1/* 2 * AAC definitions and structures 3 * Copyright (c) 2005-2006 Oded Shimon ( ods15 ods15 dyndns org ) 4 * Copyright (c) 2006-2007 Maxim Gavrilov ( maxim.gavrilov gmail com ) 5 * 6 * This file is part of FFmpeg. 7 * 8 * FFmpeg is free software; you can redistribute it and/or 9 * modify it under the terms of the GNU Lesser General Public 10 * License as published by the Free Software Foundation; either 11 * version 2.1 of the License, or (at your option) any later version. 12 * 13 * FFmpeg is distributed in the hope that it will be useful, 14 * but WITHOUT ANY WARRANTY; without even the implied warranty of 15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU 16 * Lesser General Public License for more details. 17 * 18 * You should have received a copy of the GNU Lesser General Public 19 * License along with FFmpeg; if not, write to the Free Software 20 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA 21 */ 22 23/** 24 * @file 25 * AAC definitions and structures 26 * @author Oded Shimon ( ods15 ods15 dyndns org ) 27 * @author Maxim Gavrilov ( maxim.gavrilov gmail com ) 28 */ 29 30#ifndef AVCODEC_AAC_H 31#define AVCODEC_AAC_H 32 33#include "avcodec.h" 34#include "dsputil.h" 35#include "fft.h" 36#include "mpeg4audio.h" 37#include "sbr.h" 38 39#include <stdint.h> 40 41#define AAC_INIT_VLC_STATIC(num, size) \ 42 INIT_VLC_STATIC(&vlc_spectral[num], 8, ff_aac_spectral_sizes[num], \ 43 ff_aac_spectral_bits[num], sizeof( ff_aac_spectral_bits[num][0]), sizeof( ff_aac_spectral_bits[num][0]), \ 44 ff_aac_spectral_codes[num], sizeof(ff_aac_spectral_codes[num][0]), sizeof(ff_aac_spectral_codes[num][0]), \ 45 size); 46 47#define MAX_CHANNELS 64 48#define MAX_ELEM_ID 16 49 50#define TNS_MAX_ORDER 20 51 52enum RawDataBlockType { 53 TYPE_SCE, 54 TYPE_CPE, 55 TYPE_CCE, 56 TYPE_LFE, 57 TYPE_DSE, 58 TYPE_PCE, 59 TYPE_FIL, 60 TYPE_END, 61}; 62 63enum ExtensionPayloadID { 64 EXT_FILL, 65 EXT_FILL_DATA, 66 EXT_DATA_ELEMENT, 67 EXT_DYNAMIC_RANGE = 0xb, 68 EXT_SBR_DATA = 0xd, 69 EXT_SBR_DATA_CRC = 0xe, 70}; 71 72enum WindowSequence { 73 ONLY_LONG_SEQUENCE, 74 LONG_START_SEQUENCE, 75 EIGHT_SHORT_SEQUENCE, 76 LONG_STOP_SEQUENCE, 77}; 78 79enum BandType { 80 ZERO_BT = 0, ///< Scalefactors and spectral data are all zero. 81 FIRST_PAIR_BT = 5, ///< This and later band types encode two values (rather than four) with one code word. 82 ESC_BT = 11, ///< Spectral data are coded with an escape sequence. 83 NOISE_BT = 13, ///< Spectral data are scaled white noise not coded in the bitstream. 84 INTENSITY_BT2 = 14, ///< Scalefactor data are intensity stereo positions. 85 INTENSITY_BT = 15, ///< Scalefactor data are intensity stereo positions. 86}; 87 88#define IS_CODEBOOK_UNSIGNED(x) ((x - 1) & 10) 89 90enum ChannelPosition { 91 AAC_CHANNEL_FRONT = 1, 92 AAC_CHANNEL_SIDE = 2, 93 AAC_CHANNEL_BACK = 3, 94 AAC_CHANNEL_LFE = 4, 95 AAC_CHANNEL_CC = 5, 96}; 97 98/** 99 * The point during decoding at which channel coupling is applied. 