Recherche avancée

Sur d’autres sites (4111)

• How to repair a raw H.264 stream using FFmpeg

  13 janvier 2020, par Chris Kennedy

  I apologize if I start rambling or seem incoherent.

  Preface : My house caught on fire last week.

  I have been trying to get the footage off my security camera system (Zmodo, don’t ever buy this crap) to share with the fire marshal and my insurance. I can’t get it off the "proper" way because the box doesn’t detect any drive or any file system I throw at it, and none of their viewing apps (for Android or Windows) allow video downloads.

  I was able to get the files directly from the system’s hard drive, but they’re (of course) messed up in such a way that I can’t just open up VLC (or even Zmodo’s own viewing software !) and view them. I have spent several hours scouring the web for various ways to run it through FFmpeg to see if it can repair the file, but I’ve had no luck so far and my mind is already stretched thin with everything else.

  This is the output from ffprobe :

  ffprobe version 4.2.1 Copyright (c) 2007-2019 the FFmpeg developers
  
    built with gcc 9.1.1 (GCC) 20190807
  
    configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libdav1d --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable-libvidstab --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --enable-libmfx --enable-amf --enable-ffnvcodec --enable-cuvid --enable-d3d11va --enable-nvenc --enable-nvdec --enable-dxva2 --enable-avisynth --enable-libopenmpt
  
    libavutil      56. 31.100 / 56. 31.100
  
    libavcodec     58. 54.100 / 58. 54.100
  
    libavformat    58. 29.100 / 58. 29.100
  
    libavdevice    58.  8.100 / 58.  8.100
  
    libavfilter     7. 57.100 /  7. 57.100
  
    libswscale      5.  5.100 /  5.  5.100
  
    libswresample   3.  5.100 /  3.  5.100
  
    libpostproc    55.  5.100 / 55.  5.100
  
  [h264 @ 000002479ee1c180] Format h264 detected only with low score of 1, misdetection possible!

  Then this repeats several dozen times :

      Last message repeated 1 times
  
  [h264 @ 000002479ee1de40] decode_slice_header error
  
  [h264 @ 000002479ee1de40] no frame!
  
  [h264 @ 000002479ee1de40] non-existing PPS 0 referenced

  And it finishes off with this :

  [h264 @ 000002479ee1c180] Could not find codec parameters for stream 0 (Video: h264, none): unspecified size
  
  Consider increasing the value for the 'analyzeduration' and 'probesize' options
  
  Input #0, h264, from '.\recfile_-200102-130000-135959-00001100.264':
  
    Duration: N/A, bitrate: N/A
  
      Stream #0:0: Video: h264, none, 25 fps, 25 tbr, 1200k tbn, 50 tbc

  I have tried several options with ffmpeg, including force_key_frames, setting the analyzeduration and probesize options to max_int, tried a -c copy, forcibly specified x264, tried to force a specific resolution (which it should all be standard HD video), and several others that I can’t recall at the moment.

  The video files are also accompanied by a .IDX file, but I haven’t figured out how important they are if they are at all to the video file.

  I can provide links to the smallest video and its associated IDX if needed, but if anyone can think of anything else to try on these files I’d greatly appreciate it.

• swr_convert float planar to S16

  8 janvier 2020, par Kevin Kouketsu

  How convert using libav API AV_SAMPLE_FMT_FLTP to AV_SAMPLE_FMT_S16

  I’m trying to figure out how to resample and encode PCM (captured from Microphone) 44.1KHz to AAC 48.0KHz

  That’s my resampler initializer :

  void initialize_resampler(SwrContext*& resamplerCtx, AVCodecContext* encoder, AVFrame*& rawResampledAudioFrame, AVStream* audioFormatStream)
  
  {
  
      int nb_samples = (encoder->codec->capabilities & AV_CODEC_CAP_VARIABLE_FRAME_SIZE) ? encoder->sample_rate : encoder->frame_size;
  
  
  
      int encoderFrameSize = encoder->channels * av_get_bytes_per_sample(encoder->sample_fmt) * encoder->frame_size;
  
      rawResampledAudioFrame = allocate_audioframe(encoder->sample_fmt, encoder->channel_layout, encoder->sample_rate, nb_samples);
  
  
  
      // Copy the stream parameters to the muxer
  
      check(avcodec_parameters_from_context(audioFormatStream->codecpar, encoder));
  
  
  
      // Create resampler context
  
      resamplerCtx = swr_alloc();
  
      if (resamplerCtx == nullptr)
  
          throw std::runtime_error("Could not allocate resampler context");
  
  
  
