Extract audio from video using autogen ffmpeg C# in Unity

Question

Hi I'm using ffmpeg autogen to extract audio from video in Unity, but when I following this code, the file write cannot write, it's 0Kb, so what's issue of this or someone have any examples for extract audio using this library, apologize for my English. This is github of library: https://github.com/Ruslan-B/FFmpeg.AutoGen

unsafe void TestExtractAudio()
{

    string inFile = Application.streamingAssetsPath + "/" + strFileName;
    string outFile = Application.streamingAssetsPath + "/" + strFileNameAudio;

    AVOutputFormat* outFormat = null;
    AVFormatContext* inFormatContext = null;
    AVFormatContext* outFormatContext = null;
    AVPacket packet;

    ffmpeg.av_register_all();

    inFormatContext = ffmpeg.avformat_alloc_context();
    outFormatContext = ffmpeg.avformat_alloc_context();

    if (ffmpeg.avformat_open_input(&inFormatContext, inFile, null, null) < 0)
    {
        throw new ApplicationException("Could not open input file.");
    }

    if (ffmpeg.avformat_find_stream_info(inFormatContext, null) < 0)
    {
        throw new ApplicationException("Failed to retrieve input stream info.");
    }

    ffmpeg.avformat_alloc_output_context2(&outFormatContext, null, null, outFile);
    if (outFormatContext == null)
    {
        throw new ApplicationException("Could not create output context");
    }

    outFormat = outFormatContext->oformat;

    AVStream* inStream = inFormatContext->streams[1];
    AVStream* outStream = ffmpeg.avformat_new_stream(outFormatContext, inStream->codec->codec);
    if (outStream == null)
    {
        throw new ApplicationException("Failed to allocate output stream.");
    }

    if (ffmpeg.avcodec_copy_context(outStream->codec, inStream->codec) < 0)
    {
        throw new ApplicationException("Couldn't copy input stream codec context to output stream codec context");
    }

    outFormatContext->audio_codec_id = AVCodecID.AV_CODEC_ID_MP3;

    int retcode = ffmpeg.avio_open(&outFormatContext->pb, outFile, ffmpeg.AVIO_FLAG_WRITE);
    if (retcode < 0)
    {
        throw new ApplicationException("Couldn't open output file");
    }

    int returnCode = ffmpeg.avformat_write_header(outFormatContext, null);

    if (returnCode < 0)
    {
        throw new ApplicationException("Error occurred opening output file.");
    }

    while (true)
    {
        if (ffmpeg.av_read_frame(inFormatContext, &packet) < 0)
        {
            break;
        }

        if (packet.stream_index == 1)
        {

            inStream = inFormatContext->streams[1];
            outStream = outFormatContext->streams[0];

            // TODO: Replicate log packet functionality to print out what's inside the packet.

            packet.pts = ffmpeg.av_rescale_q_rnd(packet.pts, inStream->time_base, outStream->time_base,
                AVRounding.AV_ROUND_NEAR_INF | AVRounding.AV_ROUND_PASS_MINMAX);
            packet.dts = ffmpeg.av_rescale_q_rnd(packet.dts, inStream->time_base, outStream->time_base,
                AVRounding.AV_ROUND_NEAR_INF | AVRounding.AV_ROUND_PASS_MINMAX);

            packet.duration = ffmpeg.av_rescale_q(packet.duration, inStream->time_base, outStream->time_base);

            int returncode = ffmpeg.av_interleaved_write_frame(outFormatContext, &packet);

        }

        ffmpeg.av_packet_unref(&packet);
    }

    ffmpeg.av_write_trailer(outFormatContext);


    ffmpeg.avformat_close_input(&inFormatContext);

    ffmpeg.avformat_free_context(outFormatContext);

    Console.WriteLine("Press any key to continue...");

    Console.ReadKey();
}

the value returnCode return less than 0, so someone can fix this, thanks so much for that

Answer 1

The problem is here:

inStream = inFormatContext->streams[1];
outStream = outFormatContext->streams[0];

This code:

ffmpeg.av_interleaved_write_frame

does the following validation

static int check_packet(AVFormatContext *s, AVPacket *pkt)
{
if (pkt->stream_index < 0 || pkt->stream_index >= s->nb_streams) {
    av_log(s, AV_LOG_ERROR, "Invalid packet stream index: %d\n"
}

you need to change the packet.stream_index from 1 to 0 before calling int returncode = ffmpeg.av_interleaved_write_frame(outFormatContext, &packet);

See:

if (type == AVMediaType.AVMEDIA_TYPE_AUDIO){

    inStream = inFormatContext->streams[1];
    outStream = outFormatContext->streams[0];

    // TODO: Replicate log packet functionality to print out what's inside the packet.

    ffmpeg.av_packet_rescale_ts(&packet, inStream->time_base, outStream->time_base);

    packet.stream_index = 0;

    int returncode = ffmpeg.av_write_frame(outFormatContext, &packet);

}