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 )
35 #include "mpeg4audio.h"
39 #define AAC_INIT_VLC_STATIC(num, size) \
40 INIT_VLC_STATIC(&vlc_spectral[num], 6, ff_aac_spectral_sizes[num], \
41 ff_aac_spectral_bits[num], sizeof( ff_aac_spectral_bits[num][0]), sizeof( ff_aac_spectral_bits[num][0]), \
42 ff_aac_spectral_codes[num], sizeof(ff_aac_spectral_codes[num][0]), sizeof(ff_aac_spectral_codes[num][0]), \
45 #define MAX_CHANNELS 64
46 #define MAX_ELEM_ID 16
48 #define TNS_MAX_ORDER 20
50 enum AudioObjectType {
53 AOT_AAC_MAIN, ///< Y Main
54 AOT_AAC_LC, ///< Y Low Complexity
55 AOT_AAC_SSR, ///< N (code in SoC repo) Scalable Sample Rate
56 AOT_AAC_LTP, ///< N (code in SoC repo) Long Term Prediction
57 AOT_SBR, ///< N (in progress) Spectral Band Replication
58 AOT_AAC_SCALABLE, ///< N Scalable
59 AOT_TWINVQ, ///< N Twin Vector Quantizer
60 AOT_CELP, ///< N Code Excited Linear Prediction
61 AOT_HVXC, ///< N Harmonic Vector eXcitation Coding
62 AOT_TTSI = 12, ///< N Text-To-Speech Interface
63 AOT_MAINSYNTH, ///< N Main Synthesis
64 AOT_WAVESYNTH, ///< N Wavetable Synthesis
65 AOT_MIDI, ///< N General MIDI
66 AOT_SAFX, ///< N Algorithmic Synthesis and Audio Effects
67 AOT_ER_AAC_LC, ///< N Error Resilient Low Complexity
68 AOT_ER_AAC_LTP = 19, ///< N Error Resilient Long Term Prediction
69 AOT_ER_AAC_SCALABLE, ///< N Error Resilient Scalable
70 AOT_ER_TWINVQ, ///< N Error Resilient Twin Vector Quantizer
71 AOT_ER_BSAC, ///< N Error Resilient Bit-Sliced Arithmetic Coding
72 AOT_ER_AAC_LD, ///< N Error Resilient Low Delay
73 AOT_ER_CELP, ///< N Error Resilient Code Excited Linear Prediction
74 AOT_ER_HVXC, ///< N Error Resilient Harmonic Vector eXcitation Coding
75 AOT_ER_HILN, ///< N Error Resilient Harmonic and Individual Lines plus Noise
76 AOT_ER_PARAM, ///< N Error Resilient Parametric
77 AOT_SSC, ///< N SinuSoidal Coding
80 enum RawDataBlockType {
91 enum ExtensionPayloadID {
95 EXT_DYNAMIC_RANGE = 0xb,
97 EXT_SBR_DATA_CRC = 0xe,
100 enum WindowSequence {
103 EIGHT_SHORT_SEQUENCE,
108 ZERO_BT = 0, ///< Scalefactors and spectral data are all zero.
109 FIRST_PAIR_BT = 5, ///< This and later band types encode two values (rather than four) with one code word.
110 ESC_BT = 11, ///< Spectral data are coded with an escape sequence.
111 NOISE_BT = 13, ///< Spectral data are scaled white noise not coded in the bitstream.
112 INTENSITY_BT2 = 14, ///< Scalefactor data are intensity stereo positions.
113 INTENSITY_BT = 15, ///< Scalefactor data are intensity stereo positions.
116 #define IS_CODEBOOK_UNSIGNED(x) ((x - 1) & 10)
118 enum ChannelPosition {
119 AAC_CHANNEL_FRONT = 1,
120 AAC_CHANNEL_SIDE = 2,
121 AAC_CHANNEL_BACK = 3,
127 * The point during decoding at which channel coupling is applied.
131 BETWEEN_TNS_AND_IMDCT,
147 #define MAX_PREDICTORS 672
150 * Individual Channel Stream
153 uint8_t max_sfb; ///< number of scalefactor bands per group
154 enum WindowSequence window_sequence[2];
155 uint8_t use_kb_window[2]; ///< If set, use Kaiser-Bessel window, otherwise use a sinus window.