      // Set options
  
      check(av_opt_set_int(resamplerCtx, "in_channel_count", 2, 0));
  
      check(av_opt_set_int(resamplerCtx, "in_sample_rate", 44100, 0));
  
      check(av_opt_set_sample_fmt(resamplerCtx, "in_sample_fmt", AV_SAMPLE_FMT_S16, 0));
  
      check(av_opt_set_int(resamplerCtx, "out_channel_count", encoder->channels, 0));
  
      check(av_opt_set_int(resamplerCtx, "out_sample_rate", encoder->sample_rate, 0));
  
      check(av_opt_set_sample_fmt(resamplerCtx, "out_sample_fmt", encoder->sample_fmt, 0));
  
  
  
      // initialize the resampling context
  
      check(swr_init(resamplerCtx));
  
  }

  And for resample I’ve this code :

      AVPacket pkt{};
  
      while (true)
  
      {
  
          AVPacket input_packet;
  
          av_init_packet(&input_packet);
  
  
  
          check(av_read_frame(inputContext, &input_packet));
  
  
  
          check(avcodec_send_packet(decoderContext, &input_packet));
  
          check(avcodec_receive_frame(decoderContext, decodedFrame));
  
  
  
  
  
          // WHAT DO HERE swr_convert(resamplerContext, )
  
          av_packet_unref(&input_packet);
  
  
  
          av_init_packet(&pkt);
  
          auto in_stream = inputContext->streams[pkt.stream_index];
  
          auto out_stream = outputContext->streams[pkt.stream_index];
  
  
  
          pkt.pts = av_rescale_q_rnd(pkt.pts, in_stream->time_base, out_stream->time_base, (AVRounding)(AV_ROUND_NEAR_INF | AV_ROUND_PASS_MINMAX));
  
          pkt.dts = av_rescale_q_rnd(pkt.dts, in_stream->time_base, out_stream->time_base, (AVRounding)(AV_ROUND_NEAR_INF | AV_ROUND_PASS_MINMAX));
  
          pkt.duration = av_rescale_q(pkt.duration, in_stream->time_base, out_stream->time_base);
  
          pkt.pos = -1;
  
  
  
          check(avcodec_send_frame(encoderContext, decodedFrame));
  
          check(avcodec_receive_packet(encoderContext, &pkt));
  
  
  
          check(av_interleaved_write_frame(outputContext, &pkt));
  
          av_packet_unref(&pkt);
  
      }

  REading the docs I can’t figure out what exactly I need to pass to function. I have this code to encode PCM to MP2 (output is AV_SAMPLE_FMT_S16)

  const uint8_t* inPtr[] { const_cast<const>(&amp;pcmData[0]), nullptr, nullptr,nullptr,nullptr,nullptr,nullptr,nullptr };
  
  uint8_t* outPtr[] { &amp;resampledAudioData[0], nullptr, nullptr,nullptr,nullptr,nullptr,nullptr,nullptr };
  
  
  
  int resampledSamplesCount{ swr_convert(
  
      resamplerCtx,
  
      outPtr,
  
      maxResampledSamplesCount,
  
      inPtr,
  
      inputSampleCount) };
  
  
  
  // Negativo indica erro.
  
  check(resampledSamplesCount);
  
  </const>

  pcmData is raw data from input AVPacket (PCM)

  What I get here is : MP2 isn’t planar so it uses the same outPtr[0] different from plannar which needs two valid pointers to same writable data. But what I need to pass to inPtr, for example ?

  When I try to use the same code, ffmpeg try to write on outPtr[1] which is nullptr.

• How to hide/disable ffmpeg erros when using OpenCV (python) ?

  20 décembre 2019, par Mehran

  I’m using OpenCV python to capture a video.
  This is my code

  import cv2
  
  
  
  cap = cv2.VideoCapture("vid.mp4")
  
  while True:
  
      flag, frame =  cap.read()
  
  
  
      if not flag:
  
          cv2.imshow('video', frame)
  
  
  
      if cv2.waitKey(10) == 27:
  
          break

  When a frame is not ready it produces an error like this

  

  or

  Truncating packet of size 2916 to 1536
  
  [h264 @ 0x7ffa4180be00] AVC: nal size 2912
  
  [h264 @ 0x7ffa4180be00] AVC: nal size 2912
  
  [h264 @ 0x7ffa4180be00] no frame!
  
  
  
  [mov,mp4,m4a,3gp,3g2,mj2 @ 0x7ffa41803000] stream 0, offset 0x14565: partial file

  I wanted to find a way to hide this error ! I guess that this error is being produced by ffmpeg. Is there any way to hide or disable it ?

  This error is produced when I call cap.read(). And I also tried to wrap it with try ... except ... but it doesn’t work because it doesn’t throw any exceptions.