aboutsummaryrefslogtreecommitdiffstats
path: root/libavcodec/atrac3plusdec.c
blob: dbdb3b31944943b694ff46ff96e72f6e6a203c87 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
/*
 * ATRAC3+ compatible decoder
 *
 * Copyright (c) 2010-2013 Maxim Poliakovski
 *
 * This file is part of Libav.
 *
 * Libav is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * Libav is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with Libav; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

/**
 * @file
 * Sony ATRAC3+ compatible decoder.
 *
 * Container formats used to store its data:
 * RIFF WAV (.at3) and Sony OpenMG (.oma, .aa3).
 *
 * Technical description of this codec can be found here:
 * http://wiki.multimedia.cx/index.php?title=ATRAC3plus
 *
 * Kudos to Benjamin Larsson and Michael Karcher
 * for their precious technical help!
 */

#include <stdint.h>
#include <string.h>

#include "libavutil/channel_layout.h"
#include "libavutil/float_dsp.h"
#include "avcodec.h"
#include "get_bits.h"
#include "internal.h"
#include "atrac.h"
#include "atrac3plus.h"

typedef struct ATRAC3PContext {
    GetBitContext gb;
    AVFloatDSPContext fdsp;

    DECLARE_ALIGNED(32, float, samples)[2][ATRAC3P_FRAME_SAMPLES];  ///< quantized MDCT spectrum
    DECLARE_ALIGNED(32, float, mdct_buf)[2][ATRAC3P_FRAME_SAMPLES]; ///< output of the IMDCT
    DECLARE_ALIGNED(32, float, time_buf)[2][ATRAC3P_FRAME_SAMPLES]; ///< output of the gain compensation
    DECLARE_ALIGNED(32, float, outp_buf)[2][ATRAC3P_FRAME_SAMPLES];

    AtracGCContext gainc_ctx;   ///< gain compensation context
    FFTContext mdct_ctx;
    FFTContext ipqf_dct_ctx;    ///< IDCT context used by IPQF

    Atrac3pChanUnitCtx *ch_units;   ///< global channel units

    int num_channel_blocks;     ///< number of channel blocks
    uint8_t channel_blocks[5];  ///< channel configuration descriptor
    uint64_t my_channel_layout; ///< current channel layout
} ATRAC3PContext;

static av_cold int atrac3p_decode_close(AVCodecContext *avctx)
{
    av_free(((ATRAC3PContext *)(avctx->priv_data))->ch_units);

    return 0;
}

static av_cold int set_channel_params(ATRAC3PContext *ctx,
                                      AVCodecContext *avctx)
{
    memset(ctx->channel_blocks, 0, sizeof(ctx->channel_blocks));

    switch (avctx->channel_layout) {
    case AV_CH_FRONT_LEFT:
    case AV_CH_LAYOUT_MONO:
        ctx->num_channel_blocks = 1;
        ctx->channel_blocks[0]  = CH_UNIT_MONO;
        break;
    case AV_CH_LAYOUT_STEREO:
        ctx->num_channel_blocks = 1;
        ctx->channel_blocks[0]  = CH_UNIT_STEREO;
        break;
    case AV_CH_LAYOUT_SURROUND:
        ctx->num_channel_blocks = 2;
        ctx->channel_blocks[0]  = CH_UNIT_STEREO;
        ctx->channel_blocks[1]  = CH_UNIT_MONO;
        break;
    case AV_CH_LAYOUT_4POINT0:
        ctx->num_channel_blocks = 3;
        ctx->channel_blocks[0]  = CH_UNIT_STEREO;
        ctx->channel_blocks[1]  = CH_UNIT_MONO;
        ctx->channel_blocks[2]  = CH_UNIT_MONO;
        break;
    case AV_CH_LAYOUT_5POINT1_BACK:
        ctx->num_channel_blocks = 4;
        ctx->channel_blocks[0]  = CH_UNIT_STEREO;
        ctx->channel_blocks[1]  = CH_UNIT_MONO;
        ctx->channel_blocks[2]  = CH_UNIT_STEREO;
        ctx->channel_blocks[3]  = CH_UNIT_MONO;
        break;
    case AV_CH_LAYOUT_6POINT1_BACK:
        ctx->num_channel_blocks = 5;
        ctx->channel_blocks[0]  = CH_UNIT_STEREO;
        ctx->channel_blocks[1]  = CH_UNIT_MONO;
        ctx->channel_blocks[2]  = CH_UNIT_STEREO;
        ctx->channel_blocks[3]  = CH_UNIT_MONO;
        ctx->channel_blocks[4]  = CH_UNIT_MONO;
        break;
    case AV_CH_LAYOUT_7POINT1:
        ctx->num_channel_blocks = 5;
        ctx->channel_blocks[0]  = CH_UNIT_STEREO;
        ctx->channel_blocks[1]  = CH_UNIT_MONO;
        ctx->channel_blocks[2]  = CH_UNIT_STEREO;
        ctx->channel_blocks[3]  = CH_UNIT_STEREO;
        ctx->channel_blocks[4]  = CH_UNIT_MONO;
        break;
    default:
        av_log(avctx, AV_LOG_ERROR,
               "Unsupported channel layout: %"PRIx64"!\n", avctx->channel_layout);
        return AVERROR_INVALIDDATA;
    }

