首页 > 解决方案 > 将 MKA 音频文件解码为原始数据(MKA 音频到原始音频数据)

问题描述

我的任务是打开一个mka扩展名为(Matroska容器)的现有音频文件并提取原始音频数据。示例仅显示从mp2文件中提取原始数据的示例。我不知道如何用mka容器做到这一点。我想要这样的东西:

UPD

我找到了一个以音频文件中记录的格式保存音频数据的选项。一个例子如下所示。

PS。这只是一个测试版本,很可能存在内存泄漏和其他问题。


#include <QFile>
#include <QDebug>
#include "audiodecoder.h"

int main(int argc, char* argv[])
{
    AudioDecoder decoder("/home/test/test.mka");
    bool started = decoder.start();
    if (!started) {
        return EXIT_FAILURE;
    }

    QFile file("/home/test/rawData.bin");
    file.open(QIODevice::WriteOnly);

    while (true) {
        auto data = decoder.getData(255);
        if (data.isEmpty()) {
            break;
        }
        file.write(data.data(), data.size());
    }
    file.close();
    return EXIT_SUCCESS;
}

音频解码器.h

class AudioDecoder {
public:
    AudioDecoder(const QString& fileName);
    AudioDecoder& operator=(const AudioDecoder& rhs) = delete;
    AudioDecoder& operator=(AudioDecoder&& rhs) = delete;
    AudioDecoder(const AudioDecoder& rhs) = delete;
    AudioDecoder(AudioDecoder&& rhs) = delete;
    virtual ~AudioDecoder(void);

    virtual bool start(void) noexcept;
    virtual QByteArray getData(const quint16& size) noexcept;
    virtual bool stop(void) noexcept;

protected:
    bool m_initialized;
    QString m_fileName;

    AVFrame* p_frame = nullptr;
    AVPacket* p_packet = nullptr;
    AVCodecContext* p_cdcCtx = nullptr;
    AVFormatContext* p_frmCtx = nullptr;
};

音频解码器.cpp


static void logging(const char* message)
{
    qDebug() << message;
}

AudioDecoder::AudioDecoder(const QString& fileName)
    : m_initialized(false)
    , m_fileName(fileName)
    , p_cdcCtx(nullptr)
    , p_frmCtx(nullptr)
{
    av_register_all();
}

QByteArray AudioDecoder::getData(const quint16& dataSize) noexcept
{
    QByteArray data;
    qint32 response = 0;
    if (av_read_frame(p_frmCtx, p_packet) >= 0) {
        //logging(QString("AVPacket->pts %1").arg(p_packet->pts).toStdString().c_str());
        //response = decode_packet(p_packet, p_cdcCtx, p_frame);
        response = avcodec_send_packet(p_cdcCtx, p_packet);
        if (response < 0) {
            logging("Error while sending a packet to the decoder");
            return {};
        }
        while (response >= 0) {
            response = avcodec_receive_frame(p_cdcCtx, p_frame);
            if (response == AVERROR(EAGAIN) || response == AVERROR_EOF) {
                break;
            }
            else if (response < 0) {
                logging("Error while receiving a frame from the decoder");
                return {};
            }
            if (response >= 0) {
                logging(QString("Frame %1 (type=%2, size=%3 bytes) pts %4 key_frame %5 [DTS %6], duration[%7]")
                            .arg(p_cdcCtx->frame_number)
                            .arg(av_get_picture_type_char(p_frame->pict_type))
                            .arg(p_frame->pkt_size)
                            .arg(p_frame->pts)
                            .arg(p_frame->key_frame)
                            .arg(p_frame->coded_picture_number)
                            .arg(p_frame->pkt_duration)
                            .toStdString()
                            .c_str());

                for (int i = 0; i < p_frame->linesize[0]; ++i) {
                    data.push_back(p_frame->data[0][i]);
                }
            }
        }
        av_packet_unref(p_packet);
        return data;
    }
    return {};
}

bool AudioDecoder::start(void) noexcept
{
    if (m_initialized) {
        return true;
    }

    int error;
    // Open the input file to read from it.
    if ((error = avformat_open_input(&p_frmCtx,
             m_fileName.toStdString().c_str(), nullptr, nullptr))
        < 0) {
        qDebug() << "Could not open input file: " << m_fileName;
        p_frmCtx = nullptr;
        return false;
    }
    // Get information on the input file (number of streams etc.).
    if ((error = avformat_find_stream_info(p_frmCtx, nullptr)) < 0) {
        avformat_close_input(&p_frmCtx);
        qDebug() << __LINE__;
        return false;
    }
    // Make sure that there is only one stream in the input file.
    if ((p_frmCtx)->nb_streams != 1) {
        avformat_close_input(&p_frmCtx);
        qDebug() << __LINE__;
        return false;
    }

    if (p_frmCtx->streams[0]->codecpar->codec_type != AVMEDIA_TYPE_AUDIO) {
        avformat_close_input(&p_frmCtx);
        qDebug() << __LINE__;
        return false;
    }

    // Find a decoder for the audio stream.
    AVCodec* input_codec = nullptr;
    if (!(input_codec = avcodec_find_decoder((p_frmCtx)->streams[0]->codecpar->codec_id))) {
        avformat_close_input(&p_frmCtx);
        qDebug() << __LINE__;
        return false;
    }
    // Allocate a new decoding context.
    AVCodecContext* avctx = avcodec_alloc_context3(input_codec);
    if (!avctx) {
        avformat_close_input(&p_frmCtx);
        qDebug() << __LINE__;
        return false;
    }
    // Initialize the stream parameters with demuxer information.
    error = avcodec_parameters_to_context(avctx, (p_frmCtx)->streams[0]->codecpar);
    if (error < 0) {
        avformat_close_input(&p_frmCtx);
        avcodec_free_context(&avctx);
        qDebug() << __LINE__;
        return false;
    }
    /* Open the decoder for the audio stream to use it later. */
    if ((error = avcodec_open2(avctx, input_codec, NULL)) < 0) {
        avcodec_free_context(&avctx);
        avformat_close_input(&p_frmCtx);
        qDebug() << __LINE__;
        return false;
    }
    /* Save the decoder context for easier access later. */
    p_cdcCtx = avctx;
    av_dump_format(p_frmCtx, 0, m_fileName.toStdString().c_str(), 0);

    p_frame = av_frame_alloc();
    if (!p_frame) {
        logging("failed to allocated memory for AVFrame");
        return false;
    }
    p_packet = av_packet_alloc();
    if (!p_packet) {
        logging("failed to allocated memory for AVPacket");
        return false;
    }
    return m_initialized = true;
}

bool AudioDecoder::stop(void) noexcept
{
    if (p_cdcCtx != nullptr) {
        avcodec_free_context(&p_cdcCtx);
    }
    if (p_frmCtx != nullptr) {
        avformat_close_input(&p_frmCtx);
    }
    return true;
}

AudioDecoder::~AudioDecoder(void)
{
    stop();
}

但是这个例子中的问题是我没有实现获得准确请求的音频数据大小的能力。就我而言,它只是被忽略了。

标签: c++audioffmpegdecode

解决方案


如果您想提取以从音频文件中获取的格式编码的音频数据,那么您应该查看示例。在此示例中,您需要将输出类型从 更改accalaw。压缩流将在AVPacket::buf中:

对存储数据包数据的引用计数缓冲区的引用。

我也推荐查看这篇文章。它会帮助你:)


推荐阅读