| 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 <string.h> |
| 23 | |
| 24 | #include "avcodec.h" |
| 25 | #include "psymodel.h" |
| 26 | #include "iirfilter.h" |
| 27 | #include "libavutil/mem.h" |
| 28 | |
| 29 | extern const FFPsyModel ff_aac_psy_model; |
| 30 | |
| 31 | av_cold int ff_psy_init(FFPsyContext *ctx, AVCodecContext *avctx, int num_lens, |
| 32 | const uint8_t **bands, const int* num_bands, |
| 33 | int num_groups, const uint8_t *group_map) |
| 34 | { |
| 35 | int i, j, k = 0; |
| 36 | |
| 37 | ctx->avctx = avctx; |
| 38 | ctx->ch = av_mallocz_array(sizeof(ctx->ch[0]), avctx->channels * 2); |
| 39 | ctx->group = av_mallocz_array(sizeof(ctx->group[0]), num_groups); |
| 40 | ctx->bands = av_malloc_array (sizeof(ctx->bands[0]), num_lens); |
| 41 | ctx->num_bands = av_malloc_array (sizeof(ctx->num_bands[0]), num_lens); |
| 42 | memcpy(ctx->bands, bands, sizeof(ctx->bands[0]) * num_lens); |
| 43 | memcpy(ctx->num_bands, num_bands, sizeof(ctx->num_bands[0]) * num_lens); |
| 44 | |
| 45 | /* assign channels to groups (with virtual channels for coupling) */ |
| 46 | for (i = 0; i < num_groups; i++) { |
| 47 | /* NOTE: Add 1 to handle the AAC chan_config without modification. |
| 48 | * This has the side effect of allowing an array of 0s to map |
| 49 | * to one channel per group. |
| 50 | */ |
| 51 | ctx->group[i].num_ch = group_map[i] + 1; |
| 52 | for (j = 0; j < ctx->group[i].num_ch * 2; j++) |
| 53 | ctx->group[i].ch[j] = &ctx->ch[k++]; |
| 54 | } |
| 55 | |
| 56 | switch (ctx->avctx->codec_id) { |
| 57 | case AV_CODEC_ID_AAC: |
| 58 | ctx->model = &ff_aac_psy_model; |
| 59 | break; |
| 60 | } |
| 61 | if (ctx->model->init) |
| 62 | return ctx->model->init(ctx); |
| 63 | return 0; |
| 64 | } |
| 65 | |
| 66 | FFPsyChannelGroup *ff_psy_find_group(FFPsyContext *ctx, int channel) |
| 67 | { |
| 68 | int i = 0, ch = 0; |
| 69 | |
| 70 | while (ch <= channel) |
| 71 | ch += ctx->group[i++].num_ch; |
| 72 | |
| 73 | return &ctx->group[i-1]; |
| 74 | } |
| 75 | |
| 76 | av_cold void ff_psy_end(FFPsyContext *ctx) |
| 77 | { |
| 78 | if (ctx->model && ctx->model->end) |
| 79 | ctx->model->end(ctx); |
| 80 | av_freep(&ctx->bands); |
| 81 | av_freep(&ctx->num_bands); |
| 82 | av_freep(&ctx->group); |
| 83 | av_freep(&ctx->ch); |
| 84 | } |
| 85 | |
| 86 | typedef struct FFPsyPreprocessContext{ |
| 87 | AVCodecContext *avctx; |
| 88 | float stereo_att; |
| 89 | struct FFIIRFilterCoeffs *fcoeffs; |
| 90 | struct FFIIRFilterState **fstate; |
| 91 | struct FFIIRFilterContext fiir; |
| 92 | }FFPsyPreprocessContext; |
| 93 | |
| 94 | #define FILT_ORDER 4 |
| 95 | |
| 96 | av_cold struct FFPsyPreprocessContext* ff_psy_preprocess_init(AVCodecContext *avctx) |
| 97 | { |
| 98 | FFPsyPreprocessContext *ctx; |
| 99 | int i; |
| 100 | float cutoff_coeff = 0; |
| 101 | ctx = av_mallocz(sizeof(FFPsyPreprocessContext)); |
| 102 | ctx->avctx = avctx; |
| 103 | |
| 104 | if (avctx->cutoff > 0) |
| 105 | cutoff_coeff = 2.0 * avctx->cutoff / avctx->sample_rate; |
| 106 | |
| 107 | if (!cutoff_coeff && avctx->codec_id == AV_CODEC_ID_AAC) |
| 108 | cutoff_coeff = 2.0 * AAC_CUTOFF(avctx) / avctx->sample_rate; |
| 109 | |
| 110 | if (cutoff_coeff && cutoff_coeff < 0.98) |
| 111 | ctx->fcoeffs = ff_iir_filter_init_coeffs(avctx, FF_FILTER_TYPE_BUTTERWORTH, |
| 112 | FF_FILTER_MODE_LOWPASS, FILT_ORDER, |
| 113 | cutoff_coeff, 0.0, 0.0); |
| 114 | if (ctx->fcoeffs) { |
| 115 | ctx->fstate = av_mallocz_array(sizeof(ctx->fstate[0]), avctx->channels); |
| 116 | for (i = 0; i < avctx->channels; i++) |
| 117 | ctx->fstate[i] = ff_iir_filter_init_state(FILT_ORDER); |
| 118 | } |
| 119 | |
| 120 | ff_iir_filter_init(&ctx->fiir); |
| 121 | |
| 122 | return ctx; |
| 123 | } |
| 124 | |
| 125 | void ff_psy_preprocess(struct FFPsyPreprocessContext *ctx, float **audio, int channels) |
| 126 | { |
| 127 | int ch; |
| 128 | int frame_size = ctx->avctx->frame_size; |
| 129 | FFIIRFilterContext *iir = &ctx->fiir; |
| 130 | |
| 131 | if (ctx->fstate) { |
| 132 | for (ch = 0; ch < channels; ch++) |
| 133 | iir->filter_flt(ctx->fcoeffs, ctx->fstate[ch], frame_size, |
| 134 | &audio[ch][frame_size], 1, &audio[ch][frame_size], 1); |
| 135 | } |
| 136 | } |
| 137 | |
| 138 | av_cold void ff_psy_preprocess_end(struct FFPsyPreprocessContext *ctx) |
| 139 | { |
| 140 | int i; |
| 141 | ff_iir_filter_free_coeffs(ctx->fcoeffs); |
| 142 | if (ctx->fstate) |
| 143 | for (i = 0; i < ctx->avctx->channels; i++) |
| 144 | ff_iir_filter_free_state(ctx->fstate[i]); |
| 145 | av_freep(&ctx->fstate); |
| 146 | av_free(ctx); |
| 147 | } |