Audio Recorder C++ FFmpeg 4.2 on Windows


using Visual Studio 2019

1.Install VBCable
Set up
reference: https://mediarealm.com.au/articles/stereo-mix-setup-windows-10/

2.Set up C++ property & Linker property in project property

(Linker : add Winmm.lib; in somewhere)

3.-----------------------------Souce.cpp---------------------------
#define _CRT_SECURE_NO_WARNINGS
//#include <QDebug>
//#include <QByteArray>

#include <stdio.h>
#include <iostream>

#define __STDC_CONSTANT_MACROS

extern "C" {
#include <libavutil/avassert.h>
#include <libavutil/channel_layout.h>
#include <libavutil/opt.h>
#include <libavutil/mathematics.h>
#include <libavutil/timestamp.h>
#include <libavformat/avformat.h>
//#include "libavutil/imgutils.h"
#include <libswscale/swscale.h>
#include <libswresample/swresample.h>
}

#include <windows.h>

using namespace std;


const int NUMPTS = 44100 * 5;
int sampleRate = 44100;
int channels = 2;
int bitPerSample = 16;
int bytePerSample = 4;
/* Following field are set by program */
int nb_samples;
//short int* waveIn; // BlockAlign * 441
BYTE* waveIn; // BlockAlign * 441
HWAVEIN hWaveIn;
WAVEHDR WaveInHdr;
MMRESULT result;
HWAVEOUT hWaveOut;
WAVEFORMATEX pFormat;
FILE* fp;
AVFormatContext* oc;
/****************************/

typedef struct OutputStream {
    AVStream* st;
    AVCodecContext* enc;

    /* pts of the next frame that will be generated */
    int64_t next_pts;
    int samples_count;

    AVFrame* frame;
    AVFrame* tmp_frame;

    float t, tincr, tincr2;

    struct SwsContext* sws_ctx;
    struct SwrContext* swr_ctx;
} OutputStream;

OutputStream audio_st = { 0 };

static int write_frame(AVFormatContext* fmt_ctx, const AVRational* time_base, AVStream* st, AVPacket* pkt)
{
    /* rescale output packet timestamp values from codec to stream timebase */
    av_packet_rescale_ts(pkt, *time_base, st->time_base);
    pkt->stream_index = st->index;

    /* Write the compressed frame to the media file. */
    //log_packet(fmt_ctx, pkt);
    return av_interleaved_write_frame(fmt_ctx, pkt);
}

static void add_stream(OutputStream* ost, AVFormatContext* oc, AVCodec** codec, enum AVCodecID codec_id)
{
    AVCodecContext* c;
    int i;

    /* find the encoder */
    *codec = avcodec_find_encoder(codec_id);
    if (!(*codec)) {
        fprintf(stderr, "Could not find encoder for '%s'\n",
            avcodec_get_name(codec_id));
        exit(1);
    }

    ost->st = avformat_new_stream(oc, NULL);
    if (!ost->st) {
        fprintf(stderr, "Could not allocate stream\n");
        exit(1);
    }
    ost->st->id = oc->nb_streams - 1;
    c = avcodec_alloc_context3(*codec);
    if (!c) {
        fprintf(stderr, "Could not alloc an encoding context\n");
        exit(1);
    }
    ost->enc = c;

    switch ((*codec)->type) {
    case AVMEDIA_TYPE_AUDIO:
        c->sample_fmt = (*codec)->sample_fmts ?
            (*codec)->sample_fmts[0] : AV_SAMPLE_FMT_FLTP;
        c->bit_rate = 64000;
        c->sample_rate = 44100;
        if ((*codec)->supported_samplerates) {
            c->sample_rate = (*codec)->supported_samplerates[0];
            for (i = 0; (*codec)->supported_samplerates[i]; i++) {
                if ((*codec)->supported_samplerates[i] == 44100)
                    c->sample_rate = 44100;
            }
        }
        c->channels = av_get_channel_layout_nb_channels(c->channel_layout);
        c->channel_layout = AV_CH_LAYOUT_STEREO;
        if ((*codec)->channel_layouts) {
            c->channel_layout = (*codec)->channel_layouts[0];
            for (i = 0; (*codec)->channel_layouts[i]; i++) {
                if ((*codec)->channel_layouts[i] == AV_CH_LAYOUT_STEREO)
                    c->channel_layout = AV_CH_LAYOUT_STEREO;
            }
        }
        c->channels = av_get_channel_layout_nb_channels(c->channel_layout);
        //ost->st->time_base = (AVRational){ 1, c->sample_rate };
        ost->st->time_base.num = 1;
        ost->st->time_base.den = c->sample_rate;
        break;
    default:
        break;
    }

