2 Copyright (C) 2013 Paul Brossier <piem@aubio.org>
4 This file is part of aubio.
6 aubio is free software: you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation, either version 3 of the License, or
9 (at your option) any later version.
11 aubio is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with aubio. If not, see <http://www.gnu.org/licenses/>.
27 #include <libavcodec/avcodec.h>
28 #include <libavformat/avformat.h>
29 #include <libavresample/avresample.h>
30 #include <libavutil/opt.h>
33 #include "aubio_priv.h"
36 #include "source_avcodec.h"
38 #define MAX_CHANNELS 6
40 #define MAX_SAMPLES MAX_CHANNELS * MAX_SIZE
42 #define SHORT_TO_FLOAT(x) (smpl_t)(x * 3.0517578125e-05)
44 struct _aubio_source_avcodec_t {
49 // some data about the file
55 AVFormatContext *avFormatCtx;
56 AVCodecContext *avCodecCtx;
59 AVAudioResampleContext *avr;
65 aubio_source_avcodec_t * new_aubio_source_avcodec(char_t * path, uint_t samplerate, uint_t hop_size) {
66 aubio_source_avcodec_t * s = AUBIO_NEW(aubio_source_avcodec_t);
70 AUBIO_ERR("Aborted opening null path\n");
74 s->hop_size = hop_size;
78 // try opening the file and get some info about it
79 // register all formats and codecs
83 AVFormatContext *avFormatCtx = s->avFormatCtx;
85 if ( (err = avformat_open_input(&avFormatCtx, s->path, NULL, NULL) ) < 0 ) {
86 uint8_t errorstr_len = 128;
87 char errorstr[errorstr_len];
88 if (av_strerror (err, errorstr, errorstr_len) == 0) {
89 AUBIO_ERR("Failed opening %s (%s)\n", s->path, errorstr);
91 AUBIO_ERR("Failed opening %s (unknown error)\n", s->path);
96 // retrieve stream information
97 if ( (err = avformat_find_stream_info(avFormatCtx, NULL)) < 0 ) {
98 uint8_t errorstr_len = 128;
99 char errorstr[errorstr_len];
100 if (av_strerror (err, errorstr, errorstr_len) == 0) {
101 AUBIO_ERR("Could not find stream information for %s (%s)\n", s->path, errorstr);
103 AUBIO_ERR("Could not find stream information for %s (unknown error)\n", s->path);
108 // Dump information about file onto standard error
109 //av_dump_format(avFormatCtx, 0, s->path, 0);
112 sint_t selected_stream = -1;
113 for (i = 0; i < avFormatCtx->nb_streams; i++) {
114 if (avFormatCtx->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO) {
115 if (selected_stream == -1) {
118 AUBIO_WRN("More than one audio stream in %s, taking the first one\n", s->path);
122 if (selected_stream == -1) {
123 AUBIO_ERR("No audio stream in %s\n", s->path);
127 //AUBIO_DBG("Taking stream %d in file %s\n", selected_stream, s->path);
129 AVCodecContext *avCodecCtx = s->avCodecCtx;
130 avCodecCtx = avFormatCtx->streams[selected_stream]->codec;
131 AVCodec *codec = avcodec_find_decoder(avCodecCtx->codec_id);
133 AUBIO_ERR("Could not find decoder for %s", s->path);
137 if ( ( err = avcodec_open2(avCodecCtx, codec, NULL) ) < 0) {
138 uint8_t errorstr_len = 128;
139 char errorstr[errorstr_len];
140 if (av_strerror (err, errorstr, errorstr_len) == 0) {
141 AUBIO_ERR("Could not load codec for %s (%s)\n", s->path, errorstr);
143 AUBIO_ERR("Could not load codec for %s (unknown error)\n", s->path);
148 /* get input specs */
149 s->input_samplerate = avCodecCtx->sample_rate;
150 s->input_channels = avCodecCtx->channels;
152 //AUBIO_DBG("input_samplerate: %d\n", s->input_samplerate);
153 //AUBIO_DBG("input_channels: %d\n", s->input_channels);
155 if (samplerate == 0) {
156 samplerate = s->input_samplerate;
157 //AUBIO_DBG("sampling rate set to 0, automagically adjusting to %d\n", samplerate);
159 s->samplerate = samplerate;
161 int64_t input_layout = av_get_default_channel_layout(s->input_channels);
162 int64_t mono_layout = av_get_default_channel_layout(1);
