轉(zhuǎn)碼一般流程

1. 獲取音頻數(shù)據(jù)(AAC,MP3等)
2. 解碼（獲取音頻原始采樣數(shù)據(jù)pcm）
3. 編碼（對(duì)pcm進(jìn)行編碼）

使用ffmpeg的函數(shù)表示的大概流程

//初始化輸入
avformat_open_input()           -- 打開對(duì)應(yīng)音頻文件
avformat_find_stream_info()     -- 從輸入文件中獲取到流的相關(guān)信息昧旨，例如：文件中流的數(shù)量

//初始化解碼器
avcodec_find_decoder()          -- 根據(jù)ffmpeg提供的解碼器id，找到對(duì)應(yīng)的解碼器
avcodec_open2()                     -- 打開解碼器

//初始化輸出
avformat_alloc_context()        -- 創(chuàng)建輸出信息上下文
avcodec_find_encoder()          -- 找到編碼器
avcodec_alloc_context3()            -- 根據(jù)編碼器您机，初始化編碼上下文
avcodec_open2()                 -- 打開編碼器

//初始化一個(gè)FIFO(先進(jìn)先出)
av_audio_fifo_alloc()           -- 初始化一個(gè)先進(jìn)先出的緩存订歪，用了存儲(chǔ)解碼后的pcm數(shù)據(jù)

//初始化音頻重采用器
swr_alloc_set_opts()            -- 設(shè)置轉(zhuǎn)化器參數(shù)
swr_init()                      -- 初始化轉(zhuǎn)換器

//開始音頻轉(zhuǎn)換

while(finished) {
// 解碼
av_read_frame()                     -- 讀取要進(jìn)行轉(zhuǎn)碼的數(shù)據(jù)
avcodec_decode_audio4()         -- 進(jìn)行解碼
av_samples_alloc()              -- 創(chuàng)建樣本空間
swr_convert()                       -- 數(shù)據(jù)重采樣
av_audio_fifo_write()           -- 將數(shù)據(jù)存儲(chǔ)到fifo緩存中

// 編碼
av_audio_fifo_read()            -- 從fifo緩存中讀取pcm數(shù)據(jù)
avcodec_encode_audio2()         -- 將數(shù)據(jù)進(jìn)行編碼

}

知識(shí)點(diǎn)

AVSampleFormat(樣本數(shù)據(jù)格式)

在說明該格式之前脖祈，先說一個(gè)樣本數(shù)據(jù)存儲(chǔ)的2種方式packet和planar。

packed - 將所有聲道的數(shù)據(jù)刷晋，交替的存儲(chǔ)成一維數(shù)組

planar - 每一個(gè)聲道單獨(dú)存放,一個(gè)二維數(shù)組表示撒犀，每一行代表一個(gè)聲道福压。

例如：

現(xiàn)在有一段音頻采用掏秩，左聲道用L表示或舞，右聲道用R表示

LRLRLR......LR 表示使用了packed方式存儲(chǔ)

LLLLL..L 和 RRRRR..R 表示使用了planar方式存儲(chǔ)

在ffmpeg中，采樣樣本的數(shù)據(jù)格式有如下幾種類型

enum AVSampleFormat {
    AV_SAMPLE_FMT_NONE = -1,
    AV_SAMPLE_FMT_U8,          ///< unsigned 8 bits
    AV_SAMPLE_FMT_S16,         ///< signed 16 bits
    AV_SAMPLE_FMT_S32,         ///< signed 32 bits
    AV_SAMPLE_FMT_FLT,         ///< float
    AV_SAMPLE_FMT_DBL,         ///< double

    AV_SAMPLE_FMT_U8P,         ///< unsigned 8 bits, planar
    AV_SAMPLE_FMT_S16P,        ///< signed 16 bits, planar
    AV_SAMPLE_FMT_S32P,        ///< signed 32 bits, planar
    AV_SAMPLE_FMT_FLTP,        ///< float, planar
    AV_SAMPLE_FMT_DBLP,        ///< double, planar

    AV_SAMPLE_FMT_NB           ///< Number of sample formats. DO NOT USE if linking dynamically
};

