  1879. */
  1880. int frame_skip_threshold;
  1881. /**
  1882. * frame skip factor
  1883. * - encoding: Set by user.
  1884. * - decoding: unused
  1885. */
  1886. int frame_skip_factor;
  1887. /**
  1888. * frame skip exponent
  1889. * - encoding: Set by user.
  1890. * - decoding: unused
  1891. */
  1892. int frame_skip_exp;
  1893. /**
  1894. * frame skip comparison function
  1895. * - encoding: Set by user.
  1896. * - decoding: unused
  1897. */
  1898. int frame_skip_cmp;
  1899. /**
  1900. * Border processing masking, raises the quantizer for mbs on the borders
  1901. * of the picture.
  1902. * - encoding: Set by user.
  1903. * - decoding: unused
  1904. */
  1905. float border_masking;
  1906. /**
  1907. * minimum MB lagrange multipler
  1908. * - encoding: Set by user.
  1909. * - decoding: unused
  1910. */
  1911. int mb_lmin;
  1912. /**
  1913. * maximum MB lagrange multipler
  1914. * - encoding: Set by user.
  1915. * - decoding: unused
  1916. */
  1917. int mb_lmax;
  1918. /**
  1919. *
  1920. * - encoding: Set by user.
  1921. * - decoding: unused
  1922. */
  1923. int me_penalty_compensation;
  1924. /**
  1925. *
  1926. * - encoding: unused
  1927. * - decoding: Set by user.
  1928. */
  1929. enum AVDiscard skip_loop_filter;
  1930. /**
  1931. *
  1932. * - encoding: unused
  1933. * - decoding: Set by user.
  1934. */
  1935. enum AVDiscard skip_idct;
  1936. /**
  1937. *
  1938. * - encoding: unused
  1939. * - decoding: Set by user.
  1940. */
  1941. enum AVDiscard skip_frame;
  1942. /**
  1943. *
  1944. * - encoding: Set by user.
  1945. * - decoding: unused
  1946. */
  1947. int bidir_refine;
  1948. /**
  1949. *
  1950. * - encoding: Set by user.
  1951. * - decoding: unused
  1952. */
  1953. int brd_scale;
  1954. /**
  1955. * constant rate factor - quality-based VBR - values ~correspond to qps
  1956. * - encoding: Set by user.
  1957. * - decoding: unused
  1958. */
  1959. float crf;
  1960. /**
  1961. * constant quantization parameter rate control method
  1962. * - encoding: Set by user.
  1963. * - decoding: unused
  1964. */
  1965. int cqp;
  1966. /**
  1967. * minimum GOP size
  1968. * - encoding: Set by user.
  1969. * - decoding: unused
  1970. */
  1971. int keyint_min;
  1972. /**
  1973. * number of reference frames
  1974. * - encoding: Set by user.
  1975. * - decoding: Set by lavc.
  1976. */
  1977. int refs;
  1978. /**
  1979. * chroma qp offset from luma
  1980. * - encoding: Set by user.
  1981. * - decoding: unused
  1982. */
  1983. int chromaoffset;
  1984. /**
  1985. * Influences how often B-frames are used.
  1986. * - encoding: Set by user.
  1987. * - decoding: unused
  1988. */
  1989. int bframebias;
  1990. /**
  1991. * trellis RD quantization
  1992. * - encoding: Set by user.
  1993. * - decoding: unused
  1994. */
  1995. int trellis;
  1996. /**
  1997. * Reduce fluctuations in qp (before curve compression).
  1998. * - encoding: Set by user.
  1999. * - decoding: unused
  2000. */
  2001. float complexityblur;
  2002. /**
  2003. * in-loop deblocking filter alphac0 parameter
  2004. * alpha is in the range -6...6
  2005. * - encoding: Set by user.
  2006. * - decoding: unused
  2007. */
  2008. int deblockalpha;
  2009. /**
  2010. * in-loop deblocking filter beta parameter
  2011. * beta is in the range -6...6
  2012. * - encoding: Set by user.
  2013. * - decoding: unused
  2014. */
  2015. int deblockbeta;
  2016. /**
  2017. * macroblock subpartition sizes to consider - p8x8, p4x4, b8x8, i8x8, i4x4
  2018. * - encoding: Set by user.
  2019. * - decoding: unused
  2020. */
  2021. int partitions;
  2022. #define X264_PART_I4X4 0x001 /* Analyze i4x4 */
  2023. #define X264_PART_I8X8 0x002 /* Analyze i8x8 (requires 8x8 transform) */
  2024. #define X264_PART_P8X8 0x010 /* Analyze p16x8, p8x16 and p8x8 */
  2025. #define X264_PART_P4X4 0x020 /* Analyze p8x4, p4x8, p4x4 */
  2026. #define X264_PART_B8X8 0x100 /* Analyze b16x8, b8x16 and b8x8 */
  2027. /**
  2028. * direct MV prediction mode - 0 (none), 1 (spatial), 2 (temporal), 3 (auto)
  2029. * - encoding: Set by user.
  2030. * - decoding: unused
  2031. */
  2032. int directpred;
  2033. /**
  2034. * Audio cutoff bandwidth (0 means "automatic")
  2035. * - encoding: Set by user.
  2036. * - decoding: unused
  2037. */
  2038. int cutoff;
  2039. /**
  2040. * Multiplied by qscale for each frame and added to scene_change_score.
  2041. * - encoding: Set by user.
  2042. * - decoding: unused
  2043. */
  2044. int scenechange_factor;
  2045. /**
  2046. *
  2047. * Note: Value depends upon the compare function used for fullpel ME.
  2048. * - encoding: Set by user.
  2049. * - decoding: unused
  2050. */
  2051. int mv0_threshold;
  2052. /**
  2053. * Adjusts sensitivity of b_frame_strategy 1.
  2054. * - encoding: Set by user.
  2055. * - decoding: unused
  2056. */
  2057. int b_sensitivity;
  2058. /**
  2059. * - encoding: Set by user.
  2060. * - decoding: unused
  2061. */
  2062. int compression_level;
  2063. #define FF_COMPRESSION_DEFAULT -1
  2064. /**
  2065. * Sets whether to use LPC mode - used by FLAC encoder.
  2066. * - encoding: Set by user.
  2067. * - decoding: unused
  2068. */
  2069. int use_lpc;
  2070. /**
  2071. * LPC coefficient precision - used by FLAC encoder
  2072. * - encoding: Set by user.
  2073. * - decoding: unused
  2074. */
  2075. int lpc_coeff_precision;
  2076. /**
  2077. * - encoding: Set by user.
  2078. * - decoding: unused
  2079. */
  2080. int min_prediction_order;
  2081. /**
  2082. * - encoding: Set by user.
  2083. * - decoding: unused
  2084. */
  2085. int max_prediction_order;
  2086. /**
  2087. * search method for selecting prediction order
  2088. * - encoding: Set by user.
  2089. * - decoding: unused
  2090. */
  2091. int prediction_order_method;
  2092. /**
  2093. * - encoding: Set by user.
  2094. * - decoding: unused
  2095. */
  2096. int min_partition_order;
  2097. /**
  2098. * - encoding: Set by user.
  2099. * - decoding: unused
  2100. */
  2101. int max_partition_order;
  2102. /**
  2103. * GOP timecode frame start number, in non drop frame format
  2104. * - encoding: Set by user.
  2105. * - decoding: unused
  2106. */
  2107. int64_t timecode_frame_start;
  2108. #if LIBAVCODEC_VERSION_MAJOR < 53
  2109. /**
  2110. * Decoder should decode to this many channels if it can (0 for default)
  2111. * - encoding: unused
  2112. * - decoding: Set by user.
  2113. * @deprecated Deprecated in favor of request_channel_layout.
  2114. */
  2115. int request_channels;
  2116. #endif
  2117. /**
  2118. * Percentage of dynamic range compression to be applied by the decoder.
