Skip to content

Instantly share code, notes, and snippets.

@jl2
Created January 26, 2012 06:33
Show Gist options
  • Save jl2/1681387 to your computer and use it in GitHub Desktop.
Save jl2/1681387 to your computer and use it in GitHub Desktop.
Use ffmpeg's libavcodec and libavformat to decode an audio file into an output buffer.
/*
readmp3.c
Copyright (c) 2012, Jeremiah LaRocco jeremiah.larocco@gmail.com
Permission to use, copy, modify, and/or distribute this software for any
purpose with or without fee is hereby granted, provided that the above
copyright notice and this permission notice appear in all copies.
THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*/
/* Use libavcodec and libavformat to decode an audio file.
To demonstrate it's working:
./readaudio some_file.mp3 > snd_data.txt
gnuplot -e "plot \"snd_data.txt\""
It should plot the waveform for the first channel of the given audio file.
*/
#include "libavformat/avformat.h"
#include "libavcodec/avcodec.h"
#include "libavutil/mem.h"
#include <stdio.h>
typedef struct audio_data_s {
uint8_t *samples;
size_t buffer_size;
size_t used_buffer_size;
size_t num_samples;
size_t sample_size;
size_t sample_rate;
int8_t channels;
double duration;
int8_t planar;
} audio_data_t;
int32_t get_sample(audio_data_t *ad, size_t idx, int8_t channel) {
int32_t rv = 0;
if (idx > ad->num_samples ||
channel<0 || channel > ad->channels) {
return rv;
}
int mul = 1;
int offset = 0;
if (ad->planar == 1) {
mul = 1;
offset = ad->num_samples * channel;
} else {
offset = channel;
mul = ad->channels;
}
switch (ad->sample_size) {
case 1:
{
int8_t tmp = ad->samples[mul * idx+offset];
rv = (int32_t)tmp;
break;
}
case 2:
{
int16_t tmp = ((int16_t*)ad->samples)[mul*idx/2 + offset];
rv = (int32_t)tmp;
break;
}
default:
rv = ((int32_t*)ad->samples)[mul*idx/4 + offset];
}
return rv;
}
int read_audio(char *fname, audio_data_t *ad);
int read_audio(char *fname, audio_data_t *ad) {
// It's important this be aligned correctly...
AVFormatContext *pFormatCtx __attribute__ ((aligned (16)));
if (avformat_open_input(&pFormatCtx, fname, NULL, 0) != 0) {
return -1;
}
avformat_find_stream_info(pFormatCtx, NULL);
if (pFormatCtx->streams[0]->codec->codec_type
!= AVMEDIA_TYPE_AUDIO) {
avformat_close_input(&pFormatCtx);
return -1;
}
AVPacket packet;
av_init_packet(&packet);
AVCodecContext *aCodecCtx;
aCodecCtx=pFormatCtx->streams[0]->codec;
AVCodec *aCodec;
aCodec = avcodec_find_decoder(aCodecCtx->codec_id);
if (!aCodec) {
avformat_close_input(&pFormatCtx);
return -1;
}
if (avcodec_open2(aCodecCtx, aCodec, NULL) < 0) {
avformat_close_input(&pFormatCtx);
return -1;
}
int gotit = 0;
AVFrame *frame = NULL;
if (!frame) {
if (!(frame = avcodec_alloc_frame())) {
avcodec_close(aCodecCtx);
avformat_close_input(&pFormatCtx);
return -1;
}
} else
avcodec_get_frame_defaults(frame);
double sec_duration = pFormatCtx->duration/(double)AV_TIME_BASE;
ad->duration = sec_duration;
int brate = pFormatCtx->bit_rate;
int xp = 0;
int total_data_size = 0;
int total_samples = 0;
int estimated_buff_size = brate *(int)floor(sec_duration)/2;
int allocated_buffer = estimated_buff_size;
ad->samples = malloc(allocated_buffer);
ad->channels = aCodecCtx->channels;
ad->sample_rate = aCodecCtx->sample_rate;
int rv = av_read_frame(pFormatCtx, &packet);
while (packet.size > 0) {
int len = avcodec_decode_audio4(pFormatCtx->streams[0]->codec,
frame, &gotit, &packet);
int plane_size;
int data_size = av_samples_get_buffer_size(
&plane_size,
aCodecCtx->channels,
frame->nb_samples,
aCodecCtx->sample_fmt, 1);
if (total_data_size+data_size > allocated_buffer) {
allocated_buffer = allocated_buffer*1.25;
ad->samples = realloc(ad->samples, allocated_buffer);
}
memcpy(ad->samples+total_data_size, frame->extended_data[0], data_size);
total_data_size += data_size;
total_samples += frame->nb_samples;
rv = av_read_frame(pFormatCtx, &packet);
}
// Use the last frame to fill in the info needed
ad->used_buffer_size = total_data_size;
ad->buffer_size = allocated_buffer;
ad->planar = av_sample_fmt_is_planar(aCodecCtx->sample_fmt);
ad->sample_size = av_get_bytes_per_sample(aCodecCtx->sample_fmt);
ad->samples = realloc(ad->samples, total_data_size);
ad->buffer_size = total_data_size;
ad->num_samples = total_samples;
avcodec_close(aCodecCtx);
avformat_close_input(&pFormatCtx);
return 1;
}
int main(int argc, char *argv[]) {
if (argc<2) {
printf("No file names given.\n");
exit(1);
}
av_register_all();
audio_data_t snd_data;
read_audio(argv[1], &snd_data);
for (size_t i = 0; i<snd_data.num_samples; ++i) {
int32_t val = get_sample(&snd_data, i, 0);
printf("%lu %d\n", i, val);
}
free(snd_data.samples);
return 0;
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment