ac3.h 6.96 KB
Newer Older
1
/*
2
 * Common code between the AC-3 encoder and decoder
3
 * Copyright (c) 2000, 2001, 2002 Fabrice Bellard
4
 *
5
 * This file is part of Libav.
6
 *
7
 * Libav is free software; you can redistribute it and/or
8 9
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
10
 * version 2.1 of the License, or (at your option) any later version.
11
 *
12
 * Libav is distributed in the hope that it will be useful,
13 14 15 16 17
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
18
 * License along with Libav; if not, write to the Free Software
19
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 21
 */

Michael Niedermayer's avatar
Michael Niedermayer committed
22
/**
23
 * @file
24
 * Common code between the AC-3 encoder and decoder.
Michael Niedermayer's avatar
Michael Niedermayer committed
25 26
 */

27 28
#ifndef AVCODEC_AC3_H
#define AVCODEC_AC3_H
29

30
#define AC3_MAX_CODED_FRAME_SIZE 3840 /* in bytes */
31 32
#define AC3_MAX_CHANNELS 7            /**< maximum number of channels, including coupling channel */
#define CPL_CH 0                      /**< coupling channel index */
33

34 35 36 37
#define AC3_MAX_COEFS   256
#define AC3_BLOCK_SIZE  256
#define AC3_MAX_BLOCKS    6
#define AC3_FRAME_SIZE (AC3_MAX_BLOCKS * 256)
38
#define AC3_WINDOW_SIZE (AC3_BLOCK_SIZE * 2)
39
#define AC3_CRITICAL_BANDS 50
40
#define AC3_MAX_CPL_BANDS  18
41

42 43
#include "ac3tab.h"

44 45 46 47 48 49 50 51
/* exponent encoding strategy */
#define EXP_REUSE 0
#define EXP_NEW   1

#define EXP_D15   1
#define EXP_D25   2
#define EXP_D45   3

52 53 54 55 56 57 58 59 60 61 62
/* pre-defined gain values */
#define LEVEL_PLUS_3DB          1.4142135623730950
#define LEVEL_PLUS_1POINT5DB    1.1892071150027209
#define LEVEL_MINUS_1POINT5DB   0.8408964152537145
#define LEVEL_MINUS_3DB         0.7071067811865476
#define LEVEL_MINUS_4POINT5DB   0.5946035575013605
#define LEVEL_MINUS_6DB         0.5000000000000000
#define LEVEL_MINUS_9DB         0.3535533905932738
#define LEVEL_ZERO              0.0000000000000000
#define LEVEL_ONE               1.0000000000000000

63 64 65 66 67 68 69 70 71 72
/** Delta bit allocation strategy */
typedef enum {
    DBA_REUSE = 0,
    DBA_NEW,
    DBA_NONE,
    DBA_RESERVED
} AC3DeltaStrategy;

/** Channel mode (audio coding mode) */
typedef enum {
73 74 75 76 77 78 79 80
    AC3_CHMODE_DUALMONO = 0,
    AC3_CHMODE_MONO,
    AC3_CHMODE_STEREO,
    AC3_CHMODE_3F,
    AC3_CHMODE_2F1R,
    AC3_CHMODE_3F1R,
    AC3_CHMODE_2F2R,
    AC3_CHMODE_3F2R
81 82
} AC3ChannelMode;

83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106
/** Dolby Surround mode */
typedef enum AC3DolbySurroundMode {
    AC3_DSURMOD_NOTINDICATED = 0,
    AC3_DSURMOD_OFF,
    AC3_DSURMOD_ON,
    AC3_DSURMOD_RESERVED
} AC3DolbySurroundMode;

/** Dolby Surround EX mode */
typedef enum AC3DolbySurroundEXMode {
    AC3_DSUREXMOD_NOTINDICATED = 0,
    AC3_DSUREXMOD_OFF,
    AC3_DSUREXMOD_ON,
    AC3_DSUREXMOD_PLIIZ
} AC3DolbySurroundEXMode;

/** Dolby Headphone mode */
typedef enum AC3DolbyHeadphoneMode {
    AC3_DHEADPHONMOD_NOTINDICATED = 0,
    AC3_DHEADPHONMOD_OFF,
    AC3_DHEADPHONMOD_ON,
    AC3_DHEADPHONMOD_RESERVED
} AC3DolbyHeadphoneMode;

107 108 109 110 111
/** Preferred Stereo Downmix mode */
typedef enum AC3PreferredStereoDownmixMode {
    AC3_DMIXMOD_NOTINDICATED = 0,
    AC3_DMIXMOD_LTRT,
    AC3_DMIXMOD_LORO,
112
    AC3_DMIXMOD_DPLII // reserved value in A/52, but used by encoders to indicate DPL2
113 114
} AC3PreferredStereoDownmixMode;

