blob: 5c9c37727e7df960def1b128a1099288c469975e
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 FFmpeg. |
6 | * |
7 | * FFmpeg is free software; you can redistribute it and/or |
8 | * modify it under the terms of the GNU Lesser General Public |
9 | * 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 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
15 | * Lesser General Public License for more details. |
16 | * |
17 | * 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 | |
22 | /** |
23 | * @file |
24 | * Common code between the AC-3 encoder and decoder. |
25 | */ |
26 | |
27 | #ifndef AVCODEC_AC3_H |
28 | #define AVCODEC_AC3_H |
29 | |
30 | #define AC3_MAX_CODED_FRAME_SIZE 3840 /* in bytes */ |
31 | #define AC3_MAX_CHANNELS 7 /**< maximum number of channels, including coupling channel */ |
32 | #define CPL_CH 0 /**< coupling channel index */ |
33 | |
34 | #define AC3_MAX_COEFS 256 |
35 | #define AC3_BLOCK_SIZE 256 |
36 | #define AC3_MAX_BLOCKS 6 |
37 | #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 | #include "libavutil/opt.h" |
43 | #include "avcodec.h" |
44 | #include "ac3tab.h" |
45 | |
46 | /* exponent encoding strategy */ |
47 | #define EXP_REUSE 0 |
48 | #define EXP_NEW 1 |
49 | |
50 | #define EXP_D15 1 |
51 | #define EXP_D25 2 |
52 | #define EXP_D45 3 |
53 | |
54 | #ifndef USE_FIXED |
55 | #define USE_FIXED 0 |
56 | #endif |
57 | |
58 | #if USE_FIXED |
59 | |
60 | #define FFT_FLOAT 0 |
61 | |
62 | #define FIXR(a) ((int)((a) * 0 + 0.5)) |
63 | #define FIXR12(a) ((int)((a) * 4096 + 0.5)) |
64 | #define FIXR15(a) ((int)((a) * 32768 + 0.5)) |
65 | #define ROUND15(x) ((x) + 16384) >> 15 |
66 | |
67 | #define AC3_RENAME(x) x ## _fixed |
68 | #define AC3_NORM(norm) (1<<24)/(norm) |
69 | #define AC3_MUL(a,b) ((((int64_t) (a)) * (b))>>12) |
70 | #define AC3_RANGE(x) ((x)|(((x)&128)<<1)) |
71 | #define AC3_HEAVY_RANGE(x) ((x)<<1) |
72 | #define AC3_DYNAMIC_RANGE(x) (x) |
73 | #define AC3_SPX_BLEND(x) (x) |
74 | #define AC3_DYNAMIC_RANGE1 0 |
75 | |
76 | typedef int INTFLOAT; |
77 | typedef int16_t SHORTFLOAT; |
78 | |
79 | #else /* USE_FIXED */ |
80 | |
81 | #define FIXR(x) ((float)(x)) |
82 | #define FIXR12(x) ((float)(x)) |
83 | #define FIXR15(x) ((float)(x)) |
84 | #define ROUND15(x) (x) |
85 | |
86 | #define AC3_RENAME(x) x |
87 | #define AC3_NORM(norm) (1.0f/(norm)) |
88 | #define AC3_MUL(a,b) ((a) * (b)) |
89 | #define AC3_RANGE(x) (dynamic_range_tab[(x)]) |
90 | #define AC3_HEAVY_RANGE(x) (ff_ac3_heavy_dynamic_range_tab[(x)]) |
91 | #define AC3_DYNAMIC_RANGE(x) (powf(x, s->drc_scale)) |
92 | #define AC3_SPX_BLEND(x) (x)* (1.0f/32) |
93 | #define AC3_DYNAMIC_RANGE1 1.0f |
94 | |
95 | typedef float INTFLOAT; |
96 | typedef float SHORTFLOAT; |
97 | |
98 | #endif /* USE_FIXED */ |
99 | |
100 | #define AC3_LEVEL(x) ROUND15((x) * FIXR15(M_SQRT1_2)) |
101 | |
102 | /* pre-defined gain values */ |
103 | #define LEVEL_PLUS_3DB M_SQRT2 |
104 | #define LEVEL_PLUS_1POINT5DB 1.1892071150027209 |
