Menu Home

FFMPEG + SDL音频播放分析

抽象流程:

设置SDL的音频参数 —-> 打开声音设备,播放静音 —-> ffmpeg读取音频流中数据放入队列 —-> SDL调用用户设置的函数来获取音频数据 —-> 播放音频

SDL内部维护了一个buffer来存放解码后的数据,这个buffer中的数据来源是我们注册的回调函数(audio_callback),audio_callback调用audio_decode_frame来做具体的音频解码工作,需要引起注意的是:从流中读取出的一个音频包(avpacket)可能含有多个音频桢(avframe),所以需要多次调用avcodec_decode_audio4来完成整个包的解码,解码出来的数据存放在我们自己的缓冲中(audio_buf2)。SDL每一次回调都会引起数据从audio_buf2拷贝到SDL内部缓冲区,当audio_buf2中的数据大于SDL的缓冲区大小时,需要分多次拷贝。

关键实现:

main()函数

int main(int argc, char **argv){
    SDL_Event event; //SDL事件变量
    VideoState    *is; // 纪录视频及解码器等信息的大结构体
    is = (VideoState*) av_mallocz(sizeof(VideoState));
    if(argc < 2){
        fprintf(stderr, "Usage: play <file>\n");
        exit(1);
    }
    av_register_all(); //注册所有ffmpeg的解码器
    /* 初始化SDL,这里只实用了AUDIO,如果有视频,好需要SDL_INIT_VIDEO等等 */
    if(SDL_Init(SDL_INIT_AUDIO)){
        fprintf(stderr, "Count not initialize SDL - %s\n", SDL_GetError());
        exit(1);
    }
    is_strlcpy(is->filename, argv[1], sizeof(is->filename));
    /* 创建一个SDL线程来做视频解码工作,主线程进入SDL事件循环 */
    is->parse_tid = SDL_CreateThread(decode_thread, is);
    if(!is->parse_tid){
        SDL_WaitEvent(&event);
        switch(event.type){
            case FF_QUIT_EVENT:
            case SDL_QUIT:
                 is->quit = 1;
                SDL_Quit();
                exit(0);
                break;
            default:
                 break;
        }
    }
    return 0;
}

decode_thread()读取文件信息和音频包

static int decode_thread(void *arg){
    VideoState *is = (VideoState*)arg;
    AVFormatContext *ic = NULL;
    AVPacket pkt1, *packet = &pkt1;
    int ret, i, audio_index = -1;

    is->audioStream = -1;
    global_video_state = is; 
    /*  使用ffmpeg打开视频,解码器等 常规工作 */
    if(avFormat_open_input(&ic, is->filename, NULL,  NULL) != 0)  {
        fprintf(stderr, "open file error: %s\n", is->filename);
        return -1;
    }
    is->ic = ic;
    if(avformat_find_stream_info(ic, NULL) < 0){
        fprintf(stderr, "find stream info error\n");
        return -1;
    }
    av_dump_format(ic, 0, is->filename, 0);
    for(i  = 0; i < ic->nb_streams; i++){
         if(ic->streams[i])->codec->codec_type == AVMEDIA_TYPE_AUDIO && audio_index == -1){
            audio_index = i;
            break;
        }
    }
    if(audio_index >= 0) {
        /* 所有设置SDL音频流信息的步骤都在这个函数里完成 */
        stream_component_open(is, audio_index);
    }
    if(is->audioStream < 0){
        fprintf(stderr, "could not open codecs for file: %s\n", is->filename);
        goto fail;
    }
    /* 读包的主循环, av_read_frame不停的从文件中读取数据包(这里只取音频包)*/
    for(;;){
        if(is->quit) break;
        /* 这里audioq.size是指队列中的所有数据包带的音频数据的总量,并不是包的数量 */
        if(is->audioq.size > MAX_AUDIO_SIZE){
            SDL_Delay(10); // 毫秒
            continue;
        }
         ret = av_read_frame(is->ic, packet);
         if(ret < 0){
                if(ret == AVERROR_EOF || url_feof(is->ic->pb))    break;
                if(is->ic->pb && is->ic->pb->error)    break;
                contiue;                  
          }  
          if(packet->stream_index == is->audioStream){
                    packet_queue_put(&is->audioq, packet);
           } else{
                     av_free_packet(packet);
            }
    }
     while(!is->quit)    SDL_Delay(100);
fail: {
               SDL_Event event;
               event.type = FF_QUIT_EVENT;
               event.user.data1 = is;
               SDL_PushEvent(&event);
        }
        return 0;
}