  2119. * The default value is 1.0, corresponding to full compression.
  2120. * - encoding: unused
  2121. * - decoding: Set by user.
  2122. */
  2123. float drc_scale;
  2124. /**
  2125. * opaque 64bit number (generally a PTS) that will be reordered and
  2126. * output in AVFrame.reordered_opaque
  2127. * - encoding: unused
  2128. * - decoding: Set by user.
  2129. */
  2130. int64_t reordered_opaque;
  2131. /**
  2132. * Bits per sample/pixel of internal libavcodec pixel/sample format.
  2133. * This field is applicable only when sample_fmt is SAMPLE_FMT_S32.
  2134. * - encoding: set by user.
  2135. * - decoding: set by libavcodec.
  2136. */
  2137. int bits_per_raw_sample;
  2138. /**
  2139. * Audio channel layout.
  2140. * - encoding: set by user.
  2141. * - decoding: set by libavcodec.
  2142. */
  2143. int64_t channel_layout;
  2144. /**
  2145. * Request decoder to use this channel layout if it can (0 for default)
  2146. * - encoding: unused
  2147. * - decoding: Set by user.
  2148. */
  2149. int64_t request_channel_layout;
  2150. /**
  2151. * Ratecontrol attempt to use, at maximum, <value> of what can be used without an underflow.
  2152. * - encoding: Set by user.
  2153. * - decoding: unused.
  2154. */
  2155. float rc_max_available_vbv_use;
  2156. /**
  2157. * Ratecontrol attempt to use, at least, <value> times the amount needed to prevent a vbv overflow.
  2158. * - encoding: Set by user.
  2159. * - decoding: unused.
  2160. */
  2161. float rc_min_vbv_overflow_use;
  2162. /**
  2163. * Hardware accelerator in use
  2164. * - encoding: unused.
  2165. * - decoding: Set by libavcodec
  2166. */
  2167. struct AVHWAccel *hwaccel;
  2168. /**
  2169. * For some codecs, the time base is closer to the field rate than the frame rate.
  2170. * Most notably, H.264 and MPEG-2 specify time_base as half of frame duration
  2171. * if no telecine is used ...
  2172. *
  2173. * Set to time_base ticks per frame. Default 1, e.g., H.264/MPEG-2 set it to 2.
  2174. */
  2175. int ticks_per_frame;
  2176. /**
  2177. * Hardware accelerator context.
  2178. * For some hardware accelerators, a global context needs to be
  2179. * provided by the user. In that case, this holds display-dependent
  2180. * data FFmpeg cannot instantiate itself. Please refer to the
  2181. * FFmpeg HW accelerator documentation to know how to fill this
  2182. * is. e.g. for VA API, this is a struct vaapi_context.
  2183. * - encoding: unused
  2184. * - decoding: Set by user
  2185. */
  2186. void *hwaccel_context;
  2187. } AVCodecContext;
  2188. /**
  2189. * AVCodec.
  2190. */
  2191. typedef struct AVCodec {
  2192. /**
  2193. * Name of the codec implementation.
  2194. * The name is globally unique among encoders and among decoders (but an
  2195. * encoder and a decoder can share the same name).
  2196. * This is the primary way to find a codec from the user perspective.
  2197. */
  2198. const char *name;
  2199. enum CodecType type;
  2200. enum CodecID id;
  2201. int priv_data_size;
  2202. int (*init)(AVCodecContext *);
  2203. int (*encode)(AVCodecContext *, uint8_t *buf, int buf_size, void *data);
  2204. int (*close)(AVCodecContext *);
  2205. int (*decode)(AVCodecContext *, void *outdata, int *outdata_size, AVPacket *avpkt);
  2206. /**
  2207. * Codec capabilities.
  2208. * see CODEC_CAP_*
  2209. */
  2210. int capabilities;
  2211. struct AVCodec *next;
  2212. /**
  2213. * Flush buffers.
  2214. * Will be called when seeking
  2215. */
  2216. void (*flush)(AVCodecContext *);
  2217. const AVRational *supported_framerates; ///< array of supported framerates, or NULL if any, array is terminated by {0,0}
  2218. const enum PixelFormat *pix_fmts; ///< array of supported pixel formats, or NULL if unknown, array is terminated by -1
  2219. /**
  2220. * Descriptive name for the codec, meant to be more human readable than \p name.
  2221. * You \e should use the NULL_IF_CONFIG_SMALL() macro to define it.
  2222. */
  2223. const char *long_name;
  2224. const int *supported_samplerates; ///< array of supported audio samplerates, or NULL if unknown, array is terminated by 0
  2225. const enum SampleFormat *sample_fmts; ///< array of supported sample formats, or NULL if unknown, array is terminated by -1
  2226. const int64_t *channel_layouts; ///< array of support channel layouts, or NULL if unknown. array is terminated by 0
  2227. } AVCodec;
  2228. /**
  2229. * AVHWAccel.
  2230. */
  2231. typedef struct AVHWAccel {
  2232. /**
  2233. * Name of the hardware accelerated codec.
  2234. * The name is globally unique among encoders and among decoders (but an
  2235. * encoder and a decoder can share the same name).
  2236. */
  2237. const char *name;
  2238. /**
  2239. * Type of codec implemented by the hardware accelerator.
  2240. *
  2241. * See CODEC_TYPE_xxx
  2242. */
  2243. enum CodecType type;
  2244. /**
  2245. * Codec implemented by the hardware accelerator.
  2246. *
  2247. * See CODEC_ID_xxx
  2248. */
  2249. enum CodecID id;
  2250. /**
  2251. * Supported pixel format.
  2252. *
  2253. * Only hardware accelerated formats are supported here.
  2254. */
  2255. enum PixelFormat pix_fmt;
  2256. /**
  2257. * Hardware accelerated codec capabilities.
  2258. * see FF_HWACCEL_CODEC_CAP_*
  2259. */
  2260. int capabilities;
  2261. struct AVHWAccel *next;
  2262. /**
  2263. * Called at the beginning of each frame or field picture.
  2264. *
  2265. * Meaningful frame information (codec specific) is guaranteed to
  2266. * be parsed at this point. This function is mandatory.
  2267. *
  2268. * Note that \p buf can be NULL along with \p buf_size set to 0.
  2269. * Otherwise, this means the whole frame is available at this point.
  2270. *
  2271. * @param avctx the codec context
  2272. * @param buf the frame data buffer base
  2273. * @param buf_size the size of the frame in bytes
  2274. * @return zero if successful, a negative value otherwise
  2275. */
  2276. int (*start_frame)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
  2277. /**
  2278. * Callback for each slice.
  2279. *
  2280. * Meaningful slice information (codec specific) is guaranteed to
  2281. * be parsed at this point. This function is mandatory.
  2282. *
  2283. * @param avctx the codec context
  2284. * @param buf the slice data buffer base
  2285. * @param buf_size the size of the slice in bytes
  2286. * @return zero if successful, a negative value otherwise
  2287. */
  2288. int (*decode_slice)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
  2289. /**
  2290. * Called at the end of each frame or field picture.
  2291. *
  2292. * The whole picture is parsed at this point and can now be sent
  2293. * to the hardware accelerator. This function is mandatory.
  2294. *
  2295. * @param avctx the codec context
  2296. * @return zero if successful, a negative value otherwise
  2297. */
  2298. int (*end_frame)(AVCodecContext *avctx);
  2299. /**
  2300. * Size of HW accelerator private data.
  2301. *
  2302. * Private data is allocated with av_mallocz() before
  2303. * AVCodecContext.get_buffer() and deallocated after
  2304. * AVCodecContext.release_buffer().
  2305. */
  2306. int priv_data_size;
  2307. } AVHWAccel;
  2308. /**
  2309. * four components are given, that's all.
