libavcodec/psymodel.c

   1 /*
   2  * audio encoder psychoacoustic model
   3  * Copyright (C) 2008 Konstantin Shishkov
   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 #include "avcodec.h"
  23 #include "psymodel.h"
  24 #include "iirfilter.h"
  25
  26 extern const FFPsyModel ff_aac_psy_model;
  27
  28 av_cold int ff_psy_init(FFPsyContext *ctx, AVCodecContext *avctx,
  29                         int num_lens,
  30                         const uint8_t **bands, const int* num_bands)
  31 {
  32     ctx->avctx = avctx;
  33     ctx->psy_bands = av_mallocz(sizeof(FFPsyBand) * PSY_MAX_BANDS * avctx->channels);
  34     ctx->bands     = av_malloc (sizeof(ctx->bands[0])     * num_lens);
  35     ctx->num_bands = av_malloc (sizeof(ctx->num_bands[0]) * num_lens);
  36     memcpy(ctx->bands,     bands,     sizeof(ctx->bands[0])     *  num_lens);
  37     memcpy(ctx->num_bands, num_bands, sizeof(ctx->num_bands[0]) *  num_lens);
  38     switch (ctx->avctx->codec_id) {
  39     case CODEC_ID_AAC:
  40         ctx->model = &ff_aac_psy_model;
  41         break;
  42     }
  43     if (ctx->model->init)
  44         return ctx->model->init(ctx);
  45     return 0;
  46 }
  47
  48 FFPsyWindowInfo ff_psy_suggest_window(FFPsyContext *ctx,
  49                                       const int16_t *audio, const int16_t *la,
  50                                       int channel, int prev_type)
  51 {
  52     return ctx->model->window(ctx, audio, la, channel, prev_type);
  53 }
  54
  55 void ff_psy_set_band_info(FFPsyContext *ctx, int channel,
  56                           const float *coeffs, FFPsyWindowInfo *wi)
  57 {
  58     ctx->model->analyze(ctx, channel, coeffs, wi);
  59 }
  60
  61 av_cold void ff_psy_end(FFPsyContext *ctx)
  62 {
  63     if (ctx->model->end)
  64         ctx->model->end(ctx);
  65     av_freep(&ctx->bands);
  66     av_freep(&ctx->num_bands);
  67     av_freep(&ctx->psy_bands);
  68 }
  69
  70 typedef struct FFPsyPreprocessContext{
  71     AVCodecContext *avctx;
  72     float stereo_att;
  73     struct FFIIRFilterCoeffs *fcoeffs;
  74     struct FFIIRFilterState **fstate;
  75 }FFPsyPreprocessContext;
  76
  77 #define FILT_ORDER 4
  78
  79 av_cold struct FFPsyPreprocessContext* ff_psy_preprocess_init(AVCodecContext *avctx)
  80 {
  81     FFPsyPreprocessContext *ctx;
  82     int i;
  83     float cutoff_coeff;
  84     ctx        = av_mallocz(sizeof(FFPsyPreprocessContext));
  85     ctx->avctx = avctx;
  86
  87     if (avctx->cutoff > 0)
  88         cutoff_coeff = 2.0 * avctx->cutoff / avctx->sample_rate;
  89     else if (avctx->flags & CODEC_FLAG_QSCALE)
  90         cutoff_coeff = 1.0f / av_clip(1 + avctx->global_quality / FF_QUALITY_SCALE, 1, 8);
  91     else
  92         cutoff_coeff = avctx->bit_rate / (4.0f * avctx->sample_rate * avctx->channels);
  93
  94     ctx->fcoeffs = ff_iir_filter_init_coeffs(FF_FILTER_TYPE_BUTTERWORTH, FF_FILTER_MODE_LOWPASS,
  95                                              FILT_ORDER, cutoff_coeff, 0.0, 0.0);
  96     if (ctx->fcoeffs) {
  97         ctx->fstate = av_mallocz(sizeof(ctx->fstate[0]) * avctx->channels);
  98         for (i = 0; i < avctx->channels; i++)
  99             ctx->fstate[i] = ff_iir_filter_init_state(FILT_ORDER);
 100     }
 101     return ctx;
 102 }
 103
 104 void ff_psy_preprocess(struct FFPsyPreprocessContext *ctx,
 105                        const int16_t *audio, int16_t *dest,
 106                        int tag, int channels)
 107 {
 108     int ch, i;
 109     if (ctx->fstate) {
 110         for (ch = 0; ch < channels; ch++)
 111             ff_iir_filter(ctx->fcoeffs, ctx->fstate[tag+ch], ctx->avctx->frame_size,
 112                           audio + ch, ctx->avctx->channels,
 113                           dest  + ch, ctx->avctx->channels);
 114     } else {
 115         for (ch = 0; ch < channels; ch++)
 116             for (i = 0; i < ctx->avctx->frame_size; i++)
 117                 dest[i*ctx->avctx->channels + ch] = audio[i*ctx->avctx->channels + ch];
 118     }
 119 }
 120
 121 av_cold void ff_psy_preprocess_end(struct FFPsyPreprocessContext *ctx)
 122 {
 123     int i;
 124     ff_iir_filter_free_coeffs(ctx->fcoeffs);
 125     if (ctx->fstate)
 126         for (i = 0; i < ctx->avctx->channels; i++)
 127             ff_iir_filter_free_state(ctx->fstate[i]);
 128     av_freep(&ctx->fstate);
 129 }
 130