164 AVAudioResampleContext *avr = s->avr;
165 avr = avresample_alloc_context();
166 av_opt_set_int(avr, "in_channel_layout", input_layout, 0);
167 av_opt_set_int(avr, "out_channel_layout", mono_layout, 0);
168 av_opt_set_int(avr, "in_sample_rate", s->input_samplerate, 0);
169 av_opt_set_int(avr, "out_sample_rate", s->samplerate, 0);
170 av_opt_set_int(avr, "in_sample_fmt", avCodecCtx->sample_fmt, 0);
171 av_opt_set_int(avr, "out_sample_fmt", AV_SAMPLE_FMT_S16, 0);
172 if ( ( err = avresample_open(avr) ) < 0) {
173 uint8_t errorstr_len = 128;
174 char errorstr[errorstr_len];
175 if (av_strerror (err, errorstr, errorstr_len) == 0) {
176 AUBIO_ERR("Could not open AVAudioResampleContext for %s (%s)\n", s->path, errorstr);
178 AUBIO_ERR("Could not open AVAudioResampleContext for %s (unknown error)\n", s->path);
183 AVFrame *avFrame = s->avFrame;
184 avFrame = avcodec_alloc_frame();
186 AUBIO_ERR("Could not allocate frame for (%s)\n", s->path);
188 AVPacket avPacket = s->avPacket;
189 av_init_packet(&avPacket);
191 s->output = (int16_t *)av_malloc(FF_MIN_BUFFER_SIZE * sizeof(int16_t));
199 /* allocate data for de/interleaving reallocated when needed. */
200 //s->scratch_size = s->input_hop_size * s->input_channels;
201 //s->scratch_data = AUBIO_ARRAY(float,s->scratch_size);
202 s->avFormatCtx = avFormatCtx;
203 s->avCodecCtx = avCodecCtx;
204 s->avFrame = avFrame;
205 s->avPacket = avPacket;
208 //av_log_set_level(AV_LOG_QUIET);
213 AUBIO_ERR("can not read %s at samplerate %dHz with a hop_size of %d\n",
214 s->path, s->samplerate, s->hop_size);
215 del_aubio_source_avcodec(s);
219 void aubio_source_avcodec_readframe(aubio_source_avcodec_t *s, int * read_samples) {
220 AVFormatContext *avFormatCtx = s->avFormatCtx;
221 AVCodecContext *avCodecCtx = s->avCodecCtx;
222 AVFrame *avFrame = s->avFrame;
223 AVPacket avPacket = s->avPacket;
224 AVAudioResampleContext *avr = s->avr;
225 int16_t *output = s->output;
228 int err = av_read_frame (avFormatCtx, &avPacket);
230 //AUBIO_ERR("Could not read frame for (%s)\n", s->path);
236 int len = avcodec_decode_audio4(avCodecCtx, avFrame, &got_frame, &avPacket);
239 AUBIO_ERR("Error while decoding %s\n", s->path);
243 if (got_frame == 0) {
244 AUBIO_ERR("Could not get frame for (%s)\n", s->path);
247 av_samples_get_buffer_size(NULL,
248 avCodecCtx->channels, avFrame->nb_samples,
249 avCodecCtx->sample_fmt, 1);
250 //AUBIO_WRN("Got data_size %d frame for (%s)\n", data_size, s->path);
253 int in_samples = avFrame->nb_samples;
254 for (i = 0; i < in_samples; i ++) {
255 //AUBIO_WRN("%d\n", avFrame->data[i]);
260 int in_plane_size = 0; //avFrame->linesize[0];
261 int out_plane_size = 0; //sizeof(float); //in_samples * sizeof(float);
262 int max_out_samples = FF_MIN_BUFFER_SIZE;
263 if (avresample_convert ( avr,
264 (void **)&output, out_plane_size, max_out_samples,
265 (void **)avFrame->data, in_plane_size, in_samples) < 0) {
266 AUBIO_ERR("Could not convert frame (%s)\n", s->path);
268 //AUBIO_ERR("Got in_plane_size %d frame for (%s)\n", in_plane_size, s->path);
269 //AUBIO_WRN("Delay is %d for %s\n", avresample_get_delay(avr), s->path);
270 //AUBIO_WRN("max_out_samples is %d for FF_MIN_BUFFER_SIZE %d\n",
271 // max_out_samples, FF_MIN_BUFFER_SIZE);
273 int out_samples = avresample_available(avr) + (avresample_get_delay(avr)
274 + in_samples) * s->samplerate / s->input_samplerate;
275 //AUBIO_WRN("Converted %d to %d samples\n", in_samples, out_samples);
276 for (i = 0; i < out_samples; i ++) {
277 //AUBIO_DBG("%f\n", SHORT_TO_FLOAT(output[i]));
280 int in_plane_size = 0; //avFrame->linesize[0];
281 int out_plane_size = 0; //sizeof(float); //in_samples * sizeof(float);
