3 * Copyright (c) 2012 Martin Storsjo
5 * This file is part of FFmpeg.
7 * Permission to use, copy, modify, and/or distribute this software for any
8 * purpose with or without fee is hereby granted, provided that the above
9 * copyright notice and this permission notice appear in all copies.
11 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
12 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
13 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
14 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
15 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
16 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
17 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
20 #include <fdk-aac/aacdecoder_lib.h>
22 #include "libavutil/channel_layout.h"
23 #include "libavutil/common.h"
24 #include "libavutil/opt.h"
29 CONCEAL_METHOD_SPECTRAL_MUTING = 0,
30 CONCEAL_METHOD_NOISE_SUBSTITUTION = 1,
31 CONCEAL_METHOD_ENERGY_INTERPOLATION = 2,
35 typedef struct FDKAACDecContext {
37 HANDLE_AACDECODER handle;
39 uint8_t *decoder_buffer;
41 enum ConcealMethod conceal_method;
49 #define DMX_ANC_BUFFSIZE 128
50 #define DECODER_MAX_CHANNELS 6
51 #define DECODER_BUFFSIZE 2048 * sizeof(INT_PCM)
53 #define OFFSET(x) offsetof(FDKAACDecContext, x)
54 #define AD AV_OPT_FLAG_AUDIO_PARAM | AV_OPT_FLAG_DECODING_PARAM
55 static const AVOption fdk_aac_dec_options[] = {
56 { "conceal", "Error concealment method", OFFSET(conceal_method), AV_OPT_TYPE_INT, { .i64 = CONCEAL_METHOD_NOISE_SUBSTITUTION }, CONCEAL_METHOD_SPECTRAL_MUTING, CONCEAL_METHOD_NB - 1, AD, "conceal" },
57 { "spectral", "Spectral muting", 0, AV_OPT_TYPE_CONST, { .i64 = CONCEAL_METHOD_SPECTRAL_MUTING }, INT_MIN, INT_MAX, AD, "conceal" },
58 { "noise", "Noise Substitution", 0, AV_OPT_TYPE_CONST, { .i64 = CONCEAL_METHOD_NOISE_SUBSTITUTION }, INT_MIN, INT_MAX, AD, "conceal" },
59 { "energy", "Energy Interpolation", 0, AV_OPT_TYPE_CONST, { .i64 = CONCEAL_METHOD_ENERGY_INTERPOLATION }, INT_MIN, INT_MAX, AD, "conceal" },
60 { "drc_boost", "Dynamic Range Control: boost, where [0] is none and [127] is max boost",
61 OFFSET(drc_boost), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, 127, AD, NULL },
62 { "drc_cut", "Dynamic Range Control: attenuation factor, where [0] is none and [127] is max compression",
63 OFFSET(drc_cut), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, 127, AD, NULL },
64 { "drc_level", "Dynamic Range Control: reference level, quantized to 0.25dB steps where [0] is 0dB and [127] is -31.75dB",
65 OFFSET(drc_level), AV_OPT_TYPE_INT, { .i64 = -1}, -1, 127, AD, NULL },
66 { "drc_heavy", "Dynamic Range Control: heavy compression, where [1] is on (RF mode) and [0] is off",
67 OFFSET(drc_heavy), AV_OPT_TYPE_INT, { .i64 = -1}, -1, 1, AD, NULL },
