Newest 'ffmpeg' Questions - Stack Overflow

http://stackoverflow.com/questions/tagged/ffmpeg

Les articles publiés sur le site

  • How to stream a wmv file as an mp4 file (convert on the fly) over HTTP using ffmpeg or something similar ?

    8 juin 2016, par Adam Tal

    I understand it's not possible to use the tag to play a WMV file.

    I thought about the following solution:

    • Streaming a video means sending small chunks of frames and sound
    • Reading a wmv file in chunks is surely possible
    • An html video tag can receive chunks in an mp4/webm/ogg format and display it

    Can I read a WMV file as a Steam and convert in on the fly to output it as if it was an MP4 file?

    Few points:

    • The technology is not really important. I just have a problem I need to solve.. I would prefer a solution in C# or node.js because those are environments I'm comfortable with but anything else will do.
    • I will be happy to stream the file from a different server as wmv and output it as mp4 but I'm also able to download the file completely to my server and stream it from the local hard drive.
    • The mp4 output is not a must. It can also be webm or ogg (just need the video tag to support it)

  • please help me fix it [on hold]

    8 juin 2016, par Lợi Nè

    I have a code. it's : 

    ffmpeg -i thu.mp4 -i hinhthu3.png -filter_complex "overlay=1:1" -af "atempo=1.15,equalizer=f=32:width_type=o:width=1:g=-5,equalizer=f=64:width_type=o:width=1:g=-5,equalizer=f=125:width_type=o:width=1:g=-5,equalizer=f=250:width_type=o:width=1:g=-5,equalizer=f=500:width_type=o:width=1:g=-5,equalizer=f=1000:width_type=o:width=1:g=-5,equalizer=f=2000:width_type=o:width=1:g=-5,equalizer=f=4000:width_type=o:width=1:g=-5,equalizer=f=8000:width_type=o:width=1:g=-5,equalizer=f=16000:width_type=o:width=1:g=-5,aecho=0.8:0.88:6:0.4,pan=stereo|c0code>

    The audio faster than image. So how to match it together.I had tried many times but can't

  • Muxing from Audio and Video files with FFmpeg

    8 juin 2016, par Michael IV

    I am learning how to create MP4 video from this example.The problem is that the example demonstrates audio encoding from some dummy source data generated on the fly.I need to encode audio from a file.I have checked many examples and most of them show the same or just a separate audio encoding. In my trial and error process I am using the same AVFormatContext for both audio and video frames.I am not sure if it's right thing to do, or should I rather have 2 separate contexts?So far I got Video encoding ok but audio stream fails as AVPacket can't locate correct audio stream index. Here is how I setup audio stream:

      void open_audio(AVFormatContext *oc, AVCodec **codec, AVStream **st ,enum AVCodecID codec_id){

    //    AVCodecContext *c;
    int ret;
    //    c = st->codec;

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

    }
    /* open it */



    if(avformat_open_input(&oc,_audioInName.c_str(),NULL,NULL) !=0){

        Msg::PrintErrorMsg("Error opening audio file");

    }


    AVStream* audioStream = NULL;

    // Find the audio stream (some container files can have multiple streams in them)

    for (uint32_t i = 0; i < oc->nb_streams; ++i)

    {

        if (oc->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO)

        {

            audioStream = oc->streams[i];

            break;

        }

    }

    if (audioStream == NULL)
    {
        Msg::PrintErrorMsg("Could not find any audio stream in the file");

    }
    *st =audioStream;

    AVCodecContext *c  = audioStream->codec;
    c->codec = *codec;//avcodec_find_decoder(c->codec_id);
    audioStream->id = 1;
    c->sample_fmt  = AV_SAMPLE_FMT_S16;
    c->bit_rate    = 64000;
    c->sample_rate = 44100;
    c->channels    = 1;

    if (oc->oformat->flags & AVFMT_GLOBALHEADER){
        c->flags |= CODEC_FLAG_GLOBAL_HEADER;

    }

    if (c->codec == NULL)
    {
        Msg::PrintErrorMsg("Couldn't find a proper decoder");

    }

    ret = avcodec_open2(c, *codec, NULL);
    if (ret < 0) {

        Msg::PrintErrorMsg("Could not open audio codec\n");

    }

}

    Here "oc" is the same context used to initialize video stream as well.

