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.

3506 lines
112KB

  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 23
  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_7POINT1 (CH_LAYOUT_5POINT1|CH_BACK_LEFT|CH_BACK_RIGHT)
  364. #define CH_LAYOUT_7POINT1_WIDE (CH_LAYOUT_SURROUND|CH_LOW_FREQUENCY|\
  365. CH_BACK_LEFT|CH_BACK_RIGHT|\
  366. CH_FRONT_LEFT_OF_CENTER|CH_FRONT_RIGHT_OF_CENTER)
  367. #define CH_LAYOUT_STEREO_DOWNMIX (CH_STEREO_LEFT|CH_STEREO_RIGHT)
  368. /* in bytes */
  369. #define AVCODEC_MAX_AUDIO_FRAME_SIZE 192000 // 1 second of 48khz 32bit audio
  370. /**
  371. * Required number of additionally allocated bytes at the end of the input bitstream for decoding.
  372. * This is mainly needed because some optimized bitstream readers read
  373. * 32 or 64 bit at once and could read over the end.<br>
  374. * Note: If the first 23 bits of the additional bytes are not 0, then damaged
  375. * MPEG bitstreams could cause overread and segfault.
  376. */
  377. #define FF_INPUT_BUFFER_PADDING_SIZE 8
  378. /**
  379. * minimum encoding buffer size
  380. * Used to avoid some checks during header writing.
  381. */
  382. #define FF_MIN_BUFFER_SIZE 16384
  383. /**
  384. * motion estimation type.
  385. */
  386. enum Motion_Est_ID {
  387. ME_ZERO = 1, ///< no search, that is use 0,0 vector whenever one is needed
  388. ME_FULL,
  389. ME_LOG,
  390. ME_PHODS,
  391. ME_EPZS, ///< enhanced predictive zonal search
  392. ME_X1, ///< reserved for experiments
  393. ME_HEX, ///< hexagon based search
  394. ME_UMH, ///< uneven multi-hexagon search
  395. ME_ITER, ///< iterative search
  396. ME_TESA, ///< transformed exhaustive search algorithm
  397. };
  398. enum AVDiscard{
  399. /* We leave some space between them for extensions (drop some
  400. * keyframes for intra-only or drop just some bidir frames). */
  401. AVDISCARD_NONE =-16, ///< discard nothing
  402. AVDISCARD_DEFAULT= 0, ///< discard useless packets like 0 size packets in avi
  403. AVDISCARD_NONREF = 8, ///< discard all non reference
  404. AVDISCARD_BIDIR = 16, ///< discard all bidirectional frames
  405. AVDISCARD_NONKEY = 32, ///< discard all frames except keyframes
  406. AVDISCARD_ALL = 48, ///< discard all
  407. };
  408. typedef struct RcOverride{
  409. int start_frame;
  410. int end_frame;
  411. int qscale; // If this is 0 then quality_factor will be used instead.
  412. float quality_factor;
  413. } RcOverride;
  414. #define FF_MAX_B_FRAMES 16
  415. /* encoding support
  416. These flags can be passed in AVCodecContext.flags before initialization.
  417. Note: Not everything is supported yet.
  418. */
  419. #define CODEC_FLAG_QSCALE 0x0002 ///< Use fixed qscale.
  420. #define CODEC_FLAG_4MV 0x0004 ///< 4 MV per MB allowed / advanced prediction for H.263.
  421. #define CODEC_FLAG_QPEL 0x0010 ///< Use qpel MC.
  422. #define CODEC_FLAG_GMC 0x0020 ///< Use GMC.
  423. #define CODEC_FLAG_MV0 0x0040 ///< Always try a MB with MV=<0,0>.
  424. #define CODEC_FLAG_PART 0x0080 ///< Use data partitioning.
  425. /**
  426. * The parent program guarantees that the input for B-frames containing
  427. * streams is not written to for at least s->max_b_frames+1 frames, if
  428. * this is not set the input will be copied.
  429. */
  430. #define CODEC_FLAG_INPUT_PRESERVED 0x0100
  431. #define CODEC_FLAG_PASS1 0x0200 ///< Use internal 2pass ratecontrol in first pass mode.
  432. #define CODEC_FLAG_PASS2 0x0400 ///< Use internal 2pass ratecontrol in second pass mode.
  433. #define CODEC_FLAG_EXTERN_HUFF 0x1000 ///< Use external Huffman table (for MJPEG).
  434. #define CODEC_FLAG_GRAY 0x2000 ///< Only decode/encode grayscale.
  435. #define CODEC_FLAG_EMU_EDGE 0x4000 ///< Don't draw edges.
  436. #define CODEC_FLAG_PSNR 0x8000 ///< error[?] variables will be set during encoding.
  437. #define CODEC_FLAG_TRUNCATED 0x00010000 /** Input bitstream might be truncated at a random
  438. location instead of only at frame boundaries. */
  439. #define CODEC_FLAG_NORMALIZE_AQP 0x00020000 ///< Normalize adaptive quantization.
  440. #define CODEC_FLAG_INTERLACED_DCT 0x00040000 ///< Use interlaced DCT.
  441. #define CODEC_FLAG_LOW_DELAY 0x00080000 ///< Force low delay.
  442. #define CODEC_FLAG_ALT_SCAN 0x00100000 ///< Use alternate scan.
  443. #define CODEC_FLAG_GLOBAL_HEADER 0x00400000 ///< Place global headers in extradata instead of every keyframe.
  444. #define CODEC_FLAG_BITEXACT 0x00800000 ///< Use only bitexact stuff (except (I)DCT).
  445. /* Fx : Flag for h263+ extra options */
  446. #define CODEC_FLAG_AC_PRED 0x01000000 ///< H.263 advanced intra coding / MPEG-4 AC prediction
  447. #define CODEC_FLAG_H263P_UMV 0x02000000 ///< unlimited motion vector
  448. #define CODEC_FLAG_CBP_RD 0x04000000 ///< Use rate distortion optimization for cbp.
  449. #define CODEC_FLAG_QP_RD 0x08000000 ///< Use rate distortion optimization for qp selectioon.
  450. #define CODEC_FLAG_H263P_AIV 0x00000008 ///< H.263 alternative inter VLC
  451. #define CODEC_FLAG_OBMC 0x00000001 ///< OBMC
  452. #define CODEC_FLAG_LOOP_FILTER 0x00000800 ///< loop filter
  453. #define CODEC_FLAG_H263P_SLICE_STRUCT 0x10000000
  454. #define CODEC_FLAG_INTERLACED_ME 0x20000000 ///< interlaced motion estimation
  455. #define CODEC_FLAG_SVCD_SCAN_OFFSET 0x40000000 ///< Will reserve space for SVCD scan offset user data.
  456. #define CODEC_FLAG_CLOSED_GOP 0x80000000
  457. #define CODEC_FLAG2_FAST 0x00000001 ///< Allow non spec compliant speedup tricks.
  458. #define CODEC_FLAG2_STRICT_GOP 0x00000002 ///< Strictly enforce GOP size.
  459. #define CODEC_FLAG2_NO_OUTPUT 0x00000004 ///< Skip bitstream encoding.
  460. #define CODEC_FLAG2_LOCAL_HEADER 0x00000008 ///< Place global headers at every keyframe instead of in extradata.
  461. #define CODEC_FLAG2_BPYRAMID 0x00000010 ///< H.264 allow B-frames to be used as references.
  462. #define CODEC_FLAG2_WPRED 0x00000020 ///< H.264 weighted biprediction for B-frames
  463. #define CODEC_FLAG2_MIXED_REFS 0x00000040 ///< H.264 one reference per partition, as opposed to one reference per macroblock
  464. #define CODEC_FLAG2_8X8DCT 0x00000080 ///< H.264 high profile 8x8 transform
  465. #define CODEC_FLAG2_FASTPSKIP 0x00000100 ///< H.264 fast pskip
  466. #define CODEC_FLAG2_AUD 0x00000200 ///< H.264 access unit delimiters
  467. #define CODEC_FLAG2_BRDO 0x00000400 ///< B-frame rate-distortion optimization
  468. #define CODEC_FLAG2_INTRA_VLC 0x00000800 ///< Use MPEG-2 intra VLC table.
  469. #define CODEC_FLAG2_MEMC_ONLY 0x00001000 ///< Only do ME/MC (I frames -> ref, P frame -> ME+MC).
  470. #define CODEC_FLAG2_DROP_FRAME_TIMECODE 0x00002000 ///< timecode is in drop frame format.
  471. #define CODEC_FLAG2_SKIP_RD 0x00004000 ///< RD optimal MB level residual skipping
  472. #define CODEC_FLAG2_CHUNKS 0x00008000 ///< Input bitstream might be truncated at a packet boundaries instead of only at frame boundaries.
  473. #define CODEC_FLAG2_NON_LINEAR_QUANT 0x00010000 ///< Use MPEG-2 nonlinear quantizer.
  474. #define CODEC_FLAG2_BIT_RESERVOIR 0x00020000 ///< Use a bit reservoir when encoding if possible
  475. /* Unsupported options :
  476. * Syntax Arithmetic coding (SAC)
  477. * Reference Picture Selection
  478. * Independent Segment Decoding */
  479. /* /Fx */
  480. /* codec capabilities */
  481. #define CODEC_CAP_DRAW_HORIZ_BAND 0x0001 ///< Decoder can use draw_horiz_band callback.
  482. /**
  483. * Codec uses get_buffer() for allocating buffers.
  484. * direct rendering method 1
  485. */
  486. #define CODEC_CAP_DR1 0x0002
  487. /* If 'parse_only' field is true, then avcodec_parse_frame() can be used. */
  488. #define CODEC_CAP_PARSE_ONLY 0x0004
  489. #define CODEC_CAP_TRUNCATED 0x0008
  490. /* Codec can export data for HW decoding (XvMC). */
  491. #define CODEC_CAP_HWACCEL 0x0010
  492. /**
  493. * Codec has a nonzero delay and needs to be fed with NULL at the end to get the delayed data.
  494. * If this is not set, the codec is guaranteed to never be fed with NULL data.
  495. */
  496. #define CODEC_CAP_DELAY 0x0020
  497. /**
  498. * Codec can be fed a final frame with a smaller size.
  499. * This can be used to prevent truncation of the last audio samples.
  500. */
  501. #define CODEC_CAP_SMALL_LAST_FRAME 0x0040
  502. /**
  503. * Codec can export data for HW decoding (VDPAU).
  504. */
  505. #define CODEC_CAP_HWACCEL_VDPAU 0x0080
  506. //The following defines may change, don't expect compatibility if you use them.
  507. #define MB_TYPE_INTRA4x4 0x0001
  508. #define MB_TYPE_INTRA16x16 0x0002 //FIXME H.264-specific
  509. #define MB_TYPE_INTRA_PCM 0x0004 //FIXME H.264-specific
  510. #define MB_TYPE_16x16 0x0008
  511. #define MB_TYPE_16x8 0x0010
  512. #define MB_TYPE_8x16 0x0020
  513. #define MB_TYPE_8x8 0x0040
  514. #define MB_TYPE_INTERLACED 0x0080
  515. #define MB_TYPE_DIRECT2 0x0100 //FIXME
  516. #define MB_TYPE_ACPRED 0x0200
  517. #define MB_TYPE_GMC 0x0400
  518. #define MB_TYPE_SKIP 0x0800
  519. #define MB_TYPE_P0L0 0x1000
  520. #define MB_TYPE_P1L0 0x2000
  521. #define MB_TYPE_P0L1 0x4000
  522. #define MB_TYPE_P1L1 0x8000
  523. #define MB_TYPE_L0 (MB_TYPE_P0L0 | MB_TYPE_P1L0)
  524. #define MB_TYPE_L1 (MB_TYPE_P0L1 | MB_TYPE_P1L1)
  525. #define MB_TYPE_L0L1 (MB_TYPE_L0 | MB_TYPE_L1)
  526. #define MB_TYPE_QUANT 0x00010000
  527. #define MB_TYPE_CBP 0x00020000
  528. //Note bits 24-31 are reserved for codec specific use (h264 ref0, mpeg1 0mv, ...)
  529. /**
  530. * Pan Scan area.
  531. * This specifies the area which should be displayed.
  532. * Note there may be multiple such areas for one frame.
  533. */
  534. typedef struct AVPanScan{
  535. /**
  536. * id
  537. * - encoding: Set by user.
  538. * - decoding: Set by libavcodec.
  539. */
  540. int id;
  541. /**
  542. * width and height in 1/16 pel
  543. * - encoding: Set by user.
  544. * - decoding: Set by libavcodec.
  545. */
  546. int width;
  547. int height;
  548. /**
  549. * position of the top left corner in 1/16 pel for up to 3 fields/frames
  550. * - encoding: Set by user.
  551. * - decoding: Set by libavcodec.