105 | #define LEVEL_MINUS_1POINT5DB 0.8408964152537145 |
106 | #define LEVEL_MINUS_3DB M_SQRT1_2 |
107 | #define LEVEL_MINUS_4POINT5DB 0.5946035575013605 |
108 | #define LEVEL_MINUS_6DB 0.5000000000000000 |
109 | #define LEVEL_MINUS_9DB 0.3535533905932738 |
110 | #define LEVEL_ZERO 0.0000000000000000 |
111 | #define LEVEL_ONE 1.0000000000000000 |
112 | |
113 | /** Delta bit allocation strategy */ |
114 | typedef enum { |
115 | DBA_REUSE = 0, |
116 | DBA_NEW, |
117 | DBA_NONE, |
118 | DBA_RESERVED |
119 | } AC3DeltaStrategy; |
120 | |
121 | /** Channel mode (audio coding mode) */ |
122 | typedef enum { |
123 | AC3_CHMODE_DUALMONO = 0, |
124 | AC3_CHMODE_MONO, |
125 | AC3_CHMODE_STEREO, |
126 | AC3_CHMODE_3F, |
127 | AC3_CHMODE_2F1R, |
128 | AC3_CHMODE_3F1R, |
129 | AC3_CHMODE_2F2R, |
130 | AC3_CHMODE_3F2R |
131 | } AC3ChannelMode; |
132 | |
133 | /** Dolby Surround mode */ |
134 | typedef enum AC3DolbySurroundMode { |
135 | AC3_DSURMOD_NOTINDICATED = 0, |
136 | AC3_DSURMOD_OFF, |
137 | AC3_DSURMOD_ON, |
138 | AC3_DSURMOD_RESERVED |
139 | } AC3DolbySurroundMode; |
140 | |
141 | /** Dolby Surround EX mode */ |
142 | typedef enum AC3DolbySurroundEXMode { |
143 | AC3_DSUREXMOD_NOTINDICATED = 0, |
144 | AC3_DSUREXMOD_OFF, |
145 | AC3_DSUREXMOD_ON, |
146 | AC3_DSUREXMOD_PLIIZ |
147 | } AC3DolbySurroundEXMode; |
148 | |
149 | /** Dolby Headphone mode */ |
150 | typedef enum AC3DolbyHeadphoneMode { |
151 | AC3_DHEADPHONMOD_NOTINDICATED = 0, |
152 | AC3_DHEADPHONMOD_OFF, |
153 | AC3_DHEADPHONMOD_ON, |
154 | AC3_DHEADPHONMOD_RESERVED |
155 | } AC3DolbyHeadphoneMode; |
156 | |
157 | /** Preferred Stereo Downmix mode */ |
158 | typedef enum AC3PreferredStereoDownmixMode { |
159 | AC3_DMIXMOD_NOTINDICATED = 0, |
160 | AC3_DMIXMOD_LTRT, |
161 | AC3_DMIXMOD_LORO, |
162 | AC3_DMIXMOD_DPLII // reserved value in A/52, but used by encoders to indicate DPL2 |
163 | } AC3PreferredStereoDownmixMode; |
164 | |
165 | typedef struct AC3BitAllocParameters { |
166 | int sr_code; |
167 | int sr_shift; |
168 | int slow_gain, slow_decay, fast_decay, db_per_bit, floor; |
169 | int cpl_fast_leak, cpl_slow_leak; |
170 | } AC3BitAllocParameters; |
171 | |
172 | /** |
173 | * @struct AC3HeaderInfo |
174 | * Coded AC-3 header values up to the lfeon element, plus derived values. |
175 | */ |
176 | typedef struct AC3HeaderInfo { |
177 | /** @name Coded elements |
178 | * @{ |
179 | */ |
180 | uint16_t sync_word; |
181 | uint16_t crc1; |
182 | uint8_t sr_code; |
183 | uint8_t bitstream_id; |
184 | uint8_t bitstream_mode; |
185 | uint8_t channel_mode; |
186 | uint8_t lfe_on; |
187 | uint8_t frame_type; |
188 | int substreamid; ///< substream identification |
189 | int center_mix_level; ///< Center mix level index |
190 | int surround_mix_level; ///< Surround mix level index |
191 | uint16_t channel_map; |
192 | int num_blocks; ///< number of audio blocks |
193 | int dolby_surround_mode; |
194 | /** @} */ |
195 | |
196 | /** @name Derived values |
197 | * @{ |
