libavcodec/aac.h

   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 libavcodec/aac.h
  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 "libavutil/internal.h"
  34 #include "avcodec.h"
  35 #include "dsputil.h"
  36 #include "mpeg4audio.h"
  37
  38 #include <stdint.h>
  39
  40 #define AAC_INIT_VLC_STATIC(num, size) \
  41     INIT_VLC_STATIC(&vlc_spectral[num], 6, ff_aac_spectral_sizes[num], \
  42          ff_aac_spectral_bits[num], sizeof( ff_aac_spectral_bits[num][0]), sizeof( ff_aac_spectral_bits[num][0]), \
  43         ff_aac_spectral_codes[num], sizeof(ff_aac_spectral_codes[num][0]), sizeof(ff_aac_spectral_codes[num][0]), \
  44         size);
  45
  46 #define MAX_CHANNELS 64
  47 #define MAX_ELEM_ID 16
  48
  49 #define TNS_MAX_ORDER 20
  50
  51 enum RawDataBlockType {
  52     TYPE_SCE,
  53     TYPE_CPE,
  54     TYPE_CCE,
  55     TYPE_LFE,
  56     TYPE_DSE,
  57     TYPE_PCE,
  58     TYPE_FIL,
  59     TYPE_END,
  60 };
  61
  62 enum ExtensionPayloadID {
  63     EXT_FILL,
  64     EXT_FILL_DATA,
  65     EXT_DATA_ELEMENT,
  66     EXT_DYNAMIC_RANGE = 0xb,
  67     EXT_SBR_DATA      = 0xd,
  68     EXT_SBR_DATA_CRC  = 0xe,
  69 };
  70
  71 enum WindowSequence {
  72     ONLY_LONG_SEQUENCE,
  73     LONG_START_SEQUENCE,
  74     EIGHT_SHORT_SEQUENCE,
  75     LONG_STOP_SEQUENCE,
  76 };
  77
  78 enum BandType {
  79     ZERO_BT        = 0,     ///< Scalefactors and spectral data are all zero.
  80     FIRST_PAIR_BT  = 5,     ///< This and later band types encode two values (rather than four) with one code word.
  81     ESC_BT         = 11,    ///< Spectral data are coded with an escape sequence.
  82     NOISE_BT       = 13,    ///< Spectral data are scaled white noise not coded in the bitstream.
  83     INTENSITY_BT2  = 14,    ///< Scalefactor data are intensity stereo positions.
  84     INTENSITY_BT   = 15,    ///< Scalefactor data are intensity stereo positions.
  85 };
  86
  87 #define IS_CODEBOOK_UNSIGNED(x) ((x - 1) & 10)
  88
  89 enum ChannelPosition {
  90     AAC_CHANNEL_FRONT = 1,
  91     AAC_CHANNEL_SIDE  = 2,
  92     AAC_CHANNEL_BACK  = 3,
  93     AAC_CHANNEL_LFE   = 4,
  94     AAC_CHANNEL_CC    = 5,
  95 };
  96
  97 /**
  98  * The point during decoding at which channel coupling is applied.
  99  */
 100 enum CouplingPoint {
 101     BEFORE_TNS,
 102     BETWEEN_TNS_AND_IMDCT,
 103     AFTER_IMDCT = 3,
 104 };
 105
 106 /**
 107  * Predictor State
 108  */
 109 typedef struct {
 110     float cor0;
 111     float cor1;
 112     float var0;
 113     float var1;
 114     float r0;
 115     float r1;
 116 } PredictorState;
 117
 118 #define MAX_PREDICTORS 672
 119
 120 /**
 121  * Individual Channel Stream
 122  */
 123 typedef struct {
 124     uint8_t max_sfb;            ///< number of scalefactor bands per group
 125     enum WindowSequence window_sequence[2];
 126     uint8_t use_kb_window[2];   ///< If set, use Kaiser-Bessel window, otherwise use a sinus window.
 127     int num_window_groups;
 128     uint8_t group_len[8];
 129     const uint16_t *swb_offset; ///< table of offsets to the lowest spectral coefficient of a scalefactor band, sfb, for a particular window
 130     int num_swb;                ///< number of scalefactor window bands
 131     int num_windows;
 132     int tns_max_bands;
 133     int predictor_present;
 134     int predictor_initialized;
 135     int predictor_reset_group;
 136     uint8_t prediction_used[41];
 137 } IndividualChannelStream;
 138
 139 /**
 140  * Temporal Noise Shaping
 141  */
 142 typedef struct {
 143     int present;
 144     int n_filt[8];
 145     int length[8][4];
 146     int direction[8][4];
 147     int order[8][4];
 148     float coef[8][4][TNS_MAX_ORDER];
 149 } TemporalNoiseShaping;
 150
 151 /**
 152  * Dynamic Range Control - decoded from the bitstream but not processed further.