  552. */
  553. int16_t position[3][2];
  554. }AVPanScan;
  555. #define FF_COMMON_FRAME \
  556. /**\
  557. * pointer to the picture planes.\
  558. * This might be different from the first allocated byte\
  559. * - encoding: \
  560. * - decoding: \
  561. */\
  562. uint8_t *data[4];\
  563. int linesize[4];\
  564. /**\
  565. * pointer to the first allocated byte of the picture. Can be used in get_buffer/release_buffer.\
  566. * This isn't used by libavcodec unless the default get/release_buffer() is used.\
  567. * - encoding: \
  568. * - decoding: \
  569. */\
  570. uint8_t *base[4];\
  571. /**\
  572. * 1 -> keyframe, 0-> not\
  573. * - encoding: Set by libavcodec.\
  574. * - decoding: Set by libavcodec.\
  575. */\
  576. int key_frame;\
  577. \
  578. /**\
  579. * Picture type of the frame, see ?_TYPE below.\
  580. * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
  581. * - decoding: Set by libavcodec.\
  582. */\
  583. int pict_type;\
  584. \
  585. /**\
  586. * presentation timestamp in time_base units (time when frame should be shown to user)\
  587. * If AV_NOPTS_VALUE then frame_rate = 1/time_base will be assumed.\
  588. * - encoding: MUST be set by user.\
  589. * - decoding: Set by libavcodec.\
  590. */\
  591. int64_t pts;\
  592. \
  593. /**\
  594. * picture number in bitstream order\
  595. * - encoding: set by\
  596. * - decoding: Set by libavcodec.\
  597. */\
  598. int coded_picture_number;\
  599. /**\
  600. * picture number in display order\
  601. * - encoding: set by\
  602. * - decoding: Set by libavcodec.\
  603. */\
  604. int display_picture_number;\
  605. \
  606. /**\
  607. * quality (between 1 (good) and FF_LAMBDA_MAX (bad)) \
  608. * - encoding: Set by libavcodec. for coded_picture (and set by user for input).\
  609. * - decoding: Set by libavcodec.\
  610. */\
  611. int quality; \
  612. \
  613. /**\
  614. * buffer age (1->was last buffer and dint change, 2->..., ...).\
  615. * Set to INT_MAX if the buffer has not been used yet.\
  616. * - encoding: unused\
  617. * - decoding: MUST be set by get_buffer().\
  618. */\
  619. int age;\
  620. \
  621. /**\
  622. * is this picture used as reference\
  623. * The values for this are the same as the MpegEncContext.picture_structure\
  624. * variable, that is 1->top field, 2->bottom field, 3->frame/both fields.\
  625. * Set to 4 for delayed, non-reference frames.\
  626. * - encoding: unused\
  627. * - decoding: Set by libavcodec. (before get_buffer() call)).\
  628. */\
  629. int reference;\
  630. \
  631. /**\
  632. * QP table\
  633. * - encoding: unused\
  634. * - decoding: Set by libavcodec.\
  635. */\
  636. int8_t *qscale_table;\
  637. /**\
  638. * QP store stride\
  639. * - encoding: unused\
  640. * - decoding: Set by libavcodec.\
  641. */\
  642. int qstride;\
  643. \
  644. /**\
  645. * mbskip_table[mb]>=1 if MB didn't change\
  646. * stride= mb_width = (width+15)>>4\
  647. * - encoding: unused\
  648. * - decoding: Set by libavcodec.\
  649. */\
  650. uint8_t *mbskip_table;\
  651. \
  652. /**\
  653. * motion vector table\
  654. * @code\
  655. * example:\
  656. * int mv_sample_log2= 4 - motion_subsample_log2;\
  657. * int mb_width= (width+15)>>4;\
  658. * int mv_stride= (mb_width << mv_sample_log2) + 1;\
  659. * motion_val[direction][x + y*mv_stride][0->mv_x, 1->mv_y];\
  660. * @endcode\
  661. * - encoding: Set by user.\
  662. * - decoding: Set by libavcodec.\
  663. */\
  664. int16_t (*motion_val[2])[2];\
  665. \
  666. /**\
  667. * macroblock type table\
  668. * mb_type_base + mb_width + 2\
  669. * - encoding: Set by user.\
  670. * - decoding: Set by libavcodec.\
  671. */\
  672. uint32_t *mb_type;\
  673. \
  674. /**\
  675. * log2 of the size of the block which a single vector in motion_val represents: \
  676. * (4->16x16, 3->8x8, 2-> 4x4, 1-> 2x2)\
  677. * - encoding: unused\
  678. * - decoding: Set by libavcodec.\
  679. */\
  680. uint8_t motion_subsample_log2;\
  681. \
  682. /**\
  683. * for some private data of the user\
  684. * - encoding: unused\
  685. * - decoding: Set by user.\
  686. */\
  687. void *opaque;\
  688. \
  689. /**\
  690. * error\
  691. * - encoding: Set by libavcodec. if flags&CODEC_FLAG_PSNR.\
  692. * - decoding: unused\
  693. */\
  694. uint64_t error[4];\
  695. \
  696. /**\
  697. * type of the buffer (to keep track of who has to deallocate data[*])\
  698. * - encoding: Set by the one who allocates it.\
  699. * - decoding: Set by the one who allocates it.\
  700. * Note: User allocated (direct rendering) & internal buffers cannot coexist currently.\
  701. */\
  702. int type;\
  703. \
  704. /**\
  705. * When decoding, this signals how much the picture must be delayed.\
  706. * extra_delay = repeat_pict / (2*fps)\
  707. * - encoding: unused\
  708. * - decoding: Set by libavcodec.\
  709. */\
  710. int repeat_pict;\
  711. \
  712. /**\
  713. * \
  714. */\
  715. int qscale_type;\
  716. \
  717. /**\
  718. * The content of the picture is interlaced.\
  719. * - encoding: Set by user.\
  720. * - decoding: Set by libavcodec. (default 0)\
  721. */\
  722. int interlaced_frame;\
  723. \
  724. /**\
  725. * If the content is interlaced, is top field displayed first.\
  726. * - encoding: Set by user.\
  727. * - decoding: Set by libavcodec.\
  728. */\
  729. int top_field_first;\
  730. \
  731. /**\
  732. * Pan scan.\
  733. * - encoding: Set by user.\
  734. * - decoding: Set by libavcodec.\
  735. */\
  736. AVPanScan *pan_scan;\
  737. \
  738. /**\
  739. * Tell user application that palette has changed from previous frame.\
  740. * - encoding: ??? (no palette-enabled encoder yet)\
  741. * - decoding: Set by libavcodec. (default 0).\
  742. */\
  743. int palette_has_changed;\
  744. \
  745. /**\
  746. * codec suggestion on buffer type if != 0\
  747. * - encoding: unused\
  748. * - decoding: Set by libavcodec. (before get_buffer() call)).\
  749. */\
  750. int buffer_hints;\
  751. \
  752. /**\
  753. * DCT coefficients\
  754. * - encoding: unused\
  755. * - decoding: Set by libavcodec.\
  756. */\
  757. short *dct_coeff;\
  758. \
  759. /**\
  760. * motion referece frame index\
  761. * - encoding: Set by user.\
  762. * - decoding: Set by libavcodec.\
  763. */\
  764. int8_t *ref_index[2];\
  765. \
  766. /**\
  767. * reordered opaque 64bit number (generally a PTS) from AVCodecContext.reordered_opaque\
  768. * output in AVFrame.reordered_opaque\
  769. * - encoding: unused\
  770. * - decoding: Read by user.\
  771. */\
  772. int64_t reordered_opaque;\
  773. \
  774. /**\
  775. * hardware accelerator private data (FFmpeg allocated)\
  776. * - encoding: unused\
  777. * - decoding: Set by libavcodec\
  778. */\
  779. void *hwaccel_picture_private;\
  780. #define FF_QSCALE_TYPE_MPEG1 0
  781. #define FF_QSCALE_TYPE_MPEG2 1
  782. #define FF_QSCALE_TYPE_H264 2
  783. #define FF_BUFFER_TYPE_INTERNAL 1
  784. #define FF_BUFFER_TYPE_USER 2 ///< direct rendering buffers (image is (de)allocated by user)
  785. #define FF_BUFFER_TYPE_SHARED 4 ///< Buffer from somewhere else; don't deallocate image (data/base), all other tables are not shared.
  786. #define FF_BUFFER_TYPE_COPY 8 ///< Just a (modified) copy of some other buffer, don't deallocate anything.
  787. #define FF_I_TYPE 1 ///< Intra
  788. #define FF_P_TYPE 2 ///< Predicted
  789. #define FF_B_TYPE 3 ///< Bi-dir predicted
  790. #define FF_S_TYPE 4 ///< S(GMC)-VOP MPEG4
  791. #define FF_SI_TYPE 5 ///< Switching Intra
  792. #define FF_SP_TYPE 6 ///< Switching Predicted
  793. #define FF_BI_TYPE 7
  794. #define FF_BUFFER_HINTS_VALID 0x01 // Buffer hints value is meaningful (if 0 ignore).
  795. #define FF_BUFFER_HINTS_READABLE 0x02 // Codec will read from buffer.
  796. #define FF_BUFFER_HINTS_PRESERVE 0x04 // User must not alter buffer content.
  797. #define FF_BUFFER_HINTS_REUSABLE 0x08 // Codec will reuse the buffer (update).
  798. /**
  799. * Audio Video Frame.
  800. * New fields can be added to the end of FF_COMMON_FRAME with minor version
  801. * bumps.
  802. * Removal, reordering and changes to existing fields require a major
  803. * version bump. No fields should be added into AVFrame before or after
  804. * FF_COMMON_FRAME!
  805. * sizeof(AVFrame) must not be used outside libav*.
  806. */
  807. typedef struct AVFrame {
  808. FF_COMMON_FRAME
  809. } AVFrame;
  810. /**
  811. * main external API structure.
  812. * New fields can be added to the end with minor version bumps.
  813. * Removal, reordering and changes to existing fields require a major
  814. * version bump.
  815. * sizeof(AVCodecContext) must not be used outside libav*.
  816. */
  817. typedef struct AVCodecContext {
  818. /**
  819. * information on struct for av_log
  820. * - set by avcodec_alloc_context
  821. */
  822. const AVClass *av_class;
  823. /**
  824. * the average bitrate
  825. * - encoding: Set by user; unused for constant quantizer encoding.
  826. * - decoding: Set by libavcodec. 0 or some bitrate if this info is available in the stream.
  827. */
  828. int bit_rate;
  829. /**
  830. * number of bits the bitstream is allowed to diverge from the reference.
  831. * the reference can be CBR (for CBR pass1) or VBR (for pass2)
  832. * - encoding: Set by user; unused for constant quantizer encoding.
  833. * - decoding: unused
  834. */
  835. int bit_rate_tolerance;
  836. /**
  837. * CODEC_FLAG_*.
  838. * - encoding: Set by user.
  839. * - decoding: Set by user.
  840. */
  841. int flags;
  842. /**
  843. * Some codecs need additional format info. It is stored here.
  844. * If any muxer uses this then ALL demuxers/parsers AND encoders for the
  845. * specific codec MUST set it correctly otherwise stream copy breaks.
  846. * In general use of this field by muxers is not recommanded.
  847. * - encoding: Set by libavcodec.
  848. * - decoding: Set by libavcodec. (FIXME: Is this OK?)
  849. */
  850. int sub_id;
  851. /**
  852. * Motion estimation algorithm used for video coding.
  853. * 1 (zero), 2 (full), 3 (log), 4 (phods), 5 (epzs), 6 (x1), 7 (hex),
  854. * 8 (umh), 9 (iter), 10 (tesa) [7, 8, 10 are x264 specific, 9 is snow specific]
  855. * - encoding: MUST be set by user.
  856. * - decoding: unused
  857. */
  858. int me_method;
  859. /**
  860. * some codecs need / can use extradata like Huffman tables.
  861. * mjpeg: Huffman tables
  862. * rv10: additional flags
  863. * mpeg4: global headers (they can be in the bitstream or here)
  864. * The allocated memory should be FF_INPUT_BUFFER_PADDING_SIZE bytes larger
  865. * than extradata_size to avoid prolems if it is read with the bitstream reader.
  866. * The bytewise contents of extradata must not depend on the architecture or CPU endianness.
  867. * - encoding: Set/allocated/freed by libavcodec.
  868. * - decoding: Set/allocated/freed by user.
  869. */
  870. uint8_t *extradata;
  871. int extradata_size;
  872. /**
  873. * This is the fundamental unit of time (in seconds) in terms
  874. * of which frame timestamps are represented. For fixed-fps content,
  875. * timebase should be 1/framerate and timestamp increments should be
  876. * identically 1.
  877. * - encoding: MUST be set by user.
  878. * - decoding: Set by libavcodec.
  879. */
  880. AVRational time_base;
  881. /* video only */
  882. /**
  883. * picture width / height.
  884. * - encoding: MUST be set by user.
  885. * - decoding: Set by libavcodec.
  886. * Note: For compatibility it is possible to set this instead of
  887. * coded_width/height before decoding.
  888. */
  889. int width, height;
  890. #define FF_ASPECT_EXTENDED 15
  891. /**
  892. * the number of pictures in a group of pictures, or 0 for intra_only
  893. * - encoding: Set by user.
  894. * - decoding: unused
  895. */
  896. int gop_size;
  897. /**
  898. * Pixel format, see PIX_FMT_xxx.
  899. * - encoding: Set by user.
  900. * - decoding: Set by libavcodec.
  901. */
  902. enum PixelFormat pix_fmt;
  903. /**
  904. * Frame rate emulation. If not zero, the lower layer (i.e. format handler)
  905. * has to read frames at native frame rate.
  906. * - encoding: Set by user.
  907. * - decoding: unused
  908. */
  909. int rate_emu;
  910. /**
  911. * If non NULL, 'draw_horiz_band' is called by the libavcodec
  912. * decoder to draw a horizontal band. It improves cache usage. Not
  913. * all codecs can do that. You must check the codec capabilities
  914. * beforehand.
  915. * The function is also used by hardware acceleration APIs.
  916. * It is called at least once during frame decoding to pass
  917. * the data needed for hardware render.
  918. * In that mode instead of pixel data, AVFrame points to
  919. * a structure specific to the acceleration API. The application
  920. * reads the structure and can change some fields to indicate progress
  921. * or mark state.
  922. * - encoding: unused
  923. * - decoding: Set by user.
  924. * @param height the height of the slice
  925. * @param y the y position of the slice
  926. * @param type 1->top field, 2->bottom field, 3->frame
  927. * @param offset offset into the AVFrame.data from which the slice should be read
  928. */
  929. void (*draw_horiz_band)(struct AVCodecContext *s,
  930. const AVFrame *src, int offset[4],
  931. int y, int type, int height);
  932. /* audio only */
  933. int sample_rate; ///< samples per second
  934. int channels; ///< number of audio channels
  935. /**
  936. * audio sample format
  937. * - encoding: Set by user.
  938. * - decoding: Set by libavcodec.
  939. */
  940. enum SampleFormat sample_fmt; ///< sample format, currently unused
  941. /* The following data should not be initialized. */
  942. /**
  943. * Samples per packet, initialized when calling 'init'.
  944. */
  945. int frame_size;
  946. int frame_number; ///< audio or video frame number
  947. int real_pict_num; ///< Returns the real picture number of previous encoded frame.
  948. /**
  949. * Number of frames the decoded output will be delayed relative to
  950. * the encoded input.
  951. * - encoding: Set by libavcodec.
  952. * - decoding: unused
  953. */
  954. int delay;
  955. /* - encoding parameters */
  956. float qcompress; ///< amount of qscale change between easy & hard scenes (0.0-1.0)
  957. float qblur; ///< amount of qscale smoothing over time (0.0-1.0)
  958. /**
  959. * minimum quantizer
  960. * - encoding: Set by user.
  961. * - decoding: unused
  962. */
  963. int qmin;
  964. /**
  965. * maximum quantizer
  966. * - encoding: Set by user.
  967. * - decoding: unused
  968. */
  969. int qmax;
  970. /**
  971. * maximum quantizer difference between frames
  972. * - encoding: Set by user.
  973. * - decoding: unused
  974. */
  975. int max_qdiff;
  976. /**
  977. * maximum number of B-frames between non-B-frames
  978. * Note: The output will be delayed by max_b_frames+1 relative to the input.
  979. * - encoding: Set by user.
  980. * - decoding: unused
  981. */
  982. int max_b_frames;
  983. /**
  984. * qscale factor between IP and B-frames
  985. * If > 0 then the last P-frame quantizer will be used (q= lastp_q*factor+offset).
  986. * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
  987. * - encoding: Set by user.
  988. * - decoding: unused
  989. */
  990. float b_quant_factor;
  991. /** obsolete FIXME remove */
  992. int rc_strategy;
  993. #define FF_RC_STRATEGY_XVID 1
  994. int b_frame_strategy;
  995. /**
  996. * hurry up amount
  997. * - encoding: unused
  998. * - decoding: Set by user. 1-> Skip B-frames, 2-> Skip IDCT/dequant too, 5-> Skip everything except header
  999. * @deprecated Deprecated in favor of skip_idct and skip_frame.
  1000. */
  1001. int hurry_up;
  1002. struct AVCodec *codec;
  1003. void *priv_data;
  1004. int rtp_payload_size; /* The size of the RTP payload: the coder will */
  1005. /* do its best to deliver a chunk with size */
  1006. /* below rtp_payload_size, the chunk will start */
  1007. /* with a start code on some codecs like H.263. */
  1008. /* This doesn't take account of any particular */
  1009. /* headers inside the transmitted RTP payload. */
  1010. /* The RTP callback: This function is called */
  1011. /* every time the encoder has a packet to send. */
  1012. /* It depends on the encoder if the data starts */
  1013. /* with a Start Code (it should). H.263 does. */
  1014. /* mb_nb contains the number of macroblocks */
  1015. /* encoded in the RTP payload. */
  1016. void (*rtp_callback)(struct AVCodecContext *avctx, void *data, int size, int mb_nb);
  1017. /* statistics, used for 2-pass encoding */
  1018. int mv_bits;
  1019. int header_bits;
  1020. int i_tex_bits;
  1021. int p_tex_bits;
  1022. int i_count;
  1023. int p_count;
  1024. int skip_count;
  1025. int misc_bits;
  1026. /**
  1027. * number of bits used for the previously encoded frame
  1028. * - encoding: Set by libavcodec.