    /* Some formats want stream headers to be separate. */
    if (oc->oformat->flags & AVFMT_GLOBALHEADER)
        c->flags |= AV_CODEC_FLAG_GLOBAL_HEADER;
}
static AVFrame* alloc_audio_frame(enum AVSampleFormat sample_fmt,
    uint64_t channel_layout,
    int sample_rate, int nb_samples)
{
    AVFrame* frame = av_frame_alloc();
    int ret;

    if (!frame) {
        fprintf(stderr, "Error allocating an audio frame\n");
        exit(1);
    }

    frame->format = sample_fmt;
    frame->channel_layout = channel_layout;
    frame->sample_rate = sample_rate;
    frame->nb_samples = nb_samples;

    if (nb_samples) {
        ret = av_frame_get_buffer(frame, 0);
        if (ret < 0) {
            fprintf(stderr, "Error allocating an audio buffer\n");
            exit(1);
        }
    }

    return frame;
}
static void open_audio(AVFormatContext* oc, AVCodec* codec, OutputStream* ost, AVDictionary* opt_arg)
{
    AVCodecContext* c;
    //int nb_samples;
    int ret;
    AVDictionary* opt = NULL;

    c = ost->enc;

    /* open it */
    av_dict_copy(&opt, opt_arg, 0);
    ret = avcodec_open2(c, codec, &opt);
    av_dict_free(&opt);
    if (ret < 0) {
        //fprintf(stderr, "Could not open audio codec: %s\n", av_err2str(ret));
        fprintf(stderr, "Could not open audio codec: \n");
        exit(1);
    }

    if (c->codec->capabilities & AV_CODEC_CAP_VARIABLE_FRAME_SIZE)
        nb_samples = 4410;
    else
        nb_samples = c->frame_size;

    cout << "nb_samples : " << nb_samples << endl;
    ost->frame = alloc_audio_frame(c->sample_fmt, c->channel_layout,
        c->sample_rate, nb_samples);
    ost->tmp_frame = alloc_audio_frame(AV_SAMPLE_FMT_S16, c->channel_layout,
        c->sample_rate, nb_samples);

    /* copy the stream parameters to the muxer */
    ret = avcodec_parameters_from_context(ost->st->codecpar, c);
    if (ret < 0) {
        fprintf(stderr, "Could not copy the stream parameters\n");
        exit(1);
    }

    /* create resampler context */
    ost->swr_ctx = swr_alloc();
    if (!ost->swr_ctx) {
        fprintf(stderr, "Could not allocate resampler context\n");
        exit(1);
    }

    /* set options */
    av_opt_set_int(ost->swr_ctx, "in_channel_count", c->channels, 0);
    av_opt_set_int(ost->swr_ctx, "in_sample_rate", c->sample_rate, 0);
    av_opt_set_sample_fmt(ost->swr_ctx, "in_sample_fmt", AV_SAMPLE_FMT_S16, 0);
    av_opt_set_int(ost->swr_ctx, "out_channel_count", c->channels, 0);
    av_opt_set_int(ost->swr_ctx, "out_sample_rate", c->sample_rate, 0);
    av_opt_set_sample_fmt(ost->swr_ctx, "out_sample_fmt", c->sample_fmt, 0);

    /* initialize the resampling context */
    if ((ret = swr_init(ost->swr_ctx)) < 0) {
        fprintf(stderr, "Failed to initialize the resampling context\n");
        exit(1);
    }
}
/* Prepare a 16 bit dummy audio frame of 'frame_size' samples and
 * 'nb_channels' channels. */
static AVFrame* get_audio_frame(OutputStream* ost)
{
    AVFrame* frame = ost->tmp_frame;