115
typedef struct AC3BitAllocParameters {
116 117 118 119
    int sr_code;
    int sr_shift;
    int slow_gain, slow_decay, fast_decay, db_per_bit, floor;
    int cpl_fast_leak, cpl_slow_leak;
120 121
} AC3BitAllocParameters;

122 123 124 125
/**
 * @struct AC3HeaderInfo
 * Coded AC-3 header values up to the lfeon element, plus derived values.
 */
126
typedef struct AC3HeaderInfo {
127
    /** @name Coded elements
128 129 130 131
     * @{
     */
    uint16_t sync_word;
    uint16_t crc1;
132
    uint8_t sr_code;
133
    uint8_t bitstream_id;
134
    uint8_t bitstream_mode;
135 136
    uint8_t channel_mode;
    uint8_t lfe_on;
137
    uint8_t frame_type;
138
    int substreamid;                        ///< substream identification
139 140 141
    int center_mix_level;                   ///< Center mix level index
    int surround_mix_level;                 ///< Surround mix level index
    uint16_t channel_map;
142
    int num_blocks;                         ///< number of audio blocks
143
    int dolby_surround_mode;
144 145
    /** @} */

146
    /** @name Derived values
147 148
     * @{
     */
149
    uint8_t sr_shift;
150 151 152 153
    uint16_t sample_rate;
    uint32_t bit_rate;
    uint8_t channels;
    uint16_t frame_size;
154
    uint64_t channel_layout;
155 156 157
    /** @} */
} AC3HeaderInfo;

158
typedef enum {
159 160 161 162 163
    EAC3_FRAME_TYPE_INDEPENDENT = 0,
    EAC3_FRAME_TYPE_DEPENDENT,
    EAC3_FRAME_TYPE_AC3_CONVERT,
    EAC3_FRAME_TYPE_RESERVED
} EAC3FrameType;
164

165
void ff_ac3_common_init(void);
166 167

/**
168
 * Calculate the log power-spectral density of the input signal.
169 170 171 172 173 174 175 176 177 178
 * This gives a rough estimate of signal power in the frequency domain by using
 * the spectral envelope (exponents).  The psd is also separately grouped
 * into critical bands for use in the calculating the masking curve.
 * 128 units in psd = -6 dB.  The dbknee parameter in AC3BitAllocParameters
 * determines the reference level.
 *
 * @param[in]  exp        frequency coefficient exponents
 * @param[in]  start      starting bin location
 * @param[in]  end        ending bin location
 * @param[out] psd        signal power for each frequency bin
179
 * @param[out] band_psd   signal power for each critical band
180 181
 */
void ff_ac3_bit_alloc_calc_psd(int8_t *exp, int start, int end, int16_t *psd,
182
                               int16_t *band_psd);
183 184

/**
185
 * Calculate the masking curve.
186
 * First, the excitation is calculated using parameters in s and the signal
187 188 189 190 191
 * power in each critical band.  The excitation is compared with a predefined
 * hearing threshold table to produce the masking curve.  If delta bit
 * allocation information is provided, it is used for adjusting the masking
 * curve, usually to give a closer match to a better psychoacoustic model.
 *
192
 * @param[in]  s            adjustable bit allocation parameters
193
 * @param[in]  band_psd     signal power for each critical band
194 195 196 197
 * @param[in]  start        starting bin location
 * @param[in]  end          ending bin location
 * @param[in]  fast_gain    fast gain (estimated signal-to-mask ratio)
 * @param[in]  is_lfe       whether or not the channel being processed is the LFE
198 199 200 201 202
 * @param[in]  dba_mode     delta bit allocation mode (none, reuse, or new)
 * @param[in]  dba_nsegs    number of delta segments
 * @param[in]  dba_offsets  location offsets for each segment
 * @param[in]  dba_lengths  length of each segment
 * @param[in]  dba_values   delta bit allocation for each segment
203
 * @param[out] mask         calculated masking curve
204
 * @return returns 0 for success, non-zero for error
205
 */
206
int ff_ac3_bit_alloc_calc_mask(AC3BitAllocParameters *s, int16_t *band_psd,
207 208 209 210
                               int start, int end, int fast_gain, int is_lfe,
                               int dba_mode, int dba_nsegs, uint8_t *dba_offsets,
                               uint8_t *dba_lengths, uint8_t *dba_values,
                               int16_t *mask);
211

212
#endif /* AVCODEC_AC3_H */