Recherche avancée

Sur d’autres sites (6606)

• Transcoding audio using xuggler

  23 juin 2014, par amd

  I am trying to convert an audio file with the header

  Opening audio decoder: [pcm] Uncompressed PCM audio decoder
  
  AUDIO: 44100 Hz, 2 ch, s16le, 1411.2 kbit/100.00% (ratio: 176400->176400)
  
  Selected audio codec: [pcm] afm: pcm (Uncompressed PCM)

  I want to transcode this file to mp3 format. I have following code snippet but its not working well. I have written it using XUGGLER code snippet for transcoding audio and video.

  Audio decoder is

      audioDecoder = IStreamCoder.make(IStreamCoder.Direction.DECODING, ICodec.findDecodingCodec(ICodec.ID.CODEC_ID_PCM_S16LE));
  
      audioDecoder.setSampleRate(44100);
  
      audioDecoder.setBitRate(176400);
  
      audioDecoder.setChannels(2);
  
      audioDecoder.setTimeBase(IRational.make(1,1000));
  
      if (audioDecoder.open(IMetaData.make(), IMetaData.make()) < 0)
  
          return false;
  
      return true; 

  Audio encoder is

      outContainer = IContainer.make();
  
      outContainerFormat = IContainerFormat.make();
  
      outContainerFormat.setOutputFormat("mp3", urlOut, null);
  
      int retVal = outContainer.open(urlOut, IContainer.Type.WRITE, outContainerFormat);
  
      if (retVal < 0) {
  
          System.out.println("Could not open output container");
  
          return false;
  
      }
  
      outAudioCoder = IStreamCoder.make(IStreamCoder.Direction.ENCODING, ICodec.findEncodingCodec(ICodec.ID.CODEC_ID_MP3));
  
      outAudioStream = outContainer.addNewStream(outAudioCoder);
  
      outAudioCoder.setSampleRate(new Integer(44100));
  
      outAudioCoder.setChannels(2);
  
      retVal = outAudioCoder.open(IMetaData.make(), IMetaData.make());
  
      if (retVal < 0) {
  
          System.out.println("Could not open audio coder");
  
          return false;
  
      }
  
      retVal = outContainer.writeHeader();
  
      if (retVal < 0) {
  
          System.out.println("Could not write output FLV header: ");
  
          return false;
  
      }
  
      return true;

  And here is encode method where i send packets of 32 byte to transcode

  public void encode(byte[] audioFrame){
  
      //duration of 1 video frame
  
      long lastVideoPts = 0;
  
  
  
      IPacket packet_out = IPacket.make();
  
      int lastPos = 0;
  
      int lastPos_out = 0;
  
  
  
      IAudioSamples audioSamples = IAudioSamples.make(48000, audioDecoder.getChannels());
  
      IAudioSamples audioSamples_resampled = IAudioSamples.make(48000, audioDecoder.getChannels());
  
  
  
      //we always have 32 bytes/sample
  
      int pos = 0;
  
      int audioFrameLength = audioFrame.length;
  
      int audioFrameCnt = 1;
  
      iBuffer = IBuffer.make(null, audioFrame, 0, audioFrameLength);
  
      IPacket packet = IPacket.make(iBuffer);
  
      //packet.setKeyPacket(true);
  
      packet.setTimeBase(IRational.make(1,1000));
  
      packet.setDuration(20);
  
      packet.setDts(audioFrameCnt*20);
  
      packet.setPts(audioFrameCnt*20);
  
      packet.setStreamIndex(1);
  
      packet.setPosition(lastPos);
  
      lastPos+=audioFrameLength;
  
      int pksz = packet.getSize();
  
      packet.setComplete(true, pksz);
  
      /*
  
      * A packet can actually contain multiple samples
  
      */
  
      int offset = 0;
  
      int retVal;
  
      while(offset < packet.getSize())
  
      {
  
          int bytesDecoded = audioDecoder.decodeAudio(audioSamples, packet, offset);
  
          if (bytesDecoded < 0)
  
              throw new RuntimeException("got error decoding audio ");
  
          offset += bytesDecoded;
  
          if (audioSamples.isComplete())
  
