Statistics
| Branch: | Revision:

ffmpeg / libavcodec / aac.h @ 84dc2d8a

History | View | Annotate | Download (9.85 KB)

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 "avcodec.h"
34
#include "dsputil.h"
35
#include "mpeg4audio.h"
36

    
37
#include <stdint.h>
38

    
39
#define AAC_INIT_VLC_STATIC(num, size) \
40
    INIT_VLC_STATIC(&vlc_spectral[num], 8, 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]), \
43
        size);
44

    
45
#define MAX_CHANNELS 64
46
#define MAX_ELEM_ID 16
47

    
48
#define TNS_MAX_ORDER 20
49

    
50
enum RawDataBlockType {
51
    TYPE_SCE,
52
    TYPE_CPE,
53
    TYPE_CCE,
54
    TYPE_LFE,
55
    TYPE_DSE,
56
    TYPE_PCE,
57
    TYPE_FIL,
58
    TYPE_END,
59
};
60

    
61
enum ExtensionPayloadID {
62
    EXT_FILL,
63
    EXT_FILL_DATA,
64
    EXT_DATA_ELEMENT,
65
    EXT_DYNAMIC_RANGE = 0xb,
66
    EXT_SBR_DATA      = 0xd,
67
    EXT_SBR_DATA_CRC  = 0xe,
68
};
69

    
70
enum WindowSequence {
71
    ONLY_LONG_SEQUENCE,
72
    LONG_START_SEQUENCE,
73
    EIGHT_SHORT_SEQUENCE,
74
    LONG_STOP_SEQUENCE,
75
};
76

    
77
enum BandType {
78
    ZERO_BT        = 0,     ///< Scalefactors and spectral data are all zero.
79
    FIRST_PAIR_BT  = 5,     ///< This and later band types encode two values (rather than four) with one code word.
80
    ESC_BT         = 11,    ///< Spectral data are coded with an escape sequence.
81
    NOISE_BT       = 13,    ///< Spectral data are scaled white noise not coded in the bitstream.
82
    INTENSITY_BT2  = 14,    ///< Scalefactor data are intensity stereo positions.
83
    INTENSITY_BT   = 15,    ///< Scalefactor data are intensity stereo positions.
84
};
85

    
86
#define IS_CODEBOOK_UNSIGNED(x) ((x - 1) & 10)
87

    
88
enum ChannelPosition {
89
    AAC_CHANNEL_FRONT = 1,
90
    AAC_CHANNEL_SIDE  = 2,
91
    AAC_CHANNEL_BACK  = 3,
92
    AAC_CHANNEL_LFE   = 4,
93
    AAC_CHANNEL_CC    = 5,
94
};
95

    
96
/**
97
 * The point during decoding at which channel coupling is applied.
98
 */
99
enum CouplingPoint {
100
    BEFORE_TNS,
101
    BETWEEN_TNS_AND_IMDCT,
102
    AFTER_IMDCT = 3,
103
};
104

    
105
/**
106
 * Output configuration status
107
 */
108
enum OCStatus {
109
    OC_NONE,        //< Output unconfigured
110
    OC_TRIAL_PCE,   //< Output configuration under trial specified by an inband PCE
111
    OC_TRIAL_FRAME, //< Output configuration under trial specified by a frame header
112
    OC_GLOBAL_HDR,  //< Output configuration set in a global header but not yet locked
113
    OC_LOCKED,      //< Output configuration locked in place
114
};
115

    
116
/**
117
 * Predictor State
118
 */
119
typedef struct {
120
    float cor0;
121
    float cor1;
122
    float var0;
123
    float var1;
124
    float r0;
125
    float r1;
126
} PredictorState;
127

    
128
#define MAX_PREDICTORS 672
129

    
130
#define SCALE_DIV_512    36    ///< scalefactor difference that corresponds to scale difference in 512 times
131
#define SCALE_ONE_POS   140    ///< scalefactor index that corresponds to scale=1.0
132
#define SCALE_MAX_POS   255    ///< scalefactor index maximum value
133
#define SCALE_MAX_DIFF   60    ///< maximum scalefactor difference allowed by standard
134
#define SCALE_DIFF_ZERO  60    ///< codebook index corresponding to zero scalefactor indices difference
135

    
136
/**
137
 * Individual Channel Stream
138
 */
139
typedef struct {
140
    uint8_t max_sfb;            ///< number of scalefactor bands per group
141
    enum WindowSequence window_sequence[2];
142
    uint8_t use_kb_window[2];   ///< If set, use Kaiser-Bessel window, otherwise use a sinus window.
143
    int num_window_groups;
144
    uint8_t group_len[8];
145
    const uint16_t *swb_offset; ///< table of offsets to the lowest spectral coefficient of a scalefactor band, sfb, for a particular window
146
    const uint8_t *swb_sizes;   ///< table of scalefactor band sizes for a particular window
147
    int num_swb;                ///< number of scalefactor window bands
148
    int num_windows;
149
    int tns_max_bands;
150
    int predictor_present;
151
    int predictor_initialized;
152
    int predictor_reset_group;
153
    uint8_t prediction_used[41];
154
} IndividualChannelStream;
155

    
156
/**
157
 * Temporal Noise Shaping
158
 */
159
typedef struct {
160
    int present;
161
    int n_filt[8];
162
    int length[8][4];
163
    int direction[8][4];
164
    int order[8][4];
165
    float coef[8][4][TNS_MAX_ORDER];
166
} TemporalNoiseShaping;
167

