Recherche avancée

Sur d’autres sites (9629)

• Programatically get non-overlapping images from MP4

  15 février 2013, par Carlos F

  My ultimate goal is to get meaningful snapshots from MP4 videos that are either 30 min or 1 hour long. "Meaningful" is a bit ambitious, so I have simplified my requirements.

  The image should be crisp - non-overlapping, and ideally not blurry. Initially, I thought getting a keyframe would work, but I had no idea that keyframes could have overlapping images embedded in them like this :

  Of course, some keyframe images look like this and those are much better :

  

  I was wondering if someone might have source code to :

  Take a sequence of say 10-15 continuous keyframes (jpg or png) and identify the best keyframe from all of them.

  This must happen entirely programaticaly. I found this paper : http://research.microsoft.com/pubs/68802/blur_determination_compressed.pdf

  and felt that I could "rank" a few images based on the above paper, but then I was dissuaded by this link : Extracting DCT coefficients from encoded images and video given that my source video is an MP4. Of course, this confuses me because the input into the system is just a sequence of jpg images.

  Another link that is interesting is :

  Detection of Blur in Images/Video sequences

  However, I am not sure if this will work for "overlapping" images.

  Any ideas ?

• How do I convert ADPCM to PCM using FFmpeg ?

  26 janvier 2013, par mystafer

  I have a video feed that sends me audio using the ADPCM codec. However, android only supports PCM format. How can I convert the ADPCM audio feed into a PCM audio feed ?

  The answer to this may be similar to the answer to this question.

  I have successfully decoded the frame with this code :

  int len = avcodec_decode_audio4(pAudioCodecCtx, pAudioFrame, &frameFinished, &packet);

  Is the secret here to use a reverse encode function ?

  Here is what I have so far in my audio decode function :

  if(packet_queue_get(env, javaThread, pAudioPacketQueue, &packet, 1) < 0) {
  
      LOGE("audio - after get packet failed");
  
      return;
  
  }
  
  LOGD("Dequeued audio packet");
  
  
  
  // calculate frame size
  
  int frameSize;
  
  if (pPcmAudioCodecCtx->frame_size) {
  
      frameSize = pPcmAudioCodecCtx->frame_size;
  
  } else {
  
      /* if frame_size is not set, the number of samples must be
  
       * calculated from the buffer size */
  
      int64_t nb_samples = (int64_t)AUDIO_PCM_OUTBUFF_SIZE * 8 /
  
              (av_get_bits_per_sample(pPcmAudioCodecCtx->codec_id) *
  
                      pPcmAudioCodecCtx->channels);
  
      frameSize = nb_samples;
  
  }
  
  
  
  int pcmBytesPerSample = av_get_bytes_per_sample(pPcmAudioCodecCtx->sample_fmt);
  
  int pcmFrameBytes = frameSize * pcmBytesPerSample * pPcmAudioCodecCtx->channels;
  
  
  
  uint8_t *pDataStart = packet.data;
  
  while(packet.size > 0) {
  
      int len = avcodec_decode_audio4(pAudioCodecCtx, pAudioFrame, &frameFinished, &packet);
  
      LOGD("Decoded ADPCM frame");
  
  
  
      if (len < 0) {
  
          LOGE("Error while decoding audio");
  
          return;
  
      }
  
  
  
      if (frameFinished) {
  
          // store frame data in FIFO buffer
  
          uint8_t *inputBuffer = pAudioFrame->data[0];
  
          int inputBufferSize = pAudioFrame->linesize[0];
  
          av_fifo_generic_write(fifoBuffer, inputBuffer, inputBufferSize, NULL);
  
          LOGD("Added ADPCM frame to FIFO buffer");
  
  
  
          // check if fifo buffer has enough data for a PCM frame
  
          while (av_fifo_size(fifoBuffer) >= pcmFrameBytes) {
  
              LOGI("PCM frame data in FIFO buffer");
  
  
  
              // read frame's worth of data from FIFO buffer
  
              av_fifo_generic_read(fifoBuffer, pAudioPcmOutBuffer, pcmFrameBytes, NULL);
  
              LOGD("Read data from FIFO buffer into pcm frame");
  
  
  
  
  
              avcodec_get_frame_defaults(pPcmAudioFrame);
  
              LOGD("Got frame defaults");
  
  
  
              pPcmAudioFrame->nb_samples = pcmFrameBytes / (pPcmAudioCodecCtx->channels *
  
                      pcmBytesPerSample);
  
  
  
              avcodec_fill_audio_frame(pPcmAudioFrame, pPcmAudioCodecCtx->channels,
  
                      pPcmAudioCodecCtx->sample_fmt,
  
                      pAudioPcmOutBuffer, pcmFrameBytes, 1);
  
              LOGD("Filled frame audio with data");
  
  
  
              // fill audio play buffer
  
              int dataSize = pPcmAudioFrame->linesize[0];
  
              LOGD("Data to output: %d", dataSize);
  
              jbyteArray audioPlayBuffer = (jbyteArray) env->GetObjectField(ffmpegCtx, env->GetFieldID(cls, "audioPlayBuffer", "[B"));
  
              jbyte *bytes = env->GetByteArrayElements(audioPlayBuffer, NULL);
  
              memcpy(bytes, pPcmAudioFrame->data[0], dataSize);
  
              env->ReleaseByteArrayElements(audioPlayBuffer, bytes, 0);
  
              LOGD("Copied data into Java array");
  
  
  
              env->CallVoidMethod(player, env->GetMethodID(playerCls, "updateAudio", "(I)V"), dataSize);
  
          }

• Restreaming video containing two languages live with ffmpeg

  9 novembre 2012, par user1810837

  I have a project where i need to restream a live stream which has two languages setup on the audio.
  Spanish on left and English on right

  The stream mapping is :

  Stream #0:0: Video: h264 ([7][0][0][0] / 0x0007), yuv420p, 512x288 [SAR 1:1 DAR 16:9], q=2-31, 1k tbn, 1k tbc
  
  Stream #0:1: Audio: mp3 ([2][0][0][0] / 0x0002), 44100 Hz, stereo, s16, 18 kb/s

  I need to restream this back live with just the English from the right side or just spanish from the left side, I tried looking everywhere but did not find any type of solution .

  Since this needs to be done live, I can't be using other programs to separate video and audio to get it done.

  This needs to be done through ffmpeg and I wonder if it even capable of doing so with original built or it would need some custom modification.