stream_component_open():设置音频参数和打开设备

int stream_component_open(videoState *is, int stream_index){
    AVFormatContext *ic = is->ic;
    AVCodecContext *codecCtx;
    AVCodec *codec;
    /* 在用SDL_OpenAudio()打开音频设备的时候需要这两个参数*/
    /* wanted_spec是我们期望设置的属性,spec是系统最终接受的参数 */
    /* 我们需要检查系统接受的参数是否正确 */
    SDL_AudioSpec wanted_spec, spec;
    int64_t wanted_channel_layout = 0; // 声道布局(SDL中的具体定义见“FFMPEG结构体”部分) 
    int wanted_nb_channels; // 声道数
    /*  SDL支持的声道数为 1, 2, 4, 6 */
    /*  后面我们会使用这个数组来纠正不支持的声道数目 */
    const int next_nb_channels[] = { 0, 0, 1, 6,  2, 6, 4, 6 }; 

    if(stream_index < 0 || stream_index >= ic->nb_streams)    return -1;
    codecCtx = ic->streams[stream_index]->codec;
    wanted_nb_channels = codecCtx->channels;
    if(!wanted_channel_layout || wanted_nb_channels != av_get_channel_layout_nb_channels(wanted_channel_layout)) {
        wanted_channel_layout = av_get_default_channel_lauout(wanted_channel_nb_channels);
        wanted_channel_layout &= ~AV_CH_LAYOUT_STEREO_DOWNMIX;
    }
    wanted_spec.channels = av_get_channels_layout_nb_channels(wanted_channel_layout);
    wanted_spec.freq = codecCtx->sample_rate;
    if(wanted_spec.freq <= 0 || wanted_spec.channels <=0){
           fprintf(stderr, "Invaild sample rate or channel count!\n");
            return -1;
    }
    wanted_spec.format = AUDIO_S16SYS; // 具体含义请查看“SDL宏定义”部分
    wanted_spec.silence = 0; // 0指示静音
    wanted_spec.samples = SDL_AUDIO_BUFFER_SIZE; // 自定义SDL缓冲区大小
    wanted_spec.callback = audio_callback; // 音频解码的关键回调函数
    wanted_spec.userdata = is; // 传给上面回调函数的外带数据

    /*  打开音频设备,这里使用一个while来循环尝试打开不同的声道数(由上面 */
    /*  next_nb_channels数组指定)直到成功打开,或者全部失败 */
    while(SDL_OpenAudio(&wanted_spec, &spec) < 0){
        fprintf(stderr, "SDL_OpenAudio(%d channels): %s\n", wanted_spec.channels, SDL_GetError());
        wanted_spec.channels = next_nb_channels[FFMIN(7, wanted_spec.channels)]; // FFMIN()由ffmpeg定义的宏,返回较小的数
        if(!wanted_spec.channels){
              fprintf(stderr, "No more channel to try\n");
              return -1;
        }
        wanted_channel_layout = av_get_default_channel_layout(wanted_spec.channels);
    }
    /* 检查实际使用的配置(保存在spec,由SDL_OpenAudio()填充) */
    if(spec.format != AUDIO_S16SYS){
        fprintf(stderr, "SDL advised audio format %d is not supported\n", spec.format);
        return -1;
    }
    if(spec.channels != wanted_spec.channels) {
        wanted_channel_layout = av_get_default_channel_layout(spec.channels);
        if(!wanted_channel_layout){
                fprintf(stderr, "SDL advised channel count %d is not support\n", spec.channels);
                return -1;
        }
    }
    /* 把设置好的参数保存到大结构中 */
    is->audio_src_fmt = is->audio_tgt_fmt = AV_SAMPLE_FMT_S16;
    is->audio_src_freq = is->audio_tgt_freq = spec.freq;
    is->audio_src_channel_layout = is->audio_tgt_layout = wanted_channel_layout;
    is->audio_src_channels = is->audio_tat_channels = spec.channels;