    AVRational temp;
    temp.num = 1;
    temp.den = 1;
    /* check if we want to generate more frames */
    /*if (av_compare_ts(ost->next_pts, ost->enc->time_base,
        STREAM_DURATION, temp) >= 0)
        return NULL;*/
  
    //frame->data[0] = (uint8_t*)waveIn;

    for (int i = 0; i < nb_samples * bytePerSample; i++)
        frame->data[0][i] = waveIn[i];
  
    frame->pts = ost->next_pts;
    ost->next_pts += frame->nb_samples;

    return frame;
}
static int write_audio_frame(AVFormatContext* oc, OutputStream* ost)
{
    AVCodecContext* c;
    AVPacket pkt = { 0 }; // data and size must be 0;
    AVFrame* frame;
    int ret;
    int got_packet;
    int dst_nb_samples;

    av_init_packet(&pkt);
    c = ost->enc;

    frame = get_audio_frame(ost);

    if (frame) {
        /* convert samples from native format to destination codec format, using the resampler */
            /* compute destination number of samples */
        dst_nb_samples = av_rescale_rnd(swr_get_delay(ost->swr_ctx, c->sample_rate) + frame->nb_samples,
            c->sample_rate, c->sample_rate, AV_ROUND_UP);
        av_assert0(dst_nb_samples == frame->nb_samples);

        /* when we pass a frame to the encoder, it may keep a reference to it
         * internally;
         * make sure we do not overwrite it here
         */
        ret = av_frame_make_writable(ost->frame);
        if (ret < 0)
            exit(1);

        /* convert to destination format */
        ret = swr_convert(ost->swr_ctx,
            ost->frame->data, dst_nb_samples,
            (const uint8_t**)frame->data, frame->nb_samples);
        if (ret < 0) {
            fprintf(stderr, "Error while converting\n");
            exit(1);
        }
        frame = ost->frame;

        AVRational temp;
        temp.num = 1;
        temp.den = c->sample_rate;
        //frame->pts = av_rescale_q(ost->samples_count, (AVRational) { 1, c->sample_rate }, c->time_base);
        frame->pts = av_rescale_q(ost->samples_count, temp, c->time_base);
        ost->samples_count += dst_nb_samples;
    }

    //ret = avcodec_encode_audio2(c, &pkt, frame, &got_packet);
    ret = avcodec_send_frame(c, frame);
    if (ret < 0) {
        //fprintf(stderr, "Error encoding audio frame: %s\n", av_err2str(ret));
        fprintf(stderr, "Error encoding audio frame: \n");
        if (ret == AVERROR_EOF)
            return 1;
        exit(1);
    }
    got_packet = 1;
    while (ret >= 0) {
        ret = avcodec_receive_packet(c, &pkt);
        if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF) {
            got_packet = 0;
            return 0;
        }
        if (write_frame(oc, &c->time_base, ost->st, &pkt) < 0) {
            fprintf(stderr, "Error while writing audio frame: \n");
            exit(1);
        }
    }

    return (frame || got_packet) ? 0 : 1;
}

static void close_stream(AVFormatContext* oc, OutputStream* ost)
{
    avcodec_free_context(&ost->enc);
    av_frame_free(&ost->frame);
    av_frame_free(&ost->tmp_frame);
    sws_freeContext(ost->sws_ctx);
    swr_free(&ost->swr_ctx);
}

void CALLBACK waveInProc(HWAVEIN hwi, UINT uMsg, DWORD dwInstance, DWORD dwParam1, DWORD dwParam2)
{

    //printf("(waveInProc) hWaveIn 0x%p, nMessage 0x%04X, nInstance %d, nParameter1 %d, nParameter2 %d\n", hwi, uMsg, dwInstance, dwParam1, dwParam2);