282 int max_out_samples = FF_MIN_BUFFER_SIZE;
283 if (avresample_convert ( avr,
284 NULL, out_plane_size, max_out_samples,
285 (uint8_t **)avFrame->data, in_plane_size, in_samples) < 0) {
286 AUBIO_ERR("Could not convert frame (%s)\n", s->path);
288 AUBIO_ERR("Got in_plane_size %d frame for (%s)\n", in_plane_size, s->path);
289 AUBIO_WRN("Delay is %d for %s\n", avresample_get_delay(avr), s->path);
290 AUBIO_WRN("max_out_samples is %d for FF_MIN_BUFFER_SIZE %d\n",
291 max_out_samples, FF_MIN_BUFFER_SIZE);
293 //int out_samples = avresample_available (avr) +
294 // (avresample_get_delay(avr) + in_samples) * s->samplerate / s->input_samplerate;
295 int out_samples = avresample_available(avr);
296 AUBIO_WRN("Found %d samples available for %s\n", avresample_available(avr), s->path);
298 err = avresample_read (avr, (uint8_t **)output, out_samples);
300 AUBIO_WRN("Error %d while reading %s\n", err, s->path);
302 AUBIO_WRN("Converted %d to %d samples\n", in_samples, out_samples);
304 for (i = 0; i < out_samples; i ++) {
305 AUBIO_WRN("%f\n", SHORT_TO_FLOAT(output[i]));
308 s->avFormatCtx = avFormatCtx;
309 s->avCodecCtx = avCodecCtx;
310 s->avFrame = avFrame;
311 s->avPacket = avPacket;
315 *read_samples = out_samples;
319 void aubio_source_avcodec_do(aubio_source_avcodec_t * s, fvec_t * read_data, uint_t * read){
321 //AUBIO_DBG("entering 'do' with %d, %d\n", s->read_samples, s->read_index);
323 if (s->read_samples == 0) {
324 int avcodec_read = 0;
325 aubio_source_avcodec_readframe(s, &avcodec_read);
326 s->read_samples += avcodec_read;
329 if (s->read_samples < s->hop_size) {
330 // write the end of the buffer to the beginning of read_data
331 int partial = s->read_samples;
332 for (i = 0; i < partial; i++) {
333 read_data->data[i] = SHORT_TO_FLOAT(s->output[i + s->read_index]);
338 int avcodec_read = 0;
339 aubio_source_avcodec_readframe(s, &avcodec_read);
340 s->read_samples += avcodec_read;
342 // write the beginning of the buffer to the end of read_data
343 int end = MIN((int)(s->hop_size), s->read_samples);
344 if (avcodec_read == 0) {
347 for (i = partial; i < end; i++) {
348 read_data->data[i] = SHORT_TO_FLOAT(s->output[i - partial + s->read_index]);
350 if (end < s->hop_size) {
351 for (i = end; i < s->hop_size; i++) {
352 read_data->data[i] = 0.;
355 s->read_index += partial;
356 s->read_samples -= partial;
359 for (i = 0; i < s->hop_size; i++) {
360 read_data->data[i] = SHORT_TO_FLOAT(s->output[i + s->read_index]);
362 s->read_index += s->hop_size;
363 s->read_samples -= s->hop_size;
368 void aubio_source_avcodec_do_multi(aubio_source_avcodec_t * s, fmat_t * read_data, uint_t * read){
369 //uint_t i,j, input_channels = s->input_channels;
372 uint_t aubio_source_avcodec_get_samplerate(aubio_source_avcodec_t * s) {
373 return s->samplerate;
376 uint_t aubio_source_avcodec_get_channels(aubio_source_avcodec_t * s) {
377 return s->input_channels;
380 uint_t aubio_source_avcodec_seek (aubio_source_avcodec_t * s, uint_t pos) {
381 //uint_t resampled_pos = (uint_t)ROUND(pos * s->input_samplerate * 1. / s->samplerate);
382 return 0; //sf_seek (s->handle, resampled_pos, SEEK_SET);
385 void del_aubio_source_avcodec(aubio_source_avcodec_t * s){
387 if (s->output != NULL) {
390 if (s->avr != NULL) {
391 avresample_close( s->avr );
395 if (s->avFrame != NULL) {
396 avcodec_free_frame( &(s->avFrame) );
399 if ( &(s->avPacket) != NULL) {
400 av_free_packet( &(s->avPacket) );
402 if (s->avCodecCtx != NULL) {
403 avcodec_close ( s->avCodecCtx );
405 s->avCodecCtx = NULL;
406 if (s->avFormatCtx != NULL) {
407 avformat_close_input ( &(s->avFormatCtx) );
410 s->avFormatCtx = NULL;
414 #endif /* HAVE_SNDFILE */