libavcodec/ac3.h

   1 /*
   2  * Common code between AC3 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 ac3.h
  24  * Common code between AC3 encoder and decoder.
  25  */
  26
  27 #ifndef FFMPEG_AC3_H
  28 #define FFMPEG_AC3_H
  29
  30 #include "ac3tab.h"
  31
  32 #define AC3_MAX_CODED_FRAME_SIZE 3840 /* in bytes */
  33 #define AC3_MAX_CHANNELS 6 /* including LFE channel */
  34
  35 #define NB_BLOCKS 6 /* number of PCM blocks inside an AC3 frame */
  36 #define AC3_FRAME_SIZE (NB_BLOCKS * 256)
  37
  38 /* exponent encoding strategy */
  39 #define EXP_REUSE 0
  40 #define EXP_NEW   1
  41
  42 #define EXP_D15   1
  43 #define EXP_D25   2
  44 #define EXP_D45   3
  45
  46 /** Delta bit allocation strategy */
  47 typedef enum {
  48     DBA_REUSE = 0,
  49     DBA_NEW,
  50     DBA_NONE,
  51     DBA_RESERVED
  52 } AC3DeltaStrategy;
  53
  54 /** Channel mode (audio coding mode) */
  55 typedef enum {
  56     AC3_CHMODE_DUALMONO = 0,
  57     AC3_CHMODE_MONO,
  58     AC3_CHMODE_STEREO,
  59     AC3_CHMODE_3F,
  60     AC3_CHMODE_2F1R,
  61     AC3_CHMODE_3F1R,
  62     AC3_CHMODE_2F2R,
  63     AC3_CHMODE_3F2R
  64 } AC3ChannelMode;
  65
  66 typedef struct AC3BitAllocParameters {
  67     int sr_code;
  68     int sr_shift;
  69     int slow_gain, slow_decay, fast_decay, db_per_bit, floor;
  70     int cpl_fast_leak, cpl_slow_leak;
  71 } AC3BitAllocParameters;
  72
  73 /**
  74  * @struct AC3HeaderInfo
  75  * Coded AC-3 header values up to the lfeon element, plus derived values.
  76  */
  77 typedef struct {
  78     /** @defgroup coded Coded elements
  79      * @{
  80      */
  81     uint16_t sync_word;
  82     uint16_t crc1;
  83     uint8_t sr_code;
  84     uint8_t bitstream_id;
  85     uint8_t channel_mode;
  86     uint8_t lfe_on;
  87     uint8_t frame_type;
  88     int substreamid;                        ///< substream identification
  89     int center_mix_level;                   ///< Center mix level index
  90     int surround_mix_level;                 ///< Surround mix level index
  91     uint16_t channel_map;
  92     int num_blocks;                         ///< number of audio blocks
  93     /** @} */
  94
  95     /** @defgroup derived Derived values
  96      * @{
  97      */
  98     uint8_t sr_shift;
  99     uint16_t sample_rate;
 100     uint32_t bit_rate;
 101     uint8_t channels;
 102     uint16_t frame_size;
 103     /** @} */
 104 } AC3HeaderInfo;
 105
 106 typedef enum {
 107     EAC3_FRAME_TYPE_INDEPENDENT = 0,
 108     EAC3_FRAME_TYPE_DEPENDENT,
 109     EAC3_FRAME_TYPE_AC3_CONVERT,
 110     EAC3_FRAME_TYPE_RESERVED
 111 } EAC3FrameType;
 112
 113 void ac3_common_init(void);
 114
 115 /**
 116  * Calculates the log power-spectral density of the input signal.
 117  * This gives a rough estimate of signal power in the frequency domain by using
 118  * the spectral envelope (exponents).  The psd is also separately grouped
 119  * into critical bands for use in the calculating the masking curve.
 120  * 128 units in psd = -6 dB.  The dbknee parameter in AC3BitAllocParameters
 121  * determines the reference level.
 122  *
 123  * @param[in]  exp        frequency coefficient exponents
 124  * @param[in]  start      starting bin location
 125  * @param[in]  end        ending bin location
 126  * @param[out] psd        signal power for each frequency bin
 127  * @param[out] band_psd   signal power for each critical band
 128  */
 129 void ff_ac3_bit_alloc_calc_psd(int8_t *exp, int start, int end, int16_t *psd,
 130                                int16_t *band_psd);
 131
 132 /**
 133  * Calculates the masking curve.
 134  * First, the excitation is calculated using parameters in \p s and the signal
 135  * power in each critical band.  The excitation is compared with a predefined
 136  * hearing threshold table to produce the masking curve.  If delta bit
 137  * allocation information is provided, it is used for adjusting the masking
 138  * curve, usually to give a closer match to a better psychoacoustic model.
 139  *
 140  * @param[in]  s            adjustable bit allocation parameters
 141  * @param[in]  band_psd     signal power for each critical band
 142  * @param[in]  start        starting bin location
 143  * @param[in]  end          ending bin location
 144  * @param[in]  fast_gain    fast gain (estimated signal-to-mask ratio)
 145  * @param[in]  is_lfe       whether or not the channel being processed is the LFE
 146  * @param[in]  dba_mode     delta bit allocation mode (none, reuse, or new)
 147  * @param[in]  dba_nsegs    number of delta segments
 148  * @param[in]  dba_offsets  location offsets for each segment
 149  * @param[in]  dba_lengths  length of each segment
 150  * @param[in]  dba_values   delta bit allocation for each segment
 151  * @param[out] mask         calculated masking curve
 152  */
 153 void ff_ac3_bit_alloc_calc_mask(AC3BitAllocParameters *s, int16_t *band_psd,
 154                                 int start, int end, int fast_gain, int is_lfe,
 155                                 int dba_mode, int dba_nsegs, uint8_t *dba_offsets,
 156                                 uint8_t *dba_lengths, uint8_t *dba_values,
 157                                 int16_t *mask);
 158
 159 /**
 160  * Calculates bit allocation pointers.
 161  * The SNR is the difference between the masking curve and the signal.  AC-3
 162  * uses this value for each frequency bin to allocate bits.  The \p snroffset
 163  * parameter is a global adjustment to the SNR for all bins.
 164  *
 165  * @param[in]  mask       masking curve
 166  * @param[in]  psd        signal power for each frequency bin
 167  * @param[in]  start      starting bin location
 168  * @param[in]  end        ending bin location
 169  * @param[in]  snr_offset SNR adjustment
 170  * @param[in]  floor      noise floor
 171  * @param[in]  bap_tab    look-up table for bit allocation pointers
 172  * @param[out] bap        bit allocation pointers
 173  */
 174 void ff_ac3_bit_alloc_calc_bap(int16_t *mask, int16_t *psd, int start, int end,
 175                                int snr_offset, int floor,
 176                                const uint8_t *bap_tab, uint8_t *bap);
 177
 178 void ac3_parametric_bit_allocation(AC3BitAllocParameters *s, uint8_t *bap,
 179                                    int8_t *exp, int start, int end,
 180                                    int snr_offset, int fast_gain, int is_lfe,
 181                                    int dba_mode, int dba_nsegs,
 182                                    uint8_t *dba_offsets, uint8_t *dba_lengths,
 183                                    uint8_t *dba_values);
 184
 185 #endif /* FFMPEG_AC3_H */