AVCodecContext的屬性

channel_layout

channel_layout表示聲道布局蒙幻，代表左右聲道形成的立體聲映凳，還是單聲道。

在ffmpeg中邮破，支持的聲道布局有

/**
 * @defgroup channel_masks Audio channel masks
 *
 * A channel layout is a 64-bits integer with a bit set for every channel.
 * The number of bits set must be equal to the number of channels.
 * The value 0 means that the channel layout is not known.
 * @note this data structure is not powerful enough to handle channels
 * combinations that have the same channel multiple times, such as
 * dual-mono.
 *
 * @{
 */
#define AV_CH_FRONT_LEFT             0x00000001 
#define AV_CH_FRONT_RIGHT            0x00000002
#define AV_CH_FRONT_CENTER           0x00000004
#define AV_CH_LOW_FREQUENCY          0x00000008
#define AV_CH_BACK_LEFT              0x00000010
#define AV_CH_BACK_RIGHT             0x00000020
#define AV_CH_FRONT_LEFT_OF_CENTER   0x00000040
#define AV_CH_FRONT_RIGHT_OF_CENTER  0x00000080
#define AV_CH_BACK_CENTER            0x00000100
#define AV_CH_SIDE_LEFT              0x00000200
#define AV_CH_SIDE_RIGHT             0x00000400
#define AV_CH_TOP_CENTER             0x00000800
#define AV_CH_TOP_FRONT_LEFT         0x00001000
#define AV_CH_TOP_FRONT_CENTER       0x00002000
#define AV_CH_TOP_FRONT_RIGHT        0x00004000
#define AV_CH_TOP_BACK_LEFT          0x00008000
#define AV_CH_TOP_BACK_CENTER        0x00010000
#define AV_CH_TOP_BACK_RIGHT         0x00020000
#define AV_CH_STEREO_LEFT            0x20000000  ///< Stereo downmix.
#define AV_CH_STEREO_RIGHT           0x40000000  ///< See AV_CH_STEREO_LEFT.
#define AV_CH_WIDE_LEFT              0x0000000080000000ULL
#define AV_CH_WIDE_RIGHT             0x0000000100000000ULL
#define AV_CH_SURROUND_DIRECT_LEFT   0x0000000200000000ULL
#define AV_CH_SURROUND_DIRECT_RIGHT  0x0000000400000000ULL
#define AV_CH_LOW_FREQUENCY_2        0x0000000800000000ULL

/** Channel mask value used for AVCodecContext.request_channel_layout
    to indicate that the user requests the channel order of the decoder output
    to be the native codec channel order. */
#define AV_CH_LAYOUT_NATIVE          0x8000000000000000ULL

/**
 * @}
 * @defgroup channel_mask_c Audio channel layouts
 * @{
 * */
#define AV_CH_LAYOUT_MONO              (AV_CH_FRONT_CENTER)  // 單聲道
#define AV_CH_LAYOUT_STEREO            (AV_CH_FRONT_LEFT|AV_CH_FRONT_RIGHT) // 立體聲道
#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_HEXADECAGONAL     (AV_CH_LAYOUT_OCTAGONAL|AV_CH_WIDE_LEFT|AV_CH_WIDE_RIGHT|AV_CH_TOP_BACK_LEFT|AV_CH_TOP_BACK_RIGHT|AV_CH_TOP_BACK_CENTER|AV_CH_TOP_FRONT_CENTER|AV_CH_TOP_FRONT_LEFT|AV_CH_TOP_FRONT_RIGHT)
#define AV_CH_LAYOUT_STEREO_DOWNMIX    (AV_CH_STEREO_LEFT|AV_CH_STEREO_RIGHT)

通過channels,channel_layout和sample_fmt,可以很便捷的設(shè)置聲道的相關(guān)信息

例如：

channels = 2;
channel_layout = AV_CH_LAYOUT_STEREO
sample_fmt = AV_SAMPLE_FMT_FLTP

表示 這個(gè)音頻數(shù)據(jù)有2個(gè)聲道诈豌，分別是左右聲道，聲音數(shù)據(jù)信息采用浮點(diǎn)型的planar格式進(jìn)行存儲(chǔ)抒和，即左右聲道分開存儲(chǔ)

sample_rate

sample_rate表示聲道的采樣率矫渔，表示1s內(nèi)采集的聲音樣本個(gè)數(shù)。

例如：44100表示1s內(nèi)采集了44100個(gè)聲音的樣本數(shù)據(jù)

AVFrame

AVFrame 存儲(chǔ)的數(shù)據(jù)摧莽，是解碼后的數(shù)據(jù)庙洼。即音頻中的PCM數(shù)據(jù)

data

存儲(chǔ)了音頻聲道或圖片信息

linesize

對(duì)于視頻，存儲(chǔ)了每一個(gè)圖片平面的長度

對(duì)于音頻镊辕，存儲(chǔ)了每一個(gè)聲道中數(shù)據(jù)的長度

對(duì)于音頻油够，只有linesize[0]被使用，因?yàn)橐纛l中征懈，每一個(gè)聲道的大小應(yīng)該相等

extended_data

對(duì)于視頻石咬，只是簡單的指向data[]