  2310. * the last component is alpha
  2311. */
  2312. typedef struct AVPicture {
  2313. uint8_t *data[4];
  2314. int linesize[4]; ///< number of bytes per line
  2315. } AVPicture;
  2316. #if LIBAVCODEC_VERSION_MAJOR < 53
  2317. /**
  2318. * AVPaletteControl
  2319. * This structure defines a method for communicating palette changes
  2320. * between and demuxer and a decoder.
  2321. *
  2322. * @deprecated Use AVPacket to send palette changes instead.
  2323. * This is totally broken.
  2324. */
  2325. #define AVPALETTE_SIZE 1024
  2326. #define AVPALETTE_COUNT 256
  2327. typedef struct AVPaletteControl {
  2328. /* Demuxer sets this to 1 to indicate the palette has changed;
  2329. * decoder resets to 0. */
  2330. int palette_changed;
  2331. /* 4-byte ARGB palette entries, stored in native byte order; note that
  2332. * the individual palette components should be on a 8-bit scale; if
  2333. * the palette data comes from an IBM VGA native format, the component
  2334. * data is probably 6 bits in size and needs to be scaled. */
  2335. unsigned int palette[AVPALETTE_COUNT];
  2336. } AVPaletteControl attribute_deprecated;
  2337. #endif
  2338. enum AVSubtitleType {
  2339. SUBTITLE_NONE,
  2340. SUBTITLE_BITMAP, ///< A bitmap, pict will be set
  2341. /**
  2342. * Plain text, the text field must be set by the decoder and is
  2343. * authoritative. ass and pict fields may contain approximations.
  2344. */
  2345. SUBTITLE_TEXT,
  2346. /**
  2347. * Formatted text, the ass field must be set by the decoder and is
  2348. * authoritative. pict and text fields may contain approximations.
  2349. */
  2350. SUBTITLE_ASS,
  2351. };
  2352. typedef struct AVSubtitleRect {
  2353. int x; ///< top left corner of pict, undefined when pict is not set
  2354. int y; ///< top left corner of pict, undefined when pict is not set
  2355. int w; ///< width of pict, undefined when pict is not set
  2356. int h; ///< height of pict, undefined when pict is not set
  2357. int nb_colors; ///< number of colors in pict, undefined when pict is not set
  2358. /**
  2359. * data+linesize for the bitmap of this subtitle.
  2360. * can be set for text/ass as well once they where rendered
  2361. */
  2362. AVPicture pict;
  2363. enum AVSubtitleType type;
  2364. char *text; ///< 0 terminated plain UTF-8 text
  2365. /**
  2366. * 0 terminated ASS/SSA compatible event line.
  2367. * The pressentation of this is unaffected by the other values in this
  2368. * struct.
  2369. */
  2370. char *ass;
  2371. } AVSubtitleRect;
  2372. typedef struct AVSubtitle {
  2373. uint16_t format; /* 0 = graphics */
  2374. uint32_t start_display_time; /* relative to packet pts, in ms */
  2375. uint32_t end_display_time; /* relative to packet pts, in ms */
  2376. unsigned num_rects;
  2377. AVSubtitleRect **rects;
  2378. } AVSubtitle;
  2379. /* packet functions */
  2380. void av_destruct_packet_nofree(AVPacket *pkt);
  2381. /**
  2382. * Default packet destructor.
  2383. */
  2384. void av_destruct_packet(AVPacket *pkt);
  2385. /**
  2386. * Initialize optional fields of a packet with default values.
  2387. *
  2388. * @param pkt packet
  2389. */
  2390. void av_init_packet(AVPacket *pkt);
  2391. /**
  2392. * Allocate the payload of a packet and initialize its fields with
  2393. * default values.
  2394. *
  2395. * @param pkt packet
  2396. * @param size wanted payload size
  2397. * @return 0 if OK, AVERROR_xxx otherwise
  2398. */
  2399. int av_new_packet(AVPacket *pkt, int size);
  2400. /**
  2401. * Reduce packet size, correctly zeroing padding
  2402. *
  2403. * @param pkt packet
  2404. * @param size new size
  2405. */
  2406. void av_shrink_packet(AVPacket *pkt, int size);
  2407. /**
  2408. * @warning This is a hack - the packet memory allocation stuff is broken. The
  2409. * packet is allocated if it was not really allocated.
  2410. */
  2411. int av_dup_packet(AVPacket *pkt);
  2412. /**
  2413. * Free a packet.
  2414. *
  2415. * @param pkt packet to free
  2416. */
  2417. static inline void av_free_packet(AVPacket *pkt)
  2418. {
  2419. if (pkt && pkt->destruct) {
  2420. pkt->destruct(pkt);
  2421. }
  2422. }
  2423. /* resample.c */
  2424. struct ReSampleContext;
  2425. struct AVResampleContext;
  2426. typedef struct ReSampleContext ReSampleContext;
  2427. #if LIBAVCODEC_VERSION_MAJOR < 53
  2428. /**
  2429. * @deprecated Use av_audio_resample_init() instead.
  2430. */
  2431. attribute_deprecated ReSampleContext *audio_resample_init(int output_channels, int input_channels,
  2432. int output_rate, int input_rate);
  2433. #endif
  2434. /**
  2435. * Initializes audio resampling context
  2436. *
  2437. * @param output_channels number of output channels
  2438. * @param input_channels number of input channels
  2439. * @param output_rate output sample rate
  2440. * @param input_rate input sample rate
  2441. * @param sample_fmt_out requested output sample format
  2442. * @param sample_fmt_in input sample format
  2443. * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
  2444. * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
  2445. * @param linear If 1 then the used FIR filter will be linearly interpolated
  2446. between the 2 closest, if 0 the closest will be used
  2447. * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
  2448. * @return allocated ReSampleContext, NULL if error occured
  2449. */
  2450. ReSampleContext *av_audio_resample_init(int output_channels, int input_channels,
  2451. int output_rate, int input_rate,
  2452. enum SampleFormat sample_fmt_out,
  2453. enum SampleFormat sample_fmt_in,
  2454. int filter_length, int log2_phase_count,
  2455. int linear, double cutoff);
  2456. int audio_resample(ReSampleContext *s, short *output, short *input, int nb_samples);
  2457. void audio_resample_close(ReSampleContext *s);
  2458. /**
  2459. * Initializes an audio resampler.
  2460. * Note, if either rate is not an integer then simply scale both rates up so they are.
  2461. * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
  2462. * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
  2463. * @param linear If 1 then the used FIR filter will be linearly interpolated
  2464. between the 2 closest, if 0 the closest will be used
  2465. * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
  2466. */
  2467. struct AVResampleContext *av_resample_init(int out_rate, int in_rate, int filter_length, int log2_phase_count, int linear, double cutoff);
  2468. /**
  2469. * resamples.
  2470. * @param src an array of unconsumed samples
  2471. * @param consumed the number of samples of src which have been consumed are returned here
  2472. * @param src_size the number of unconsumed samples available
  2473. * @param dst_size the amount of space in samples available in dst
  2474. * @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.
  2475. * @return the number of samples written in dst or -1 if an error occurred
  2476. */
  2477. int av_resample(struct AVResampleContext *c, short *dst, short *src, int *consumed, int src_size, int dst_size, int update_ctx);
  2478. /**
  2479. * Compensates samplerate/timestamp drift. The compensation is done by changing
  2480. * the resampler parameters, so no audible clicks or similar distortions occur
  2481. * @param compensation_distance distance in output samples over which the compensation should be performed
  2482. * @param sample_delta number of output samples which should be output less
  2483. *
  2484. * example: av_resample_compensate(c, 10, 500)
  2485. * here instead of 510 samples only 500 samples would be output
  2486. *
  2487. * note, due to rounding the actual compensation might be slightly different,
  2488. * especially if the compensation_distance is large and the in_rate used during init is small
  2489. */
  2490. void av_resample_compensate(struct AVResampleContext *c, int sample_delta, int compensation_distance);
  2491. void av_resample_close(struct AVResampleContext *c);
  2492. /**
  2493. * Allocate memory for a picture. Call avpicture_free to free it.
