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 )
6 * This file is part of FFmpeg.
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.
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.
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
25 * AAC definitions and structures
26 * @author Oded Shimon ( ods15 ods15 dyndns org )
27 * @author Maxim Gavrilov ( maxim.gavrilov gmail com )
36 #include "mpeg4audio.h"
38 #include "fmtconvert.h"
42 #define MAX_CHANNELS 64
43 #define MAX_ELEM_ID 16
45 #define TNS_MAX_ORDER 20
46 #define MAX_LTP_LONG_SFB 40
48 enum RawDataBlockType {
59 enum ExtensionPayloadID {
63 EXT_DYNAMIC_RANGE = 0xb,
65 EXT_SBR_DATA_CRC = 0xe,
76 ZERO_BT = 0, ///< Scalefactors and spectral data are all zero.
77 FIRST_PAIR_BT = 5, ///< This and later band types encode two values (rather than four) with one code word.
78 ESC_BT = 11, ///< Spectral data are coded with an escape sequence.
79 NOISE_BT = 13, ///< Spectral data are scaled white noise not coded in the bitstream.
80 INTENSITY_BT2 = 14, ///< Scalefactor data are intensity stereo positions.
81 INTENSITY_BT = 15, ///< Scalefactor data are intensity stereo positions.
84 #define IS_CODEBOOK_UNSIGNED(x) ((x - 1) & 10)
86 enum ChannelPosition {
87 AAC_CHANNEL_FRONT = 1,
95 * The point during decoding at which channel coupling is applied.
99 BETWEEN_TNS_AND_IMDCT,
104 * Output configuration status
107 OC_NONE, ///< Output unconfigured
108 OC_TRIAL_PCE, ///< Output configuration under trial specified by an inband PCE
109 OC_TRIAL_FRAME, ///< Output configuration under trial specified by a frame header
110 OC_GLOBAL_HDR, ///< Output configuration set in a global header but not yet locked
111 OC_LOCKED, ///< Output configuration locked in place
126 #define MAX_PREDICTORS 672
128 #define SCALE_DIV_512 36 ///< scalefactor difference that corresponds to scale difference in 512 times
129 #define SCALE_ONE_POS 140 ///< scalefactor index that corresponds to scale=1.0
130 #define SCALE_MAX_POS 255 ///< scalefactor index maximum value
131 #define SCALE_MAX_DIFF 60 ///< maximum scalefactor difference allowed by standard
132 #define SCALE_DIFF_ZERO 60 ///< codebook index corresponding to zero scalefactor indices difference
133 #define POW_SF2_ZERO 200 ///< ff_aac_pow2sf_tab index corresponding to pow(2, 0);
136 * Long Term Prediction
142 int8_t used[MAX_LTP_LONG_SFB];
143 } LongTermPrediction;
146 * Individual Channel Stream
149 uint8_t max_sfb; ///< number of scalefactor bands per group
150 enum WindowSequence window_sequence[2];
151 uint8_t use_kb_window[2]; ///< If set, use Kaiser-Bessel window, otherwise use a sinus window.
152 int num_window_groups;
153 uint8_t group_len[8];
154 LongTermPrediction ltp;
155 const uint16_t *swb_offset; ///< table of offsets to the lowest spectral coefficient of a scalefactor band, sfb, for a particular window
156 const uint8_t *swb_sizes; ///< table of scalefactor band sizes for a particular window
157 int num_swb; ///< number of scalefactor window bands
160 int predictor_present;
161 int predictor_initialized;
162 int predictor_reset_group;
163 uint8_t prediction_used[41];
164 } IndividualChannelStream;
167 * Temporal Noise Shaping
175 float coef[8][4][TNS_MAX_ORDER];
176 } TemporalNoiseShaping;
179 * Dynamic Range Control - decoded from the bitstream but not processed further.
182 int pce_instance_tag; ///< Indicates with which program the DRC info is associated.
183 int dyn_rng_sgn[17]; ///< DRC sign information; 0 - positive, 1 - negative
184 int dyn_rng_ctl[17]; ///< DRC magnitude information
185 int exclude_mask[MAX_CHANNELS]; ///< Channels to be excluded from DRC processing.
186 int band_incr; ///< Number of DRC bands greater than 1 having DRC info.
187 int interpolation_scheme; ///< Indicates the interpolation scheme used in the SBR QMF domain.
188 int band_top[17]; ///< Indicates the top of the i-th DRC band in units of 4 spectral lines.
189 int prog_ref_level; /**< A reference level for the long-term program audio level for all
192 } DynamicRangeControl;
202 * coupling parameters
205 enum CouplingPoint coupling_point; ///< The point during decoding at which coupling is applied.
206 int num_coupled; ///< number of target elements
207 enum RawDataBlockType type[8]; ///< Type of channel element to be coupled - SCE or CPE.
208 int id_select[8]; ///< element id
209 int ch_select[8]; /**< [0] shared list of gains; [1] list of gains for right channel;
210 * [2] list of gains for left channel; [3] lists of gains for both channels
216 * Single Channel Element - used for both SCE and LFE elements.
219 IndividualChannelStream ics;
220 TemporalNoiseShaping tns;
222 enum BandType band_type[128]; ///< band types
223 int band_type_run_end[120]; ///< band type run end points
224 float sf[120]; ///< scalefactors
225 int sf_idx[128]; ///< scalefactor indices (used by encoder)
226 uint8_t zeroes[128]; ///< band is not coded (used by encoder)
227 DECLARE_ALIGNED(32, float, coeffs)[1024]; ///< coefficients for IMDCT
228 DECLARE_ALIGNED(32, float, saved)[1024]; ///< overlap
229 DECLARE_ALIGNED(32, float, ret)[2048]; ///< PCM output
230 DECLARE_ALIGNED(16, float, ltp_state)[3072]; ///< time signal for LTP
231 PredictorState predictor_state[MAX_PREDICTORS];
232 } SingleChannelElement;
235 * channel element - generic struct for SCE/CPE/CCE/LFE
239 int common_window; ///< Set if channels share a common 'IndividualChannelStream' in bitstream.
240 int ms_mode; ///< Signals mid/side stereo flags coding mode (used by encoder)
241 uint8_t ms_mask[128]; ///< Set if mid/side stereo is used for each scalefactor window band
243 SingleChannelElement ch[2];
245 ChannelCoupling coup;
246 SpectralBandReplication sbr;
253 AVCodecContext *avctx;
255 MPEG4AudioConfig m4ac;
257 int is_saved; ///< Set if elements have stored overlap from previous frame.
258 DynamicRangeControl che_drc;
261 * @name Channel element related data
264 enum ChannelPosition che_pos[4][MAX_ELEM_ID]; /**< channel element channel mapping with the
265 * first index as the first 4 raw data block types
267 ChannelElement *che[4][MAX_ELEM_ID];
268 ChannelElement *tag_che_map[4][MAX_ELEM_ID];
273 * @name temporary aligned temporary buffers
274 * (We do not want to have these on the stack.)
277 DECLARE_ALIGNED(32, float, buf_mdct)[1024];
281 * @name Computed / set up during initialization
285 FFTContext mdct_small;
288 FmtConvertContext fmt_conv;
293 * @name Members used for output interleaving
296 float *output_data[MAX_CHANNELS]; ///< Points to each element's 'ret' buffer (PCM output).
299 DECLARE_ALIGNED(32, float, temp)[128];
301 enum OCStatus output_configured;
304 #endif /* AVCODEC_AAC_H */