    codec = avcodec_find_decoder(codecCtx>codec_id);
    if(!codec || (avcodec_open2(codecCtx, codec, NULL) < 0)){
        fprintf(stderr, "Unsupported codec!\n");
        return -1;
    }
    ic->streams[stream_index]->discard = AVDISCARD_DEFAULT; //具体含义请查看“FFMPEG宏定义”部分
    is->audioStream = stream_index;
    is->audio_st = ic->streams[stream_index];
    is->audio_buf_size = 0;
    is->audio_buf_index = 0;
    memset(&is->audio_pkt, 0, sizeof(is->audio_pkt));
    packet_queue_init(&is->audioq);
    SDL_PauseAudio(0); // 开始播放静音
}

audio_callback(): 回调函数,向SDL缓冲区填充数据

void audio_callback(void *userdata, Uint8 *stream, int len){
    VideoState *is = (VideoState*)userdata;
    int len1, audio_data_size;

    /*   len是由SDL传入的SDL缓冲区的大小,如果这个缓冲未满,我们就一直往里填充数据 */
    while(len > 0){
        /*  audio_buf_index 和 audio_buf_size 标示我们自己用来放置解码出来的数据的缓冲区,*/
        /*   这些数据待copy到SDL缓冲区, 当audio_buf_index >= audio_buf_size的时候意味着我*/
        /*   们的缓冲为空,没有数据可供copy,这时候需要调用audio_decode_frame来解码出更
        /*   多的桢数据 */
        if(is->audio_buf_index >= is->audio_buf_size){
                audio_data_size = audio_decode_frame(is);
                /* audio_data_size < 0 标示没能解码出数据,我们默认播放静音 */
                is(audio_data_size < 0){
                         is->audio_buf_size = 1024;
                         /* 清零,静音 */
			 memset(is->audio_buf, 0, is->audio_buf_size);
                } else{
                          is->audio_buf_size = audio_data_size;
                 }
                 is->audio_buf_index = 0;
        }
        /*  查看stream可用空间,决定一次copy多少数据,剩下的下次继续copy */
        len1 = is->audio_buf_size - is->audio_buf_index;
        if(len1 > len)    len1 = len;

        memcpy(stream, (uint8_t*)is->audio_buf + is->audio_buf_index, len1);
        len -= len1;
        stream += len1;
        is->audio_buf_index += len1;
    }
}

audio_decode_frame():解码音频

int audio_decode_frame(VideoState *is){
    int len1, len2, decoded_data_size;
    AVPacket *pkt = &is->audio_pkt;
    int got_frame = 0;
    int64_t dec_channel_layout;
    int wanted_nb_samples, resampled_data_size;

    for(;;){
      while(is->audio_pkt_size > 0){
        if(!is->audio_frame){
            if(!(is->audio_frame = avacodec_alloc_frame())){
                return AVERROR(ENOMEM);
            }
        } else
          avcodec_get_frame_defaults(is->audio_frame);