71 static const AVClass fdk_aac_dec_class = {
72 "libfdk-aac decoder", av_default_item_name, fdk_aac_dec_options, LIBAVUTIL_VERSION_INT
75 static int get_stream_info(AVCodecContext *avctx)
77 FDKAACDecContext *s = avctx->priv_data;
78 CStreamInfo *info = aacDecoder_GetStreamInfo(s->handle);
79 int channel_counts[9] = { 0 };
81 uint64_t ch_layout = 0;
84 av_log(avctx, AV_LOG_ERROR, "Unable to get stream info\n");
85 return AVERROR_UNKNOWN;
88 if (info->sampleRate <= 0) {
89 av_log(avctx, AV_LOG_ERROR, "Stream info not initialized\n");
90 return AVERROR_UNKNOWN;
92 avctx->sample_rate = info->sampleRate;
93 avctx->frame_size = info->frameSize;
95 for (i = 0; i < info->numChannels; i++) {
96 AUDIO_CHANNEL_TYPE ctype = info->pChannelType[i];
97 if (ctype <= ACT_NONE || ctype > ACT_TOP) {
98 av_log(avctx, AV_LOG_WARNING, "unknown channel type\n");
101 channel_counts[ctype]++;
103 av_log(avctx, AV_LOG_DEBUG,
104 "%d channels - front:%d side:%d back:%d lfe:%d top:%d\n",
106 channel_counts[ACT_FRONT], channel_counts[ACT_SIDE],
107 channel_counts[ACT_BACK], channel_counts[ACT_LFE],
108 channel_counts[ACT_FRONT_TOP] + channel_counts[ACT_SIDE_TOP] +
109 channel_counts[ACT_BACK_TOP] + channel_counts[ACT_TOP]);
111 switch (channel_counts[ACT_FRONT]) {
113 ch_layout |= AV_CH_LAYOUT_STEREO | AV_CH_FRONT_LEFT_OF_CENTER |
114 AV_CH_FRONT_RIGHT_OF_CENTER;
117 ch_layout |= AV_CH_LAYOUT_STEREO | AV_CH_FRONT_CENTER;
120 ch_layout |= AV_CH_LAYOUT_STEREO;
123 ch_layout |= AV_CH_FRONT_CENTER;
126 av_log(avctx, AV_LOG_WARNING,
127 "unsupported number of front channels: %d\n",
128 channel_counts[ACT_FRONT]);
132 if (channel_counts[ACT_SIDE] > 0) {
133 if (channel_counts[ACT_SIDE] == 2) {
134 ch_layout |= AV_CH_SIDE_LEFT | AV_CH_SIDE_RIGHT;
136 av_log(avctx, AV_LOG_WARNING,
137 "unsupported number of side channels: %d\n",
138 channel_counts[ACT_SIDE]);
142 if (channel_counts[ACT_BACK] > 0) {
143 switch (channel_counts[ACT_BACK]) {
145 ch_layout |= AV_CH_BACK_LEFT | AV_CH_BACK_RIGHT | AV_CH_BACK_CENTER;
148 ch_layout |= AV_CH_BACK_LEFT | AV_CH_BACK_RIGHT;
151 ch_layout |= AV_CH_BACK_CENTER;
154 av_log(avctx, AV_LOG_WARNING,
155 "unsupported number of back channels: %d\n",
156 channel_counts[ACT_BACK]);
161 if (channel_counts[ACT_LFE] > 0) {
162 if (channel_counts[ACT_LFE] == 1) {
163 ch_layout |= AV_CH_LOW_FREQUENCY;
165 av_log(avctx, AV_LOG_WARNING,
166 "unsupported number of LFE channels: %d\n",
167 channel_counts[ACT_LFE]);
172 av_get_channel_layout_nb_channels(ch_layout) != info->numChannels) {
173 av_log(avctx, AV_LOG_WARNING, "unsupported channel configuration\n");
177 avctx->channel_layout = 0;
179 avctx->channel_layout = ch_layout;
181 avctx->channels = info->numChannels;
186 static av_cold int fdk_aac_decode_close(AVCodecContext *avctx)