    return 0;
}

static av_cold int atrac3p_decode_init(AVCodecContext *avctx)
{
    ATRAC3PContext *ctx = avctx->priv_data;
    int i, ch, ret;

    if (!avctx->block_align) {
        av_log(avctx, AV_LOG_ERROR, "block_align is not set\n");
        return AVERROR(EINVAL);
    }

    avpriv_float_dsp_init(&ctx->fdsp, avctx->flags & CODEC_FLAG_BITEXACT);

    /* initialize IPQF */
    ff_mdct_init(&ctx->ipqf_dct_ctx, 5, 1, 32.0 / 32768.0);

    ff_atrac3p_init_imdct(avctx, &ctx->mdct_ctx);

    ff_atrac_init_gain_compensation(&ctx->gainc_ctx, 6, 2);

    ff_atrac3p_init_wave_synth();

    if ((ret = set_channel_params(ctx, avctx)) < 0)
        return ret;

    ctx->my_channel_layout = avctx->channel_layout;

    ctx->ch_units = av_mallocz(sizeof(*ctx->ch_units) *
                               ctx->num_channel_blocks);
    if (!ctx->ch_units) {
        atrac3p_decode_close(avctx);
        return AVERROR(ENOMEM);
    }

    for (i = 0; i < ctx->num_channel_blocks; i++) {
        for (ch = 0; ch < 2; ch++) {
            ctx->ch_units[i].channels[ch].ch_num          = ch;
            ctx->ch_units[i].channels[ch].wnd_shape       = &ctx->ch_units[i].channels[ch].wnd_shape_hist[0][0];
            ctx->ch_units[i].channels[ch].wnd_shape_prev  = &ctx->ch_units[i].channels[ch].wnd_shape_hist[1][0];
            ctx->ch_units[i].channels[ch].gain_data       = &ctx->ch_units[i].channels[ch].gain_data_hist[0][0];
            ctx->ch_units[i].channels[ch].gain_data_prev  = &ctx->ch_units[i].channels[ch].gain_data_hist[1][0];
            ctx->ch_units[i].channels[ch].tones_info      = &ctx->ch_units[i].channels[ch].tones_info_hist[0][0];
            ctx->ch_units[i].channels[ch].tones_info_prev = &ctx->ch_units[i].channels[ch].tones_info_hist[1][0];
        }

        ctx->ch_units[i].waves_info      = &ctx->ch_units[i].wave_synth_hist[0];
        ctx->ch_units[i].waves_info_prev = &ctx->ch_units[i].wave_synth_hist[1];
    }

    avctx->sample_fmt = AV_SAMPLE_FMT_FLTP;

    return 0;
}

static void decode_residual_spectrum(Atrac3pChanUnitCtx *ctx,
                                     float out[2][ATRAC3P_FRAME_SAMPLES],
                                     int num_channels,
                                     AVCodecContext *avctx)
{
    int i, sb, ch, qu, nspeclines, RNG_index;
    float *dst, q;
    int16_t *src;
    /* calculate RNG table index for each subband */
    int sb_RNG_index[ATRAC3P_SUBBANDS] = { 0 };

    if (ctx->mute_flag) {
        for (ch = 0; ch < num_channels; ch++)
            memset(out[ch], 0, ATRAC3P_FRAME_SAMPLES * sizeof(*out[ch]));
        return;
    }

    for (qu = 0, RNG_index = 0; qu < ctx->used_quant_units; qu++)
        RNG_index += ctx->channels[0].qu_sf_idx[qu] +
                     ctx->channels[1].qu_sf_idx[qu];

    for (sb = 0; sb < ctx->num_coded_subbands; sb++, RNG_index += 128)
        sb_RNG_index[sb] = RNG_index & 0x3FC;