156 int num_window_groups;
157 uint8_t group_len[8];
158 const uint16_t *swb_offset; ///< table of offsets to the lowest spectral coefficient of a scalefactor band, sfb, for a particular window
159 int num_swb; ///< number of scalefactor window bands
162 int predictor_present;
163 int predictor_initialized;
164 int predictor_reset_group;
165 uint8_t prediction_used[41];
166 } IndividualChannelStream;
169 * Temporal Noise Shaping
177 float coef[8][4][TNS_MAX_ORDER];
178 } TemporalNoiseShaping;
181 * Dynamic Range Control - decoded from the bitstream but not processed further.
184 int pce_instance_tag; ///< Indicates with which program the DRC info is associated.
185 int dyn_rng_sgn[17]; ///< DRC sign information; 0 - positive, 1 - negative
186 int dyn_rng_ctl[17]; ///< DRC magnitude information
187 int exclude_mask[MAX_CHANNELS]; ///< Channels to be excluded from DRC processing.
188 int band_incr; ///< Number of DRC bands greater than 1 having DRC info.
189 int interpolation_scheme; ///< Indicates the interpolation scheme used in the SBR QMF domain.
190 int band_top[17]; ///< Indicates the top of the i-th DRC band in units of 4 spectral lines.
191 int prog_ref_level; /**< A reference level for the long-term program audio level for all
194 } DynamicRangeControl;
203 * coupling parameters
206 enum CouplingPoint coupling_point; ///< The point during decoding at which coupling is applied.
207 int num_coupled; ///< number of target elements
208 enum RawDataBlockType type[8]; ///< Type of channel element to be coupled - SCE or CPE.
209 int id_select[8]; ///< element id
210 int ch_select[8]; /**< [0] shared list of gains; [1] list of gains for right channel;
211 * [2] list of gains for left channel; [3] lists of gains for both channels
217 * Single Channel Element - used for both SCE and LFE elements.
220 IndividualChannelStream ics;
221 TemporalNoiseShaping tns;
222 enum BandType band_type[120]; ///< band types
223 int band_type_run_end[120]; ///< band type run end points
224 float sf[120]; ///< scalefactors
225 DECLARE_ALIGNED_16(float, coeffs[1024]); ///< coefficients for IMDCT
226 DECLARE_ALIGNED_16(float, saved[512]); ///< overlap
227 DECLARE_ALIGNED_16(float, ret[1024]); ///< PCM output
228 PredictorState predictor_state[MAX_PREDICTORS];
229 } SingleChannelElement;
232 * channel element - generic struct for SCE/CPE/CCE/LFE
236 uint8_t ms_mask[120]; ///< Set if mid/side stereo is used for each scalefactor window band
238 SingleChannelElement ch[2];
240 ChannelCoupling coup;
247 AVCodecContext * avccontext;
249 MPEG4AudioConfig m4ac;
251 int is_saved; ///< Set if elements have stored overlap from previous frame.
252 DynamicRangeControl che_drc;
258 enum ChannelPosition che_pos[4][MAX_ELEM_ID]; /**< channel element channel mapping with the
259 * first index as the first 4 raw data block types
261 ChannelElement * che[4][MAX_ELEM_ID];
265 * @defgroup temporary aligned temporary buffers (We do not want to have these on the stack.)
268 DECLARE_ALIGNED_16(float, buf_mdct[1024]);
272 * @defgroup tables Computed / set up during initialization.
276 MDCTContext mdct_small;
282 * @defgroup output Members used for output interleaving.
285 float *output_data[MAX_CHANNELS]; ///< Points to each element's 'ret' buffer (PCM output).
286 float add_bias; ///< offset for dsp.float_to_int16
287 float sf_scale; ///< Pre-scale for correct IMDCT and dsp.float_to_int16.
288 int sf_offset; ///< offset into pow2sf_tab as appropriate for dsp.float_to_int16
291 DECLARE_ALIGNED(16, float, temp[128]);
294 #endif /* AVCODEC_AAC_H */