100 */ 101enum CouplingPoint { 102 BEFORE_TNS, 103 BETWEEN_TNS_AND_IMDCT, 104 AFTER_IMDCT = 3, 105}; 106 107/** 108 * Output configuration status 109 */ 110enum OCStatus { 111 OC_NONE, //< Output unconfigured 112 OC_TRIAL_PCE, //< Output configuration under trial specified by an inband PCE 113 OC_TRIAL_FRAME, //< Output configuration under trial specified by a frame header 114 OC_GLOBAL_HDR, //< Output configuration set in a global header but not yet locked 115 OC_LOCKED, //< Output configuration locked in place 116}; 117 118/** 119 * Predictor State 120 */ 121typedef struct { 122 float cor0; 123 float cor1; 124 float var0; 125 float var1; 126 float r0; 127 float r1; 128} PredictorState; 129 130#define MAX_PREDICTORS 672 131 132#define SCALE_DIV_512 36 ///< scalefactor difference that corresponds to scale difference in 512 times 133#define SCALE_ONE_POS 140 ///< scalefactor index that corresponds to scale=1.0 134#define SCALE_MAX_POS 255 ///< scalefactor index maximum value 135#define SCALE_MAX_DIFF 60 ///< maximum scalefactor difference allowed by standard 136#define SCALE_DIFF_ZERO 60 ///< codebook index corresponding to zero scalefactor indices difference 137 138/** 139 * Individual Channel Stream 140 */ 141typedef struct { 142 uint8_t max_sfb; ///< number of scalefactor bands per group 143 enum WindowSequence window_sequence[2]; 144 uint8_t use_kb_window[2]; ///< If set, use Kaiser-Bessel window, otherwise use a sinus window. 145 int num_window_groups; 146 uint8_t group_len[8]; 147 const uint16_t *swb_offset; ///< table of offsets to the lowest spectral coefficient of a scalefactor band, sfb, for a particular window 148 const uint8_t *swb_sizes; ///< table of scalefactor band sizes for a particular window 149 int num_swb; ///< number of scalefactor window bands 150 int num_windows; 151 int tns_max_bands; 152 int predictor_present; 153 int predictor_initialized; 154 int predictor_reset_group; 155 uint8_t prediction_used[41]; 156} IndividualChannelStream; 157 158/** 159 * Temporal Noise Shaping 160 */ 161typedef struct { 162 int present; 163 int n_filt[8]; 164 int length[8][4]; 165 int direction[8][4]; 166 int order[8][4]; 167 float coef[8][4][TNS_MAX_ORDER]; 168} TemporalNoiseShaping; 169 170/** 171 * Dynamic Range Control - decoded from the bitstream but not processed further. 172 */ 173typedef struct { 174 int pce_instance_tag; ///< Indicates with which program the DRC info is associated. 175 int dyn_rng_sgn[17]; ///< DRC sign information; 0 - positive, 1 - negative 176 int dyn_rng_ctl[17]; ///< DRC magnitude information 177 int exclude_mask[MAX_CHANNELS]; ///< Channels to be excluded from DRC processing. 178 int band_incr; ///< Number of DRC bands greater than 1 having DRC info. 179 int interpolation_scheme; ///< Indicates the interpolation scheme used in the SBR QMF domain. 180 int band_top[17]; ///< Indicates the top of the i-th DRC band in units of 4 spectral lines. 181 int prog_ref_level; /**< A reference level for the long-term program audio level for all 182 * channels combined. 183 */ 184} DynamicRangeControl; 185 186typedef struct { 187 int num_pulse; 188 int start; 189 int pos[4]; 190 int amp[4]; 191} Pulse; 192 193/** 194 * coupling parameters 195 */ 196typedef struct { 197 enum CouplingPoint coupling_point; ///< The point during decoding at which coupling is applied. 