  1029. * - decoding: unused
  1030. */
  1031. int frame_bits;
  1032. /**
  1033. * Private data of the user, can be used to carry app specific stuff.
  1034. * - encoding: Set by user.
  1035. * - decoding: Set by user.
  1036. */
  1037. void *opaque;
  1038. char codec_name[32];
  1039. enum CodecType codec_type; /* see CODEC_TYPE_xxx */
  1040. enum CodecID codec_id; /* see CODEC_ID_xxx */
  1041. /**
  1042. * fourcc (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
  1043. * This is used to work around some encoder bugs.
  1044. * A demuxer should set this to what is stored in the field used to identify the codec.
  1045. * If there are multiple such fields in a container then the demuxer should choose the one
  1046. * which maximizes the information about the used codec.
  1047. * If the codec tag field in a container is larger then 32 bits then the demuxer should
  1048. * remap the longer ID to 32 bits with a table or other structure. Alternatively a new
  1049. * extra_codec_tag + size could be added but for this a clear advantage must be demonstrated
  1050. * first.
  1051. * - encoding: Set by user, if not then the default based on codec_id will be used.
  1052. * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
  1053. */
  1054. unsigned int codec_tag;
  1055. /**
  1056. * Work around bugs in encoders which sometimes cannot be detected automatically.
  1057. * - encoding: Set by user
  1058. * - decoding: Set by user
  1059. */
  1060. int workaround_bugs;
  1061. #define FF_BUG_AUTODETECT 1 ///< autodetection
  1062. #define FF_BUG_OLD_MSMPEG4 2
  1063. #define FF_BUG_XVID_ILACE 4
  1064. #define FF_BUG_UMP4 8
  1065. #define FF_BUG_NO_PADDING 16
  1066. #define FF_BUG_AMV 32
  1067. #define FF_BUG_AC_VLC 0 ///< Will be removed, libavcodec can now handle these non-compliant files by default.
  1068. #define FF_BUG_QPEL_CHROMA 64
  1069. #define FF_BUG_STD_QPEL 128
  1070. #define FF_BUG_QPEL_CHROMA2 256
  1071. #define FF_BUG_DIRECT_BLOCKSIZE 512
  1072. #define FF_BUG_EDGE 1024
  1073. #define FF_BUG_HPEL_CHROMA 2048
  1074. #define FF_BUG_DC_CLIP 4096
  1075. #define FF_BUG_MS 8192 ///< Work around various bugs in Microsoft's broken decoders.
  1076. //#define FF_BUG_FAKE_SCALABILITY 16 //Autodetection should work 100%.
  1077. /**
  1078. * luma single coefficient elimination threshold
  1079. * - encoding: Set by user.
  1080. * - decoding: unused
  1081. */
  1082. int luma_elim_threshold;
  1083. /**
  1084. * chroma single coeff elimination threshold
  1085. * - encoding: Set by user.
  1086. * - decoding: unused
  1087. */
  1088. int chroma_elim_threshold;
  1089. /**
  1090. * strictly follow the standard (MPEG4, ...).
  1091. * - encoding: Set by user.
  1092. * - decoding: Set by user.
  1093. * Setting this to STRICT or higher means the encoder and decoder will
  1094. * generally do stupid things. While setting it to inofficial or lower
  1095. * will mean the encoder might use things that are not supported by all
  1096. * spec compliant decoders. Decoders make no difference between normal,
  1097. * inofficial and experimental, that is they always try to decode things
  1098. * when they can unless they are explicitly asked to behave stupid
  1099. * (=strictly conform to the specs)
  1100. */
  1101. int strict_std_compliance;
  1102. #define FF_COMPLIANCE_VERY_STRICT 2 ///< Strictly conform to a older more strict version of the spec or reference software.
  1103. #define FF_COMPLIANCE_STRICT 1 ///< Strictly conform to all the things in the spec no matter what consequences.
  1104. #define FF_COMPLIANCE_NORMAL 0
  1105. #define FF_COMPLIANCE_INOFFICIAL -1 ///< Allow inofficial extensions.
  1106. #define FF_COMPLIANCE_EXPERIMENTAL -2 ///< Allow nonstandardized experimental things.
  1107. /**
  1108. * qscale offset between IP and B-frames
  1109. * - encoding: Set by user.
  1110. * - decoding: unused
  1111. */
  1112. float b_quant_offset;
  1113. /**
  1114. * Error recognization; higher values will detect more errors but may
  1115. * misdetect some more or less valid parts as errors.
  1116. * - encoding: unused
  1117. * - decoding: Set by user.
  1118. */
  1119. int error_recognition;
  1120. #define FF_ER_CAREFUL 1
  1121. #define FF_ER_COMPLIANT 2
  1122. #define FF_ER_AGGRESSIVE 3
  1123. #define FF_ER_VERY_AGGRESSIVE 4
  1124. /**
  1125. * Called at the beginning of each frame to get a buffer for it.
  1126. * If pic.reference is set then the frame will be read later by libavcodec.
  1127. * avcodec_align_dimensions() should be used to find the required width and
  1128. * height, as they normally need to be rounded up to the next multiple of 16.
  1129. * - encoding: unused
  1130. * - decoding: Set by libavcodec., user can override.
  1131. */
  1132. int (*get_buffer)(struct AVCodecContext *c, AVFrame *pic);
  1133. /**
  1134. * Called to release buffers which were allocated with get_buffer.
  1135. * A released buffer can be reused in get_buffer().
  1136. * pic.data[*] must be set to NULL.
  1137. * - encoding: unused
  1138. * - decoding: Set by libavcodec., user can override.
  1139. */
  1140. void (*release_buffer)(struct AVCodecContext *c, AVFrame *pic);
  1141. /**
  1142. * Size of the frame reordering buffer in the decoder.
  1143. * For MPEG-2 it is 1 IPB or 0 low delay IP.
  1144. * - encoding: Set by libavcodec.
  1145. * - decoding: Set by libavcodec.
  1146. */
  1147. int has_b_frames;
  1148. /**
  1149. * number of bytes per packet if constant and known or 0
  1150. * Used by some WAV based audio codecs.
  1151. */
  1152. int block_align;
  1153. int parse_only; /* - decoding only: If true, only parsing is done
  1154. (function avcodec_parse_frame()). The frame
  1155. data is returned. Only MPEG codecs support this now. */
  1156. /**
  1157. * 0-> h263 quant 1-> mpeg quant
  1158. * - encoding: Set by user.
  1159. * - decoding: unused
  1160. */
  1161. int mpeg_quant;
  1162. /**
  1163. * pass1 encoding statistics output buffer
  1164. * - encoding: Set by libavcodec.
  1165. * - decoding: unused
  1166. */
  1167. char *stats_out;
  1168. /**
  1169. * pass2 encoding statistics input buffer
  1170. * Concatenated stuff from stats_out of pass1 should be placed here.
  1171. * - encoding: Allocated/set/freed by user.
  1172. * - decoding: unused
  1173. */
  1174. char *stats_in;
  1175. /**
  1176. * ratecontrol qmin qmax limiting method
  1177. * 0-> clipping, 1-> use a nice continous function to limit qscale wthin qmin/qmax.
  1178. * - encoding: Set by user.
  1179. * - decoding: unused
  1180. */
  1181. float rc_qsquish;
  1182. float rc_qmod_amp;
  1183. int rc_qmod_freq;
  1184. /**
  1185. * ratecontrol override, see RcOverride
  1186. * - encoding: Allocated/set/freed by user.
  1187. * - decoding: unused
  1188. */
  1189. RcOverride *rc_override;
  1190. int rc_override_count;
  1191. /**
  1192. * rate control equation
  1193. * - encoding: Set by user
  1194. * - decoding: unused
  1195. */
  1196. const char *rc_eq;
  1197. /**
  1198. * maximum bitrate
  1199. * - encoding: Set by user.
  1200. * - decoding: unused
  1201. */
  1202. int rc_max_rate;
  1203. /**
  1204. * minimum bitrate
  1205. * - encoding: Set by user.
  1206. * - decoding: unused
  1207. */
  1208. int rc_min_rate;
  1209. /**
  1210. * decoder bitstream buffer size
  1211. * - encoding: Set by user.
  1212. * - decoding: unused
  1213. */
  1214. int rc_buffer_size;
  1215. float rc_buffer_aggressivity;
  1216. /**
  1217. * qscale factor between P and I-frames
  1218. * If > 0 then the last p frame quantizer will be used (q= lastp_q*factor+offset).
  1219. * If < 0 then normal ratecontrol will be done (q= -normal_q*factor+offset).
  1220. * - encoding: Set by user.
  1221. * - decoding: unused
  1222. */
  1223. float i_quant_factor;
  1224. /**
  1225. * qscale offset between P and I-frames
  1226. * - encoding: Set by user.
  1227. * - decoding: unused
  1228. */
  1229. float i_quant_offset;
  1230. /**
  1231. * initial complexity for pass1 ratecontrol
  1232. * - encoding: Set by user.
  1233. * - decoding: unused
  1234. */
  1235. float rc_initial_cplx;
  1236. /**
  1237. * DCT algorithm, see FF_DCT_* below
  1238. * - encoding: Set by user.
  1239. * - decoding: unused
  1240. */
  1241. int dct_algo;
  1242. #define FF_DCT_AUTO 0
  1243. #define FF_DCT_FASTINT 1
  1244. #define FF_DCT_INT 2
  1245. #define FF_DCT_MMX 3
  1246. #define FF_DCT_MLIB 4
  1247. #define FF_DCT_ALTIVEC 5
  1248. #define FF_DCT_FAAN 6
  1249. /**
  1250. * luminance masking (0-> disabled)
  1251. * - encoding: Set by user.
  1252. * - decoding: unused
  1253. */
  1254. float lumi_masking;
  1255. /**
  1256. * temporary complexity masking (0-> disabled)
  1257. * - encoding: Set by user.
  1258. * - decoding: unused
  1259. */
  1260. float temporal_cplx_masking;
  1261. /**
  1262. * spatial complexity masking (0-> disabled)
  1263. * - encoding: Set by user.
  1264. * - decoding: unused
  1265. */
  1266. float spatial_cplx_masking;
  1267. /**
  1268. * p block masking (0-> disabled)
  1269. * - encoding: Set by user.
  1270. * - decoding: unused
  1271. */
  1272. float p_masking;
  1273. /**
  1274. * darkness masking (0-> disabled)
  1275. * - encoding: Set by user.
  1276. * - decoding: unused
  1277. */
  1278. float dark_masking;
  1279. /**
  1280. * IDCT algorithm, see FF_IDCT_* below.
  1281. * - encoding: Set by user.
  1282. * - decoding: Set by user.
  1283. */
  1284. int idct_algo;
  1285. #define FF_IDCT_AUTO 0
  1286. #define FF_IDCT_INT 1
  1287. #define FF_IDCT_SIMPLE 2
  1288. #define FF_IDCT_SIMPLEMMX 3
  1289. #define FF_IDCT_LIBMPEG2MMX 4
  1290. #define FF_IDCT_PS2 5
  1291. #define FF_IDCT_MLIB 6
  1292. #define FF_IDCT_ARM 7
  1293. #define FF_IDCT_ALTIVEC 8
  1294. #define FF_IDCT_SH4 9
  1295. #define FF_IDCT_SIMPLEARM 10
  1296. #define FF_IDCT_H264 11
  1297. #define FF_IDCT_VP3 12
  1298. #define FF_IDCT_IPP 13
  1299. #define FF_IDCT_XVIDMMX 14
  1300. #define FF_IDCT_CAVS 15
  1301. #define FF_IDCT_SIMPLEARMV5TE 16
  1302. #define FF_IDCT_SIMPLEARMV6 17
  1303. #define FF_IDCT_SIMPLEVIS 18
  1304. #define FF_IDCT_WMV2 19
  1305. #define FF_IDCT_FAAN 20
  1306. #define FF_IDCT_EA 21
  1307. #define FF_IDCT_SIMPLENEON 22
  1308. #define FF_IDCT_SIMPLEALPHA 23
  1309. /**
  1310. * slice count
  1311. * - encoding: Set by libavcodec.
  1312. * - decoding: Set by user (or 0).
  1313. */
  1314. int slice_count;
  1315. /**
  1316. * slice offsets in the frame in bytes
  1317. * - encoding: Set/allocated by libavcodec.
  1318. * - decoding: Set/allocated by user (or NULL).
  1319. */
  1320. int *slice_offset;
  1321. /**
  1322. * error concealment flags
  1323. * - encoding: unused
  1324. * - decoding: Set by user.
  1325. */
  1326. int error_concealment;
  1327. #define FF_EC_GUESS_MVS 1
  1328. #define FF_EC_DEBLOCK 2
  1329. /**
  1330. * dsp_mask could be add used to disable unwanted CPU features
  1331. * CPU features (i.e. MMX, SSE. ...)
  1332. *
  1333. * With the FORCE flag you may instead enable given CPU features.
  1334. * (Dangerous: Usable in case of misdetection, improper usage however will
  1335. * result into program crash.)
  1336. */
  1337. unsigned dsp_mask;
  1338. #define FF_MM_FORCE 0x80000000 /* Force usage of selected flags (OR) */
  1339. /* lower 16 bits - CPU features */
  1340. #define FF_MM_MMX 0x0001 ///< standard MMX
  1341. #define FF_MM_3DNOW 0x0004 ///< AMD 3DNOW
  1342. #define FF_MM_MMXEXT 0x0002 ///< SSE integer functions or AMD MMX ext
  1343. #define FF_MM_SSE 0x0008 ///< SSE functions
  1344. #define FF_MM_SSE2 0x0010 ///< PIV SSE2 functions
  1345. #define FF_MM_3DNOWEXT 0x0020 ///< AMD 3DNowExt
  1346. #define FF_MM_SSE3 0x0040 ///< Prescott SSE3 functions
  1347. #define FF_MM_SSSE3 0x0080 ///< Conroe SSSE3 functions
  1348. #define FF_MM_SSE4 0x0100 ///< Penryn SSE4.1 functions
  1349. #define FF_MM_SSE42 0x0200 ///< Nehalem SSE4.2 functions
  1350. #define FF_MM_IWMMXT 0x0100 ///< XScale IWMMXT
  1351. #define FF_MM_ALTIVEC 0x0001 ///< standard AltiVec
  1352. /**
  1353. * bits per sample/pixel from the demuxer (needed for huffyuv).
  1354. * - encoding: Set by libavcodec.
  1355. * - decoding: Set by user.
  1356. */
  1357. int bits_per_coded_sample;
  1358. /**
  1359. * prediction method (needed for huffyuv)
  1360. * - encoding: Set by user.
  1361. * - decoding: unused
  1362. */
  1363. int prediction_method;
  1364. #define FF_PRED_LEFT 0
  1365. #define FF_PRED_PLANE 1
  1366. #define FF_PRED_MEDIAN 2
  1367. /**
  1368. * sample aspect ratio (0 if unknown)
  1369. * That is the width of a pixel divided by the height of the pixel.
  1370. * Numerator and denominator must be relatively prime and smaller than 256 for some video standards.
  1371. * - encoding: Set by user.
  1372. * - decoding: Set by libavcodec.
  1373. */
  1374. AVRational sample_aspect_ratio;
  1375. /**
  1376. * the picture in the bitstream
  1377. * - encoding: Set by libavcodec.
  1378. * - decoding: Set by libavcodec.
  1379. */
  1380. AVFrame *coded_frame;
  1381. /**
  1382. * debug
  1383. * - encoding: Set by user.
  1384. * - decoding: Set by user.