 153  */
 154 typedef struct {
 155     int pce_instance_tag;                           ///< Indicates with which program the DRC info is associated.
 156     int dyn_rng_sgn[17];                            ///< DRC sign information; 0 - positive, 1 - negative
 157     int dyn_rng_ctl[17];                            ///< DRC magnitude information
 158     int exclude_mask[MAX_CHANNELS];                 ///< Channels to be excluded from DRC processing.
 159     int band_incr;                                  ///< Number of DRC bands greater than 1 having DRC info.
 160     int interpolation_scheme;                       ///< Indicates the interpolation scheme used in the SBR QMF domain.
 161     int band_top[17];                               ///< Indicates the top of the i-th DRC band in units of 4 spectral lines.
 162     int prog_ref_level;                             /**< A reference level for the long-term program audio level for all
 163                                                      *   channels combined.
 164                                                      */
 165 } DynamicRangeControl;
 166
 167 typedef struct {
 168     int num_pulse;
 169     int pos[4];
 170     int amp[4];
 171 } Pulse;
 172
 173 /**
 174  * coupling parameters
 175  */
 176 typedef struct {
 177     enum CouplingPoint coupling_point;  ///< The point during decoding at which coupling is applied.
 178     int num_coupled;       ///< number of target elements
 179     enum RawDataBlockType type[8];   ///< Type of channel element to be coupled - SCE or CPE.
 180     int id_select[8];      ///< element id
 181     int ch_select[8];      /**< [0] shared list of gains; [1] list of gains for right channel;
 182                             *   [2] list of gains for left channel; [3] lists of gains for both channels
 183                             */
 184     float gain[16][120];
 185 } ChannelCoupling;
 186
 187 /**
 188  * Single Channel Element - used for both SCE and LFE elements.
 189  */
 190 typedef struct {
 191     IndividualChannelStream ics;
 192     TemporalNoiseShaping tns;
 193     enum BandType band_type[120];             ///< band types
 194     int band_type_run_end[120];               ///< band type run end points
 195     float sf[120];                            ///< scalefactors
 196     DECLARE_ALIGNED_16(float, coeffs[1024]);  ///< coefficients for IMDCT
 197     DECLARE_ALIGNED_16(float, saved[512]);    ///< overlap
 198     DECLARE_ALIGNED_16(float, ret[1024]);     ///< PCM output
 199     PredictorState predictor_state[MAX_PREDICTORS];
 200 } SingleChannelElement;
 201
 202 /**
 203  * channel element - generic struct for SCE/CPE/CCE/LFE
 204  */
 205 typedef struct {
 206     // CPE specific
 207     uint8_t ms_mask[120];     ///< Set if mid/side stereo is used for each scalefactor window band
 208     // shared
 209     SingleChannelElement ch[2];
 210     // CCE specific
 211     ChannelCoupling coup;
 212 } ChannelElement;
 213
 214 /**
 215  * main AAC context
 216  */
 217 typedef struct {
 218     AVCodecContext * avccontext;
 219
 220     MPEG4AudioConfig m4ac;
 221
 222     int is_saved;                 ///< Set if elements have stored overlap from previous frame.
 223     DynamicRangeControl che_drc;
 224
 225     /**
 226      * @defgroup elements Channel element related data.
 227      * @{
 228      */
 229     enum ChannelPosition che_pos[4][MAX_ELEM_ID]; /**< channel element channel mapping with the
 230                                                    *   first index as the first 4 raw data block types
 231                                                    */
 232     ChannelElement * che[4][MAX_ELEM_ID];
 233     ChannelElement * tag_che_map[4][MAX_ELEM_ID];
 234     int tags_mapped;
 235     /** @} */
 236
 237     /**
 238      * @defgroup temporary aligned temporary buffers (We do not want to have these on the stack.)
 239      * @{
 240      */
 241     DECLARE_ALIGNED_16(float, buf_mdct[1024]);
 242     /** @} */
 243
 244     /**
 245      * @defgroup tables   Computed / set up during initialization.
 246      * @{
 247      */
 248     MDCTContext mdct;
 249     MDCTContext mdct_small;
 250     DSPContext dsp;
 251     int random_state;
 252     /** @} */
 253
 254     /**
 255      * @defgroup output   Members used for output interleaving.
 256      * @{
 257      */
 258     float *output_data[MAX_CHANNELS];                 ///< Points to each element's 'ret' buffer (PCM output).
 259     float add_bias;                                   ///< offset for dsp.float_to_int16
 260     float sf_scale;                                   ///< Pre-scale for correct IMDCT and dsp.float_to_int16.
 261     int sf_offset;                                    ///< offset into pow2sf_tab as appropriate for dsp.float_to_int16
 262     /** @} */
 263
 264     DECLARE_ALIGNED(16, float, temp[128]);
 265 } AACContext;
 266
 267 #endif /* AVCODEC_AAC_H */