  2494. *
  2495. * @param picture the picture to be filled in
  2496. * @param pix_fmt the format of the picture
  2497. * @param width the width of the picture
  2498. * @param height the height of the picture
  2499. * @return zero if successful, a negative value if not
  2500. */
  2501. int avpicture_alloc(AVPicture *picture, enum PixelFormat pix_fmt, int width, int height);
  2502. /**
  2503. * Free a picture previously allocated by avpicture_alloc().
  2504. *
  2505. * @param picture the AVPicture to be freed
  2506. */
  2507. void avpicture_free(AVPicture *picture);
  2508. /**
  2509. * Fill in the AVPicture fields.
  2510. * The fields of the given AVPicture are filled in by using the 'ptr' address
  2511. * which points to the image data buffer. Depending on the specified picture
  2512. * format, one or multiple image data pointers and line sizes will be set.
  2513. * If a planar format is specified, several pointers will be set pointing to
  2514. * the different picture planes and the line sizes of the different planes
  2515. * will be stored in the lines_sizes array.
  2516. * Call with ptr == NULL to get the required size for the ptr buffer.
  2517. *
  2518. * @param picture AVPicture whose fields are to be filled in
  2519. * @param ptr Buffer which will contain or contains the actual image data
  2520. * @param pix_fmt The format in which the picture data is stored.
  2521. * @param width the width of the image in pixels
  2522. * @param height the height of the image in pixels
  2523. * @return size of the image data in bytes
  2524. */
  2525. int avpicture_fill(AVPicture *picture, uint8_t *ptr,
  2526. int pix_fmt, int width, int height);
  2527. int avpicture_layout(const AVPicture* src, enum PixelFormat pix_fmt, int width, int height,
  2528. unsigned char *dest, int dest_size);
  2529. /**
  2530. * Calculate the size in bytes that a picture of the given width and height
  2531. * would occupy if stored in the given picture format.
  2532. * Note that this returns the size of a compact representation as generated
  2533. * by avpicture_layout, which can be smaller than the size required for e.g.
  2534. * avpicture_fill.
  2535. *
  2536. * @param pix_fmt the given picture format
  2537. * @param width the width of the image
  2538. * @param height the height of the image
  2539. * @return Image data size in bytes or -1 on error (e.g. too large dimensions).
  2540. */
  2541. int avpicture_get_size(enum PixelFormat pix_fmt, int width, int height);
  2542. void avcodec_get_chroma_sub_sample(enum PixelFormat pix_fmt, int *h_shift, int *v_shift);
  2543. const char *avcodec_get_pix_fmt_name(enum PixelFormat pix_fmt);
  2544. void avcodec_set_dimensions(AVCodecContext *s, int width, int height);
  2545. /**
  2546. * Returns the pixel format corresponding to the name \p name.
  2547. *
  2548. * If there is no pixel format with name \p name, then looks for a
  2549. * pixel format with the name corresponding to the native endian
  2550. * format of \p name.
  2551. * For example in a little-endian system, first looks for "gray16",
  2552. * then for "gray16le".
  2553. *
  2554. * Finally if no pixel format has been found, returns \c PIX_FMT_NONE.
  2555. */
  2556. enum PixelFormat avcodec_get_pix_fmt(const char* name);
  2557. unsigned int avcodec_pix_fmt_to_codec_tag(enum PixelFormat p);
  2558. #define FF_LOSS_RESOLUTION 0x0001 /**< loss due to resolution change */
  2559. #define FF_LOSS_DEPTH 0x0002 /**< loss due to color depth change */
  2560. #define FF_LOSS_COLORSPACE 0x0004 /**< loss due to color space conversion */
  2561. #define FF_LOSS_ALPHA 0x0008 /**< loss of alpha bits */
  2562. #define FF_LOSS_COLORQUANT 0x0010 /**< loss due to color quantization */
  2563. #define FF_LOSS_CHROMA 0x0020 /**< loss of chroma (e.g. RGB to gray conversion) */
  2564. /**
  2565. * Computes what kind of losses will occur when converting from one specific
  2566. * pixel format to another.
  2567. * When converting from one pixel format to another, information loss may occur.
  2568. * For example, when converting from RGB24 to GRAY, the color information will
  2569. * be lost. Similarly, other losses occur when converting from some formats to
  2570. * other formats. These losses can involve loss of chroma, but also loss of
  2571. * resolution, loss of color depth, loss due to the color space conversion, loss
  2572. * of the alpha bits or loss due to color quantization.
  2573. * avcodec_get_fix_fmt_loss() informs you about the various types of losses
  2574. * which will occur when converting from one pixel format to another.
  2575. *
  2576. * @param[in] dst_pix_fmt destination pixel format
  2577. * @param[in] src_pix_fmt source pixel format
  2578. * @param[in] has_alpha Whether the source pixel format alpha channel is used.
  2579. * @return Combination of flags informing you what kind of losses will occur.
  2580. */
  2581. int avcodec_get_pix_fmt_loss(enum PixelFormat dst_pix_fmt, enum PixelFormat src_pix_fmt,
  2582. int has_alpha);
  2583. /**
  2584. * Finds the best pixel format to convert to given a certain source pixel
  2585. * format. When converting from one pixel format to another, information loss
  2586. * may occur. For example, when converting from RGB24 to GRAY, the color
  2587. * information will be lost. Similarly, other losses occur when converting from
  2588. * some formats to other formats. avcodec_find_best_pix_fmt() searches which of
  2589. * the given pixel formats should be used to suffer the least amount of loss.
  2590. * The pixel formats from which it chooses one, are determined by the
  2591. * \p pix_fmt_mask parameter.
  2592. *
  2593. * @code
  2594. * src_pix_fmt = PIX_FMT_YUV420P;
  2595. * pix_fmt_mask = (1 << PIX_FMT_YUV422P) || (1 << PIX_FMT_RGB24);
  2596. * dst_pix_fmt = avcodec_find_best_pix_fmt(pix_fmt_mask, src_pix_fmt, alpha, &loss);
  2597. * @endcode
  2598. *
  2599. * @param[in] pix_fmt_mask bitmask determining which pixel format to choose from
  2600. * @param[in] src_pix_fmt source pixel format
  2601. * @param[in] has_alpha Whether the source pixel format alpha channel is used.
  2602. * @param[out] loss_ptr Combination of flags informing you what kind of losses will occur.
  2603. * @return The best pixel format to convert to or -1 if none was found.
  2604. */
  2605. enum PixelFormat avcodec_find_best_pix_fmt(int64_t pix_fmt_mask, enum PixelFormat src_pix_fmt,
  2606. int has_alpha, int *loss_ptr);
  2607. /**
  2608. * Print in buf the string corresponding to the pixel format with
  2609. * number pix_fmt, or an header if pix_fmt is negative.
  2610. *
  2611. * @param[in] buf the buffer where to write the string
  2612. * @param[in] buf_size the size of buf
  2613. * @param[in] pix_fmt the number of the pixel format to print the corresponding info string, or
  2614. * a negative value to print the corresponding header.
  2615. * Meaningful values for obtaining a pixel format info vary from 0 to PIX_FMT_NB -1.
  2616. */
  2617. void avcodec_pix_fmt_string (char *buf, int buf_size, enum PixelFormat pix_fmt);
  2618. #define FF_ALPHA_TRANSP 0x0001 /* image has some totally transparent pixels */
  2619. #define FF_ALPHA_SEMI_TRANSP 0x0002 /* image has some transparent pixels */
  2620. /**
  2621. * Tell if an image really has transparent alpha values.