    /* inverse quant and power compensation */
    for (ch = 0; ch < num_channels; ch++) {
        /* clear channel's residual spectrum */
        memset(out[ch], 0, ATRAC3P_FRAME_SAMPLES * sizeof(*out[ch]));

        for (qu = 0; qu < ctx->used_quant_units; qu++) {
            src        = &ctx->channels[ch].spectrum[ff_atrac3p_qu_to_spec_pos[qu]];
            dst        = &out[ch][ff_atrac3p_qu_to_spec_pos[qu]];
            nspeclines = ff_atrac3p_qu_to_spec_pos[qu + 1] -
                         ff_atrac3p_qu_to_spec_pos[qu];

            if (ctx->channels[ch].qu_wordlen[qu] > 0) {
                q = ff_atrac3p_sf_tab[ctx->channels[ch].qu_sf_idx[qu]] *
                    ff_atrac3p_mant_tab[ctx->channels[ch].qu_wordlen[qu]];
                for (i = 0; i < nspeclines; i++)
                    dst[i] = src[i] * q;
            }
        }

        for (sb = 0; sb < ctx->num_coded_subbands; sb++)
            ff_atrac3p_power_compensation(ctx, ch, &out[ch][0],
                                          sb_RNG_index[sb], sb);
    }

    if (ctx->unit_type == CH_UNIT_STEREO) {
        for (sb = 0; sb < ctx->num_coded_subbands; sb++) {
            if (ctx->swap_channels[sb]) {
                for (i = 0; i < ATRAC3P_SUBBAND_SAMPLES; i++)
                    FFSWAP(float, out[0][sb * ATRAC3P_SUBBAND_SAMPLES + i],
                                  out[1][sb * ATRAC3P_SUBBAND_SAMPLES + i]);
            }

            /* flip coefficients' sign if requested */
            if (ctx->negate_coeffs[sb])
                for (i = 0; i < ATRAC3P_SUBBAND_SAMPLES; i++)
                    out[1][sb * ATRAC3P_SUBBAND_SAMPLES + i] = -(out[1][sb * ATRAC3P_SUBBAND_SAMPLES + i]);
        }
    }
}

static void reconstruct_frame(ATRAC3PContext *ctx, Atrac3pChanUnitCtx *ch_unit,
                              int num_channels, AVCodecContext *avctx)
{
    int ch, sb;

    for (ch = 0; ch < num_channels; ch++) {
        for (sb = 0; sb < ch_unit->num_subbands; sb++) {
            /* inverse transform and windowing */
            ff_atrac3p_imdct(&ctx->fdsp, &ctx->mdct_ctx,
                             &ctx->samples[ch][sb * ATRAC3P_SUBBAND_SAMPLES],
                             &ctx->mdct_buf[ch][sb * ATRAC3P_SUBBAND_SAMPLES],
                             (ch_unit->channels[ch].wnd_shape_prev[sb] << 1) +
                             ch_unit->channels[ch].wnd_shape[sb], sb);

            /* gain compensation and overlapping */
            ff_atrac_gain_compensation(&ctx->gainc_ctx,
                                       &ctx->mdct_buf[ch][sb * ATRAC3P_SUBBAND_SAMPLES],
                                       &ch_unit->prev_buf[ch][sb * ATRAC3P_SUBBAND_SAMPLES],
                                       &ch_unit->channels[ch].gain_data_prev[sb],
                                       &ch_unit->channels[ch].gain_data[sb],
                                       ATRAC3P_SUBBAND_SAMPLES,
                                       &ctx->time_buf[ch][sb * ATRAC3P_SUBBAND_SAMPLES]);
        }

        /* zero unused subbands in both output and overlapping buffers */
        memset(&ch_unit->prev_buf[ch][ch_unit->num_subbands * ATRAC3P_SUBBAND_SAMPLES],
               0,
               (ATRAC3P_SUBBANDS - ch_unit->num_subbands) *
               ATRAC3P_SUBBAND_SAMPLES *
               sizeof(ch_unit->prev_buf[ch][ch_unit->num_subbands * ATRAC3P_SUBBAND_SAMPLES]));
        memset(&ctx->time_buf[ch][ch_unit->num_subbands * ATRAC3P_SUBBAND_SAMPLES],
               0,
               (ATRAC3P_SUBBANDS - ch_unit->num_subbands) *
               ATRAC3P_SUBBAND_SAMPLES *
               sizeof(ctx->time_buf[ch][ch_unit->num_subbands * ATRAC3P_SUBBAND_SAMPLES]));