    Then I am trying to write audio frame like this:

      void write_audio_frame(AVFormatContext *oc, AVStream *st){
    AVCodecContext *c;
    AVPacket pkt = { 0 }; // data and size must be 0;
    AVFrame *frame = avcodec_alloc_frame();
    int got_packet, ret;
    av_init_packet(&pkt);
    c = st->codec;
    /////
    //  get_audio_frame(samples, audio_input_frame_size, c->channels);

    ////Read the packet:
    while(av_read_frame(oc,&pkt) == 0 ){

        if(pkt.stream_index ==st->index){

        // Try to decode the packet into a frame
        int frameFinished = 0;
        avcodec_decode_audio4(c, frame, &frameFinished, &pkt);

        // Some frames rely on multiple packets, so we have to make sure the frame is finished before
        // we can use it
        if (frameFinished){
            assert(frameFinished);
            ret = avcodec_encode_audio2(c, &pkt, frame, &got_packet);
            if (ret < 0) {
                Msg::PrintErrorMsg("Error encoding audio frame\n");

            }
            if (!got_packet){
                printf("failed to aquire packet");
            }
            pkt.stream_index = st->index;
            /* Write the compressed frame to the media file. */
            ret = av_interleaved_write_frame(oc, &pkt);
            if (ret != 0) {

                Msg::PrintErrorMsg("Error while writing audio frame.");
            }

          }
        }

       }
    }
    av_free_packet(&pkt);
    avcodec_free_frame(&frame);
}

    The thing is I never pass this statement: "if(pkt.stream_index ==st->index)".Packet stream index is never equal to the audio stream index.Anyone can point out where I am wrong?

    UPDATE:

    I did managed to open input audio stream for encoding but I can't encode audio and video streams into single output.From what I see PTS and DTS are probably the source of the problem.Currently I calculate pts based on muxing.c example but it doesn't work for audio at all.

    Here is how I use it :

       while(frame_count < _streamDurationNBFrames-1){

        uint8_t *frameToWrite =_frames.front();


        // Compute current audio and video time. ///

        if (audio_st){
            audio_pts = (double)audioIn_st->pts.val * audioIn_st->time_base.num / audioIn_st->time_base.den;
        }
        else{

            audio_pts = 0.0;
        }
        if (video_st){

            video_pts = (double)video_st->pts.val * video_st->time_base.num /   video_st->time_base.den;

        }else{
            video_pts = 0.0;
        }


        if ((!audio_st || audio_pts >= _streamDuration) && (!video_st || video_pts >= _streamDuration)){

            break;

        }


        if (audio_st && audio_pts < video_pts) {
            av_read_frame(informat, &pkt);//read audio from input stream
             Msg::PrintMsg("Encode audio here...");

          //==================   AUDIO ENCODE HERE   


           outpkt.data = pkt.data;
           outpkt.size = pkt.size;
           outpkt.stream_index = pkt.stream_index;
           outpkt.flags |= AV_PKT_FLAG_KEY;
           outpkt.pts = pkt.pts;
           outpkt.dts =pkt.dts;
           if(av_interleaved_write_frame(oc, &outpkt) < 0)
           {
            Msg::PrintErrorMsg("Fail Audio Write ");
           }
           else
           {
               audio_st->codec->frame_number++;
           }
           av_free_packet(&outpkt);
           av_free_packet(&pkt);



         }else{
          //==================   VIDEO  ENCODE HERE   

            write_video_frame(oc, video_st,frameToWrite);

            frame->pts += av_rescale_q(1, video_st->codec->time_base, video_st->time_base);
         }

        ///at last delete this frame:
        _frames.pop();
        delete frameToWrite; ///deallocate the written frame!
    }

    Somehow ,once I am in the audio encoding loop the audio_pts never reaches the video_pts and is always zero:

    audio_pts = (double)audio_st->pts.val * audio_st->time_base.num / audio_st->time_base.den; is always zero because  (double)audio_st->pts.val  returns zero.

    So basically I am asking the same question again:How to do the muxing when audio comes from external file?

    Btw,the answer below doesn't help as it presumes both audio and video streams come from the same file, whereas in my case only audio comes from the external source.

  • ffmpeg 0.9 won't upgrade or uninstall ehvOVJ

    8 juin 2016, par Jase

    I'm trying to upgrade ffmpeg on centos6 but everything I try doesn't work . I've tried reinstalling to overwrite it but no luck. this is the version and when it was installed. It installed 2 weeks ago and it installed as 0.9. I actually need it to be up to date (ffmpeg 3.0). Any suggestions would be great

  • Is it possible to save an audio stream with pure JavaScript in the browser ?

    8 juin 2016, par Konstantin

    I would like to make a webpage where visitors can save audio streams by clicking on links (not live streams, but links from a radio archive which uses streams), I want to do this without a server backend with pure JavaScript in the browser. I read somewhere about the JavaScript port of FFMpeg, which can encode and save video / audio in the browser utilizing so called blobs. However download library is huge, as far as I remember 17 MB. In fact I would need only stream copying the audio streams, not a real encoding process. I usually use similar commands to save a programme:

    ffmpeg -i http://stream.example.com/stream_20160518_0630.mp3 -c copy -t 3600 programme.mp3

    I wonder, is it possible to compile a subset of FFMpeg into JavaScript which provides only the really needed stream copying?

Navigation

Sites Web