        len1 = avcodec_decode_audio4(is->audio_st_codec, is->audio_frame, got_frame, pkt);
        /* 解码错误,跳过整个包 */
        if(len1 < 0){
           is->audio_pkt_size = 0;
           break;
        }
        is->audio_pkt_data += len1;
        is->audio_pkt_size -= len1;
        if(!got_frame)   continue;
        /* 计算解码出来的桢需要的缓冲大小 */
        decoded_data_size = av_samples_get_buffer_size(NULL,
                            is->audio_frame_channels,
                            is->audio_frame_nb_samples,
                            is->audio_frame_format, 1);
        dec_channel_layout = (is->audio_frame->channel_layout && is->audio_frame->channels
                   == av_get_channel_layout_nb_channels(is->audio_frame->channel_layout))
                   ? is->audio_frame->channel_layout : av_get_default_channel_layout(is->audio_frame->channels);                       
        wanted_nb_samples =  is->audio_frame->nb_samples;
        if (is->audio_frame->format != is->audio_src_fmt || 
            dec_channel_layout != is->audio_src_channel_layout ||
            is->audio_frame->sample_rate != is->audio_src_freq || 
            (wanted_nb_samples != is->audio_frame->nb_samples && !is->swr_ctx)) {
                if (is->swr_ctx) swr_free(&is->swr_ctx);
                is->swr_ctx = swr_alloc_set_opts(NULL,
                                                 is->audio_tgt_channel_layout,
                                                 is->audio_tgt_fmt,
                                                 is->audio_tgt_freq,
                                                 dec_channel_layout,
                                                 is->audio_frame->format,
                                                 is->audio_frame->sample_rate,
                                                 0, NULL);
                 if (!is->swr_ctx || swr_init(is->swr_ctx) < 0) {
                     fprintf(stderr, "swr_init() failed\n");
                     break;
                 }
                 is->audio_src_channel_layout = dec_channel_layout;
                 is->audio_src_channels = is->audio_st->codec->channels;
                 is->audio_src_freq = is->audio_st->codec->sample_rate;
                 is->audio_src_fmt = is->audio_st->codec->sample_fmt;
         }
         /* 这里我们可以对采样数进行调整,增加或者减少,一般可以用来做声画同步 */
         if (is->swr_ctx) {
             const uint8_t **in = (const uint8_t **)is->audio_frame->extended_data;
             uint8_t *out[] = { is->audio_buf2 };
             if (wanted_nb_samples != is->audio_frame->nb_samples) {
                if(swr_set_compensation(is->swr_ctx, 
                  (wanted_nb_samples - is->audio_frame->nb_samples)*is->audio_tgt_freq/is->audio_frame->sample_rate,
                   wanted_nb_samples * is->audio_tgt_freq/is->audio_frame->sample_rate) < 0) {
                        fprintf(stderr, "swr_set_compensation() failed\n");
                        break;
                   }
             }
             len2 = swr_convert(is->swr_ctx, out,  
                  sizeof(is->audio_buf2)/is->audio_tgt_channels/av_get_bytes_per_sample(is->audio_tgt_fmt),  
                  in, is->audio_frame->nb_samples);
             if (len2 < 0) {
                  fprintf(stderr, "swr_convert() failed\n");
                  break;
             }
             if(len2 == sizeof(is->audio_buf2)/is->audio_tgt_channels/av_get_bytes_per_sample(is->audio_tgt_fmt)) {
                 fprintf(stderr, "warning: audio buffer is probably too small\n");
                 swr_init(is->swr_ctx);
             }
             is->audio_buf = is->audio_buf2;
             resampled_data_size = len2*is->audio_tgt_channels*av_get_bytes_per_sample(is->audio_tgt_fmt);
           } else {
             resampled_data_size = decoded_data_size;
             is->audio_buf = is->audio_frame->data[0];
           }
           /*  返回得到的数据 */
           return resampled_data_size;
       }
       if (pkt->data) av_free_packet(pkt);
       memset(pkt, 0, sizeof(*pkt));
       if (is->quit) return -1;
       if (packet_queue_get(&is->audioq, pkt, 1) < 0) return -1;
       is->audio_pkt_data = pkt->data;
       is->audio_pkt_size = pkt->size;