  1385. */
  1386. int debug;
  1387. #define FF_DEBUG_PICT_INFO 1
  1388. #define FF_DEBUG_RC 2
  1389. #define FF_DEBUG_BITSTREAM 4
  1390. #define FF_DEBUG_MB_TYPE 8
  1391. #define FF_DEBUG_QP 16
  1392. #define FF_DEBUG_MV 32
  1393. #define FF_DEBUG_DCT_COEFF 0x00000040
  1394. #define FF_DEBUG_SKIP 0x00000080
  1395. #define FF_DEBUG_STARTCODE 0x00000100
  1396. #define FF_DEBUG_PTS 0x00000200
  1397. #define FF_DEBUG_ER 0x00000400
  1398. #define FF_DEBUG_MMCO 0x00000800
  1399. #define FF_DEBUG_BUGS 0x00001000
  1400. #define FF_DEBUG_VIS_QP 0x00002000
  1401. #define FF_DEBUG_VIS_MB_TYPE 0x00004000
  1402. #define FF_DEBUG_BUFFERS 0x00008000
  1403. /**
  1404. * debug
  1405. * - encoding: Set by user.
  1406. * - decoding: Set by user.
  1407. */
  1408. int debug_mv;
  1409. #define FF_DEBUG_VIS_MV_P_FOR 0x00000001 //visualize forward predicted MVs of P frames
  1410. #define FF_DEBUG_VIS_MV_B_FOR 0x00000002 //visualize forward predicted MVs of B frames
  1411. #define FF_DEBUG_VIS_MV_B_BACK 0x00000004 //visualize backward predicted MVs of B frames
  1412. /**
  1413. * error
  1414. * - encoding: Set by libavcodec if flags&CODEC_FLAG_PSNR.
  1415. * - decoding: unused
  1416. */
  1417. uint64_t error[4];
  1418. /**
  1419. * minimum MB quantizer
  1420. * - encoding: unused
  1421. * - decoding: unused
  1422. */
  1423. int mb_qmin;
  1424. /**
  1425. * maximum MB quantizer
  1426. * - encoding: unused
  1427. * - decoding: unused
  1428. */
  1429. int mb_qmax;
  1430. /**
  1431. * motion estimation comparison function
  1432. * - encoding: Set by user.
  1433. * - decoding: unused
  1434. */
  1435. int me_cmp;
  1436. /**
  1437. * subpixel motion estimation comparison function
  1438. * - encoding: Set by user.
  1439. * - decoding: unused
  1440. */
  1441. int me_sub_cmp;
  1442. /**
  1443. * macroblock comparison function (not supported yet)
  1444. * - encoding: Set by user.
  1445. * - decoding: unused
  1446. */
  1447. int mb_cmp;
  1448. /**
  1449. * interlaced DCT comparison function
  1450. * - encoding: Set by user.
  1451. * - decoding: unused
  1452. */
  1453. int ildct_cmp;
  1454. #define FF_CMP_SAD 0
  1455. #define FF_CMP_SSE 1
  1456. #define FF_CMP_SATD 2
  1457. #define FF_CMP_DCT 3
  1458. #define FF_CMP_PSNR 4
  1459. #define FF_CMP_BIT 5
  1460. #define FF_CMP_RD 6
  1461. #define FF_CMP_ZERO 7
  1462. #define FF_CMP_VSAD 8
  1463. #define FF_CMP_VSSE 9
  1464. #define FF_CMP_NSSE 10
  1465. #define FF_CMP_W53 11
  1466. #define FF_CMP_W97 12
  1467. #define FF_CMP_DCTMAX 13
  1468. #define FF_CMP_DCT264 14
  1469. #define FF_CMP_CHROMA 256
  1470. /**
  1471. * ME diamond size & shape
  1472. * - encoding: Set by user.
  1473. * - decoding: unused
  1474. */
  1475. int dia_size;
  1476. /**
  1477. * amount of previous MV predictors (2a+1 x 2a+1 square)
  1478. * - encoding: Set by user.
  1479. * - decoding: unused
  1480. */
  1481. int last_predictor_count;
  1482. /**
  1483. * prepass for motion estimation
  1484. * - encoding: Set by user.
  1485. * - decoding: unused
  1486. */
  1487. int pre_me;
  1488. /**
  1489. * motion estimation prepass comparison function
  1490. * - encoding: Set by user.
  1491. * - decoding: unused
  1492. */
  1493. int me_pre_cmp;
  1494. /**
  1495. * ME prepass diamond size & shape
  1496. * - encoding: Set by user.
  1497. * - decoding: unused
  1498. */
  1499. int pre_dia_size;
  1500. /**
  1501. * subpel ME quality
  1502. * - encoding: Set by user.
  1503. * - decoding: unused
  1504. */
  1505. int me_subpel_quality;
  1506. /**
  1507. * callback to negotiate the pixelFormat
  1508. * @param fmt is the list of formats which are supported by the codec,
  1509. * it is terminated by -1 as 0 is a valid format, the formats are ordered by quality.
  1510. * The first is always the native one.
  1511. * @return the chosen format
  1512. * - encoding: unused
  1513. * - decoding: Set by user, if not set the native format will be chosen.
  1514. */
  1515. enum PixelFormat (*get_format)(struct AVCodecContext *s, const enum PixelFormat * fmt);
  1516. /**
  1517. * DTG active format information (additional aspect ratio
  1518. * information only used in DVB MPEG-2 transport streams)
  1519. * 0 if not set.
  1520. *
  1521. * - encoding: unused
  1522. * - decoding: Set by decoder.
  1523. */
  1524. int dtg_active_format;
  1525. #define FF_DTG_AFD_SAME 8
  1526. #define FF_DTG_AFD_4_3 9
  1527. #define FF_DTG_AFD_16_9 10
  1528. #define FF_DTG_AFD_14_9 11
  1529. #define FF_DTG_AFD_4_3_SP_14_9 13
  1530. #define FF_DTG_AFD_16_9_SP_14_9 14
  1531. #define FF_DTG_AFD_SP_4_3 15
  1532. /**
  1533. * maximum motion estimation search range in subpel units
  1534. * If 0 then no limit.
  1535. *
  1536. * - encoding: Set by user.
  1537. * - decoding: unused
  1538. */
  1539. int me_range;
  1540. /**
  1541. * intra quantizer bias
  1542. * - encoding: Set by user.
  1543. * - decoding: unused
  1544. */
  1545. int intra_quant_bias;
  1546. #define FF_DEFAULT_QUANT_BIAS 999999
  1547. /**
  1548. * inter quantizer bias
  1549. * - encoding: Set by user.
  1550. * - decoding: unused
  1551. */
  1552. int inter_quant_bias;
  1553. /**
  1554. * color table ID
  1555. * - encoding: unused
  1556. * - decoding: Which clrtable should be used for 8bit RGB images.
  1557. * Tables have to be stored somewhere. FIXME
  1558. */
  1559. int color_table_id;
  1560. /**
  1561. * internal_buffer count
  1562. * Don't touch, used by libavcodec default_get_buffer().
  1563. */
  1564. int internal_buffer_count;
  1565. /**
  1566. * internal_buffers
  1567. * Don't touch, used by libavcodec default_get_buffer().
  1568. */
  1569. void *internal_buffer;
  1570. #define FF_LAMBDA_SHIFT 7
  1571. #define FF_LAMBDA_SCALE (1<<FF_LAMBDA_SHIFT)
  1572. #define FF_QP2LAMBDA 118 ///< factor to convert from H.263 QP to lambda
  1573. #define FF_LAMBDA_MAX (256*128-1)
  1574. #define FF_QUALITY_SCALE FF_LAMBDA_SCALE //FIXME maybe remove
  1575. /**
  1576. * Global quality for codecs which cannot change it per frame.
  1577. * This should be proportional to MPEG-1/2/4 qscale.
  1578. * - encoding: Set by user.
  1579. * - decoding: unused
  1580. */
  1581. int global_quality;
  1582. #define FF_CODER_TYPE_VLC 0
  1583. #define FF_CODER_TYPE_AC 1
  1584. #define FF_CODER_TYPE_RAW 2
  1585. #define FF_CODER_TYPE_RLE 3
  1586. #define FF_CODER_TYPE_DEFLATE 4
  1587. /**
  1588. * coder type
  1589. * - encoding: Set by user.
  1590. * - decoding: unused
  1591. */
  1592. int coder_type;
  1593. /**
  1594. * context model
  1595. * - encoding: Set by user.
  1596. * - decoding: unused
  1597. */
  1598. int context_model;
  1599. #if 0
  1600. /**
  1601. *
  1602. * - encoding: unused
  1603. * - decoding: Set by user.
  1604. */
  1605. uint8_t * (*realloc)(struct AVCodecContext *s, uint8_t *buf, int buf_size);
  1606. #endif
  1607. /**
  1608. * slice flags
  1609. * - encoding: unused
  1610. * - decoding: Set by user.
  1611. */
  1612. int slice_flags;
  1613. #define SLICE_FLAG_CODED_ORDER 0x0001 ///< draw_horiz_band() is called in coded order instead of display
  1614. #define SLICE_FLAG_ALLOW_FIELD 0x0002 ///< allow draw_horiz_band() with field slices (MPEG2 field pics)
  1615. #define SLICE_FLAG_ALLOW_PLANE 0x0004 ///< allow draw_horiz_band() with 1 component at a time (SVQ1)
  1616. /**
  1617. * XVideo Motion Acceleration
  1618. * - encoding: forbidden
  1619. * - decoding: set by decoder
  1620. */
  1621. int xvmc_acceleration;
  1622. /**
  1623. * macroblock decision mode
  1624. * - encoding: Set by user.
  1625. * - decoding: unused
  1626. */
  1627. int mb_decision;
  1628. #define FF_MB_DECISION_SIMPLE 0 ///< uses mb_cmp
  1629. #define FF_MB_DECISION_BITS 1 ///< chooses the one which needs the fewest bits
  1630. #define FF_MB_DECISION_RD 2 ///< rate distortion
  1631. /**
  1632. * custom intra quantization matrix
  1633. * - encoding: Set by user, can be NULL.
  1634. * - decoding: Set by libavcodec.
  1635. */
  1636. uint16_t *intra_matrix;
  1637. /**
  1638. * custom inter quantization matrix
  1639. * - encoding: Set by user, can be NULL.
  1640. * - decoding: Set by libavcodec.
  1641. */
  1642. uint16_t *inter_matrix;
  1643. /**
  1644. * fourcc from the AVI stream header (LSB first, so "ABCD" -> ('D'<<24) + ('C'<<16) + ('B'<<8) + 'A').
  1645. * This is used to work around some encoder bugs.
  1646. * - encoding: unused
  1647. * - decoding: Set by user, will be converted to uppercase by libavcodec during init.
  1648. */
  1649. unsigned int stream_codec_tag;
  1650. /**
  1651. * scene change detection threshold
  1652. * 0 is default, larger means fewer detected scene changes.
  1653. * - encoding: Set by user.
  1654. * - decoding: unused
  1655. */
  1656. int scenechange_threshold;
  1657. /**
  1658. * minimum Lagrange multipler
  1659. * - encoding: Set by user.
  1660. * - decoding: unused
  1661. */
  1662. int lmin;
  1663. /**
  1664. * maximum Lagrange multipler
  1665. * - encoding: Set by user.
  1666. * - decoding: unused
  1667. */
  1668. int lmax;
  1669. /**
  1670. * palette control structure
  1671. * - encoding: ??? (no palette-enabled encoder yet)
  1672. * - decoding: Set by user.
  1673. */
  1674. struct AVPaletteControl *palctrl;
  1675. /**
  1676. * noise reduction strength
  1677. * - encoding: Set by user.
  1678. * - decoding: unused
  1679. */
  1680. int noise_reduction;
  1681. /**
  1682. * Called at the beginning of a frame to get cr buffer for it.
  1683. * Buffer type (size, hints) must be the same. libavcodec won't check it.
  1684. * libavcodec will pass previous buffer in pic, function should return
  1685. * same buffer or new buffer with old frame "painted" into it.
  1686. * If pic.data[0] == NULL must behave like get_buffer().
  1687. * - encoding: unused
  1688. * - decoding: Set by libavcodec., user can override
  1689. */
  1690. int (*reget_buffer)(struct AVCodecContext *c, AVFrame *pic);
  1691. /**
  1692. * Number of bits which should be loaded into the rc buffer before decoding starts.
  1693. * - encoding: Set by user.
  1694. * - decoding: unused
  1695. */
  1696. int rc_initial_buffer_occupancy;
  1697. /**
  1698. *
  1699. * - encoding: Set by user.
  1700. * - decoding: unused
  1701. */
  1702. int inter_threshold;
  1703. /**
  1704. * CODEC_FLAG2_*
  1705. * - encoding: Set by user.
  1706. * - decoding: Set by user.
  1707. */
  1708. int flags2;
  1709. /**
  1710. * Simulates errors in the bitstream to test error concealment.
  1711. * - encoding: Set by user.
  1712. * - decoding: unused
  1713. */
  1714. int error_rate;
  1715. /**
  1716. * MP3 antialias algorithm, see FF_AA_* below.
  1717. * - encoding: unused
  1718. * - decoding: Set by user.
  1719. */
  1720. int antialias_algo;
  1721. #define FF_AA_AUTO 0
  1722. #define FF_AA_FASTINT 1 //not implemented yet
  1723. #define FF_AA_INT 2
  1724. #define FF_AA_FLOAT 3
  1725. /**
  1726. * quantizer noise shaping
  1727. * - encoding: Set by user.
  1728. * - decoding: unused
  1729. */
  1730. int quantizer_noise_shaping;
  1731. /**
  1732. * thread count
  1733. * is used to decide how many independent tasks should be passed to execute()
  1734. * - encoding: Set by user.
  1735. * - decoding: Set by user.
  1736. */
  1737. int thread_count;
  1738. /**
  1739. * The codec may call this to execute several independent things.
  1740. * It will return only after finishing all tasks.
  1741. * The user may replace this with some multithreaded implementation,
  1742. * the default implementation will execute the parts serially.
  1743. * @param count the number of things to execute
  1744. * - encoding: Set by libavcodec, user can override.
  1745. * - decoding: Set by libavcodec, user can override.
  1746. */
  1747. int (*execute)(struct AVCodecContext *c, int (*func)(struct AVCodecContext *c2, void *arg), void *arg2, int *ret, int count, int size);
  1748. /**
  1749. * thread opaque
  1750. * Can be used by execute() to store some per AVCodecContext stuff.
  1751. * - encoding: set by execute()
  1752. * - decoding: set by execute()
  1753. */
  1754. void *thread_opaque;
  1755. /**
  1756. * Motion estimation threshold below which no motion estimation is
  1757. * performed, but instead the user specified motion vectors are used.
  1758. *
  1759. * - encoding: Set by user.
  1760. * - decoding: unused
  1761. */
  1762. int me_threshold;
  1763. /**
  1764. * Macroblock threshold below which the user specified macroblock types will be used.
  1765. * - encoding: Set by user.
  1766. * - decoding: unused
  1767. */
  1768. int mb_threshold;
  1769. /**
  1770. * precision of the intra DC coefficient - 8
  1771. * - encoding: Set by user.
  1772. * - decoding: unused
  1773. */
  1774. int intra_dc_precision;
  1775. /**
  1776. * noise vs. sse weight for the nsse comparsion function
  1777. * - encoding: Set by user.
  1778. * - decoding: unused
  1779. */
  1780. int nsse_weight;
  1781. /**
  1782. * Number of macroblock rows at the top which are skipped.
  1783. * - encoding: unused
  1784. * - decoding: Set by user.
