ac3.h 8.46 KB
Newer Older
1
/*
2
 * Common code between the AC-3 encoder and decoder
3
 * Copyright (c) 2000, 2001, 2002 Fabrice Bellard
4
 *
5 6 7
 * This file is part of FFmpeg.
 *
 * FFmpeg 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
 * FFmpeg 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 FFmpeg; 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 "libavutil/opt.h"
#include "avcodec.h"
44 45
#include "ac3tab.h"

46 47 48 49 50 51 52 53
/* exponent encoding strategy */
#define EXP_REUSE 0
#define EXP_NEW   1

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

54 55
#ifndef USE_FIXED
#define USE_FIXED 0
56 57
#endif

58
#if USE_FIXED
59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77

#define FFT_FLOAT 0

#define FIXR(a)                 ((int)((a) * 0 + 0.5))
#define FIXR12(a)               ((int)((a) * 4096 + 0.5))
#define FIXR15(a)               ((int)((a) * 32768 + 0.5))
#define ROUND15(x)              ((x) + 16384) >> 15

#define AC3_RENAME(x)           x ## _fixed
#define AC3_NORM(norm)          (1<<24)/(norm)
#define AC3_MUL(a,b)            ((((int64_t) (a)) * (b))>>12)
#define AC3_RANGE(x)            (x)
#define AC3_DYNAMIC_RANGE(x)    (x)
#define AC3_SPX_BLEND(x)        (x)
#define AC3_DYNAMIC_RANGE1      0

#define INTFLOAT                int
#define SHORTFLOAT              int16_t

78
#else /* USE_FIXED */
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95

#define FIXR(x)                 ((float)(x))
#define FIXR12(x)               ((float)(x))
#define FIXR15(x)               ((float)(x))
#define ROUND15(x)              (x)

#define AC3_RENAME(x)           x
#define AC3_NORM(norm)          (1.0f/(norm))
#define AC3_MUL(a,b)            ((a) * (b))
#define AC3_RANGE(x)            (dynamic_range_tab[(x)])
#define AC3_DYNAMIC_RANGE(x)    (powf(x,  s->drc_scale))
#define AC3_SPX_BLEND(x)        (x)* (1.0f/32)
#define AC3_DYNAMIC_RANGE1      1.0f

#define INTFLOAT                float
#define SHORTFLOAT              float

96
#endif /* USE_FIXED */
97 98 99

#define AC3_LEVEL(x)            ROUND15((x) * FIXR15(0.7071067811865476))

100 101 102 103 104 105 106 107 108 109 110
/* 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

111 112 113 114 115 116 117 118 119 120
/** Delta bit allocation strategy */
typedef enum {
    DBA_REUSE = 0,
    DBA_NEW,
    DBA_NONE,
    DBA_RESERVED
} AC3DeltaStrategy;

/** Channel mode (audio coding mode) */
typedef enum {
121 122 123 124 125 126 127 128
    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
129 130
} AC3ChannelMode;

131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
/** 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;

155 156 157 158 159
/** Preferred Stereo Downmix mode */
typedef enum AC3PreferredStereoDownmixMode {
    AC3_DMIXMOD_NOTINDICATED = 0,
    AC3_DMIXMOD_LTRT,
    AC3_DMIXMOD_LORO,
160
    AC3_DMIXMOD_DPLII // reserved value in A/52, but used by encoders to indicate DPL2
161 162
} AC3PreferredStereoDownmixMode;

163
typedef struct AC3BitAllocParameters {
164 165 166 167
    int sr_code;
    int sr_shift;
    int slow_gain, slow_decay, fast_decay, db_per_bit, floor;
    int cpl_fast_leak, cpl_slow_leak;
168 169
} AC3BitAllocParameters;

170 171 172 173
/**
 * @struct AC3HeaderInfo
 * Coded AC-3 header values up to the lfeon element, plus derived values.
 */
174
typedef struct AC3HeaderInfo {
175
    /** @name Coded elements
176 177 178 179
     * @{
     */
    uint16_t sync_word;
    uint16_t crc1;
180
    uint8_t sr_code;
181
    uint8_t bitstream_id;
182
    uint8_t bitstream_mode;
183 184
    uint8_t channel_mode;
    uint8_t lfe_on;
185
    uint8_t frame_type;
186
    int substreamid;                        ///< substream identification
187 188 189
    int center_mix_level;                   ///< Center mix level index
    int surround_mix_level;                 ///< Surround mix level index
    uint16_t channel_map;
190
    int num_blocks;                         ///< number of audio blocks
191
#if AV_HAVE_INCOMPATIBLE_LIBAV_ABI
192
    int dolby_surround_mode;
193
#endif
194 195
    /** @} */

196
    /** @name Derived values
197 198
     * @{
     */
199
    uint8_t sr_shift;
200 201 202 203
    uint16_t sample_rate;
    uint32_t bit_rate;
    uint8_t channels;
    uint16_t frame_size;
204
    uint64_t channel_layout;
205
    /** @} */
206 207 208
#if !AV_HAVE_INCOMPATIBLE_LIBAV_ABI
    int dolby_surround_mode;
#endif
209 210
} AC3HeaderInfo;

211
typedef enum {
212 213 214 215 216
    EAC3_FRAME_TYPE_INDEPENDENT = 0,
    EAC3_FRAME_TYPE_DEPENDENT,
    EAC3_FRAME_TYPE_AC3_CONVERT,
    EAC3_FRAME_TYPE_RESERVED
} EAC3FrameType;
217

218
void ff_ac3_common_init(void);
219 220

/**
221
 * Calculate the log power-spectral density of the input signal.
222 223 224 225 226 227 228 229 230 231
 * 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
232
 * @param[out] band_psd   signal power for each critical band
233 234
 */
void ff_ac3_bit_alloc_calc_psd(int8_t *exp, int start, int end, int16_t *psd,
235
                               int16_t *band_psd);
236 237

/**
238
 * Calculate the masking curve.
239
 * First, the excitation is calculated using parameters in s and the signal
240 241 242 243 244
 * 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.
 *
245
 * @param[in]  s            adjustable bit allocation parameters
246
 * @param[in]  band_psd     signal power for each critical band
247 248 249 250
 * @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
251 252 253 254 255
 * @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
256
 * @param[out] mask         calculated masking curve
257
 * @return returns 0 for success, non-zero for error
258
 */
259
int ff_ac3_bit_alloc_calc_mask(AC3BitAllocParameters *s, int16_t *band_psd,
260 261 262 263
                               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);
264

265
#endif /* AVCODEC_AC3_H */