     }
}

FFMPEG结构体

channel_layout_map

static const struct {
const char *name;
int nb_channels;
uint64_t layout;
} channel_layout_map[] = {
{ "mono", 1, AV_CH_LAYOUT_MONO },
{ "stereo", 2, AV_CH_LAYOUT_STEREO },
{ "2.1", 3, AV_CH_LAYOUT_2POINT1 },
{ "3.0", 3, AV_CH_LAYOUT_SURROUND },
{ "3.0(back)", 3, AV_CH_LAYOUT_2_1 },
{ "4.0", 4, AV_CH_LAYOUT_4POINT0 },
{ "quad", 4, AV_CH_LAYOUT_QUAD },
{ "quad(side)", 4, AV_CH_LAYOUT_2_2 },
{ "3.1", 4, AV_CH_LAYOUT_3POINT1 },
{ "5.0", 5, AV_CH_LAYOUT_5POINT0_BACK },
{ "5.0(side)", 5, AV_CH_LAYOUT_5POINT0 },
{ "4.1", 5, AV_CH_LAYOUT_4POINT1 },
{ "5.1", 6, AV_CH_LAYOUT_5POINT1_BACK },
{ "5.1(side)", 6, AV_CH_LAYOUT_5POINT1 },
{ "6.0", 6, AV_CH_LAYOUT_6POINT0 },
{ "6.0(front)", 6, AV_CH_LAYOUT_6POINT0_FRONT },
{ "hexagonal", 6, AV_CH_LAYOUT_HEXAGONAL },
{ "6.1", 7, AV_CH_LAYOUT_6POINT1 },
{ "6.1", 7, AV_CH_LAYOUT_6POINT1_BACK },
{ "6.1(front)", 7, AV_CH_LAYOUT_6POINT1_FRONT },
{ "7.0", 7, AV_CH_LAYOUT_7POINT0 },
{ "7.0(front)", 7, AV_CH_LAYOUT_7POINT0_FRONT },
{ "7.1", 8, AV_CH_LAYOUT_7POINT1 },
{ "7.1(wide)", 8, AV_CH_LAYOUT_7POINT1_WIDE },
{ "octagonal", 8, AV_CH_LAYOUT_OCTAGONAL },
{ "downmix", 2, AV_CH_LAYOUT_STEREO_DOWNMIX, },
};