  2622. * @return ored mask of FF_ALPHA_xxx constants
  2623. */
  2624. int img_get_alpha_info(const AVPicture *src,
  2625. enum PixelFormat pix_fmt, int width, int height);
  2626. /* deinterlace a picture */
  2627. /* deinterlace - if not supported return -1 */
  2628. int avpicture_deinterlace(AVPicture *dst, const AVPicture *src,
  2629. enum PixelFormat pix_fmt, int width, int height);
  2630. /* external high level API */
  2631. /**
  2632. * If c is NULL, returns the first registered codec,
  2633. * if c is non-NULL, returns the next registered codec after c,
  2634. * or NULL if c is the last one.
  2635. */
  2636. AVCodec *av_codec_next(AVCodec *c);
  2637. /**
  2638. * Returns the LIBAVCODEC_VERSION_INT constant.
  2639. */
  2640. unsigned avcodec_version(void);
  2641. /**
  2642. * Initializes libavcodec.
  2643. *
  2644. * @warning This function \e must be called before any other libavcodec
  2645. * function.
  2646. */
  2647. void avcodec_init(void);
  2648. #if LIBAVCODEC_VERSION_MAJOR < 53
  2649. /**
  2650. * @deprecated Deprecated in favor of avcodec_register().
  2651. */
  2652. attribute_deprecated void register_avcodec(AVCodec *codec);
  2653. #endif
  2654. /**
  2655. * Register the codec \p codec and initialize libavcodec.
  2656. *
  2657. * @see avcodec_init()
  2658. */
  2659. void avcodec_register(AVCodec *codec);
  2660. /**
  2661. * Finds a registered encoder with a matching codec ID.
  2662. *
  2663. * @param id CodecID of the requested encoder
  2664. * @return An encoder if one was found, NULL otherwise.
  2665. */
  2666. AVCodec *avcodec_find_encoder(enum CodecID id);
  2667. /**
  2668. * Finds a registered encoder with the specified name.
  2669. *
  2670. * @param name name of the requested encoder
  2671. * @return An encoder if one was found, NULL otherwise.
  2672. */
  2673. AVCodec *avcodec_find_encoder_by_name(const char *name);
  2674. /**
  2675. * Finds a registered decoder with a matching codec ID.
  2676. *
  2677. * @param id CodecID of the requested decoder
  2678. * @return A decoder if one was found, NULL otherwise.
  2679. */
  2680. AVCodec *avcodec_find_decoder(enum CodecID id);
  2681. /**
  2682. * Finds a registered decoder with the specified name.
  2683. *
  2684. * @param name name of the requested decoder
  2685. * @return A decoder if one was found, NULL otherwise.
  2686. */
  2687. AVCodec *avcodec_find_decoder_by_name(const char *name);
  2688. void avcodec_string(char *buf, int buf_size, AVCodecContext *enc, int encode);
  2689. /**
  2690. * Sets the fields of the given AVCodecContext to default values.
  2691. *
  2692. * @param s The AVCodecContext of which the fields should be set to default values.
  2693. */
  2694. void avcodec_get_context_defaults(AVCodecContext *s);
  2695. /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
  2696. * we WILL change its arguments and name a few times! */
  2697. void avcodec_get_context_defaults2(AVCodecContext *s, enum CodecType);
  2698. /**
  2699. * Allocates an AVCodecContext and sets its fields to default values. The
  2700. * resulting struct can be deallocated by simply calling av_free().
  2701. *
  2702. * @return An AVCodecContext filled with default values or NULL on failure.
  2703. * @see avcodec_get_context_defaults
  2704. */
  2705. AVCodecContext *avcodec_alloc_context(void);
  2706. /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
  2707. * we WILL change its arguments and name a few times! */
  2708. AVCodecContext *avcodec_alloc_context2(enum CodecType);
  2709. /**
  2710. * Sets the fields of the given AVFrame to default values.
  2711. *
  2712. * @param pic The AVFrame of which the fields should be set to default values.
  2713. */
  2714. void avcodec_get_frame_defaults(AVFrame *pic);
  2715. /**
  2716. * Allocates an AVFrame and sets its fields to default values. The resulting
  2717. * struct can be deallocated by simply calling av_free().
  2718. *
  2719. * @return An AVFrame filled with default values or NULL on failure.
  2720. * @see avcodec_get_frame_defaults
  2721. */
  2722. AVFrame *avcodec_alloc_frame(void);
  2723. int avcodec_default_get_buffer(AVCodecContext *s, AVFrame *pic);
  2724. void avcodec_default_release_buffer(AVCodecContext *s, AVFrame *pic);
  2725. int avcodec_default_reget_buffer(AVCodecContext *s, AVFrame *pic);
  2726. void avcodec_align_dimensions(AVCodecContext *s, int *width, int *height);
  2727. /**
  2728. * Checks if the given dimension of a picture is valid, meaning that all
  2729. * bytes of the picture can be addressed with a signed int.
  2730. *
  2731. * @param[in] w Width of the picture.
  2732. * @param[in] h Height of the picture.
  2733. * @return Zero if valid, a negative value if invalid.
  2734. */
  2735. int avcodec_check_dimensions(void *av_log_ctx, unsigned int w, unsigned int h);
  2736. enum PixelFormat avcodec_default_get_format(struct AVCodecContext *s, const enum PixelFormat * fmt);
  2737. int avcodec_thread_init(AVCodecContext *s, int thread_count);
  2738. void avcodec_thread_free(AVCodecContext *s);
  2739. int avcodec_thread_execute(AVCodecContext *s, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
  2740. int avcodec_default_execute(AVCodecContext *c, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
  2741. //FIXME func typedef
  2742. /**
  2743. * Initializes the AVCodecContext to use the given AVCodec. Prior to using this
  2744. * function the context has to be allocated.
  2745. *
  2746. * The functions avcodec_find_decoder_by_name(), avcodec_find_encoder_by_name(),
  2747. * avcodec_find_decoder() and avcodec_find_encoder() provide an easy way for
  2748. * retrieving a codec.
  2749. *
  2750. * @warning This function is not thread safe!
  2751. *
  2752. * @code
  2753. * avcodec_register_all();
  2754. * codec = avcodec_find_decoder(CODEC_ID_H264);
  2755. * if (!codec)
  2756. * exit(1);
  2757. *
  2758. * context = avcodec_alloc_context();
  2759. *
  2760. * if (avcodec_open(context, codec) < 0)
  2761. * exit(1);
  2762. * @endcode
  2763. *
  2764. * @param avctx The context which will be set up to use the given codec.
  2765. * @param codec The codec to use within the context.
  2766. * @return zero on success, a negative value on error
  2767. * @see avcodec_alloc_context, avcodec_find_decoder, avcodec_find_encoder
  2768. */
  2769. int avcodec_open(AVCodecContext *avctx, AVCodec *codec);
  2770. #if LIBAVCODEC_VERSION_MAJOR < 53
  2771. /**
  2772. * Decodes an audio frame from \p buf into \p samples.
  2773. * Wrapper function which calls avcodec_decode_audio3.
  2774. *
  2775. * @deprecated Use avcodec_decode_audio3 instead.
  2776. * @param avctx the codec context
  2777. * @param[out] samples the output buffer
  2778. * @param[in,out] frame_size_ptr the output buffer size in bytes
  2779. * @param[in] buf the input buffer
  2780. * @param[in] buf_size the input buffer size in bytes
  2781. * @return On error a negative value is returned, otherwise the number of bytes
  2782. * used or zero if no frame could be decompressed.
  2783. */
  2784. attribute_deprecated int avcodec_decode_audio2(AVCodecContext *avctx, int16_t *samples,
  2785. int *frame_size_ptr,
  2786. const uint8_t *buf, int buf_size);
  2787. #endif
  2788. /**
  2789. * Decodes an audio frame from \p avpkt->data into \p samples.
  2790. * The avcodec_decode_audio3() function decodes an audio frame from the input
  2791. * buffer \p avpkt->data of size \p avpkt->size. To decode it, it makes use of the
  2792. * audio codec which was coupled with \p avctx using avcodec_open(). The
  2793. * resulting decoded frame is stored in output buffer \p samples. If no frame
  2794. * could be decompressed, \p frame_size_ptr is zero. Otherwise, it is the
  2795. * decompressed frame size in \e bytes.