  1785. */
  1786. int skip_top;
  1787. /**
  1788. * Number of macroblock rows at the bottom which are skipped.
  1789. * - encoding: unused
  1790. * - decoding: Set by user.
  1791. */
  1792. int skip_bottom;
  1793. /**
  1794. * profile
  1795. * - encoding: Set by user.
  1796. * - decoding: Set by libavcodec.
  1797. */
  1798. int profile;
  1799. #define FF_PROFILE_UNKNOWN -99
  1800. #define FF_PROFILE_AAC_MAIN 0
  1801. #define FF_PROFILE_AAC_LOW 1
  1802. #define FF_PROFILE_AAC_SSR 2
  1803. #define FF_PROFILE_AAC_LTP 3
  1804. /**
  1805. * level
  1806. * - encoding: Set by user.
  1807. * - decoding: Set by libavcodec.
  1808. */
  1809. int level;
  1810. #define FF_LEVEL_UNKNOWN -99
  1811. /**
  1812. * low resolution decoding, 1-> 1/2 size, 2->1/4 size
  1813. * - encoding: unused
  1814. * - decoding: Set by user.
  1815. */
  1816. int lowres;
  1817. /**
  1818. * Bitstream width / height, may be different from width/height if lowres
  1819. * or other things are used.
  1820. * - encoding: unused
  1821. * - decoding: Set by user before init if known. Codec should override / dynamically change if needed.
  1822. */
  1823. int coded_width, coded_height;
  1824. /**
  1825. * frame skip threshold
  1826. * - encoding: Set by user.
  1827. * - decoding: unused
  1828. */
  1829. int frame_skip_threshold;
  1830. /**
  1831. * frame skip factor
  1832. * - encoding: Set by user.
  1833. * - decoding: unused
  1834. */
  1835. int frame_skip_factor;
  1836. /**
  1837. * frame skip exponent
  1838. * - encoding: Set by user.
  1839. * - decoding: unused
  1840. */
  1841. int frame_skip_exp;
  1842. /**
  1843. * frame skip comparison function
  1844. * - encoding: Set by user.
  1845. * - decoding: unused
  1846. */
  1847. int frame_skip_cmp;
  1848. /**
  1849. * Border processing masking, raises the quantizer for mbs on the borders
  1850. * of the picture.
  1851. * - encoding: Set by user.
  1852. * - decoding: unused
  1853. */
  1854. float border_masking;
  1855. /**
  1856. * minimum MB lagrange multipler
  1857. * - encoding: Set by user.
  1858. * - decoding: unused
  1859. */
  1860. int mb_lmin;
  1861. /**
  1862. * maximum MB lagrange multipler
  1863. * - encoding: Set by user.
  1864. * - decoding: unused
  1865. */
  1866. int mb_lmax;
  1867. /**
  1868. *
  1869. * - encoding: Set by user.
  1870. * - decoding: unused
  1871. */
  1872. int me_penalty_compensation;
  1873. /**
  1874. *
  1875. * - encoding: unused
  1876. * - decoding: Set by user.
  1877. */
  1878. enum AVDiscard skip_loop_filter;
  1879. /**
  1880. *
  1881. * - encoding: unused
  1882. * - decoding: Set by user.
  1883. */
  1884. enum AVDiscard skip_idct;
  1885. /**
  1886. *
  1887. * - encoding: unused
  1888. * - decoding: Set by user.
  1889. */
  1890. enum AVDiscard skip_frame;
  1891. /**
  1892. *
  1893. * - encoding: Set by user.
  1894. * - decoding: unused
  1895. */
  1896. int bidir_refine;
  1897. /**
  1898. *
  1899. * - encoding: Set by user.
  1900. * - decoding: unused
  1901. */
  1902. int brd_scale;
  1903. /**
  1904. * constant rate factor - quality-based VBR - values ~correspond to qps
  1905. * - encoding: Set by user.
  1906. * - decoding: unused
  1907. */
  1908. float crf;
  1909. /**
  1910. * constant quantization parameter rate control method
  1911. * - encoding: Set by user.
  1912. * - decoding: unused
  1913. */
  1914. int cqp;
  1915. /**
  1916. * minimum GOP size
  1917. * - encoding: Set by user.
  1918. * - decoding: unused
  1919. */
  1920. int keyint_min;
  1921. /**
  1922. * number of reference frames
  1923. * - encoding: Set by user.
  1924. * - decoding: Set by lavc.
  1925. */
  1926. int refs;
  1927. /**
  1928. * chroma qp offset from luma
  1929. * - encoding: Set by user.
  1930. * - decoding: unused
  1931. */
  1932. int chromaoffset;
  1933. /**
  1934. * Influences how often B-frames are used.
  1935. * - encoding: Set by user.
  1936. * - decoding: unused
  1937. */
  1938. int bframebias;
  1939. /**
  1940. * trellis RD quantization
  1941. * - encoding: Set by user.
  1942. * - decoding: unused
  1943. */
  1944. int trellis;
  1945. /**
  1946. * Reduce fluctuations in qp (before curve compression).
  1947. * - encoding: Set by user.
  1948. * - decoding: unused
  1949. */
  1950. float complexityblur;
  1951. /**
  1952. * in-loop deblocking filter alphac0 parameter
  1953. * alpha is in the range -6...6
  1954. * - encoding: Set by user.
  1955. * - decoding: unused
  1956. */
  1957. int deblockalpha;
  1958. /**
  1959. * in-loop deblocking filter beta parameter
  1960. * beta is in the range -6...6
  1961. * - encoding: Set by user.
  1962. * - decoding: unused
  1963. */
  1964. int deblockbeta;
  1965. /**
  1966. * macroblock subpartition sizes to consider - p8x8, p4x4, b8x8, i8x8, i4x4
  1967. * - encoding: Set by user.
  1968. * - decoding: unused
  1969. */
  1970. int partitions;
  1971. #define X264_PART_I4X4 0x001 /* Analyze i4x4 */
  1972. #define X264_PART_I8X8 0x002 /* Analyze i8x8 (requires 8x8 transform) */
  1973. #define X264_PART_P8X8 0x010 /* Analyze p16x8, p8x16 and p8x8 */
  1974. #define X264_PART_P4X4 0x020 /* Analyze p8x4, p4x8, p4x4 */
  1975. #define X264_PART_B8X8 0x100 /* Analyze b16x8, b8x16 and b8x8 */
  1976. /**
  1977. * direct MV prediction mode - 0 (none), 1 (spatial), 2 (temporal), 3 (auto)
  1978. * - encoding: Set by user.
  1979. * - decoding: unused
  1980. */
  1981. int directpred;
  1982. /**
  1983. * Audio cutoff bandwidth (0 means "automatic")
  1984. * - encoding: Set by user.
  1985. * - decoding: unused
  1986. */
  1987. int cutoff;
  1988. /**
  1989. * Multiplied by qscale for each frame and added to scene_change_score.
  1990. * - encoding: Set by user.
  1991. * - decoding: unused
  1992. */
  1993. int scenechange_factor;
  1994. /**
  1995. *
  1996. * Note: Value depends upon the compare function used for fullpel ME.
  1997. * - encoding: Set by user.
  1998. * - decoding: unused
  1999. */
  2000. int mv0_threshold;
  2001. /**
  2002. * Adjusts sensitivity of b_frame_strategy 1.
  2003. * - encoding: Set by user.
  2004. * - decoding: unused
  2005. */
  2006. int b_sensitivity;
  2007. /**
  2008. * - encoding: Set by user.
  2009. * - decoding: unused
  2010. */
  2011. int compression_level;
  2012. #define FF_COMPRESSION_DEFAULT -1
  2013. /**
  2014. * Sets whether to use LPC mode - used by FLAC encoder.
  2015. * - encoding: Set by user.
  2016. * - decoding: unused
  2017. */
  2018. int use_lpc;
  2019. /**
  2020. * LPC coefficient precision - used by FLAC encoder
  2021. * - encoding: Set by user.
  2022. * - decoding: unused
  2023. */
  2024. int lpc_coeff_precision;
  2025. /**
  2026. * - encoding: Set by user.
  2027. * - decoding: unused
  2028. */
  2029. int min_prediction_order;
  2030. /**
  2031. * - encoding: Set by user.
  2032. * - decoding: unused
  2033. */
  2034. int max_prediction_order;
  2035. /**
  2036. * search method for selecting prediction order
  2037. * - encoding: Set by user.
  2038. * - decoding: unused
  2039. */
  2040. int prediction_order_method;
  2041. /**
  2042. * - encoding: Set by user.
  2043. * - decoding: unused
  2044. */
  2045. int min_partition_order;
  2046. /**
  2047. * - encoding: Set by user.
  2048. * - decoding: unused
  2049. */
  2050. int max_partition_order;
  2051. /**
  2052. * GOP timecode frame start number, in non drop frame format
  2053. * - encoding: Set by user.
  2054. * - decoding: unused
  2055. */
  2056. int64_t timecode_frame_start;
  2057. #if LIBAVCODEC_VERSION_MAJOR < 53
  2058. /**
  2059. * Decoder should decode to this many channels if it can (0 for default)
  2060. * - encoding: unused
  2061. * - decoding: Set by user.
  2062. * @deprecated Deprecated in favor of request_channel_layout.
  2063. */
  2064. int request_channels;
  2065. #endif
  2066. /**
  2067. * Percentage of dynamic range compression to be applied by the decoder.
  2068. * The default value is 1.0, corresponding to full compression.
  2069. * - encoding: unused
  2070. * - decoding: Set by user.
  2071. */
  2072. float drc_scale;
  2073. /**
  2074. * opaque 64bit number (generally a PTS) that will be reordered and
  2075. * output in AVFrame.reordered_opaque
  2076. * - encoding: unused
  2077. * - decoding: Set by user.
  2078. */
  2079. int64_t reordered_opaque;
  2080. /**
  2081. * Bits per sample/pixel of internal libavcodec pixel/sample format.
  2082. * This field is applicable only when sample_fmt is SAMPLE_FMT_S32.
  2083. * - encoding: set by user.
  2084. * - decoding: set by libavcodec.
  2085. */
  2086. int bits_per_raw_sample;
  2087. /**
  2088. * Audio channel layout.
  2089. * - encoding: set by user.
  2090. * - decoding: set by libavcodec.
  2091. */
  2092. int64_t channel_layout;
  2093. /**
  2094. * Request decoder to use this channel layout if it can (0 for default)
  2095. * - encoding: unused
  2096. * - decoding: Set by user.
  2097. */
  2098. int64_t request_channel_layout;
  2099. /**
  2100. * Ratecontrol attempt to use, at maximum, <value> of what can be used without an underflow.
  2101. * - encoding: Set by user.
  2102. * - decoding: unused.
  2103. */
  2104. float rc_max_available_vbv_use;
  2105. /**
  2106. * Ratecontrol attempt to use, at least, <value> times the amount needed to prevent a vbv overflow.
  2107. * - encoding: Set by user.
  2108. * - decoding: unused.
  2109. */
  2110. float rc_min_vbv_overflow_use;
  2111. /**
  2112. * Hardware accelerator in use
  2113. * - encoding: unused.
  2114. * - decoding: Set by libavcodec
  2115. */
  2116. struct AVHWAccel *hwaccel;
  2117. /**
  2118. * For some codecs, the time base is closer to the field rate than the frame rate.
  2119. * Most notably, H.264 and MPEG-2 specify time_base as half of frame duration
  2120. * if no telecine is used ...
  2121. *
  2122. * Set to time_base ticks per frame. Default 1, e.g., H.264/MPEG-2 set it to 2.
  2123. */
  2124. int ticks_per_frame;
  2125. /**
  2126. * Hardware accelerator context.
  2127. * For some hardware accelerators, a global context needs to be
  2128. * provided by the user. In that case, this holds display-dependent
  2129. * data FFmpeg cannot instantiate itself. Please refer to the
  2130. * FFmpeg HW accelerator documentation to know how to fill this
  2131. * is. e.g. for VA API, this is a struct vaapi_context.
  2132. * - encoding: unused
  2133. * - decoding: Set by user
  2134. */
  2135. void *hwaccel_context;
  2136. } AVCodecContext;
  2137. /**
  2138. * AVCodec.
  2139. */
  2140. typedef struct AVCodec {
  2141. /**
  2142. * Name of the codec implementation.
  2143. * The name is globally unique among encoders and among decoders (but an
  2144. * encoder and a decoder can share the same name).
  2145. * This is the primary way to find a codec from the user perspective.
  2146. */
  2147. const char *name;
  2148. enum CodecType type;
  2149. enum CodecID id;
  2150. int priv_data_size;
  2151. int (*init)(AVCodecContext *);
  2152. int (*encode)(AVCodecContext *, uint8_t *buf, int buf_size, void *data);
  2153. int (*close)(AVCodecContext *);
  2154. int (*decode)(AVCodecContext *, void *outdata, int *outdata_size,
  2155. const uint8_t *buf, int buf_size);
  2156. /**
  2157. * Codec capabilities.
  2158. * see CODEC_CAP_*
  2159. */
  2160. int capabilities;
  2161. struct AVCodec *next;
  2162. /**
  2163. * Flush buffers.
  2164. * Will be called when seeking
  2165. */
  2166. void (*flush)(AVCodecContext *);
  2167. const AVRational *supported_framerates; ///< array of supported framerates, or NULL if any, array is terminated by {0,0}
  2168. const enum PixelFormat *pix_fmts; ///< array of supported pixel formats, or NULL if unknown, array is terminated by -1
  2169. /**
  2170. * Descriptive name for the codec, meant to be more human readable than \p name.
  2171. * You \e should use the NULL_IF_CONFIG_SMALL() macro to define it.
  2172. */
  2173. const char *long_name;
  2174. const int *supported_samplerates; ///< array of supported audio samplerates, or NULL if unknown, array is terminated by 0
  2175. const enum SampleFormat *sample_fmts; ///< array of supported sample formats, or NULL if unknown, array is terminated by -1
  2176. const int64_t *channel_layouts; ///< array of support channel layouts, or NULL if unknown. array is terminated by 0
  2177. } AVCodec;
  2178. /**
  2179. * AVHWAccel.
  2180. */
  2181. typedef struct AVHWAccel {
  2182. /**
  2183. * Name of the hardware accelerated codec.
  2184. * The name is globally unique among encoders and among decoders (but an
  2185. * encoder and a decoder can share the same name).
  2186. */
  2187. const char *name;
  2188. /**
  2189. * Type of codec implemented by the hardware accelerator.
  2190. *
  2191. * See CODEC_TYPE_xxx
  2192. */
  2193. enum CodecType type;
  2194. /**
  2195. * Codec implemented by the hardware accelerator.
  2196. *
  2197. * See CODEC_ID_xxx
  2198. */
  2199. enum CodecID id;
  2200. /**
  2201. * Supported pixel format.
  2202. *
  2203. * Only hardware accelerated formats are supported here.
  2204. */
  2205. enum PixelFormat pix_fmt;
  2206. /**
  2207. * Hardware accelerated codec capabilities.
  2208. * see FF_HWACCEL_CODEC_CAP_*
  2209. */
  2210. int capabilities;
  2211. struct AVHWAccel *next;
  2212. /**
  2213. * Called at the beginning of each frame or field picture.
  2214. *
  2215. * Meaningful frame information (codec specific) is guaranteed to
  2216. * be parsed at this point. This function is mandatory.
  2217. *
  2218. * Note that \p buf can be NULL along with \p buf_size set to 0.
  2219. * Otherwise, this means the whole frame is available at this point.