FFMPEG宏定义

Audio channel convenience macros

 #define AV_CH_LAYOUT_MONO              (AV_CH_FRONT_CENTER)
 #define AV_CH_LAYOUT_STEREO            (AV_CH_FRONT_LEFT|AV_CH_FRONT_RIGHT)
 #define AV_CH_LAYOUT_2POINT1           (AV_CH_LAYOUT_STEREO|AV_CH_LOW_FREQUENCY)
 #define AV_CH_LAYOUT_2_1               (AV_CH_LAYOUT_STEREO|AV_CH_BACK_CENTER)
 #define AV_CH_LAYOUT_SURROUND          (AV_CH_LAYOUT_STEREO|AV_CH_FRONT_CENTER)
 #define AV_CH_LAYOUT_3POINT1           (AV_CH_LAYOUT_SURROUND|AV_CH_LOW_FREQUENCY)
 #define AV_CH_LAYOUT_4POINT0           (AV_CH_LAYOUT_SURROUND|AV_CH_BACK_CENTER)
 #define AV_CH_LAYOUT_4POINT1           (AV_CH_LAYOUT_4POINT0|AV_CH_LOW_FREQUENCY)
 #define AV_CH_LAYOUT_2_2               (AV_CH_LAYOUT_STEREO|AV_CH_SIDE_LEFT|AV_CH_SIDE_RIGHT)
 #define AV_CH_LAYOUT_QUAD              (AV_CH_LAYOUT_STEREO|AV_CH_BACK_LEFT|AV_CH_BACK_RIGHT)
 #define AV_CH_LAYOUT_5POINT0           (AV_CH_LAYOUT_SURROUND|AV_CH_SIDE_LEFT|AV_CH_SIDE_RIGHT)
 #define AV_CH_LAYOUT_5POINT1           (AV_CH_LAYOUT_5POINT0|AV_CH_LOW_FREQUENCY)
 #define AV_CH_LAYOUT_5POINT0_BACK      (AV_CH_LAYOUT_SURROUND|AV_CH_BACK_LEFT|AV_CH_BACK_RIGHT)
 #define AV_CH_LAYOUT_5POINT1_BACK      (AV_CH_LAYOUT_5POINT0_BACK|AV_CH_LOW_FREQUENCY)
 #define AV_CH_LAYOUT_6POINT0           (AV_CH_LAYOUT_5POINT0|AV_CH_BACK_CENTER)
 #define AV_CH_LAYOUT_6POINT0_FRONT     (AV_CH_LAYOUT_2_2|AV_CH_FRONT_LEFT_OF_CENTER|AV_CH_FRONT_RIGHT_OF_CENTER)
 #define AV_CH_LAYOUT_HEXAGONAL         (AV_CH_LAYOUT_5POINT0_BACK|AV_CH_BACK_CENTER)
 #define AV_CH_LAYOUT_6POINT1           (AV_CH_LAYOUT_5POINT1|AV_CH_BACK_CENTER)
 #define AV_CH_LAYOUT_6POINT1_BACK      (AV_CH_LAYOUT_5POINT1_BACK|AV_CH_BACK_CENTER)
 #define AV_CH_LAYOUT_6POINT1_FRONT     (AV_CH_LAYOUT_6POINT0_FRONT|AV_CH_LOW_FREQUENCY)
 #define AV_CH_LAYOUT_7POINT0           (AV_CH_LAYOUT_5POINT0|AV_CH_BACK_LEFT|AV_CH_BACK_RIGHT)
 #define AV_CH_LAYOUT_7POINT0_FRONT     (AV_CH_LAYOUT_5POINT0|AV_CH_FRONT_LEFT_OF_CENTER|AV_CH_FRONT_RIGHT_OF_CENTER)
 #define AV_CH_LAYOUT_7POINT1           (AV_CH_LAYOUT_5POINT1|AV_CH_BACK_LEFT|AV_CH_BACK_RIGHT)
#define AV_CH_LAYOUT_7POINT1_WIDE      (AV_CH_LAYOUT_5POINT1|AV_CH_FRONT_LEFT_OF_CENTER|AV_CH_FRONT_RIGHT_OF_CENTER)
#define AV_CH_LAYOUT_7POINT1_WIDE_BACK (AV_CH_LAYOUT_5POINT1_BACK|AV_CH_FRONT_LEFT_OF_CENTER|AV_CH_FRONT_RIGHT_OF_CENTER)
#define AV_CH_LAYOUT_OCTAGONAL         (AV_CH_LAYOUT_5POINT0|AV_CH_BACK_LEFT|AV_CH_BACK_CENTER|AV_CH_BACK_RIGHT)
#define AV_CH_LAYOUT_STEREO_DOWNMIX    (AV_CH_STEREO_LEFT|AV_CH_STEREO_RIGHT)

SDL宏定义

SDL_AudioSpec format

AUDIO_U8           Unsigned 8-bit samples
AUDIO_S8            Signed 8-bit samples
AUDIO_U16LSB    Unsigned 16-bit samples, in little-endian byte order
AUDIO_S16LSB    Signed 16-bit samples, in little-endian byte order
AUDIO_U16MSB    Unsigned 16-bit samples, in big-endian byte order
AUDIO_S16MSB    Signed 16-bit samples, in big-endian byte order
AUDIO_U16           same as AUDIO_U16LSB (for backwards compatability probably)
AUDIO_S16           same as AUDIO_S16LSB (for backwards compatability probably)
AUDIO_U16SYS    Unsigned 16-bit samples, in system byte order
AUDIO_S16SYS     Signed 16-bit samples, in system byte order

git clone https://github.com/lnmcc/musicPlayer.git

Categories: ffmpeg sdl 视频编解码

Tagged as:

lnmcc

1 reply

Leave a Reply

Your email address will not be published.