188 FDKAACDecContext *s = avctx->priv_data;
191 aacDecoder_Close(s->handle);
192 av_freep(&s->decoder_buffer);
193 av_freep(&s->anc_buffer);
198 static av_cold int fdk_aac_decode_init(AVCodecContext *avctx)
200 FDKAACDecContext *s = avctx->priv_data;
201 AAC_DECODER_ERROR err;
204 s->handle = aacDecoder_Open(avctx->extradata_size ? TT_MP4_RAW : TT_MP4_ADTS, 1);
206 av_log(avctx, AV_LOG_ERROR, "Error opening decoder\n");
207 return AVERROR_UNKNOWN;
210 if (avctx->extradata_size) {
211 if ((err = aacDecoder_ConfigRaw(s->handle, &avctx->extradata,
212 &avctx->extradata_size)) != AAC_DEC_OK) {
213 av_log(avctx, AV_LOG_ERROR, "Unable to set extradata\n");
214 return AVERROR_INVALIDDATA;
218 if ((err = aacDecoder_SetParam(s->handle, AAC_CONCEAL_METHOD,
219 s->conceal_method)) != AAC_DEC_OK) {
220 av_log(avctx, AV_LOG_ERROR, "Unable to set error concealment method\n");
221 return AVERROR_UNKNOWN;
224 if (avctx->request_channel_layout > 0 &&
225 avctx->request_channel_layout != AV_CH_LAYOUT_NATIVE) {
226 int downmix_channels = -1;
228 switch (avctx->request_channel_layout) {
229 case AV_CH_LAYOUT_STEREO:
230 case AV_CH_LAYOUT_STEREO_DOWNMIX:
231 downmix_channels = 2;
233 case AV_CH_LAYOUT_MONO:
234 downmix_channels = 1;
237 av_log(avctx, AV_LOG_WARNING, "Invalid request_channel_layout\n");
241 if (downmix_channels != -1) {
242 if (aacDecoder_SetParam(s->handle, AAC_PCM_OUTPUT_CHANNELS,
243 downmix_channels) != AAC_DEC_OK) {
244 av_log(avctx, AV_LOG_WARNING, "Unable to set output channels in the decoder\n");
246 s->anc_buffer = av_malloc(DMX_ANC_BUFFSIZE);
247 if (!s->anc_buffer) {
248 av_log(avctx, AV_LOG_ERROR, "Unable to allocate ancillary buffer for the decoder\n");
249 ret = AVERROR(ENOMEM);
252 if (aacDecoder_AncDataInit(s->handle, s->anc_buffer, DMX_ANC_BUFFSIZE)) {
253 av_log(avctx, AV_LOG_ERROR, "Unable to register downmix ancillary buffer in the decoder\n");
254 ret = AVERROR_UNKNOWN;
261 if (s->drc_boost != -1) {
262 if (aacDecoder_SetParam(s->handle, AAC_DRC_BOOST_FACTOR, s->drc_boost) != AAC_DEC_OK) {
263 av_log(avctx, AV_LOG_ERROR, "Unable to set DRC boost factor in the decoder\n");
264 return AVERROR_UNKNOWN;
268 if (s->drc_cut != -1) {
269 if (aacDecoder_SetParam(s->handle, AAC_DRC_ATTENUATION_FACTOR, s->drc_cut) != AAC_DEC_OK) {
270 av_log(avctx, AV_LOG_ERROR, "Unable to set DRC attenuation factor in the decoder\n");
271 return AVERROR_UNKNOWN;
275 if (s->drc_level != -1) {
276 if (aacDecoder_SetParam(s->handle, AAC_DRC_REFERENCE_LEVEL, s->drc_level) != AAC_DEC_OK) {
277 av_log(avctx, AV_LOG_ERROR, "Unable to set DRC reference level in the decoder\n");
278 return AVERROR_UNKNOWN;
282 if (s->drc_heavy != -1) {
283 if (aacDecoder_SetParam(s->handle, AAC_DRC_HEAVY_COMPRESSION, s->drc_heavy) != AAC_DEC_OK) {