  2220. *
  2221. * @param avctx the codec context
  2222. * @param buf the frame data buffer base
  2223. * @param buf_size the size of the frame in bytes
  2224. * @return zero if successful, a negative value otherwise
  2225. */
  2226. int (*start_frame)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
  2227. /**
  2228. * Callback for each slice.
  2229. *
  2230. * Meaningful slice information (codec specific) is guaranteed to
  2231. * be parsed at this point. This function is mandatory.
  2232. *
  2233. * @param avctx the codec context
  2234. * @param buf the slice data buffer base
  2235. * @param buf_size the size of the slice in bytes
  2236. * @return zero if successful, a negative value otherwise
  2237. */
  2238. int (*decode_slice)(AVCodecContext *avctx, const uint8_t *buf, uint32_t buf_size);
  2239. /**
  2240. * Called at the end of each frame or field picture.
  2241. *
  2242. * The whole picture is parsed at this point and can now be sent
  2243. * to the hardware accelerator. This function is mandatory.
  2244. *
  2245. * @param avctx the codec context
  2246. * @return zero if successful, a negative value otherwise
  2247. */
  2248. int (*end_frame)(AVCodecContext *avctx);
  2249. /**
  2250. * Size of HW accelerator private data.
  2251. *
  2252. * Private data is allocated with av_mallocz() before
  2253. * AVCodecContext.get_buffer() and deallocated after
  2254. * AVCodecContext.release_buffer().
  2255. */
  2256. int priv_data_size;
  2257. } AVHWAccel;
  2258. /**
  2259. * four components are given, that's all.
  2260. * the last component is alpha
  2261. */
  2262. typedef struct AVPicture {
  2263. uint8_t *data[4];
  2264. int linesize[4]; ///< number of bytes per line
  2265. } AVPicture;
  2266. #if LIBAVCODEC_VERSION_MAJOR < 53
  2267. /**
  2268. * AVPaletteControl
  2269. * This structure defines a method for communicating palette changes
  2270. * between and demuxer and a decoder.
  2271. *
  2272. * @deprecated Use AVPacket to send palette changes instead.
  2273. * This is totally broken.
  2274. */
  2275. #define AVPALETTE_SIZE 1024
  2276. #define AVPALETTE_COUNT 256
  2277. typedef struct AVPaletteControl {
  2278. /* Demuxer sets this to 1 to indicate the palette has changed;
  2279. * decoder resets to 0. */
  2280. int palette_changed;
  2281. /* 4-byte ARGB palette entries, stored in native byte order; note that
  2282. * the individual palette components should be on a 8-bit scale; if
  2283. * the palette data comes from an IBM VGA native format, the component
  2284. * data is probably 6 bits in size and needs to be scaled. */
  2285. unsigned int palette[AVPALETTE_COUNT];
  2286. } AVPaletteControl attribute_deprecated;
  2287. #endif
  2288. typedef struct AVPacket {
  2289. /**
  2290. * Presentation timestamp in time_base units; the time at which the
  2291. * decompressed packet will be presented to the user.
  2292. * Can be AV_NOPTS_VALUE if it is not stored in the file.
  2293. * pts MUST be larger or equal to dts as presentation cannot happen before
  2294. * decompression, unless one wants to view hex dumps. Some formats misuse
  2295. * the terms dts and pts/cts to mean something different. Such timestamps
  2296. * must be converted to true pts/dts before they are stored in AVPacket.
  2297. */
  2298. int64_t pts;
  2299. /**
  2300. * Decompression timestamp in time_base units; the time at which the
  2301. * packet is decompressed.
  2302. * Can be AV_NOPTS_VALUE if it is not stored in the file.
  2303. */
  2304. int64_t dts;
  2305. uint8_t *data;
  2306. int size;
  2307. int stream_index;
  2308. int flags;
  2309. /**
  2310. * Duration of this packet in time_base units, 0 if unknown.
  2311. * Equals next_pts - this_pts in presentation order.
  2312. */
  2313. int duration;
  2314. void (*destruct)(struct AVPacket *);
  2315. void *priv;
  2316. int64_t pos; ///< byte position in stream, -1 if unknown
  2317. /**
  2318. * Time difference in stream time base units from the pts of this
  2319. * packet to the point at which the output from the decoder has converged
  2320. * independent from the availability of previous frames. That is, the
  2321. * frames are virtually identical no matter if decoding started from
  2322. * the very first frame or from this keyframe.
  2323. * Is AV_NOPTS_VALUE if unknown.
  2324. * This field is not the display duration of the current packet.
  2325. *
  2326. * The purpose of this field is to allow seeking in streams that have no
  2327. * keyframes in the conventional sense. It corresponds to the
  2328. * recovery point SEI in H.264 and match_time_delta in NUT. It is also
  2329. * essential for some types of subtitle streams to ensure that all
  2330. * subtitles are correctly displayed after seeking.
  2331. */
  2332. int64_t convergence_duration;
  2333. } AVPacket;
  2334. #define PKT_FLAG_KEY 0x0001
  2335. enum AVSubtitleType {
  2336. SUBTITLE_NONE,
  2337. SUBTITLE_BITMAP, ///< A bitmap, pict will be set
  2338. /**
  2339. * Plain text, the text field must be set by the decoder and is
  2340. * authoritative. ass and pict fields may contain approximations.
  2341. */
  2342. SUBTITLE_TEXT,
  2343. /**
  2344. * Formatted text, the ass field must be set by the decoder and is
  2345. * authoritative. pict and text fields may contain approximations.
  2346. */
  2347. SUBTITLE_ASS,
  2348. };
  2349. typedef struct AVSubtitleRect {
  2350. int x; ///< top left corner of pict, undefined when pict is not set
  2351. int y; ///< top left corner of pict, undefined when pict is not set
  2352. int w; ///< width of pict, undefined when pict is not set
  2353. int h; ///< height of pict, undefined when pict is not set
  2354. int nb_colors; ///< number of colors in pict, undefined when pict is not set
  2355. /**
  2356. * data+linesize for the bitmap of this subtitle.
  2357. * can be set for text/ass as well once they where rendered
  2358. */
  2359. AVPicture pict;
  2360. enum AVSubtitleType type;
  2361. char *text; ///< 0 terminated plain UTF-8 text
  2362. /**
  2363. * 0 terminated ASS/SSA compatible event line.
  2364. * The pressentation of this is unaffected by the other values in this
  2365. * struct.
  2366. */
  2367. char *ass;
  2368. } AVSubtitleRect;
  2369. typedef struct AVSubtitle {
  2370. uint16_t format; /* 0 = graphics */
  2371. uint32_t start_display_time; /* relative to packet pts, in ms */
  2372. uint32_t end_display_time; /* relative to packet pts, in ms */
  2373. unsigned num_rects;
  2374. AVSubtitleRect **rects;
  2375. } AVSubtitle;
  2376. /* resample.c */
  2377. struct ReSampleContext;
  2378. struct AVResampleContext;
  2379. typedef struct ReSampleContext ReSampleContext;
  2380. #if LIBAVCODEC_VERSION_MAJOR < 53
  2381. /**
  2382. * @deprecated Use av_audio_resample_init() instead.
  2383. */
  2384. attribute_deprecated ReSampleContext *audio_resample_init(int output_channels, int input_channels,
  2385. int output_rate, int input_rate);
  2386. #endif
  2387. /**
  2388. * Initializes audio resampling context
  2389. *
  2390. * @param output_channels number of output channels
  2391. * @param input_channels number of input channels
  2392. * @param output_rate output sample rate
  2393. * @param input_rate input sample rate
  2394. * @param sample_fmt_out requested output sample format
  2395. * @param sample_fmt_in input sample format
  2396. * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
  2397. * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
  2398. * @param linear If 1 then the used FIR filter will be linearly interpolated
  2399. between the 2 closest, if 0 the closest will be used
  2400. * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
  2401. * @return allocated ReSampleContext, NULL if error occured
  2402. */
  2403. ReSampleContext *av_audio_resample_init(int output_channels, int input_channels,
  2404. int output_rate, int input_rate,
  2405. enum SampleFormat sample_fmt_out,
  2406. enum SampleFormat sample_fmt_in,
  2407. int filter_length, int log2_phase_count,
  2408. int linear, double cutoff);
  2409. int audio_resample(ReSampleContext *s, short *output, short *input, int nb_samples);
  2410. void audio_resample_close(ReSampleContext *s);
  2411. /**
  2412. * Initializes an audio resampler.
  2413. * Note, if either rate is not an integer then simply scale both rates up so they are.
  2414. * @param filter_length length of each FIR filter in the filterbank relative to the cutoff freq
  2415. * @param log2_phase_count log2 of the number of entries in the polyphase filterbank
  2416. * @param linear If 1 then the used FIR filter will be linearly interpolated
  2417. between the 2 closest, if 0 the closest will be used
  2418. * @param cutoff cutoff frequency, 1.0 corresponds to half the output sampling rate
  2419. */
  2420. struct AVResampleContext *av_resample_init(int out_rate, int in_rate, int filter_length, int log2_phase_count, int linear, double cutoff);
  2421. /**
  2422. * resamples.
  2423. * @param src an array of unconsumed samples
  2424. * @param consumed the number of samples of src which have been consumed are returned here
  2425. * @param src_size the number of unconsumed samples available
  2426. * @param dst_size the amount of space in samples available in dst
  2427. * @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.
  2428. * @return the number of samples written in dst or -1 if an error occurred
  2429. */
  2430. int av_resample(struct AVResampleContext *c, short *dst, short *src, int *consumed, int src_size, int dst_size, int update_ctx);
  2431. /**
  2432. * Compensates samplerate/timestamp drift. The compensation is done by changing
  2433. * the resampler parameters, so no audible clicks or similar distortions occur
  2434. * @param compensation_distance distance in output samples over which the compensation should be performed
  2435. * @param sample_delta number of output samples which should be output less
  2436. *
  2437. * example: av_resample_compensate(c, 10, 500)
  2438. * here instead of 510 samples only 500 samples would be output
  2439. *
  2440. * note, due to rounding the actual compensation might be slightly different,
  2441. * especially if the compensation_distance is large and the in_rate used during init is small
  2442. */
  2443. void av_resample_compensate(struct AVResampleContext *c, int sample_delta, int compensation_distance);
  2444. void av_resample_close(struct AVResampleContext *c);
  2445. /**
  2446. * Allocate memory for a picture. Call avpicture_free to free it.
  2447. *
  2448. * @param picture the picture to be filled in
  2449. * @param pix_fmt the format of the picture
  2450. * @param width the width of the picture
  2451. * @param height the height of the picture
  2452. * @return zero if successful, a negative value if not
  2453. */
  2454. int avpicture_alloc(AVPicture *picture, enum PixelFormat pix_fmt, int width, int height);
  2455. /**
  2456. * Free a picture previously allocated by avpicture_alloc().
  2457. *
  2458. * @param picture the AVPicture to be freed
  2459. */
  2460. void avpicture_free(AVPicture *picture);
  2461. /**
  2462. * Fill in the AVPicture fields.
  2463. * The fields of the given AVPicture are filled in by using the 'ptr' address
  2464. * which points to the image data buffer. Depending on the specified picture
  2465. * format, one or multiple image data pointers and line sizes will be set.
  2466. * If a planar format is specified, several pointers will be set pointing to
  2467. * the different picture planes and the line sizes of the different planes
  2468. * will be stored in the lines_sizes array.
  2469. * Call with ptr == NULL to get the required size for the ptr buffer.
  2470. *
  2471. * @param picture AVPicture whose fields are to be filled in
  2472. * @param ptr Buffer which will contain or contains the actual image data
  2473. * @param pix_fmt The format in which the picture data is stored.
  2474. * @param width the width of the image in pixels
  2475. * @param height the height of the image in pixels
  2476. * @return size of the image data in bytes
  2477. */
  2478. int avpicture_fill(AVPicture *picture, uint8_t *ptr,
  2479. int pix_fmt, int width, int height);
  2480. int avpicture_layout(const AVPicture* src, enum PixelFormat pix_fmt, int width, int height,
  2481. unsigned char *dest, int dest_size);
  2482. /**
  2483. * Calculate the size in bytes that a picture of the given width and height
  2484. * would occupy if stored in the given picture format.
  2485. * Note that this returns the size of a compact representation as generated
  2486. * by avpicture_layout, which can be smaller than the size required for e.g.
  2487. * avpicture_fill.
  2488. *
  2489. * @param pix_fmt the given picture format
  2490. * @param width the width of the image
  2491. * @param height the height of the image
  2492. * @return Image data size in bytes or -1 on error (e.g. too large dimensions).
  2493. */
  2494. int avpicture_get_size(enum PixelFormat pix_fmt, int width, int height);
  2495. void avcodec_get_chroma_sub_sample(enum PixelFormat pix_fmt, int *h_shift, int *v_shift);
  2496. const char *avcodec_get_pix_fmt_name(enum PixelFormat pix_fmt);
  2497. void avcodec_set_dimensions(AVCodecContext *s, int width, int height);
  2498. /**
  2499. * Returns the pixel format corresponding to the name \p name.
  2500. *
  2501. * If there is no pixel format with name \p name, then looks for a
  2502. * pixel format with the name corresponding to the native endian
  2503. * format of \p name.
  2504. * For example in a little-endian system, first looks for "gray16",
  2505. * then for "gray16le".
  2506. *
  2507. * Finally if no pixel format has been found, returns \c PIX_FMT_NONE.
  2508. */
  2509. enum PixelFormat avcodec_get_pix_fmt(const char* name);
  2510. unsigned int avcodec_pix_fmt_to_codec_tag(enum PixelFormat p);
  2511. #define FF_LOSS_RESOLUTION 0x0001 /**< loss due to resolution change */
  2512. #define FF_LOSS_DEPTH 0x0002 /**< loss due to color depth change */
  2513. #define FF_LOSS_COLORSPACE 0x0004 /**< loss due to color space conversion */
  2514. #define FF_LOSS_ALPHA 0x0008 /**< loss of alpha bits */
  2515. #define FF_LOSS_COLORQUANT 0x0010 /**< loss due to color quantization */
  2516. #define FF_LOSS_CHROMA 0x0020 /**< loss of chroma (e.g. RGB to gray conversion) */
  2517. /**
  2518. * Computes what kind of losses will occur when converting from one specific
  2519. * pixel format to another.
  2520. * When converting from one pixel format to another, information loss may occur.
  2521. * For example, when converting from RGB24 to GRAY, the color information will
  2522. * be lost. Similarly, other losses occur when converting from some formats to
  2523. * other formats. These losses can involve loss of chroma, but also loss of
  2524. * resolution, loss of color depth, loss due to the color space conversion, loss
  2525. * of the alpha bits or loss due to color quantization.
  2526. * avcodec_get_fix_fmt_loss() informs you about the various types of losses
  2527. * which will occur when converting from one pixel format to another.
  2528. *
  2529. * @param[in] dst_pix_fmt destination pixel format
  2530. * @param[in] src_pix_fmt source pixel format
  2531. * @param[in] has_alpha Whether the source pixel format alpha channel is used.
  2532. * @return Combination of flags informing you what kind of losses will occur.
  2533. */
  2534. int avcodec_get_pix_fmt_loss(enum PixelFormat dst_pix_fmt, enum PixelFormat src_pix_fmt,
  2535. int has_alpha);
  2536. /**
  2537. * Finds the best pixel format to convert to given a certain source pixel
  2538. * format. When converting from one pixel format to another, information loss
  2539. * may occur. For example, when converting from RGB24 to GRAY, the color
  2540. * information will be lost. Similarly, other losses occur when converting from
  2541. * some formats to other formats. avcodec_find_best_pix_fmt() searches which of
  2542. * the given pixel formats should be used to suffer the least amount of loss.