對(duì)于planar格式的音頻，每一個(gè)聲道有一個(gè)獨(dú)立的數(shù)據(jù)指針卖哎，并且linesize[0]包含了每一個(gè)聲道存儲(chǔ)數(shù)據(jù)的大小

nb_samples

表示這一幀中鬼悠，每個(gè)聲道中有多少個(gè)采樣點(diǎn)

詳細(xì)代碼講解

- (void)tranformateToAAC {
    
    // 創(chuàng)建文件夾
    NSArray *paths = NSSearchPathForDirectoriesInDomains(NSDocumentDirectory, NSUserDomainMask, YES);
    NSString *docDir = [paths objectAtIndex:0];

    NSFileManager *fileManager = [NSFileManager defaultManager];
    
    NSString *h264FileName = @"testMp4ToAAC.aac";
    NSString *pcmFileName = @"testMp4ToPcm.pcm";

    
    NSString *h264Path = [docDir stringByAppendingPathComponent:h264FileName];
    NSString *pcmPath = [docDir stringByAppendingPathComponent:pcmFileName];

    
    [fileManager removeItemAtPath:h264Path error:nil];
    [fileManager removeItemAtPath:pcmPath error:nil];

    
    [fileManager createFileAtPath:h264Path contents:nil attributes:nil];
    [fileManager createFileAtPath:pcmPath contents:nil attributes:nil];

    
    fileHandle = [NSFileHandle fileHandleForWritingAtPath:h264Path];
    pcmfileHandle = [NSFileHandle fileHandleForWritingAtPath:pcmPath];

    
    NSLog(@"dic = %@",docDir);
    NSString *fileName = [docDir stringByAppendingPathComponent:@"video.mp4"];

    
    av_register_all();
    avcodec_register_all();
    
    AVFormatContext *inputFormatCtx = NULL;

    // 打開輸入音頻文件
    int ret = avformat_open_input(&inputFormatCtx, [fileName UTF8String], NULL, 0);
    
    if (ret != 0) {
        NSLog(@"打開文件失敗");
        return;
    }
    
    //獲取音頻中流的相關(guān)信息
    ret = avformat_find_stream_info(inputFormatCtx, 0);
    
    if (ret != 0) {
        NSLog(@"不能獲取流信息");
        return;
    }
    
    // 獲取數(shù)據(jù)中音頻流的序列號(hào)，這是一個(gè)標(biāo)識(shí)符
    int  index = 0,audioStream = -1;
    AVCodecContext *inputCodecCtx;

    for (index = 0; index <inputFormatCtx->nb_streams; index++) {
        
        AVStream *stream = inputFormatCtx->streams[index];
        AVCodecContext *code = stream->codec;
        if (code->codec_type == AVMEDIA_TYPE_AUDIO){
            audioStream = index;
            break;
        }
    }
    
    //從音頻流中獲取輸入編解碼相關(guān)的上下文
    inputCodecCtx = inputFormatCtx->streams[audioStream]->codec;
    //查找解碼器
    AVCodec *pCodec = avcodec_find_decoder(inputCodecCtx->codec_id);
    // 打開解碼器
    int result =  avcodec_open2(inputCodecCtx, pCodec, nil);
    if (result < 0) {
        NSLog(@"打開音頻解碼器失敗");
        return;
    }
    
    // 創(chuàng)建aac編碼器
    AVCodec *aacCodec = avcodec_find_encoder(AV_CODEC_ID_AAC);
    
    if (!aacCodec){
        printf("Can not find encoder!\n");
        return ;
    }

    //常見aac編碼相關(guān)上下文信息
    AVCodecContext *aacCodeContex = avcodec_alloc_context3(aacCodec);
    // 設(shè)置編碼相關(guān)信息
    aacCodeContex->sample_fmt = aacCodec->sample_fmts[0];
    aacCodeContex->sample_rate= inputCodecCtx->sample_rate;             // 音頻的采樣率
    aacCodeContex->channel_layout = av_get_default_channel_layout(2);
    aacCodeContex->channels = inputCodecCtx->channels;
    aacCodeContex->strict_std_compliance = FF_COMPLIANCE_EXPERIMENTAL;
    
    //打開編碼器
    AVDictionary *opts = NULL;
    result = avcodec_open2(aacCodeContex, aacCodec, &opts);
    
    if (result < 0) {
        NSLog(@"failure open code");
        return;
    }
    
    //初始化先進(jìn)先出緩存隊(duì)列
    AVAudioFifo *fifo = av_audio_fifo_alloc(AV_SAMPLE_FMT_FLTP,aacCodeContex->channels, aacCodeContex->frame_size);

    //獲取編碼每幀的最大取樣數(shù)
    int output_frame_size = aacCodeContex->frame_size;
    