198 int num_coupled; ///< number of target elements 199 enum RawDataBlockType type[8]; ///< Type of channel element to be coupled - SCE or CPE. 200 int id_select[8]; ///< element id 201 int ch_select[8]; /**< [0] shared list of gains; [1] list of gains for right channel; 202 * [2] list of gains for left channel; [3] lists of gains for both channels 203 */ 204 float gain[16][120]; 205} ChannelCoupling; 206 207/** 208 * Single Channel Element - used for both SCE and LFE elements. 209 */ 210typedef struct { 211 IndividualChannelStream ics; 212 TemporalNoiseShaping tns; 213 Pulse pulse; 214 enum BandType band_type[128]; ///< band types 215 int band_type_run_end[120]; ///< band type run end points 216 float sf[120]; ///< scalefactors 217 int sf_idx[128]; ///< scalefactor indices (used by encoder) 218 uint8_t zeroes[128]; ///< band is not coded (used by encoder) 219 DECLARE_ALIGNED(16, float, coeffs)[1024]; ///< coefficients for IMDCT 220 DECLARE_ALIGNED(16, float, saved)[1024]; ///< overlap 221 DECLARE_ALIGNED(16, float, ret)[2048]; ///< PCM output 222 PredictorState predictor_state[MAX_PREDICTORS]; 223} SingleChannelElement; 224 225/** 226 * channel element - generic struct for SCE/CPE/CCE/LFE 227 */ 228typedef struct { 229 // CPE specific 230 int common_window; ///< Set if channels share a common 'IndividualChannelStream' in bitstream. 231 int ms_mode; ///< Signals mid/side stereo flags coding mode (used by encoder) 232 uint8_t ms_mask[128]; ///< Set if mid/side stereo is used for each scalefactor window band 233 // shared 234 SingleChannelElement ch[2]; 235 // CCE specific 236 ChannelCoupling coup; 237 SpectralBandReplication sbr; 238} ChannelElement; 239 240/** 241 * main AAC context 242 */ 243typedef struct { 244 AVCodecContext * avccontext; 245 246 MPEG4AudioConfig m4ac; 247 248 int is_saved; ///< Set if elements have stored overlap from previous frame. 249 DynamicRangeControl che_drc; 250 251 /** 252 * @defgroup elements Channel element related data. 253 * @{ 254 */ 255 enum ChannelPosition che_pos[4][MAX_ELEM_ID]; /**< channel element channel mapping with the 256 * first index as the first 4 raw data block types 257 */ 258 ChannelElement * che[4][MAX_ELEM_ID]; 259 ChannelElement * tag_che_map[4][MAX_ELEM_ID]; 260 int tags_mapped; 261 /** @} */ 262 263 /** 264 * @defgroup temporary aligned temporary buffers (We do not want to have these on the stack.) 265 * @{ 266 */ 267 DECLARE_ALIGNED(16, float, buf_mdct)[1024]; 268 /** @} */ 269 270 /** 271 * @defgroup tables Computed / set up during initialization. 272 * @{ 273 */ 274 FFTContext mdct; 275 FFTContext mdct_small; 276 DSPContext dsp; 277 int random_state; 278 /** @} */ 279 280 /** 281 * @defgroup output Members used for output interleaving. 282 * @{ 283 */ 284 float *output_data[MAX_CHANNELS]; ///< Points to each element's 'ret' buffer (PCM output). 285 float add_bias; ///< offset for dsp.float_to_int16 286 float sf_scale; ///< Pre-scale for correct IMDCT and dsp.float_to_int16. 287 int sf_offset; ///< offset into pow2sf_tab as appropriate for dsp.float_to_int16 288 /** @} */ 289 290 DECLARE_ALIGNED(16, float, temp)[128]; 291 292 enum OCStatus output_configured; 293} AACContext; 294 295#endif /* AVCODEC_AAC_H */ 296