  2543. * The pixel formats from which it chooses one, are determined by the
  2544. * \p pix_fmt_mask parameter.
  2545. *
  2546. * @code
  2547. * src_pix_fmt = PIX_FMT_YUV420P;
  2548. * pix_fmt_mask = (1 << PIX_FMT_YUV422P) || (1 << PIX_FMT_RGB24);
  2549. * dst_pix_fmt = avcodec_find_best_pix_fmt(pix_fmt_mask, src_pix_fmt, alpha, &loss);
  2550. * @endcode
  2551. *
  2552. * @param[in] pix_fmt_mask bitmask determining which pixel format to choose from
  2553. * @param[in] src_pix_fmt source pixel format
  2554. * @param[in] has_alpha Whether the source pixel format alpha channel is used.
  2555. * @param[out] loss_ptr Combination of flags informing you what kind of losses will occur.
  2556. * @return The best pixel format to convert to or -1 if none was found.
  2557. */
  2558. enum PixelFormat avcodec_find_best_pix_fmt(int64_t pix_fmt_mask, enum PixelFormat src_pix_fmt,
  2559. int has_alpha, int *loss_ptr);
  2560. /**
  2561. * Print in buf the string corresponding to the pixel format with
  2562. * number pix_fmt, or an header if pix_fmt is negative.
  2563. *
  2564. * @param[in] buf the buffer where to write the string
  2565. * @param[in] buf_size the size of buf
  2566. * @param[in] pix_fmt the number of the pixel format to print the corresponding info string, or
  2567. * a negative value to print the corresponding header.
  2568. * Meaningful values for obtaining a pixel format info vary from 0 to PIX_FMT_NB -1.
  2569. */
  2570. void avcodec_pix_fmt_string (char *buf, int buf_size, enum PixelFormat pix_fmt);
  2571. #define FF_ALPHA_TRANSP 0x0001 /* image has some totally transparent pixels */
  2572. #define FF_ALPHA_SEMI_TRANSP 0x0002 /* image has some transparent pixels */
  2573. /**
  2574. * Tell if an image really has transparent alpha values.
  2575. * @return ored mask of FF_ALPHA_xxx constants
  2576. */
  2577. int img_get_alpha_info(const AVPicture *src,
  2578. enum PixelFormat pix_fmt, int width, int height);
  2579. /* deinterlace a picture */
  2580. /* deinterlace - if not supported return -1 */
  2581. int avpicture_deinterlace(AVPicture *dst, const AVPicture *src,
  2582. enum PixelFormat pix_fmt, int width, int height);
  2583. /* external high level API */
  2584. /**
  2585. * If c is NULL, returns the first registered codec,
  2586. * if c is non-NULL, returns the next registered codec after c,
  2587. * or NULL if c is the last one.
  2588. */
  2589. AVCodec *av_codec_next(AVCodec *c);
  2590. /**
  2591. * Returns the LIBAVCODEC_VERSION_INT constant.
  2592. */
  2593. unsigned avcodec_version(void);
  2594. /**
  2595. * Initializes libavcodec.
  2596. *
  2597. * @warning This function \e must be called before any other libavcodec
  2598. * function.
  2599. */
  2600. void avcodec_init(void);
  2601. #if LIBAVCODEC_VERSION_MAJOR < 53
  2602. /**
  2603. * @deprecated Deprecated in favor of avcodec_register().
  2604. */
  2605. attribute_deprecated void register_avcodec(AVCodec *codec);
  2606. #endif
  2607. /**
  2608. * Register the codec \p codec and initialize libavcodec.
  2609. *
  2610. * @see avcodec_init()
  2611. */
  2612. void avcodec_register(AVCodec *codec);
  2613. /**
  2614. * Finds a registered encoder with a matching codec ID.
  2615. *
  2616. * @param id CodecID of the requested encoder
  2617. * @return An encoder if one was found, NULL otherwise.
  2618. */
  2619. AVCodec *avcodec_find_encoder(enum CodecID id);
  2620. /**
  2621. * Finds a registered encoder with the specified name.
  2622. *
  2623. * @param name name of the requested encoder
  2624. * @return An encoder if one was found, NULL otherwise.
  2625. */
  2626. AVCodec *avcodec_find_encoder_by_name(const char *name);
  2627. /**
  2628. * Finds a registered decoder with a matching codec ID.
  2629. *
  2630. * @param id CodecID of the requested decoder
  2631. * @return A decoder if one was found, NULL otherwise.
  2632. */
  2633. AVCodec *avcodec_find_decoder(enum CodecID id);
  2634. /**
  2635. * Finds a registered decoder with the specified name.
  2636. *
  2637. * @param name name of the requested decoder
  2638. * @return A decoder if one was found, NULL otherwise.
  2639. */
  2640. AVCodec *avcodec_find_decoder_by_name(const char *name);
  2641. void avcodec_string(char *buf, int buf_size, AVCodecContext *enc, int encode);
  2642. /**
  2643. * Sets the fields of the given AVCodecContext to default values.
  2644. *
  2645. * @param s The AVCodecContext of which the fields should be set to default values.
  2646. */
  2647. void avcodec_get_context_defaults(AVCodecContext *s);
  2648. /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
  2649. * we WILL change its arguments and name a few times! */
  2650. void avcodec_get_context_defaults2(AVCodecContext *s, enum CodecType);
  2651. /**
  2652. * Allocates an AVCodecContext and sets its fields to default values. The
  2653. * resulting struct can be deallocated by simply calling av_free().
  2654. *
  2655. * @return An AVCodecContext filled with default values or NULL on failure.
  2656. * @see avcodec_get_context_defaults
  2657. */
  2658. AVCodecContext *avcodec_alloc_context(void);
  2659. /** THIS FUNCTION IS NOT YET PART OF THE PUBLIC API!
  2660. * we WILL change its arguments and name a few times! */
  2661. AVCodecContext *avcodec_alloc_context2(enum CodecType);
  2662. /**
  2663. * Sets the fields of the given AVFrame to default values.
  2664. *
  2665. * @param pic The AVFrame of which the fields should be set to default values.
  2666. */
  2667. void avcodec_get_frame_defaults(AVFrame *pic);
  2668. /**
  2669. * Allocates an AVFrame and sets its fields to default values. The resulting
  2670. * struct can be deallocated by simply calling av_free().
  2671. *
  2672. * @return An AVFrame filled with default values or NULL on failure.
  2673. * @see avcodec_get_frame_defaults
  2674. */
  2675. AVFrame *avcodec_alloc_frame(void);
  2676. int avcodec_default_get_buffer(AVCodecContext *s, AVFrame *pic);
  2677. void avcodec_default_release_buffer(AVCodecContext *s, AVFrame *pic);
  2678. int avcodec_default_reget_buffer(AVCodecContext *s, AVFrame *pic);
  2679. void avcodec_align_dimensions(AVCodecContext *s, int *width, int *height);
  2680. /**
  2681. * Checks if the given dimension of a picture is valid, meaning that all
  2682. * bytes of the picture can be addressed with a signed int.
  2683. *
  2684. * @param[in] w Width of the picture.
  2685. * @param[in] h Height of the picture.
  2686. * @return Zero if valid, a negative value if invalid.
  2687. */
  2688. int avcodec_check_dimensions(void *av_log_ctx, unsigned int w, unsigned int h);
  2689. enum PixelFormat avcodec_default_get_format(struct AVCodecContext *s, const enum PixelFormat * fmt);
  2690. int avcodec_thread_init(AVCodecContext *s, int thread_count);
  2691. void avcodec_thread_free(AVCodecContext *s);
  2692. int avcodec_thread_execute(AVCodecContext *s, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
  2693. int avcodec_default_execute(AVCodecContext *c, int (*func)(AVCodecContext *c2, void *arg2),void *arg, int *ret, int count, int size);
  2694. //FIXME func typedef
  2695. /**
  2696. * Initializes the AVCodecContext to use the given AVCodec. Prior to using this
  2697. * function the context has to be allocated.
  2698. *
  2699. * The functions avcodec_find_decoder_by_name(), avcodec_find_encoder_by_name(),
  2700. * avcodec_find_decoder() and avcodec_find_encoder() provide an easy way for
  2701. * retrieving a codec.
  2702. *
  2703. * @warning This function is not thread safe!
  2704. *
  2705. * @code
  2706. * avcodec_register_all();
  2707. * codec = avcodec_find_decoder(CODEC_ID_H264);
  2708. * if (!codec)
  2709. * exit(1);
  2710. *
  2711. * context = avcodec_alloc_context();
  2712. *
  2713. * if (avcodec_open(context, codec) < 0)
  2714. * exit(1);
  2715. * @endcode
  2716. *
  2717. * @param avctx The context which will be set up to use the given codec.
  2718. * @param codec The codec to use within the context.
  2719. * @return zero on success, a negative value on error
  2720. * @see avcodec_alloc_context, avcodec_find_decoder, avcodec_find_encoder
  2721. */
  2722. int avcodec_open(AVCodecContext *avctx, AVCodec *codec);
  2723. /**
  2724. * Decodes an audio frame from \p buf into \p samples.
  2725. * The avcodec_decode_audio2() function decodes an audio frame from the input
  2726. * buffer \p buf of size \p buf_size. To decode it, it makes use of the
  2727. * audio codec which was coupled with \p avctx using avcodec_open(). The
  2728. * resulting decoded frame is stored in output buffer \p samples. If no frame
  2729. * could be decompressed, \p frame_size_ptr is zero. Otherwise, it is the
  2730. * decompressed frame size in \e bytes.
  2731. *
  2732. * @warning You \e must set \p frame_size_ptr to the allocated size of the
  2733. * output buffer before calling avcodec_decode_audio2().
  2734. *
  2735. * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
  2736. * the actual read bytes because some optimized bitstream readers read 32 or 64
  2737. * bits at once and could read over the end.
  2738. *
  2739. * @warning The end of the input buffer \p buf should be set to 0 to ensure that
  2740. * no overreading happens for damaged MPEG streams.
  2741. *
  2742. * @note You might have to align the input buffer \p buf and output buffer \p
  2743. * samples. The alignment requirements depend on the CPU: On some CPUs it isn't
  2744. * necessary at all, on others it won't work at all if not aligned and on others
  2745. * it will work but it will have an impact on performance. In practice, the
  2746. * bitstream should have 4 byte alignment at minimum and all sample data should
  2747. * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
  2748. * the linesize is not a multiple of 16 then there's no sense in aligning the
  2749. * start of the buffer to 16.
  2750. *
  2751. * @param avctx the codec context
  2752. * @param[out] samples the output buffer
  2753. * @param[in,out] frame_size_ptr the output buffer size in bytes
  2754. * @param[in] buf the input buffer
  2755. * @param[in] buf_size the input buffer size in bytes
  2756. * @return On error a negative value is returned, otherwise the number of bytes
  2757. * used or zero if no frame could be decompressed.
  2758. */
  2759. int avcodec_decode_audio2(AVCodecContext *avctx, int16_t *samples,
  2760. int *frame_size_ptr,
  2761. const uint8_t *buf, int buf_size);
  2762. /**
  2763. * Decodes a video frame from \p buf into \p picture.
  2764. * The avcodec_decode_video() function decodes a video frame from the input
  2765. * buffer \p buf of size \p buf_size. To decode it, it makes use of the
  2766. * video codec which was coupled with \p avctx using avcodec_open(). The
  2767. * resulting decoded frame is stored in \p picture.
  2768. *
  2769. * @warning The input buffer must be \c FF_INPUT_BUFFER_PADDING_SIZE larger than
  2770. * the actual read bytes because some optimized bitstream readers read 32 or 64
  2771. * bits at once and could read over the end.
  2772. *
  2773. * @warning The end of the input buffer \p buf should be set to 0 to ensure that
  2774. * no overreading happens for damaged MPEG streams.
  2775. *
  2776. * @note You might have to align the input buffer \p buf and output buffer \p
  2777. * samples. The alignment requirements depend on the CPU: on some CPUs it isn't
  2778. * necessary at all, on others it won't work at all if not aligned and on others
  2779. * it will work but it will have an impact on performance. In practice, the
  2780. * bitstream should have 4 byte alignment at minimum and all sample data should
  2781. * be 16 byte aligned unless the CPU doesn't need it (AltiVec and SSE do). If
  2782. * the linesize is not a multiple of 16 then there's no sense in aligning the
  2783. * start of the buffer to 16.
  2784. *
  2785. * @note Some codecs have a delay between input and output, these need to be
  2786. * feeded with buf=NULL, buf_size=0 at the end to return the remaining frames.
  2787. *
  2788. * @param avctx the codec context
  2789. * @param[out] picture The AVFrame in which the decoded video frame will be stored.
  2790. * @param[in] buf the input buffer
  2791. * @param[in] buf_size the size of the input buffer in bytes
  2792. * @param[in,out] got_picture_ptr Zero if no frame could be decompressed, otherwise, it is nonzero.
  2793. * @return On error a negative value is returned, otherwise the number of bytes
  2794. * used or zero if no frame could be decompressed.
  2795. */
  2796. int avcodec_decode_video(AVCodecContext *avctx, AVFrame *picture,
  2797. int *got_picture_ptr,
  2798. const uint8_t *buf, int buf_size);
  2799. /* Decode a subtitle message. Return -1 if error, otherwise return the
  2800. * number of bytes used. If no subtitle could be decompressed,
  2801. * got_sub_ptr is zero. Otherwise, the subtitle is stored in *sub. */
  2802. int avcodec_decode_subtitle(AVCodecContext *avctx, AVSubtitle *sub,
  2803. int *got_sub_ptr,
  2804. const uint8_t *buf, int buf_size);
  2805. int avcodec_parse_frame(AVCodecContext *avctx, uint8_t **pdata,
  2806. int *data_size_ptr,
  2807. uint8_t *buf, int buf_size);
  2808. /**
  2809. * Encodes an audio frame from \p samples into \p buf.
  2810. * The avcodec_encode_audio() function encodes an audio frame from the input
  2811. * buffer \p samples. To encode it, it makes use of the audio codec which was
  2812. * coupled with \p avctx using avcodec_open(). The resulting encoded frame is
  2813. * stored in output buffer \p buf.
  2814. *
  2815. * @note The output buffer should be at least \c FF_MIN_BUFFER_SIZE bytes large.
  2816. *
  2817. * @param avctx the codec context
  2818. * @param[out] buf the output buffer
  2819. * @param[in] buf_size the output buffer size
  2820. * @param[in] samples the input buffer containing the samples
  2821. * The number of samples read from this buffer is frame_size*channels,
  2822. * both of which are defined in \p avctx.
  2823. * For PCM audio the number of samples read from \p samples is equal to
  2824. * \p buf_size * input_sample_size / output_sample_size.
  2825. * @return On error a negative value is returned, on success zero or the number
  2826. * of bytes used to encode the data read from the input buffer.
  2827. */
  2828. int avcodec_encode_audio(AVCodecContext *avctx, uint8_t *buf, int buf_size,
  2829. const short *samples);
  2830. /**
  2831. * Encodes a video frame from \p pict into \p buf.
  2832. * The avcodec_encode_video() function encodes a video frame from the input
  2833. * \p pict. To encode it, it makes use of the video codec which was coupled with
  2834. * \p avctx using avcodec_open(). The resulting encoded bytes representing the
  2835. * frame are stored in the output buffer \p buf. The input picture should be
  2836. * stored using a specific format, namely \c avctx.pix_fmt.