  2796. *
  2797. * @warning You \e must set \p frame_size_ptr to the allocated size of the
  2798. * output buffer before calling avcodec_decode_audio3().
  2799. *
  2800. * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
  2801. * the actual read bytes because some optimized bitstream readers read 32 or 64
  2802. * bits at once and could read over the end.
  2803. *
  2804. * @warning The end of the input buffer \p avpkt->data should be set to 0 to ensure that
  2805. * no overreading happens for damaged MPEG streams.
  2806. *
  2807. * @note You might have to align the input buffer \p avpkt->data and output buffer \p
  2808. * samples. The alignment requirements depend on the CPU: On some CPUs it isn't
  2809. * necessary at all, on others it won't work at all if not aligned and on others
  2810. * it will work but it will have an impact on performance. In practice, the
  2811. * bitstream should have 4 byte alignment at minimum and all sample data should
  2812. * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
  2813. * the linesize is not a multiple of 16 then there's no sense in aligning the
  2814. * start of the buffer to 16.
  2815. *
  2816. * @param avctx the codec context
  2817. * @param[out] samples the output buffer
  2818. * @param[in,out] frame_size_ptr the output buffer size in bytes
  2819. * @param[in] avpkt The input AVPacket containing the input buffer.
  2820. * @return On error a negative value is returned, otherwise the number of bytes
  2821. * used or zero if no frame could be decompressed.
  2822. */
  2823. int avcodec_decode_audio3(AVCodecContext *avctx, int16_t *samples,
  2824. int *frame_size_ptr,
  2825. AVPacket *avpkt);
  2826. #if LIBAVCODEC_VERSION_MAJOR < 53
  2827. /**
  2828. * Decodes a video frame from \p buf into \p picture.
  2829. * Wrapper function which calls avcodec_decode_video2.
  2830. *
  2831. * @deprecated Use avcodec_decode_video2 instead.
  2832. * @param avctx the codec context
  2833. * @param[out] picture The AVFrame in which the decoded video frame will be stored.
  2834. * @param[in] buf the input buffer
  2835. * @param[in] buf_size the size of the input buffer in bytes
  2836. * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
  2837. * @return On error a negative value is returned, otherwise the number of bytes
  2838. * used or zero if no frame could be decompressed.
  2839. */
  2840. attribute_deprecated int avcodec_decode_video(AVCodecContext *avctx, AVFrame *picture,
  2841. int *got_picture_ptr,
  2842. const uint8_t *buf, int buf_size);
  2843. #endif
  2844. /**
  2845. * Decodes a video frame from \p avpkt->data into \p picture.
  2846. * The avcodec_decode_video2() function decodes a video frame from the input
  2847. * buffer \p avpkt->data of size \p avpkt->size. To decode it, it makes use of the
  2848. * video codec which was coupled with \p avctx using avcodec_open(). The
  2849. * resulting decoded frame is stored in \p picture.
  2850. *
  2851. * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
  2852. * the actual read bytes because some optimized bitstream readers read 32 or 64
  2853. * bits at once and could read over the end.
  2854. *
  2855. * @warning The end of the input buffer \p buf should be set to 0 to ensure that
  2856. * no overreading happens for damaged MPEG streams.
  2857. *
  2858. * @note You might have to align the input buffer \p avpkt->data and output buffer \p
  2859. * samples. The alignment requirements depend on the CPU: on some CPUs it isn't
  2860. * necessary at all, on others it won't work at all if not aligned and on others
  2861. * it will work but it will have an impact on performance. In practice, the
  2862. * bitstream should have 4 byte alignment at minimum and all sample data should
  2863. * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
  2864. * the linesize is not a multiple of 16 then there's no sense in aligning the
  2865. * start of the buffer to 16.
  2866. *
  2867. * @note Some codecs have a delay between input and output, these need to be
  2868. * feeded with avpkt->data=NULL, avpkt->size=0 at the end to return the remaining frames.
  2869. *
  2870. * @param avctx the codec context
  2871. * @param[out] picture The AVFrame in which the decoded video frame will be stored.
  2872. * @param[in] avpkt The input AVpacket containing the input buffer.
  2873. * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
  2874. * @return On error a negative value is returned, otherwise the number of bytes
  2875. * used or zero if no frame could be decompressed.
  2876. */
  2877. int avcodec_decode_video2(AVCodecContext *avctx, AVFrame *picture,
  2878. int *got_picture_ptr,
  2879. AVPacket *avpkt);
  2880. #if LIBAVCODEC_VERSION_MAJOR < 53
  2881. /* Decode a subtitle message. Return -1 if error, otherwise return the
  2882. * number of bytes used. If no subtitle could be decompressed,
  2883. * got_sub_ptr is zero. Otherwise, the subtitle is stored in *sub. */
  2884. attribute_deprecated int avcodec_decode_subtitle(AVCodecContext *avctx, AVSubtitle *sub,
  2885. int *got_sub_ptr,
  2886. const uint8_t *buf, int buf_size);
  2887. #endif
  2888. /**
  2889. * Decodes a subtitle message.
  2890. * Returns -1 if error, otherwise returns the number of bytes used.
  2891. * If no subtitle could be decompressed, \p got_sub_ptr is zero.
  2892. * Otherwise, the subtitle is stored in \p *sub.
  2893. *
  2894. * @param avctx the codec context
  2895. * @param[out] sub The AVSubtitle in which the decoded subtitle will be stored.
  2896. * @param[in,out] got_sub_ptr Zero if no subtitle could be decompressed, otherwise, it is nonzero.
  2897. * @param[in] avpkt The input AVPacket containing the input buffer.
  2898. */
  2899. int avcodec_decode_subtitle2(AVCodecContext *avctx, AVSubtitle *sub,
  2900. int *got_sub_ptr,
  2901. AVPacket *avpkt);
  2902. int avcodec_parse_frame(AVCodecContext *avctx, uint8_t **pdata,
  2903. int *data_size_ptr,
  2904. uint8_t *buf, int buf_size);
  2905. /**
  2906. * Encodes an audio frame from \p samples into \p buf.
  2907. * The avcodec_encode_audio() function encodes an audio frame from the input
  2908. * buffer \p samples. To encode it, it makes use of the audio codec which was
  2909. * coupled with \p avctx using avcodec_open(). The resulting encoded frame is
  2910. * stored in output buffer \p buf.
  2911. *
  2912. * @note The output buffer should be at least \c FF_MIN_BUFFER_SIZE bytes large.
  2913. *
  2914. * @param avctx the codec context
  2915. * @param[out] buf the output buffer
  2916. * @param[in] buf_size the output buffer size
  2917. * @param[in] samples the input buffer containing the samples
  2918. * The number of samples read from this buffer is frame_size*channels,
  2919. * both of which are defined in \p avctx.
  2920. * For PCM audio the number of samples read from \p samples is equal to
  2921. * \p buf_size * input_sample_size / output_sample_size.
  2922. * @return On error a negative value is returned, on success zero or the number
  2923. * of bytes used to encode the data read from the input buffer.
  2924. */
  2925. int avcodec_encode_audio(AVCodecContext *avctx, uint8_t *buf, int buf_size,
  2926. const short *samples);
  2927. /**
  2928. * Encodes a video frame from \p pict into \p buf.
  2929. * The avcodec_encode_video() function encodes a video frame from the input
  2930. * \p pict. To encode it, it makes use of the video codec which was coupled with
  2931. * \p avctx using avcodec_open(). The resulting encoded bytes representing the
  2932. * frame are stored in the output buffer \p buf. The input picture should be
  2933. * stored using a specific format, namely \c avctx.pix_fmt.
