2 * audio encoder psychoacoustic model
3 * Copyright (C) 2008 Konstantin Shishkov
5 * This file is part of Libav.
7 * Libav 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.
12 * Libav 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.
17 * 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
22 #ifndef AVCODEC_PSYMODEL_H
23 #define AVCODEC_PSYMODEL_H
27 /** maximum possible number of bands */
28 #define PSY_MAX_BANDS 128
29 /** maximum number of channels */
30 #define PSY_MAX_CHANS 20
33 * single band psychoacoustic information
35 typedef struct FFPsyBand {
40 float perceptual_weight;
44 * single channel psychoacoustic information
46 typedef struct FFPsyChannel {
47 FFPsyBand psy_bands[PSY_MAX_BANDS]; ///< channel bands information
48 float entropy; ///< total PE for this channel
52 * psychoacoustic information for an arbitrary group of channels
54 typedef struct FFPsyChannelGroup {
55 FFPsyChannel *ch[PSY_MAX_CHANS]; ///< pointers to the individual channels in the group
56 uint8_t num_ch; ///< number of channels in this group
57 uint8_t coupling[PSY_MAX_BANDS]; ///< allow coupling for this band in the group
61 * windowing related information
63 typedef struct FFPsyWindowInfo {
64 int window_type[3]; ///< window type (short/long/transitional, etc.) - current, previous and next
65 int window_shape; ///< window shape (sine/KBD/whatever)
66 int num_windows; ///< number of windows in a frame
67 int grouping[8]; ///< window grouping (for e.g. AAC)
68 int *window_sizes; ///< sequence of window sizes inside one frame (for eg. WMA)
72 * context used by psychoacoustic model
74 typedef struct FFPsyContext {
75 AVCodecContext *avctx; ///< encoder context
76 const struct FFPsyModel *model; ///< encoder-specific model functions
78 FFPsyChannel *ch; ///< single channel information
79 FFPsyChannelGroup *group; ///< channel group information
80 int num_groups; ///< number of channel groups
82 uint8_t **bands; ///< scalefactor band sizes for possible frame sizes
83 int *num_bands; ///< number of scalefactor bands for possible frame sizes
84 int num_lens; ///< number of scalefactor band sets
87 int size; ///< size of the bitresevoir in bits
88 int bits; ///< number of bits used in the bitresevoir
91 void* model_priv_data; ///< psychoacoustic model implementation private data
95 * codec-specific psychoacoustic model implementation
97 typedef struct FFPsyModel {
99 int (*init) (FFPsyContext *apc);
102 * Suggest window sequence for channel.
104 * @param ctx model context
105 * @param audio samples for the current frame
106 * @param la lookahead samples (NULL when unavailable)
107 * @param channel number of channel element to analyze
108 * @param prev_type previous window type
110 * @return suggested window information in a structure
112 FFPsyWindowInfo (*window)(FFPsyContext *ctx, const int16_t *audio, const int16_t *la, int channel, int prev_type);
115 * Perform psychoacoustic analysis and set band info (threshold, energy) for a group of channels.
117 * @param ctx model context
118 * @param channel channel number of the first channel in the group to perform analysis on
119 * @param coeffs array of pointers to the transformed coefficients
120 * @param wi window information for the channels in the group
122 void (*analyze)(FFPsyContext *ctx, int channel, const float **coeffs, const FFPsyWindowInfo *wi);
124 void (*end) (FFPsyContext *apc);
128 * Initialize psychoacoustic model.
130 * @param ctx model context
131 * @param avctx codec context
132 * @param num_lens number of possible frame lengths
133 * @param bands scalefactor band lengths for all frame lengths
134 * @param num_bands number of scalefactor bands for all frame lengths
135 * @param num_groups number of channel groups
136 * @param group_map array with # of channels in group - 1, for each group
138 * @return zero if successful, a negative value if not
140 av_cold int ff_psy_init(FFPsyContext *ctx, AVCodecContext *avctx, int num_lens,
141 const uint8_t **bands, const int* num_bands,
142 int num_groups, const uint8_t *group_map);
145 * Determine what group a channel belongs to.
147 * @param ctx psymodel context
148 * @param channel channel to locate the group for
150 * @return pointer to the FFPsyChannelGroup this channel belongs to
152 FFPsyChannelGroup *ff_psy_find_group(FFPsyContext *ctx, int channel);
155 * Cleanup model context at the end.
157 * @param ctx model context
159 av_cold void ff_psy_end(FFPsyContext *ctx);
162 /**************************************************************************
163 * Audio preprocessing stuff. *
164 * This should be moved into some audio filter eventually. *
165 **************************************************************************/
166 struct FFPsyPreprocessContext;
169 * psychoacoustic model audio preprocessing initialization
171 av_cold struct FFPsyPreprocessContext* ff_psy_preprocess_init(AVCodecContext *avctx);
174 * Preprocess several channel in audio frame in order to compress it better.
176 * @param ctx preprocessing context
177 * @param audio samples to preprocess
178 * @param dest place to put filtered samples
179 * @param tag channel number
180 * @param channels number of channel to preprocess (some additional work may be done on stereo pair)
182 void ff_psy_preprocess(struct FFPsyPreprocessContext *ctx,
183 const int16_t *audio, int16_t *dest,
184 int tag, int channels);
187 * Cleanup audio preprocessing module.
189 av_cold void ff_psy_preprocess_end(struct FFPsyPreprocessContext *ctx);
191 #endif /* AVCODEC_PSYMODEL_H */