        /* resynthesize and add tonal signal */
        if (ch_unit->waves_info->tones_present ||
            ch_unit->waves_info_prev->tones_present) {
            for (sb = 0; sb < ch_unit->num_subbands; sb++)
                if (ch_unit->channels[ch].tones_info[sb].num_wavs ||
                    ch_unit->channels[ch].tones_info_prev[sb].num_wavs) {
                    ff_atrac3p_generate_tones(ch_unit, &ctx->fdsp, ch, sb,
                                              &ctx->time_buf[ch][sb * 128]);
                }
        }

        /* subband synthesis and acoustic signal output */
        ff_atrac3p_ipqf(&ctx->ipqf_dct_ctx, &ch_unit->ipqf_ctx[ch],
                        &ctx->time_buf[ch][0], &ctx->outp_buf[ch][0]);
    }

    /* swap window shape and gain control buffers. */
    for (ch = 0; ch < num_channels; ch++) {
        FFSWAP(uint8_t *, ch_unit->channels[ch].wnd_shape,
               ch_unit->channels[ch].wnd_shape_prev);
        FFSWAP(AtracGainInfo *, ch_unit->channels[ch].gain_data,
               ch_unit->channels[ch].gain_data_prev);
        FFSWAP(Atrac3pWavesData *, ch_unit->channels[ch].tones_info,
               ch_unit->channels[ch].tones_info_prev);
    }

    FFSWAP(Atrac3pWaveSynthParams *, ch_unit->waves_info, ch_unit->waves_info_prev);
}

static int atrac3p_decode_frame(AVCodecContext *avctx, void *data,
                                int *got_frame_ptr, AVPacket *avpkt)
{
    ATRAC3PContext *ctx = avctx->priv_data;
    AVFrame *frame      = data;
    int i, ret, ch_unit_id, ch_block = 0, out_ch_index = 0, channels_to_process;
    float **samples_p = (float **)frame->extended_data;

    frame->nb_samples = ATRAC3P_FRAME_SAMPLES;
    if ((ret = ff_get_buffer(avctx, frame, 0)) < 0) {
        av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
        return ret;
    }

    if ((ret = init_get_bits8(&ctx->gb, avpkt->data, avpkt->size)) < 0)
        return ret;

    if (get_bits1(&ctx->gb)) {
        av_log(avctx, AV_LOG_ERROR, "Invalid start bit!\n");
        return AVERROR_INVALIDDATA;
    }

    while (get_bits_left(&ctx->gb) >= 2 &&
           (ch_unit_id = get_bits(&ctx->gb, 2)) != CH_UNIT_TERMINATOR) {
        if (ch_unit_id == CH_UNIT_EXTENSION) {
            avpriv_report_missing_feature(avctx, "Channel unit extension");
            return AVERROR_PATCHWELCOME;
        }
        if (ch_block >= ctx->num_channel_blocks ||
            ctx->channel_blocks[ch_block] != ch_unit_id) {
            av_log(avctx, AV_LOG_ERROR,
                   "Frame data doesn't match channel configuration!\n");
            return AVERROR_INVALIDDATA;
        }

        ctx->ch_units[ch_block].unit_type = ch_unit_id;
        channels_to_process               = ch_unit_id + 1;

        if ((ret = ff_atrac3p_decode_channel_unit(&ctx->gb,
                                                  &ctx->ch_units[ch_block],
                                                  channels_to_process,
                                                  avctx)) < 0)
            return ret;

        decode_residual_spectrum(&ctx->ch_units[ch_block], ctx->samples,
                                 channels_to_process, avctx);
        reconstruct_frame(ctx, &ctx->ch_units[ch_block],
                          channels_to_process, avctx);

        for (i = 0; i < channels_to_process; i++)
            memcpy(samples_p[out_ch_index + i], ctx->outp_buf[i],
                   ATRAC3P_FRAME_SAMPLES * sizeof(**samples_p));

        ch_block++;
        out_ch_index += channels_to_process;
    }

    *got_frame_ptr = 1;

    return avctx->block_align;
}

AVCodec ff_atrac3p_decoder = {
    .name             = "atrac3plus",
    .long_name        = NULL_IF_CONFIG_SMALL("ATRAC3+ (Adaptive TRansform Acoustic Coding 3+)"),
    .type             = AVMEDIA_TYPE_AUDIO,
    .id               = AV_CODEC_ID_ATRAC3P,
    .priv_data_size   = sizeof(ATRAC3PContext),
    .init             = atrac3p_decode_init,
    .init_static_data = ff_atrac3p_init_vlcs,
    .close            = atrac3p_decode_close,
    .decode           = atrac3p_decode_frame,
};