    // 初始化重采樣上下文
    SwrContext *resample_context = NULL;
    if (init_resampler(inputCodecCtx, aacCodeContex,
                       &resample_context)){
    }
       
    BOOL finished  = NO;
    while (1) {
        
        if (finished){
            break;
        }
        
        // 查看fifo隊(duì)列中的大小是否超過可以編碼的一幀的大小
        while (av_audio_fifo_size(fifo) < output_frame_size) {
            
            // 如果沒超過亏娜，則繼續(xù)進(jìn)行解碼
            
            if (finished)
            {
                break;
            }
            
            AVFrame *audioFrame = av_frame_alloc();
            AVPacket packet;
            packet.data = NULL;
            packet.size = 0;
            int data_present;
            
            // 讀取出一幀未解碼數(shù)據(jù)
            finished =  (av_read_frame(inputFormatCtx, &packet) == AVERROR_EOF);
            
            // 判斷該幀數(shù)據(jù)是否為音頻數(shù)據(jù)
            if (packet.stream_index != audioStream) {
                continue;
            }
            
            // 開始進(jìn)行解碼
            if ( avcodec_decode_audio4(inputCodecCtx, audioFrame, &data_present, &packet) < 0) {
                NSLog(@"音頻解碼失敗");
                return ;
            }
            
            if (data_present)
            {
                //將pcm數(shù)據(jù)寫入文件
                for(int i = 0 ; i <audioFrame->channels;i++)
                {
                    NSData *data = [NSData dataWithBytes:audioFrame->data[i] length:audioFrame->linesize[0]];
                    [pcmfileHandle writeData:data];
                    
                }
            }
          
            // 初始化進(jìn)行重采樣的存儲(chǔ)空間
            uint8_t **converted_input_samples = NULL;
            if (init_converted_samples(&converted_input_samples, aacCodeContex,
                                       audioFrame->nb_samples))
            {
                return;
            }
            
            // 進(jìn)行重采樣
            if (convert_samples((const uint8_t**)audioFrame->extended_data, converted_input_samples,
                                audioFrame->nb_samples, resample_context))
            {
                return;
            }
            
            //將采樣結(jié)果加入進(jìn)fifo中
            add_samples_to_fifo(fifo, converted_input_samples,audioFrame->nb_samples);
            

            // 釋放重采樣存儲(chǔ)空間
            if (converted_input_samples)
            {
                av_freep(&converted_input_samples[0]);
                free(converted_input_samples);
            }
        }
        
        // 從fifo隊(duì)列中讀入數(shù)據(jù)
        while (av_audio_fifo_size(fifo) >= output_frame_size || finished) {
            
            AVFrame *frame;
            
            frame = av_frame_alloc();
            
            const int frame_size = FFMIN(av_audio_fifo_size(fifo),aacCodeContex->frame_size);
            
            // 設(shè)置輸入幀的相關(guān)參數(shù)
            (frame)->nb_samples     = frame_size;
            (frame)->channel_layout = aacCodeContex->channel_layout;
            (frame)->format         = aacCodeContex->sample_fmt;
            (frame)->sample_rate    = aacCodeContex->sample_rate;
            
            int error;
            
            //根據(jù)幀的相關(guān)參數(shù)焕窝，獲取數(shù)據(jù)存儲(chǔ)空間
            if ((error = av_frame_get_buffer(frame, 0)) < 0)
            {
                av_frame_free(&frame);
                return ;
            }
            
            // 從fifo中讀取frame_size個(gè)樣本數(shù)據(jù)
            if (av_audio_fifo_read(fifo, (void **)frame->data, frame_size) < frame_size)
            {
                av_frame_free(&frame);
                return ;
            }
            
            AVPacket pkt;
            av_init_packet(&pkt);
            pkt.data = NULL;
            pkt.size = 0;
            
            int data_present = 0;
            
            frame->pts = av_frame_get_best_effort_timestamp(frame);
            frame->pict_type=AV_PICTURE_TYPE_NONE;
            
            // 將pcm數(shù)據(jù)進(jìn)行編碼
            if ((error = avcodec_encode_audio2(aacCodeContex, &pkt,frame, &data_present)) < 0)
            {
                av_free_packet(&pkt);
                return ;
            }
            av_frame_free(&frame);
            
            // 如果編碼成功，寫入文件
            if (data_present) {
                NSData *data = [NSData dataWithBytes:pkt.data length:pkt.size];
                NSLog(@"pkt length = %d",pkt.size);
                [fileHandle writeData:[self adtsDataForPacketLength:pkt.size]];
                [fileHandle writeData:data];
            }
            
            av_free_packet(&pkt);
        }
        
    }
    
    NSLog(@"***************************************end");
}

參考資料

ffmpeg官方例子
代碼下載