    //cout << "waveInProc()..." << endl;
    WAVEHDR* pHdr = NULL;
    WAVEHDR pHdr2;
    switch (uMsg)
    {
    case WIM_CLOSE:
        cout << "waveInProc()... WIM_CLOSE" << endl;
        break;

    case WIM_DATA:
    {
        pHdr = (WAVEHDR*)dwParam1;
        pHdr2.lpData = (LPSTR)pHdr->lpData;
        pHdr2.dwBufferLength = pHdr->dwBufferLength;
        pHdr2.dwBytesRecorded = pHdr->dwBytesRecorded;
        pHdr2.dwUser = 0;
        pHdr2.dwFlags = 0;
        pHdr2.dwLoops = 0;

        cout << "waveInProc()... WIM_DATA : " << "pHdr->dwBufferLength:" << pHdr->dwBufferLength << "  pHdr->dwBytesRecorded:" << pHdr->dwBytesRecorded << endl;//!
        //cout << QByteArray(pHdr->lpData, pHdr->dwBytesRecorded).toHex().constData() << " : " << QByteArray(pHdr->lpData, pHdr->dwBytesRecorded).length() << endl;

        //waveOutPrepareHeader(hWaveOut, &pHdr2, sizeof(WAVEHDR));//!
        //waveOutWrite(hWaveOut, &pHdr2, sizeof(WAVEHDR));//!

        //cout << "buffer : " << pHdr->lpData[0]  << pHdr->lpData[1] << endl;

        write_audio_frame(oc, &audio_st);
        fwrite(pHdr->lpData, 1, pHdr->dwBytesRecorded, fp);

        waveInPrepareHeader(hwi, pHdr, sizeof(WAVEHDR));
        waveInAddBuffer(hwi, pHdr, sizeof(WAVEHDR));

    }
    break;

    case WIM_OPEN:
        cout << "waveInProc()... WIM_OPEN" << endl;
        break;

    default:
        break;
    }
}

int main(int argv, char** args)
{
    const char* filename = "output.mp3";
    AVOutputFormat* fmt;
    //AVFormatContext* oc;
    AVCodec* audio_codec = NULL;
    int ret;
    int have_audio = 0;
    //int encode_video = 0, encode_audio = 0;
    AVDictionary* opt = NULL;

    fp = fopen("capture.pcm", "ab+");

    if (!fp) {
        cout << "fail to open file.\n" << endl;
        return 1;
    }

    pFormat.wFormatTag = WAVE_FORMAT_PCM;
    pFormat.nChannels = channels;
    pFormat.nSamplesPerSec = sampleRate;
    pFormat.wBitsPerSample = bitPerSample;
    //pFormat.nAvgBytesPerSec = 2 * sampleRate;
    pFormat.nAvgBytesPerSec = pFormat.nSamplesPerSec * pFormat.nChannels * pFormat.wBitsPerSample / 8;
    //pFormat.nBlockAlign = 2;
    pFormat.nBlockAlign = pFormat.nChannels * pFormat.wBitsPerSample / 8;
    pFormat.cbSize = 0;

    result = waveInOpen(&hWaveIn, WAVE_MAPPER, &pFormat, (DWORD_PTR)waveInProc, NULL, CALLBACK_FUNCTION);

    if (result)
    {
        char fault[256];
        waveInGetErrorTextA(result, fault, 256);
        MessageBoxA(NULL, fault, "Failed to open waveform input device.", MB_OK | MB_ICONEXCLAMATION);
        return 1;
    }

    //--------------------------------------
    avformat_alloc_output_context2(&oc, NULL, NULL, filename);
    if (!oc) {
        printf("Could not deduce output format from file extension: using MPEG.\n");
        avformat_alloc_output_context2(&oc, NULL, "mpeg", filename);
    }
    if (!oc)
        return 1;

    fmt = oc->oformat;

    /* Add the audio and video streams using the default format codecs
     * and initialize the codecs. */
    if (fmt->audio_codec != AV_CODEC_ID_NONE) {
        add_stream(&audio_st, oc, &audio_codec, fmt->audio_codec);
        have_audio = 1;
        //encode_audio = 1;
    }

    /* Now that all the parameters are set, we can open the audio and
     * video codecs and allocate the necessary encode buffers. */
    if (have_audio)
        open_audio(oc, audio_codec, &audio_st, opt);

    av_dump_format(oc, 0, filename, 1);

