ffmpeg 编程常用 pcm 转 aac aac 转 pcm mp4 h264解码

ffmpeg 是现在开源的全能编解码器，基本上全格式都支持，纯 c 语言作成，相对比其它的 VLC ，GStreamer glib2 写的，开发更简单些，文档很棒，就是 examples 比较少。

常用的功能有：

AVFrame 数据帧
AVCodecContext 编解码器
AVPacket 数据帧
swr_convert 格式转换器

ffmpeg 的使用都差不多，查找解码器，准备数据，解码，拿结果。

基本上会了一种，其它的也就能会，新版的 4.x 的代码较之前的有些变化，现在大部分之前的代码也是兼容的。有一些 定义成了 enum 。

介绍几个非常实用的例子：

1, pcm 编码 aac (aac 和 m4a 是一种类型)

需要 libfdk_aac 库自行安装配置好，使用 ubuntu 16.0.4 x64 g++ 编译

g++ -g main.cpp -lavcodec -lavformat -lswresample -lavutil -std=c++11 -o wav_to_m4a

用法 ./wav_to_m4a ../xxx.wav ，需要说明的是，有些网站下载的 wav 根本不能用，最好是用 ffmpeg 命令转换。

现实的需求中，没有人让你做一个格式转换器。可能是从 ALSA 读取原始PCM 在编码成 AAC 或通过网络发走，或保存文件。

下面的例子，仅是编码成了 AAC 但是未添加 AAC 头信息。有空在更新。但是可以用 播放器放的。

#include <stdio.h>
#include <stdlib.h>
#include <iostream>
#include <fstream>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <sys/mman.h>

#ifdef __cplusplus
extern "C" 
{
#endif
#include <libavformat/avformat.h>
#include <libavcodec/avcodec.h>
#include <libavutil/channel_layout.h>
#include <libavutil/common.h>
#include <libavutil/frame.h>
#include <libavutil/samplefmt.h>
#include <libavutil/mem.h>
#ifdef __cplusplus
}
#endif

#ifndef WORD
#define WORD unsigned short
#endif

#ifndef DWORD
#define DWORD unsigned int
#endif
    
struct RIFF_HEADER
{
    char szRiffID[4];  // 'R','I','F','F'
    DWORD dwRiffSize;
    char szRiffFormat[4]; // 'W','A','V','E'
};

struct WAVE_FORMAT
{
    WORD wFormatTag;
    WORD wChannels;
    DWORD dwSamplesPerSec;
    DWORD dwAvgBytesPerSec;
    WORD wBlockAlign;
    WORD wBitsPerSample;
};

struct FMT_BLOCK
{
    char  szFmtID[4]; // 'f','m','t',' '
    DWORD  dwFmtSize;
    struct WAVE_FORMAT wavFormat;
};

struct DATA_BLOCK
{
    char szDataID[4]; // 'd','a','t','a'
    DWORD dwDataSize;
};

using namespace std;

void read_wav(uint8_t *wav_buf, int *fs, int *channels, int *bits_per_sample, int *wav_size, int *file_size)
{
    struct RIFF_HEADER *headblk;
    struct FMT_BLOCK   *fmtblk;
    struct DATA_BLOCK  *datblk;

    headblk = (struct RIFF_HEADER *) wav_buf;
    fmtblk  = (struct FMT_BLOCK *) &headblk[1];
    datblk  = (struct DATA_BLOCK *) &fmtblk[1];
    
    *file_size = headblk->dwRiffSize;

    //采样频率
    *fs        = fmtblk->wavFormat.dwSamplesPerSec;
    //通道数
    *channels  = fmtblk->wavFormat.wChannels;
    *wav_size  = datblk->dwDataSize;
    //采样bit数 16 24
    *bits_per_sample = fmtblk->wavFormat.wBitsPerSample;
}

int main(int argc, char **argv)
{
    int fd;
    int ret;
    struct stat stat;
    int fs, channels, bits_per_sample, wav_size, file_size;
    uint8_t *wav_buf;
    uint8_t *audio_buf;    
    const char           *out_file = "out.m4a";
    const AVCodec        *codec;
    AVFrame              *frame;
    AVPacket             *encodePacket;
    AVCodecContext       *codecContext;
    
