ffmpeg 转换语音（部分参考）

#pragma warning(push)
#pragma warning(disable : 4101 4244)
int ConverAudio(const char* input_file, const char* output_file, int samples_rate, int channel)
{
AVFormatContext *infmt_ctx;
unsigned int i = 0;
//////////////////////// 初始化 ////////////////////////
//     avcodec_register_all();
//     avdevice_register_all();
av_register_all();
infmt_ctx = av_alloc_format_context();
//打开输入文件
if(av_open_input_file(&infmt_ctx, input_file, NULL, 0, NULL)!=0)
{
//debug_string("can't open input file/n");
return -1;
}
//取出流信息
if(av_find_stream_info(infmt_ctx) <0)
{
//debug_string("can't find suitable codec parameters/n");
return -2;
}
//dump_format(infmt_ctx, 0, input_file, 0); //列出输入文件的相关流信息
// 查找音频流信息
int audioindex=-1;
for(unsigned int j = 0; j < infmt_ctx->nb_streams; j++)
{
if(infmt_ctx->streams[j]->codec->codec_type == CODEC_TYPE_AUDIO)
{
audioindex=j;
break;
}
}

if(audioindex == -1) //没有找到音频流
{
//debug_string("can't find audio stream/n");
return -3;
}
AVCodecContext *incode_ctx;
AVCodec *incodec;
incode_ctx = infmt_ctx->streams[audioindex]->codec;
//找到合适的音频解码器
incodec = avcodec_find_decoder(incode_ctx->codec_id);
if(incodec == NULL)
{
//debug_string("can't find suitable audio decoder/n");
return -4;
}

//打开该音频解码器
if(avcodec_open(incode_ctx, incodec) < 0)
{
//debug_string("can't open the audio decoder/n");
return -5;
}

//////////////////////// 输出 //////////////////////// 
/* 解析输出文件的格式 */   
AVOutputFormat *outfmt = guess_format(NULL, output_file, NULL); 
if (!outfmt)
{ 
//printf("Could not deduce output format from file extension: using MPEG./n"); 
outfmt = guess_format("mpeg", NULL, NULL); 
} 
if (!outfmt)
{ 
//debug_string("Could not find suitable output format/n"); 
return -6; 
} 
outfmt->audio_codec = CODEC_ID_MP3;
/* allocate the output media context */   
AVFormatContext *outfmt_ctx = av_alloc_format_context(); 
if (!outfmt_ctx)
{ 
//debug_string("Memory error/n"); 
return -7; 
}

/* 保存输出文件的格式 */   
outfmt_ctx->oformat = outfmt; 
snprintf(outfmt_ctx->filename, sizeof(outfmt_ctx->filename), "%s", output_file); 
/*add the audio and video streams using the default format codecs and initialize the codecs*/
/* 输出文件的视频流编码器ID */
AVCodecContext *outcode_ctx;
AVCodec *outcodec;
AVStream *audio_st = NULL/*, *video_st*/;
//double audio_pts, video_pts;
//if (outfmt->video_codec != CODEC_ID_NONE) {
//  video_st = add_video_stream(outfmt_ctx, outfmt->video_codec);
//}
/* 输出文件的音频流编码器ID */
if (outfmt->audio_codec != CODEC_ID_NONE)
{
audio_st = av_new_stream(outfmt_ctx, 1);

outcode_ctx = audio_st->codec;
outcode_ctx->codec_id = outfmt->audio_codec;
outcode_ctx->codec_type = CODEC_TYPE_AUDIO;
//oAcc->bit_rate =inCodecCtx->bit_rate ;
outcode_ctx->sample_rate = samples_rate > 0 ? samples_rate : incode_ctx->sample_rate;
//outcode_ctx->time_base.den = 1;
//outcode_ctx->time_base.num = outcode_ctx->sample_rate;
outcode_ctx->channels = channel > 0 ? channel : incode_ctx->channels;
outcode_ctx->block_align = incode_ctx->block_align;
if(outcode_ctx->block_align == 1 && outcode_ctx->codec_id == CODEC_ID_MP3)
outcode_ctx->block_align = 0;
}
/* 设置输出参数 */
if (av_set_parameters(outfmt_ctx, NULL) < 0)
{
//debug_string("Invalid output format parameters/n");
return -8; 
}

/* 列出输出文件的格式信息 */   
dump_format(outfmt_ctx, 0, output_file, 1); 
strcpy(outfmt_ctx->title, infmt_ctx->title); 
strcpy(outfmt_ctx->author, infmt_ctx->author); 
strcpy(outfmt_ctx->copyright, infmt_ctx->copyright); 
strcpy(outfmt_ctx->comment, infmt_ctx->comment); 
strcpy(outfmt_ctx->album, infmt_ctx->album); 
outfmt_ctx->year = infmt_ctx->year; 
outfmt_ctx->track = infmt_ctx->track; 
strcpy(outfmt_ctx->genre, infmt_ctx->genre); 
//dump_format(oc,0,output_file_name,1);//列出输出文件的相关流信息 
//找到合适的音频编码器 
outcodec = avcodec_find_encoder(outfmt_ctx->oformat->audio_codec);//CODEC_ID_AAC
if(!outcodec)
{
//debug_string("can't find suitable audio encoder/n");
return -9;
}
if(avcodec_open(outcode_ctx, outcodec) <0)
{
//debug_string("can't open the output audio codec");
return -10;
}