284 av_log(avctx, AV_LOG_ERROR, "Unable to set DRC heavy compression in the decoder\n");
285 return AVERROR_UNKNOWN;
289 avctx->sample_fmt = AV_SAMPLE_FMT_S16;
293 fdk_aac_decode_close(avctx);
297 static int fdk_aac_decode_frame(AVCodecContext *avctx, void *data,
298 int *got_frame_ptr, AVPacket *avpkt)
300 FDKAACDecContext *s = avctx->priv_data;
301 AVFrame *frame = data;
303 AAC_DECODER_ERROR err;
304 UINT valid = avpkt->size;
305 uint8_t *buf, *tmpptr = NULL;
308 err = aacDecoder_Fill(s->handle, &avpkt->data, &avpkt->size, &valid);
309 if (err != AAC_DEC_OK) {
310 av_log(avctx, AV_LOG_ERROR, "aacDecoder_Fill() failed: %x\n", err);
311 return AVERROR_INVALIDDATA;
314 if (s->initialized) {
315 frame->nb_samples = avctx->frame_size;
316 if ((ret = ff_get_buffer(avctx, frame, 0)) < 0)
320 buf_size = DECODER_BUFFSIZE * DECODER_MAX_CHANNELS;
321 buf = s->decoder_buffer;
323 buf = frame->extended_data[0];
324 buf_size = avctx->channels * frame->nb_samples *
325 av_get_bytes_per_sample(avctx->sample_fmt);
328 buf_size = DECODER_BUFFSIZE * DECODER_MAX_CHANNELS;
330 if (!s->decoder_buffer)
331 s->decoder_buffer = av_malloc(buf_size);
332 if (!s->decoder_buffer)
333 return AVERROR(ENOMEM);
335 buf = tmpptr = s->decoder_buffer;
338 err = aacDecoder_DecodeFrame(s->handle, (INT_PCM *) buf, buf_size, 0);
339 if (err == AAC_DEC_NOT_ENOUGH_BITS) {
340 ret = avpkt->size - valid;
343 if (err != AAC_DEC_OK) {
344 av_log(avctx, AV_LOG_ERROR,
345 "aacDecoder_DecodeFrame() failed: %x\n", err);
346 ret = AVERROR_UNKNOWN;
350 if (!s->initialized) {
351 if ((ret = get_stream_info(avctx)) < 0)
354 frame->nb_samples = avctx->frame_size;
358 frame->nb_samples = avctx->frame_size;
359 if ((ret = ff_get_buffer(avctx, frame, 0)) < 0)
362 if (s->decoder_buffer) {
363 memcpy(frame->extended_data[0], buf,
364 avctx->channels * avctx->frame_size *
365 av_get_bytes_per_sample(avctx->sample_fmt));
368 av_freep(&s->decoder_buffer);
372 ret = avpkt->size - valid;
378 static av_cold void fdk_aac_decode_flush(AVCodecContext *avctx)
380 FDKAACDecContext *s = avctx->priv_data;
381 AAC_DECODER_ERROR err;
386 if ((err = aacDecoder_SetParam(s->handle,
387 AAC_TPDEC_CLEAR_BUFFER, 1)) != AAC_DEC_OK)
388 av_log(avctx, AV_LOG_WARNING, "failed to clear buffer when flushing\n");
391 AVCodec ff_libfdk_aac_decoder = {
392 .name = "libfdk_aac",
393 .long_name = NULL_IF_CONFIG_SMALL("Fraunhofer FDK AAC"),
394 .type = AVMEDIA_TYPE_AUDIO,
395 .id = AV_CODEC_ID_AAC,
396 .priv_data_size = sizeof(FDKAACDecContext),
397 .init = fdk_aac_decode_init,
398 .decode = fdk_aac_decode_frame,
399 .close = fdk_aac_decode_close,
400 .flush = fdk_aac_decode_flush,
401 .capabilities = CODEC_CAP_DR1 | CODEC_CAP_CHANNEL_CONF,
402 .priv_class = &fdk_aac_dec_class,