    //打开文件进行 mmap 
    fd = open(argv[1], O_RDONLY);
    fstat(fd, &stat);
    wav_buf = (uint8_t*)mmap(NULL, stat.st_size, PROT_READ, MAP_SHARED, fd, 0);
    read_wav(wav_buf, &fs, &channels, &bits_per_sample, &wav_size, &file_size);
    printf("wav format: fs = %d, channels = %d, bits_per_sample = %d, wav_size = %d file_size = %d

", fs, channels, bits_per_sample, wav_size, file_size);
    
    //真实wav 跳过头部
    audio_buf = wav_buf + sizeof(struct RIFF_HEADER) + sizeof(struct FMT_BLOCK) + sizeof(struct DATA_BLOCK);

    avcodec_register_all(); //ffmpeg 4.x 已经不需要此函数
    
    codec =  avcodec_find_encoder_by_name("libfdk_aac");

    if(! codec)
    {
        printf("avcodec_find_encoder error 
");
        return -1;
    }

    codecContext = avcodec_alloc_context3(codec);
    if(! codecContext)
    {
        printf("Could not allocate audio codec context 
");
        return -1;
    }

    codecContext->bit_rate       = 64000;
    codecContext->sample_fmt     = AV_SAMPLE_FMT_S16; 
    codecContext->sample_rate    = 44100;
    codecContext->channel_layout = AV_CH_LAYOUT_STEREO;
    codecContext->channels       = av_get_channel_layout_nb_channels(codecContext->channel_layout);

    printf("sample_fmt:%d sample_rate:%d channel_layout:%d channels:%d 
", codecContext->sample_fmt, codecContext->sample_rate, codecContext->channel_layout, codecContext->channels);

    /* open avcodec */
    if(0 > avcodec_open2(codecContext, codec, NULL))
    {
        printf("Could not open codec 
");
        return -1;
    }

    /* frame containing input raw audio */
    frame = av_frame_alloc();
    if(! frame)
    {
        printf("Could not allocate audio frame 
");
        return -1;
    }

    frame->nb_samples     = codecContext->frame_size;
    frame->format         = codecContext->sample_fmt;
    frame->channel_layout = codecContext->channel_layout;

    printf("nb_samples:%d format:%d channel_layout:%d 
", frame->nb_samples, frame->format, frame->channel_layout);

    encodePacket = av_packet_alloc();
    
    int size = av_samples_get_buffer_size(NULL, codecContext->channels,codecContext->frame_size,codecContext->sample_fmt, 1);
    uint8_t *frame_buf = (uint8_t *)av_malloc(size);
    avcodec_fill_audio_frame(frame, codecContext->channels, codecContext->sample_fmt,(const uint8_t*)frame_buf, size, 1);
    
    ofstream acc_file(out_file, ios::binary | ios::out | ios::trunc);

    int pos = 0;
    while(pos <= file_size)
    {
        frame->data[0] = audio_buf + pos;

        /* send the frame for encoding */
        ret = avcodec_send_frame(codecContext, frame);
        if(0 > ret)
        {
            printf("Error sending the frame to the encoder 
");
            return -1;
        }

        /* read all the available output packets (in general there may be any
         * number of them */
        while(0 <= ret)
        {
            ret = avcodec_receive_packet(codecContext, encodePacket);
            if(ret == AVERROR(EAGAIN) || ret == AVERROR_EOF)
                break;
            else if(0 > ret)
            {
                printf("Error encoding audio frame 
");
                return -1;
            }

            //c++ 写文件流
            acc_file.write((const char*)encodePacket->data, encodePacket->size);
            av_packet_unref(encodePacket);
        }

        pos += size;
    }

    cout << "encode done" << endl;
    return 0;
}

// main.cpp

2, acc 解码 pcm

3, mp4 解码 h264 acc 并调用 SDL 播放视频

更新中。。