  2934. *
  2935. * @param avctx the codec context
  2936. * @param[out] buf the output buffer for the bitstream of encoded frame
  2937. * @param[in] buf_size the size of the output buffer in bytes
  2938. * @param[in] pict the input picture to encode
  2939. * @return On error a negative value is returned, on success zero or the number
  2940. * of bytes used from the output buffer.
  2941. */
  2942. int avcodec_encode_video(AVCodecContext *avctx, uint8_t *buf, int buf_size,
  2943. const AVFrame *pict);
  2944. int avcodec_encode_subtitle(AVCodecContext *avctx, uint8_t *buf, int buf_size,
  2945. const AVSubtitle *sub);
  2946. int avcodec_close(AVCodecContext *avctx);
  2947. /**
  2948. * Register all the codecs, parsers and bitstream filters which were enabled at
  2949. * configuration time. If you do not call this function you can select exactly
  2950. * which formats you want to support, by using the individual registration
  2951. * functions.
  2952. *
  2953. * @see avcodec_register
  2954. * @see av_register_codec_parser
  2955. * @see av_register_bitstream_filter
  2956. */
  2957. void avcodec_register_all(void);
  2958. /**
  2959. * Flush buffers, should be called when seeking or when switching to a different stream.
  2960. */
  2961. void avcodec_flush_buffers(AVCodecContext *avctx);
  2962. void avcodec_default_free_buffers(AVCodecContext *s);
  2963. /* misc useful functions */
  2964. /**
  2965. * Returns a single letter to describe the given picture type \p pict_type.
  2966. *
  2967. * @param[in] pict_type the picture type
  2968. * @return A single character representing the picture type.
  2969. */
  2970. char av_get_pict_type_char(int pict_type);
  2971. /**
  2972. * Returns codec bits per sample.
  2973. *
  2974. * @param[in] codec_id the codec
  2975. * @return Number of bits per sample or zero if unknown for the given codec.
  2976. */
  2977. int av_get_bits_per_sample(enum CodecID codec_id);
  2978. /**
  2979. * Returns sample format bits per sample.
  2980. *
  2981. * @param[in] sample_fmt the sample format
  2982. * @return Number of bits per sample or zero if unknown for the given sample format.
  2983. */
  2984. int av_get_bits_per_sample_format(enum SampleFormat sample_fmt);
  2985. /* frame parsing */
  2986. typedef struct AVCodecParserContext {
  2987. void *priv_data;
  2988. struct AVCodecParser *parser;
  2989. int64_t frame_offset; /* offset of the current frame */
  2990. int64_t cur_offset; /* current offset
  2991. (incremented by each av_parser_parse()) */
  2992. int64_t next_frame_offset; /* offset of the next frame */
  2993. /* video info */
  2994. int pict_type; /* XXX: Put it back in AVCodecContext. */
  2995. /**
  2996. * This field is used for proper frame duration computation in lavf.
  2997. * It signals, how much longer the frame duration of the current frame
  2998. * is compared to normal frame duration.
  2999. *
  3000. * frame_duration = (1 + repeat_pict) * time_base
  3001. *
  3002. * It is used by codecs like H.264 to display telecined material.
  3003. */
  3004. int repeat_pict; /* XXX: Put it back in AVCodecContext. */
  3005. int64_t pts; /* pts of the current frame */
  3006. int64_t dts; /* dts of the current frame */
  3007. /* private data */
  3008. int64_t last_pts;
  3009. int64_t last_dts;
  3010. int fetch_timestamp;
  3011. #define AV_PARSER_PTS_NB 4
  3012. int cur_frame_start_index;
  3013. int64_t cur_frame_offset[AV_PARSER_PTS_NB];
  3014. int64_t cur_frame_pts[AV_PARSER_PTS_NB];
  3015. int64_t cur_frame_dts[AV_PARSER_PTS_NB];
  3016. int flags;
  3017. #define PARSER_FLAG_COMPLETE_FRAMES 0x0001
  3018. int64_t offset; ///< byte offset from starting packet start
  3019. int64_t cur_frame_end[AV_PARSER_PTS_NB];
  3020. /*!
  3021. * Set by parser to 1 for key frames and 0 for non-key frames.
  3022. * It is initialized to -1, so if the parser doesn't set this flag,
  3023. * old-style fallback using FF_I_TYPE picture type as key frames
  3024. * will be used.
  3025. */
  3026. int key_frame;
  3027. /**
  3028. * Time difference in stream time base units from the pts of this
  3029. * packet to the point at which the output from the decoder has converged
  3030. * independent from the availability of previous frames. That is, the
  3031. * frames are virtually identical no matter if decoding started from
  3032. * the very first frame or from this keyframe.
  3033. * Is AV_NOPTS_VALUE if unknown.
  3034. * This field is not the display duration of the current frame.
  3035. *
  3036. * The purpose of this field is to allow seeking in streams that have no
  3037. * keyframes in the conventional sense. It corresponds to the
  3038. * recovery point SEI in H.264 and match_time_delta in NUT. It is also
  3039. * essential for some types of subtitle streams to ensure that all
  3040. * subtitles are correctly displayed after seeking.
  3041. */
  3042. int64_t convergence_duration;
  3043. // Timestamp generation support:
  3044. /**
  3045. * Synchronization point for start of timestamp generation.
  3046. *
  3047. * Set to >0 for sync point, 0 for no sync point and <0 for undefined
  3048. * (default).
  3049. *
  3050. * For example, this corresponds to presence of H.264 buffering period
  3051. * SEI message.
  3052. */
  3053. int dts_sync_point;
  3054. /**
  3055. * Offset of the current timestamp against last timestamp sync point in
  3056. * units of AVCodecContext.time_base.
  3057. *
  3058. * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
  3059. * contain a valid timestamp offset.
  3060. *
  3061. * Note that the timestamp of sync point has usually a nonzero
  3062. * dts_ref_dts_delta, which refers to the previous sync point. Offset of
  3063. * the next frame after timestamp sync point will be usually 1.
  3064. *
  3065. * For example, this corresponds to H.264 cpb_removal_delay.
  3066. */
  3067. int dts_ref_dts_delta;
  3068. /**
  3069. * Presentation delay of current frame in units of AVCodecContext.time_base.
  3070. *
  3071. * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
  3072. * contain valid non-negative timestamp delta (presentation time of a frame
  3073. * must not lie in the past).
  3074. *
  3075. * This delay represents the difference between decoding and presentation
  3076. * time of the frame.
  3077. *
  3078. * For example, this corresponds to H.264 dpb_output_delay.
  3079. */
  3080. int pts_dts_delta;
  3081. /**
  3082. * Position of the packet in file.
  3083. *
  3084. * Analogous to cur_frame_pts/dts
  3085. */
  3086. int64_t cur_frame_pos[AV_PARSER_PTS_NB];
  3087. /**
  3088. * Byte position of currently parsed frame in stream.
  3089. */
  3090. int64_t pos;
  3091. /**
  3092. * Previous frame byte position.
  3093. */
  3094. int64_t last_pos;
  3095. } AVCodecParserContext;
  3096. typedef struct AVCodecParser {
  3097. int codec_ids[5]; /* several codec IDs are permitted */
  3098. int priv_data_size;
  3099. int (*parser_init)(AVCodecParserContext *s);
  3100. int (*parser_parse)(AVCodecParserContext *s,
  3101. AVCodecContext *avctx,
  3102. const uint8_t **poutbuf, int *poutbuf_size,
  3103. const uint8_t *buf, int buf_size);
  3104. void (*parser_close)(AVCodecParserContext *s);
  3105. int (*split)(AVCodecContext *avctx, const uint8_t *buf, int buf_size);
  3106. struct AVCodecParser *next;
  3107. } AVCodecParser;
  3108. AVCodecParser *av_parser_next(AVCodecParser *c);
  3109. void av_register_codec_parser(AVCodecParser *parser);
  3110. AVCodecParserContext *av_parser_init(int codec_id);
  3111. #if LIBAVCODEC_VERSION_MAJOR < 53
  3112. attribute_deprecated
  3113. int av_parser_parse(AVCodecParserContext *s,
  3114. AVCodecContext *avctx,
  3115. uint8_t **poutbuf, int *poutbuf_size,
  3116. const uint8_t *buf, int buf_size,
  3117. int64_t pts, int64_t dts);
  3118. #endif
  3119. /**
  3120. * Parse a packet.