198 | */ |
199 | uint8_t sr_shift; |
200 | uint16_t sample_rate; |
201 | uint32_t bit_rate; |
202 | uint8_t channels; |
203 | uint16_t frame_size; |
204 | uint64_t channel_layout; |
205 | /** @} */ |
206 | } AC3HeaderInfo; |
207 | |
208 | typedef enum { |
209 | EAC3_FRAME_TYPE_INDEPENDENT = 0, |
210 | EAC3_FRAME_TYPE_DEPENDENT, |
211 | EAC3_FRAME_TYPE_AC3_CONVERT, |
212 | EAC3_FRAME_TYPE_RESERVED |
213 | } EAC3FrameType; |
214 | |
215 | void ff_ac3_common_init(void); |
216 | |
217 | /** |
218 | * Calculate the log power-spectral density of the input signal. |
219 | * This gives a rough estimate of signal power in the frequency domain by using |
220 | * the spectral envelope (exponents). The psd is also separately grouped |
221 | * into critical bands for use in the calculating the masking curve. |
222 | * 128 units in psd = -6 dB. The dbknee parameter in AC3BitAllocParameters |
223 | * determines the reference level. |
224 | * |
225 | * @param[in] exp frequency coefficient exponents |
226 | * @param[in] start starting bin location |
227 | * @param[in] end ending bin location |
228 | * @param[out] psd signal power for each frequency bin |
229 | * @param[out] band_psd signal power for each critical band |
230 | */ |
231 | void ff_ac3_bit_alloc_calc_psd(int8_t *exp, int start, int end, int16_t *psd, |
232 | int16_t *band_psd); |
233 | |
234 | /** |
235 | * Calculate the masking curve. |
236 | * First, the excitation is calculated using parameters in s and the signal |
237 | * power in each critical band. The excitation is compared with a predefined |
238 | * hearing threshold table to produce the masking curve. If delta bit |
239 | * allocation information is provided, it is used for adjusting the masking |
240 | * curve, usually to give a closer match to a better psychoacoustic model. |
241 | * |
242 | * @param[in] s adjustable bit allocation parameters |
243 | * @param[in] band_psd signal power for each critical band |
244 | * @param[in] start starting bin location |
245 | * @param[in] end ending bin location |
246 | * @param[in] fast_gain fast gain (estimated signal-to-mask ratio) |
247 | * @param[in] is_lfe whether or not the channel being processed is the LFE |
248 | * @param[in] dba_mode delta bit allocation mode (none, reuse, or new) |
249 | * @param[in] dba_nsegs number of delta segments |
250 | * @param[in] dba_offsets location offsets for each segment |
251 | * @param[in] dba_lengths length of each segment |
252 | * @param[in] dba_values delta bit allocation for each segment |
253 | * @param[out] mask calculated masking curve |
254 | * @return returns 0 for success, non-zero for error |
255 | */ |
256 | int ff_ac3_bit_alloc_calc_mask(AC3BitAllocParameters *s, int16_t *band_psd, |
257 | int start, int end, int fast_gain, int is_lfe, |
258 | int dba_mode, int dba_nsegs, uint8_t *dba_offsets, |
259 | uint8_t *dba_lengths, uint8_t *dba_values, |
260 | int16_t *mask); |
261 | |
262 | #endif /* AVCODEC_AC3_H */ |
263 |