Recherche avancée

Sur d’autres sites (8428)

• FFMPEG silencedetect output does not match Audacity Sound Finder

  12 avril 2017, par tngo

  I’m currently working on labeling sound sections in an audio file.

  When I ran my audio file through Audacity Sound Finder with the following settings this is what I get

  0.448980    0.498866    1
  
  0.698413    0.927891    2
  
  1.147392    1.426757    3
  
  1.566440    1.815873    4

  These are the first 4 labels for the audio file where the sounds were found to be louder than -12dB.

  However when I ran the same audio file through FFMPEG silencedetect filter with n=-12dB:d=0.1 this is the output

  [silencedetect @ 000000000272a8c0] silence_start: -0.00712018
  
  [silencedetect @ 000000000272a8c0] silence_end: 0.371519 | silence_duration: 0.378639
  
  [silencedetect @ 000000000272a8c0] silence_start: 0.457279
  
  [silencedetect @ 000000000272a8c0] silence_end: 0.650159 | silence_duration: 0.19288
  
  [silencedetect @ 000000000272a8c0] silence_start: 0.828798
  
  [silencedetect @ 000000000272a8c0] silence_end: 1.11456 | silence_duration: 0.28576
  
  [silencedetect @ 000000000272a8c0] silence_start: 1.38608
  
  [silencedetect @ 000000000272a8c0] silence_end: 1.8576 | silence_duration: 0.1
  
  [silencedetect @ 000000000272a8c0] silence_start: 2.31488

  For FFMPEG to detect sound I’m considering silence_end to be sound_start and silence_start to be sound_end. With that logic the output from FFMPEG actually is

  0.371519 0.457279 1
  
  0.650159 0.828798 2
  
  1.11456  1.38608  3
  
  1.8576   2.31488  4

  To me it seems like all the timestamps from FFMPEG are shifted to the left compare to Audacity Sound Finder. Is there something that I’m missing ?

  Any help would be much appreciated.
  Thanks,

• Revision d115dbc24c : Adjust style to match Google Coding Style a little more closely. Most of these

  30 octobre 2012, par Ronald S. Bultje

  Changed Paths : Modify /vp8/common/onyx.h Modify /vp8/encoder/bitstream.c Modify /vp8/encoder/dct.c Modify /vp8/encoder/encodeframe.c Modify /vp8/encoder/encodeintra.c Modify /vp8/encoder/firstpass.c Modify /vp8/encoder/generic/csystemdependent.c (...)

• Google Speech API "Sample rate in request does not match FLAC header"

  13 février 2017, par kjdion84

  I’m trying to convert an mp4 video clip into a FLAC audio file and then have google speech spit out the words from the video so that I can detect if specific words were said.

  I have everything working except that I am getting an error from the Speech API :

  {
  
    "error": {
  
      "code": 400,
  
      "message": "Sample rate in request does not match FLAC header.",
  
      "status": "INVALID_ARGUMENT"
  
    }
  
  }

  I am using FFMPEG in order to convert the mp4 into a FLAC file. I am specifying that the FLAC file be 16 bits in the command, but when I right click on the FLAC file Windows is telling me it is 302kbps.

  Here is my PHP code :

  // convert mp4 video to 16 bit flac audio file
  
  $cmd = 'C:/wamp/www/ffmpeg/bin/ffmpeg.exe -i C:/wamp/www/test.mp4 -c:a flac -sample_fmt s16 C:/wamp/www/test.flac';
  
  exec($cmd, $output);
  
  
  
  // convert flac to text so we can detect if certain words were said
  
  $data = array(
  
      "config" => array(
  
          "encoding" => "FLAC",
  
          "sampleRate" => 16000,
  
          "languageCode" => "en-US"
  
      ),
  
      "audio" => array(
  
          "content" => base64_encode(file_get_contents("test.flac")),
  
      )
  
  );
  
  
  
  $json_data = json_encode($data);
  
  
  
  $ch = curl_init();
  
  curl_setopt($ch, CURLOPT_URL, 'https://speech.googleapis.com/v1beta1/speech:syncrecognize?key=MY_API_KEY');
  
  curl_setopt($ch, CURLOPT_HTTPHEADER, array("Content-Type: application/json"));
  
  curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
  
  curl_setopt($ch, CURLOPT_POST, true);
  
  curl_setopt($ch, CURLOPT_POSTFIELDS, $json_data);
  
  curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false);
  
  
  
  $result = curl_exec($ch);