/* 如果输出文件不存在, 则创建输出文件 */   
if (!(outfmt->flags & AVFMT_NOFILE))
{ 
if (url_fopen(&outfmt_ctx->pb, output_file, URL_WRONLY) < 0)
{
//debug_string("Could not open '%s'/n", output_file); 
return -11; 
}
}

int ret = 0; 

uint8_t* outbuf; 
uint8_t* audiobuf; 
uint8_t* resamplebuf; 
uint8_t* databuf; 
int outbuf_len = AVCODEC_MAX_AUDIO_FRAME_SIZE;; 
int audiobuf_len = AVCODEC_MAX_AUDIO_FRAME_SIZE; 
int resamplebuf_len = AVCODEC_MAX_AUDIO_FRAME_SIZE; 
int64_t pt; 
outbuf = (uint8_t*)av_malloc(outbuf_len); 
audiobuf = (uint8_t*)av_malloc(audiobuf_len); 
resamplebuf = (uint8_t*)av_malloc(resamplebuf_len); 
databuf = (uint8_t*)av_malloc(AVCODEC_MAX_AUDIO_FRAME_SIZE); 
AVFifoBuffer fifo; 
av_fifo_init(&fifo, av_fifo_size(&fifo)+AVCODEC_MAX_AUDIO_FRAME_SIZE); 
ReSampleContext* resample = audio_resample_init(outcode_ctx->channels,
incode_ctx->channels, outcode_ctx->sample_rate, incode_ctx->sample_rate);
av_write_header(outfmt_ctx); 
AVPacket packet; 
while(av_read_frame(infmt_ctx, &packet) >= 0)
{ 
pt = packet.pts; 
if(outcode_ctx->codec_id != CODEC_ID_NONE
&& packet.stream_index == audio_st->index)
{
if(incode_ctx->codec_id == outcode_ctx->codec_id
&& incode_ctx->sample_rate == outcode_ctx->sample_rate
&& incode_ctx->channels == incode_ctx->channels)
{ 
av_write_frame(outfmt_ctx, &packet); 
} 
else
{ 
if (packet.size > 0) 
{ 
avcodec_decode_audio(incode_ctx, (short *)audiobuf, &audiobuf_len,
packet.data, packet.size); 
if(incode_ctx->sample_rate != outcode_ctx->sample_rate
|| incode_ctx->channels != outcode_ctx->channels)
{ 
// resample the audio 
resamplebuf_len = audio_resample(resample, (short*)resamplebuf,
(short*)audiobuf, audiobuf_len / (incode_ctx->channels * 2)); 
//av_fifo_realloc(&fifo, av_fifo_size(&fifo)+audiobuf_len);  
av_fifo_write(&fifo,resamplebuf,resamplebuf_len*outcode_ctx->channels*2);
}
else
{ 
//av_fifo_realloc(&fifo, av_fifo_size(&fifo)+audiobuf_len);  
av_fifo_write(&fifo, audiobuf, audiobuf_len); 
} 
if(audiobuf_len > 0)
{ 
int frame_bytes = outcode_ctx->frame_size * 2 * outcode_ctx->channels; 
while(av_fifo_read(&fifo, databuf, frame_bytes) == 0)
{ 
AVPacket pkt; 
av_init_packet(&pkt); 
pkt.size = avcodec_encode_audio(outcode_ctx, outbuf, outbuf_len, (short*)databuf); 
if (outcode_ctx->coded_frame && outcode_ctx->coded_frame->pts != AV_NOPTS_VALUE)   
pkt.pts = av_rescale_q(outcode_ctx->coded_frame->pts,
outcode_ctx->time_base, audio_st->time_base);
pkt.stream_index = audio_st->index; 
pkt.data = outbuf; 
pkt.flags |= PKT_FLAG_KEY; 
// write the compressed frame in the media file 
if (pkt.size > 0) 
{ 
ret = av_write_frame(outfmt_ctx, &pkt); 
} 
av_free_packet(&pkt); 
} 
} 
} //End if 
} 
} 
av_free_packet(&packet); 
} 

av_fifo_free(&fifo); 
/* write the trailer, if any */
av_write_trailer(outfmt_ctx); 
av_free(outbuf); 
av_free(audiobuf); 
av_free(resamplebuf); 
av_free(databuf); 

audio_resample_close(resample); 
avcodec_close(outcode_ctx); 
/* free the streams */
for(i = 0; i < outfmt_ctx->nb_streams; i++) { 
av_freep(&outfmt_ctx->streams[i]->codec); 
av_freep(&outfmt_ctx->streams[i]); 
} 
if (!(outfmt->flags & AVFMT_NOFILE)) { 
/* close the output file */
url_fclose(outfmt_ctx->pb); 
} 
av_free(outfmt_ctx); 

avcodec_close(incode_ctx); 
av_close_input_file(infmt_ctx); 
return ret; 


return 1;
}
#pragma warning(pop)