  3121. *
  3122. * @param s parser context.
  3123. * @param avctx codec context.
  3124. * @param poutbuf set to pointer to parsed buffer or NULL if not yet finished.
  3125. * @param poutbuf_size set to size of parsed buffer or zero if not yet finished.
  3126. * @param buf input buffer.
  3127. * @param buf_size input length, to signal EOF, this should be 0 (so that the last frame can be output).
  3128. * @param pts input presentation timestamp.
  3129. * @param dts input decoding timestamp.
  3130. * @param pos input byte position in stream.
  3131. * @return the number of bytes of the input bitstream used.
  3132. *
  3133. * Example:
  3134. * @code
  3135. * while(in_len){
  3136. * len = av_parser_parse2(myparser, AVCodecContext, &data, &size,
  3137. * in_data, in_len,
  3138. * pts, dts, pos);
  3139. * in_data += len;
  3140. * in_len -= len;
  3141. *
  3142. * if(size)
  3143. * decode_frame(data, size);
  3144. * }
  3145. * @endcode
  3146. */
  3147. int av_parser_parse2(AVCodecParserContext *s,
  3148. AVCodecContext *avctx,
  3149. uint8_t **poutbuf, int *poutbuf_size,
  3150. const uint8_t *buf, int buf_size,
  3151. int64_t pts, int64_t dts,
  3152. int64_t pos);
  3153. int av_parser_change(AVCodecParserContext *s,
  3154. AVCodecContext *avctx,
  3155. uint8_t **poutbuf, int *poutbuf_size,
  3156. const uint8_t *buf, int buf_size, int keyframe);
  3157. void av_parser_close(AVCodecParserContext *s);
  3158. typedef struct AVBitStreamFilterContext {
  3159. void *priv_data;
  3160. struct AVBitStreamFilter *filter;
  3161. AVCodecParserContext *parser;
  3162. struct AVBitStreamFilterContext *next;
  3163. } AVBitStreamFilterContext;
  3164. typedef struct AVBitStreamFilter {
  3165. const char *name;
  3166. int priv_data_size;
  3167. int (*filter)(AVBitStreamFilterContext *bsfc,
  3168. AVCodecContext *avctx, const char *args,
  3169. uint8_t **poutbuf, int *poutbuf_size,
  3170. const uint8_t *buf, int buf_size, int keyframe);
  3171. void (*close)(AVBitStreamFilterContext *bsfc);
  3172. struct AVBitStreamFilter *next;
  3173. } AVBitStreamFilter;
  3174. void av_register_bitstream_filter(AVBitStreamFilter *bsf);
  3175. AVBitStreamFilterContext *av_bitstream_filter_init(const char *name);
  3176. int av_bitstream_filter_filter(AVBitStreamFilterContext *bsfc,
  3177. AVCodecContext *avctx, const char *args,
  3178. uint8_t **poutbuf, int *poutbuf_size,
  3179. const uint8_t *buf, int buf_size, int keyframe);
  3180. void av_bitstream_filter_close(AVBitStreamFilterContext *bsf);
  3181. AVBitStreamFilter *av_bitstream_filter_next(AVBitStreamFilter *f);
  3182. /* memory */
  3183. /**
  3184. * Reallocates the given block if it is not large enough, otherwise it
  3185. * does nothing.
  3186. *
  3187. * @see av_realloc
  3188. */
  3189. void *av_fast_realloc(void *ptr, unsigned int *size, unsigned int min_size);
  3190. /**
  3191. * Copy image 'src' to 'dst'.
  3192. */
  3193. void av_picture_copy(AVPicture *dst, const AVPicture *src,
  3194. enum PixelFormat pix_fmt, int width, int height);
  3195. /**
  3196. * Crop image top and left side.
  3197. */
  3198. int av_picture_crop(AVPicture *dst, const AVPicture *src,
  3199. enum PixelFormat pix_fmt, int top_band, int left_band);
  3200. /**
  3201. * Pad image.
  3202. */
  3203. int av_picture_pad(AVPicture *dst, const AVPicture *src, int height, int width, enum PixelFormat pix_fmt,
  3204. int padtop, int padbottom, int padleft, int padright, int *color);
  3205. unsigned int av_xiphlacing(unsigned char *s, unsigned int v);
  3206. /**
  3207. * Parses \p str and put in \p width_ptr and \p height_ptr the detected values.
  3208. *
  3209. * @return 0 in case of a successful parsing, a negative value otherwise
  3210. * @param[in] str the string to parse: it has to be a string in the format
  3211. * <width>x<height> or a valid video frame size abbreviation.
  3212. * @param[in,out] width_ptr pointer to the variable which will contain the detected
  3213. * frame width value
  3214. * @param[in,out] height_ptr pointer to the variable which will contain the detected
  3215. * frame height value
  3216. */
  3217. int av_parse_video_frame_size(int *width_ptr, int *height_ptr, const char *str);
  3218. /**
  3219. * Parses \p str and put in \p frame_rate the detected values.
  3220. *
  3221. * @return 0 in case of a successful parsing, a negative value otherwise
  3222. * @param[in] str the string to parse: it has to be a string in the format
  3223. * <frame_rate_num>/<frame_rate_den>, a float number or a valid video rate abbreviation
  3224. * @param[in,out] frame_rate pointer to the AVRational which will contain the detected
  3225. * frame rate
  3226. */
  3227. int av_parse_video_frame_rate(AVRational *frame_rate, const char *str);
  3228. /* error handling */
  3229. #if EINVAL > 0
  3230. #define AVERROR(e) (-(e)) /**< Returns a negative error code from a POSIX error code, to return from library functions. */
  3231. #define AVUNERROR(e) (-(e)) /**< Returns a POSIX error code from a library function error return value. */
  3232. #else
  3233. /* Some platforms have E* and errno already negated. */
  3234. #define AVERROR(e) (e)
  3235. #define AVUNERROR(e) (e)
  3236. #endif
  3237. #define AVERROR_UNKNOWN AVERROR(EINVAL) /**< unknown error */
  3238. #define AVERROR_IO AVERROR(EIO) /**< I/O error */
  3239. #define AVERROR_NUMEXPECTED AVERROR(EDOM) /**< Number syntax expected in filename. */
  3240. #define AVERROR_INVALIDDATA AVERROR(EINVAL) /**< invalid data found */
  3241. #define AVERROR_NOMEM AVERROR(ENOMEM) /**< not enough memory */
  3242. #define AVERROR_NOFMT AVERROR(EILSEQ) /**< unknown format */
  3243. #define AVERROR_NOTSUPP AVERROR(ENOSYS) /**< Operation not supported. */
  3244. #define AVERROR_NOENT AVERROR(ENOENT) /**< No such file or directory. */
  3245. #define AVERROR_EOF AVERROR(EPIPE) /**< End of file. */
  3246. #define AVERROR_PATCHWELCOME -MKTAG('P','A','W','E') /**< Not yet implemented in FFmpeg. Patches welcome. */
  3247. /**
  3248. * Registers the hardware accelerator \p hwaccel.
  3249. */
  3250. void av_register_hwaccel(AVHWAccel *hwaccel);
  3251. /**
  3252. * If hwaccel is NULL, returns the first registered hardware accelerator,
  3253. * if hwaccel is non-NULL, returns the next registered hardware accelerator
  3254. * after hwaccel, or NULL if hwaccel is the last one.
  3255. */
  3256. AVHWAccel *av_hwaccel_next(AVHWAccel *hwaccel);
  3257. #endif /* AVCODEC_AVCODEC_H */