  2837. *
  2838. * @param avctx the codec context
  2839. * @param[out] buf the output buffer for the bitstream of encoded frame
  2840. * @param[in] buf_size the size of the output buffer in bytes
  2841. * @param[in] pict the input picture to encode
  2842. * @return On error a negative value is returned, on success zero or the number
  2843. * of bytes used from the output buffer.
  2844. */
  2845. int avcodec_encode_video(AVCodecContext *avctx, uint8_t *buf, int buf_size,
  2846. const AVFrame *pict);
  2847. int avcodec_encode_subtitle(AVCodecContext *avctx, uint8_t *buf, int buf_size,
  2848. const AVSubtitle *sub);
  2849. int avcodec_close(AVCodecContext *avctx);
  2850. /**
  2851. * Register all the codecs, parsers and bitstream filters which were enabled at
  2852. * configuration time. If you do not call this function you can select exactly
  2853. * which formats you want to support, by using the individual registration
  2854. * functions.
  2855. *
  2856. * @see avcodec_register
  2857. * @see av_register_codec_parser
  2858. * @see av_register_bitstream_filter
  2859. */
  2860. void avcodec_register_all(void);
  2861. /**
  2862. * Flush buffers, should be called when seeking or when switching to a different stream.
  2863. */
  2864. void avcodec_flush_buffers(AVCodecContext *avctx);
  2865. void avcodec_default_free_buffers(AVCodecContext *s);
  2866. /* misc useful functions */
  2867. /**
  2868. * Returns a single letter to describe the given picture type \p pict_type.
  2869. *
  2870. * @param[in] pict_type the picture type
  2871. * @return A single character representing the picture type.
  2872. */
  2873. char av_get_pict_type_char(int pict_type);
  2874. /**
  2875. * Returns codec bits per sample.
  2876. *
  2877. * @param[in] codec_id the codec
  2878. * @return Number of bits per sample or zero if unknown for the given codec.
  2879. */
  2880. int av_get_bits_per_sample(enum CodecID codec_id);
  2881. /**
  2882. * Returns sample format bits per sample.
  2883. *
  2884. * @param[in] sample_fmt the sample format
  2885. * @return Number of bits per sample or zero if unknown for the given sample format.
  2886. */
  2887. int av_get_bits_per_sample_format(enum SampleFormat sample_fmt);
  2888. /* frame parsing */
  2889. typedef struct AVCodecParserContext {
  2890. void *priv_data;
  2891. struct AVCodecParser *parser;
  2892. int64_t frame_offset; /* offset of the current frame */
  2893. int64_t cur_offset; /* current offset
  2894. (incremented by each av_parser_parse()) */
  2895. int64_t next_frame_offset; /* offset of the next frame */
  2896. /* video info */
  2897. int pict_type; /* XXX: Put it back in AVCodecContext. */
  2898. /**
  2899. * This field is used for proper frame duration computation in lavf.
  2900. * It signals, how much longer the frame duration of the current frame
  2901. * is compared to normal frame duration.
  2902. *
  2903. * frame_duration = (1 + repeat_pict) * time_base
  2904. *
  2905. * It is used by codecs like H.264 to display telecined material.
  2906. */
  2907. int repeat_pict; /* XXX: Put it back in AVCodecContext. */
  2908. int64_t pts; /* pts of the current frame */
  2909. int64_t dts; /* dts of the current frame */
  2910. /* private data */
  2911. int64_t last_pts;
  2912. int64_t last_dts;
  2913. int fetch_timestamp;
  2914. #define AV_PARSER_PTS_NB 4
  2915. int cur_frame_start_index;
  2916. int64_t cur_frame_offset[AV_PARSER_PTS_NB];
  2917. int64_t cur_frame_pts[AV_PARSER_PTS_NB];
  2918. int64_t cur_frame_dts[AV_PARSER_PTS_NB];
  2919. int flags;
  2920. #define PARSER_FLAG_COMPLETE_FRAMES 0x0001
  2921. int64_t offset; ///< byte offset from starting packet start
  2922. int64_t cur_frame_end[AV_PARSER_PTS_NB];
  2923. /*!
  2924. * Set by parser to 1 for key frames and 0 for non-key frames.
  2925. * It is initialized to -1, so if the parser doesn't set this flag,
  2926. * old-style fallback using FF_I_TYPE picture type as key frames
  2927. * will be used.
  2928. */
  2929. int key_frame;
  2930. /**
  2931. * Time difference in stream time base units from the pts of this
  2932. * packet to the point at which the output from the decoder has converged
  2933. * independent from the availability of previous frames. That is, the
  2934. * frames are virtually identical no matter if decoding started from
  2935. * the very first frame or from this keyframe.
  2936. * Is AV_NOPTS_VALUE if unknown.
  2937. * This field is not the display duration of the current frame.
  2938. *
  2939. * The purpose of this field is to allow seeking in streams that have no
  2940. * keyframes in the conventional sense. It corresponds to the
  2941. * recovery point SEI in H.264 and match_time_delta in NUT. It is also
  2942. * essential for some types of subtitle streams to ensure that all
  2943. * subtitles are correctly displayed after seeking.
  2944. */
  2945. int64_t convergence_duration;
  2946. // Timestamp generation support:
  2947. /**
  2948. * Synchronization point for start of timestamp generation.
  2949. *
  2950. * Set to >0 for sync point, 0 for no sync point and <0 for undefined
  2951. * (default).
  2952. *
  2953. * For example, this corresponds to presence of H.264 buffering period
  2954. * SEI message.
  2955. */
  2956. int dts_sync_point;
  2957. /**
  2958. * Offset of the current timestamp against last timestamp sync point in
  2959. * units of AVCodecContext.time_base.
  2960. *
  2961. * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
  2962. * contain a valid timestamp offset.
  2963. *
  2964. * Note that the timestamp of sync point has usually a nonzero
  2965. * dts_ref_dts_delta, which refers to the previous sync point. Offset of
  2966. * the next frame after timestamp sync point will be usually 1.
  2967. *
  2968. * For example, this corresponds to H.264 cpb_removal_delay.
  2969. */
  2970. int dts_ref_dts_delta;
  2971. /**
  2972. * Presentation delay of current frame in units of AVCodecContext.time_base.
  2973. *
  2974. * Set to INT_MIN when dts_sync_point unused. Otherwise, it must
  2975. * contain valid non-negative timestamp delta (presentation time of a frame
  2976. * must not lie in the past).
  2977. *
  2978. * This delay represents the difference between decoding and presentation
  2979. * time of the frame.
  2980. *
  2981. * For example, this corresponds to H.264 dpb_output_delay.
  2982. */
  2983. int pts_dts_delta;
  2984. /**
  2985. * Position of the packet in file.
  2986. *
  2987. * Analogous to cur_frame_pts/dts
  2988. */
  2989. int64_t cur_frame_pos[AV_PARSER_PTS_NB];
  2990. /**
  2991. * Byte position of currently parsed frame in stream.
  2992. */
  2993. int64_t pos;
  2994. /**
  2995. * Previous frame byte position.
  2996. */
  2997. int64_t last_pos;
  2998. } AVCodecParserContext;
  2999. typedef struct AVCodecParser {
  3000. int codec_ids[5]; /* several codec IDs are permitted */
  3001. int priv_data_size;
  3002. int (*parser_init)(AVCodecParserContext *s);
  3003. int (*parser_parse)(AVCodecParserContext *s,
  3004. AVCodecContext *avctx,
  3005. const uint8_t **poutbuf, int *poutbuf_size,
  3006. const uint8_t *buf, int buf_size);
  3007. void (*parser_close)(AVCodecParserContext *s);
  3008. int (*split)(AVCodecContext *avctx, const uint8_t *buf, int buf_size);
  3009. struct AVCodecParser *next;
  3010. } AVCodecParser;
  3011. AVCodecParser *av_parser_next(AVCodecParser *c);
  3012. void av_register_codec_parser(AVCodecParser *parser);
  3013. AVCodecParserContext *av_parser_init(int codec_id);
  3014. #if LIBAVCODEC_VERSION_MAJOR < 53
  3015. attribute_deprecated
  3016. int av_parser_parse(AVCodecParserContext *s,
  3017. AVCodecContext *avctx,
  3018. uint8_t **poutbuf, int *poutbuf_size,
  3019. const uint8_t *buf, int buf_size,
  3020. int64_t pts, int64_t dts);
  3021. #endif
  3022. /**
  3023. * Parse a packet.
  3024. *
  3025. * @param s parser context.
  3026. * @param avctx codec context.
  3027. * @param poutbuf set to pointer to parsed buffer or NULL if not yet finished.
  3028. * @param poutbuf_size set to size of parsed buffer or zero if not yet finished.
  3029. * @param buf input buffer.
  3030. * @param buf_size input length, to signal EOF, this should be 0 (so that the last frame can be output).
  3031. * @param pts input presentation timestamp.
  3032. * @param dts input decoding timestamp.
  3033. * @param pos input byte position in stream.
  3034. * @return the number of bytes of the input bitstream used.
  3035. *
  3036. * Example:
  3037. * @code
  3038. * while(in_len){
  3039. * len = av_parser_parse2(myparser, AVCodecContext, &data, &size,
  3040. * in_data, in_len,
  3041. * pts, dts, pos);
  3042. * in_data += len;
  3043. * in_len -= len;
  3044. *
  3045. * if(size)
  3046. * decode_frame(data, size);
  3047. * }
  3048. * @endcode
  3049. */
  3050. int av_parser_parse2(AVCodecParserContext *s,
  3051. AVCodecContext *avctx,
  3052. uint8_t **poutbuf, int *poutbuf_size,
  3053. const uint8_t *buf, int buf_size,
  3054. int64_t pts, int64_t dts,
  3055. int64_t pos);
  3056. int av_parser_change(AVCodecParserContext *s,
  3057. AVCodecContext *avctx,
  3058. uint8_t **poutbuf, int *poutbuf_size,
  3059. const uint8_t *buf, int buf_size, int keyframe);
  3060. void av_parser_close(AVCodecParserContext *s);
  3061. typedef struct AVBitStreamFilterContext {
  3062. void *priv_data;
  3063. struct AVBitStreamFilter *filter;
  3064. AVCodecParserContext *parser;
  3065. struct AVBitStreamFilterContext *next;
  3066. } AVBitStreamFilterContext;
  3067. typedef struct AVBitStreamFilter {
  3068. const char *name;
  3069. int priv_data_size;
  3070. int (*filter)(AVBitStreamFilterContext *bsfc,
  3071. AVCodecContext *avctx, const char *args,
  3072. uint8_t **poutbuf, int *poutbuf_size,
  3073. const uint8_t *buf, int buf_size, int keyframe);
  3074. void (*close)(AVBitStreamFilterContext *bsfc);
  3075. struct AVBitStreamFilter *next;
  3076. } AVBitStreamFilter;
  3077. void av_register_bitstream_filter(AVBitStreamFilter *bsf);
  3078. AVBitStreamFilterContext *av_bitstream_filter_init(const char *name);
  3079. int av_bitstream_filter_filter(AVBitStreamFilterContext *bsfc,
  3080. AVCodecContext *avctx, const char *args,
  3081. uint8_t **poutbuf, int *poutbuf_size,
  3082. const uint8_t *buf, int buf_size, int keyframe);
  3083. void av_bitstream_filter_close(AVBitStreamFilterContext *bsf);
  3084. AVBitStreamFilter *av_bitstream_filter_next(AVBitStreamFilter *f);
  3085. /* memory */
  3086. /**
  3087. * Reallocates the given block if it is not large enough, otherwise it
  3088. * does nothing.
  3089. *
  3090. * @see av_realloc
  3091. */
  3092. void *av_fast_realloc(void *ptr, unsigned int *size, unsigned int min_size);
  3093. /**
  3094. * Copy image 'src' to 'dst'.
  3095. */
  3096. void av_picture_copy(AVPicture *dst, const AVPicture *src,
  3097. enum PixelFormat pix_fmt, int width, int height);
  3098. /**
  3099. * Crop image top and left side.
  3100. */
  3101. int av_picture_crop(AVPicture *dst, const AVPicture *src,
  3102. enum PixelFormat pix_fmt, int top_band, int left_band);
  3103. /**
  3104. * Pad image.
  3105. */
  3106. int av_picture_pad(AVPicture *dst, const AVPicture *src, int height, int width, enum PixelFormat pix_fmt,
  3107. int padtop, int padbottom, int padleft, int padright, int *color);
  3108. unsigned int av_xiphlacing(unsigned char *s, unsigned int v);
  3109. /**
  3110. * Parses \p str and put in \p width_ptr and \p height_ptr the detected values.
  3111. *
  3112. * @return 0 in case of a successful parsing, a negative value otherwise
  3113. * @param[in] str the string to parse: it has to be a string in the format
  3114. * <width>x<height> or a valid video frame size abbreviation.
  3115. * @param[in,out] width_ptr pointer to the variable which will contain the detected
  3116. * frame width value
  3117. * @param[in,out] height_ptr pointer to the variable which will contain the detected
  3118. * frame height value
  3119. */
  3120. int av_parse_video_frame_size(int *width_ptr, int *height_ptr, const char *str);
  3121. /**
  3122. * Parses \p str and put in \p frame_rate the detected values.
  3123. *
  3124. * @return 0 in case of a successful parsing, a negative value otherwise
  3125. * @param[in] str the string to parse: it has to be a string in the format
  3126. * <frame_rate_num>/<frame_rate_den>, a float number or a valid video rate abbreviation
  3127. * @param[in,out] frame_rate pointer to the AVRational which will contain the detected
  3128. * frame rate
  3129. */
  3130. int av_parse_video_frame_rate(AVRational *frame_rate, const char *str);
  3131. /* error handling */
  3132. #if EINVAL > 0
  3133. #define AVERROR(e) (-(e)) /**< Returns a negative error code from a POSIX error code, to return from library functions. */
  3134. #define AVUNERROR(e) (-(e)) /**< Returns a POSIX error code from a library function error return value. */
  3135. #else
  3136. /* Some platforms have E* and errno already negated. */
  3137. #define AVERROR(e) (e)
  3138. #define AVUNERROR(e) (e)
  3139. #endif
  3140. #define AVERROR_UNKNOWN AVERROR(EINVAL) /**< unknown error */
  3141. #define AVERROR_IO AVERROR(EIO) /**< I/O error */
  3142. #define AVERROR_NUMEXPECTED AVERROR(EDOM) /**< Number syntax expected in filename. */
  3143. #define AVERROR_INVALIDDATA AVERROR(EINVAL) /**< invalid data found */
  3144. #define AVERROR_NOMEM AVERROR(ENOMEM) /**< not enough memory */
  3145. #define AVERROR_NOFMT AVERROR(EILSEQ) /**< unknown format */
  3146. #define AVERROR_NOTSUPP AVERROR(ENOSYS) /**< Operation not supported. */
  3147. #define AVERROR_NOENT AVERROR(ENOENT) /**< No such file or directory. */
  3148. #define AVERROR_EOF AVERROR(EPIPE) /**< End of file. */
  3149. #define AVERROR_PATCHWELCOME -MKTAG('P','A','W','E') /**< Not yet implemented in FFmpeg. Patches welcome. */
  3150. /**
  3151. * Registers the hardware accelerator \p hwaccel.
  3152. */
  3153. void av_register_hwaccel(AVHWAccel *hwaccel);
  3154. /**
  3155. * If hwaccel is NULL, returns the first registered hardware accelerator,
  3156. * if hwaccel is non-NULL, returns the next registered hardware accelerator
  3157. * after hwaccel, or NULL if hwaccel is the last one.
  3158. */
  3159. AVHWAccel *av_hwaccel_next(AVHWAccel *hwaccel);
  3160. #endif /* AVCODEC_AVCODEC_H */