    /* open the output file, if needed */
    if (!(fmt->flags & AVFMT_NOFILE)) {
        ret = avio_open(&oc->pb, filename, AVIO_FLAG_WRITE);
        if (ret < 0) {
            //fprintf(stderr, "Could not open '%s': %s\n", filename, av_err2str(ret));
            fprintf(stderr, "Could not open '%s': \n", filename);
            return 1;
        }
    }


    /* Write the stream header, if any. */
    ret = avformat_write_header(oc, &opt);
    if (ret < 0) {
        //fprintf(stderr, "Error occurred when opening output file: %s\n", av_err2str(ret));
        fprintf(stderr, "Error occurred when opening output file: \n");
        return 1;
    }
    //--------------------------------------
    //waveIn = new short int[nb_samples * bytePerSample];
    waveIn = new BYTE[nb_samples * bytePerSample];
    WaveInHdr.lpData = (LPSTR)waveIn;
    WaveInHdr.dwBufferLength = nb_samples * bytePerSample;
    WaveInHdr.dwBytesRecorded = 0;
    WaveInHdr.dwUser = 0;
    WaveInHdr.dwFlags = 0;
    WaveInHdr.dwLoops = 0;
    waveInPrepareHeader(hWaveIn, &WaveInHdr, sizeof(WAVEHDR));

    cout << "WaveInHdr.dwBufferLength : " << WaveInHdr.dwBufferLength << endl;

    result = waveInAddBuffer(hWaveIn, &WaveInHdr, sizeof(WAVEHDR));
    if (result)
    {
        MessageBoxA(NULL, "Failed to read block from device", NULL, MB_OK | MB_ICONEXCLAMATION);
        return 1;
    }

    result = waveInStart(hWaveIn);
    if (result)
    {
        MessageBoxA(NULL, "Failed to start recording", NULL, MB_OK | MB_ICONEXCLAMATION);
        return 1;
    }

    /*if (waveOutOpen(&hWaveOut, WAVE_MAPPER, &pFormat, 0, 0, WAVE_FORMAT_DIRECT))
    {
        MessageBoxA(NULL, "Failed to replay", NULL, MB_OK | MB_ICONEXCLAMATION);
    }*/ //!
  
    Sleep((NUMPTS / sampleRate) * 1000);

    //waveOutUnprepareHeader(hWaveOut, &WaveInHdr, sizeof(WAVEHDR));//!
    //waveOutClose(hWaveOut);//!

    waveInUnprepareHeader(hWaveIn, &WaveInHdr, sizeof(WAVEHDR));
    waveInStop(hWaveIn);
    waveInClose(hWaveIn);

    AVPacket pkt = { 0 }; // data and size must be 0;
    av_init_packet(&pkt);

    ret = avcodec_send_frame(audio_st.enc, NULL);
    /*if (ret < 0) {
        //fprintf(stderr, "Error encoding audio frame: %s\n", av_err2str(ret));
        fprintf(stderr, "Error encoding audio frame: \n");
        if (ret == AVERROR_EOF)
            return 1;
        exit(1);
    }*/
    while (ret >= 0) {
        ret = avcodec_receive_packet(audio_st.enc, &pkt);
        if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF) {
          
        }
        if (write_frame(oc, &audio_st.enc->time_base, audio_st.st, &pkt) < 0) {
            fprintf(stderr, "Error while writing audio frame: \n");
            exit(1);
        }
    }


    av_write_trailer(oc);


    /* Close each codec. */
    if (have_audio)
        close_stream(oc, &audio_st);

    if (!(fmt->flags & AVFMT_NOFILE))
        /* Close the output file. */
        avio_closep(&oc->pb);

    /* free the stream */
    avformat_free_context(oc);

    fclose(fp);

    return 0;
}
---------------------------------------------------------------------------------------------

reference: https://kldp.org/node/135620

4. Usage : play sound and run it, it will record what you're playing and store it to capture.pcm & output.mp3 file

5.enter commend to convert .pcm to .wav
>ffmpeg -f s16le -ar 44.1k -ac 2 -i capture.pcm output.wav


留言

這個網誌中的熱門文章

組合語言 Assembly Language