          {
  
              int samplesConsumed = 0;
  
              while (samplesConsumed < audioSamples.getNumSamples()) {
  
                  retVal = outAudioCoder.encodeAudio(packet_out, audioSamples, samplesConsumed);
  
                  if (retVal <= 0)
  
                      throw new RuntimeException("Could not encode audio");
  
                  samplesConsumed += retVal;
  
                  if (packet_out.isComplete()) {
  
                      packet_out.setPosition(lastPos_out);
  
                      packet_out.setStreamIndex(1);
  
                      lastPos_out+=packet_out.getSize();
  
                      retVal = outContainer.writePacket(packet_out);
  
                      if(retVal < 0){
  
                          throw new RuntimeException("Could not write data packet");
  
                      }
  
                  }
  
              }
  
          }
  
  
  
      }
  
  
  
  }

  I get an output file but it doesnt get played. I have very little experience of audio encoding and sampling. Thanks in advance.

• libavcodec get video duration and framerate

  17 septembre 2013, par Tishu

  I have a video encoded in .3gp h.264 and I am looking to get its framerate and duration in C. Here is the code I use after opening the file and finding the appropriate codecs :

  AVRational rational = gVideoCodecCtx->time_base;
  
  
  
  LOGI(10, "numerator is %i", rational.num);
  
  LOGI(10, "denominator is %i", rational.den);
  
  LOGI(10, "duration is %d", gFormatCtx->duration);
  
  LOGI(10, "fps is %d", (double)av_q2d(rational));

  And here is the output :

  12-02 12:30:19.819: I/FFmpegTest(23903): numerator is 1
  
  12-02 12:30:19.819: I/FFmpegTest(23903): denominator is 180000
  
  12-02 12:30:19.819: I/FFmpegTest(23903): duration is 6594490
  
  12-02 12:30:19.819: I/FFmpegTest(23903): fps is 1692926992

  From the documentation I understand that the duration is meant to be "duration/time_base" which gives me 6594490 / 180000 = 36.6. The duration of my video file is 6 seconds and I do not know where this factor of 6 would come from.

  Also the framerate seems to be completely off.

  It is currenlty hard to find help as a lot of tutorials use deprecated methods and the documentation does not give examples.

  Any help would be appreciated.

  Thanks

  Edit :
  Thanks to the comment below I managed to print the following

  12-02 18:59:36.279: I/FFmpegTest(435): numerator is 1
  
  12-02 18:59:36.279: I/FFmpegTest(435): denominator is 180000
  
  12-02 18:59:36.279: I/FFmpegTest(435): duration is 6594490
  
  12-02 18:59:36.279: I/FFmpegTest(435): fps is 0.000006

  I also managed to find out a frame's timestamp in msec with this :

  int msec = 1000*(packet.pts * timeBase * gVideoCodecCtx->ticks_per_frame);

  This returns me something that's roughly 33fps (I expect 30). But I can't figure out how to retrieve the duration. The documentation says "duration of the stream, in AV_TIME_BASE fractional seconds" but 6594490 * 0.000006 = 39.5 - the correct duration is 6.3 seconds). Also the exact fps is 30 but nor sure how to get from 0.000006 to 30 with the above figures)

  Thanks

• how to explain this C language code from ffmpeg remuxing

  20 avril 2017, par Fidona

  can someone help me what the meaning of this code ? this is a code from ffmpeg remuxing code.

  static void log_packet(const AVFormatContext *fmt_ctx, const AVPacket *pkt, const char *tag, int hours, int mins, int secs, int us, int *time )
  
  {
  
      AVRational *time_base = &fmt_ctx->streams[pkt->stream_index]->time_base;
  
      total = total + pkt->duration;
  
    //  printf( "%d:%d:%d.%d \n", hours,  mins,  secs, us );
  
      *time = av_q2d(*time_base) * pkt->pts ;
  
  }

  I’m kind of new to C language. thank you !