    
168
/**
169
 * Dynamic Range Control - decoded from the bitstream but not processed further.
170
 */
171
typedef struct {
172
    int pce_instance_tag;                           ///< Indicates with which program the DRC info is associated.
173
    int dyn_rng_sgn[17];                            ///< DRC sign information; 0 - positive, 1 - negative
174
    int dyn_rng_ctl[17];                            ///< DRC magnitude information
175
    int exclude_mask[MAX_CHANNELS];                 ///< Channels to be excluded from DRC processing.
176
    int band_incr;                                  ///< Number of DRC bands greater than 1 having DRC info.
177
    int interpolation_scheme;                       ///< Indicates the interpolation scheme used in the SBR QMF domain.
178
    int band_top[17];                               ///< Indicates the top of the i-th DRC band in units of 4 spectral lines.
179
    int prog_ref_level;                             /**< A reference level for the long-term program audio level for all
180
                                                     *   channels combined.
181
                                                     */
182
} DynamicRangeControl;
183

    
184
typedef struct {
185
    int num_pulse;
186
    int start;
187
    int pos[4];
188
    int amp[4];
189
} Pulse;
190

    
191
/**
192
 * coupling parameters
193
 */
194
typedef struct {
195
    enum CouplingPoint coupling_point;  ///< The point during decoding at which coupling is applied.
196
    int num_coupled;       ///< number of target elements
197
    enum RawDataBlockType type[8];   ///< Type of channel element to be coupled - SCE or CPE.
198
    int id_select[8];      ///< element id
199
    int ch_select[8];      /**< [0] shared list of gains; [1] list of gains for right channel;
200
                            *   [2] list of gains for left channel; [3] lists of gains for both channels
201
                            */
202
    float gain[16][120];
203
} ChannelCoupling;
204

    
205
/**
206
 * Single Channel Element - used for both SCE and LFE elements.
207
 */
208
typedef struct {
209
    IndividualChannelStream ics;
210
    TemporalNoiseShaping tns;
211
    Pulse pulse;
212
    enum BandType band_type[128];             ///< band types
213
    int band_type_run_end[120];               ///< band type run end points
214
    float sf[120];                            ///< scalefactors
215
    int sf_idx[128];                          ///< scalefactor indices (used by encoder)
216
    uint8_t zeroes[128];                      ///< band is not coded (used by encoder)
217
    DECLARE_ALIGNED(16, float, coeffs)[1024]; ///< coefficients for IMDCT
218
    DECLARE_ALIGNED(16, float, saved)[1024];  ///< overlap
219
    DECLARE_ALIGNED(16, float, ret)[1024];    ///< PCM output
220
    PredictorState predictor_state[MAX_PREDICTORS];
221
} SingleChannelElement;
222

    
223
/**
224
 * channel element - generic struct for SCE/CPE/CCE/LFE
225
 */
226
typedef struct {
227
    // CPE specific
228
    int common_window;        ///< Set if channels share a common 'IndividualChannelStream' in bitstream.
229
    int     ms_mode;          ///< Signals mid/side stereo flags coding mode (used by encoder)
230
    uint8_t ms_mask[128];     ///< Set if mid/side stereo is used for each scalefactor window band
231
    // shared
232
    SingleChannelElement ch[2];
233
    // CCE specific
234
    ChannelCoupling coup;
235
} ChannelElement;
236

    
237
/**
238
 * main AAC context
239
 */
240
typedef struct {
241
    AVCodecContext * avccontext;
242

    
243
    MPEG4AudioConfig m4ac;
244

    
245
    int is_saved;                 ///< Set if elements have stored overlap from previous frame.
246
    DynamicRangeControl che_drc;
247

    
248
    /**
249
     * @defgroup elements Channel element related data.
250
     * @{
251
     */
252
    enum ChannelPosition che_pos[4][MAX_ELEM_ID]; /**< channel element channel mapping with the
253
                                                   *   first index as the first 4 raw data block types
254
                                                   */
255
    ChannelElement * che[4][MAX_ELEM_ID];
256
    ChannelElement * tag_che_map[4][MAX_ELEM_ID];
257
    int tags_mapped;
258
    /** @} */
259

    
260
    /**
261
     * @defgroup temporary aligned temporary buffers (We do not want to have these on the stack.)
262
     * @{
263
     */
264
    DECLARE_ALIGNED(16, float, buf_mdct)[1024];
265
    /** @} */
266

    
267
    /**
268
     * @defgroup tables   Computed / set up during initialization.
269
     * @{
270
     */
271
    FFTContext mdct;
272
    FFTContext mdct_small;
273
    DSPContext dsp;
274
    int random_state;
275
    /** @} */
276

    
277
    /**
278
     * @defgroup output   Members used for output interleaving.
279
     * @{
280
     */
281
    float *output_data[MAX_CHANNELS];                 ///< Points to each element's 'ret' buffer (PCM output).
282
    float add_bias;                                   ///< offset for dsp.float_to_int16
283
    float sf_scale;                                   ///< Pre-scale for correct IMDCT and dsp.float_to_int16.
284
    int sf_offset;                                    ///< offset into pow2sf_tab as appropriate for dsp.float_to_int16
285
    /** @} */
286

    
287
    DECLARE_ALIGNED(16, float, temp)[128];
288

    
289
    enum OCStatus output_configured;
290
} AACContext;
291

    
292
#endif /* AVCODEC_AAC_H */