Recherche avancée

Sur d’autres sites (12943)

• Revision 30797 : On passe en version 0.6 avec tout le code revu et plus ou moins corrigé

  10 août 2009, par kent1@… — Log

  On passe en version 0.6 avec tout le code revu et plus ou moins corrigé

• Unable to convert mov to mp4 using ffmpeg carrierwave in ruby

  9 avril 2021, par LearningROR

  I am unable to convert MOV files to MP4 using carrierwave and FFmpeg.

  


  It does upload mp4 file but does not convert it to mp4 and so on browser only sound gets played not the video.

  


  What I am missing here ? I tried most of the solutions.

  


  Code :

  


  class VideoUploader < CarrierWave::Uploader::Base
  include CarrierWave::MiniMagick
  include CarrierWave::Video
  include CarrierWave::Video::Thumbnailer
  include CarrierWave::FFmpeg
  # Choose what kind of storage to use for this uploader:
  storage :file
  # Override the directory where uploaded files will be stored.
  # This is a sensible default for uploaders that are meant to be mounted:
  def store_dir
    "uploads/#{model.class.to_s.underscore}/#{mounted_as}/#{model.id}"
  end
  version :video, :if => :video? do
    process :encode
  end
  version :thumb do
    process thumbnail: [{format: 'jpg', quality: 8, size: 360, logger: Rails.logger, square: false}]
    def full_filename for_file
      jpg_name for_file, version_name
    end
  end
  def encode
    tmp_path = File.join( File.dirname(current_path), "tmpfile.mp4" )
    movie = FFMPEG::Movie.new(current_path)
    movie.transcode(tmp_path, custom: %w(-c:v libx264 -c:a aac -vf format=yuv420p -movflags +faststart)) do |progress|
      puts progress
    end
    File.rename tmp_path, current_path
  end
  protected
  def video?(new_file)
    new_file.content_type.include? 'video'
  end
end


  


  Logs :

  


  I, [2021-04-09T15:33:24.783337 #9397]  INFO -- : Running transcoding...
["/usr/local/bin/ffmpeg", "-y", "-i", "/Users/osx/workspace_ror/xxx-react/tmp/1617964403-459168728101935-0021-9213/video/sample_iTunes__1_.mov", "-c:v", "libx264", "-c:a", "aac", "-vf", "format=yuv420p", "-movflags", "+faststart", "/Users/osx/workspace_ror/xxx-react/tmp/1617964403-459168728101935-0021-9213/video/tmpfile.mp4"]

0.0
0.08304093567251461
0.1672514619883041
0.2327485380116959
0.3229239766081871
0.4008187134502924
0.45426900584795327
0.5033918128654971
0.5403508771929825
0.6100584795321637
0.663859649122807
0.711812865497076
0.7657309941520468
0.8196491228070175
0.87953216374269
0.9394152046783625
1.0
1.0
I, [2021-04-09T15:33:33.555508 #9397]  INFO -- : Transcoding of /Users/osx/workspace_ror/xxx-react/tmp/1617964403-459168728101935-0021-9213/video/sample_iTunes__1_.mov to /Users/osx/workspace_ror/xxx-react/tmp/1617964403-459168728101935-0021-9213/video/tmpfile.mp4 succeeded

Running....ffmpegthumbnailer -i /Users/osx/workspace_ror/xxx-react/tmp/1617964403-459168728101935-0021-9213/thumb/sample_iTunes__1_.mov -o /Users/osx/workspace_ror/xxx-react/tmp/1617964403-459168728101935-0021-9213/thumb/tmpfile.jpg -c jpg -q 8 -s 360
Success!


  


• Audio extracted using ffmpeg giving nan and very low/high (2.6283893e-37) values

  17 février 2023, par Shreya Gupta

  I'm extracting raw audio from a .flac file(youtube link - https://youtu.be/—BfvyPmVMo , extracting 10sec from this) and storing it as np-array float32. Getting nan, and extremely low and high values = [-8.63468980e+30 -9.76911766e+34 -1.93298665e+37 -2.69400132e+38 .... 2.33467619e-38  5.71834450e-39  6.70703932e-38  7.58865403e-36].

  


  On this when I try to get melspectrogram then it gives an error -

  


  TypeError: array([nan, nan, nan, ..., nan, nan, nan], dtype=float32) has type numpy.ndarray, but expected one of: int, long, float


  


  To extract raw audio using(this is similar to what is done in dmvr lib, https://github.com/deepmind/dmvr/blob/master/examples/generate_from_file.py)

  


  cmd = (
      ffmpeg
      .input(video_path, ss=start, t=end-start)
      .output("pipe:", ac=1, ar=sampling_rate, format="s32le")
  )

  audio, _ = cmd.run(capture_stdout=True, quiet=True)

  audio = np.frombuffer(audio, np.float32) "


  


  To get melspectrogram -

  


  mel_spec = librosa.feature.melspectrogram(audio,sr=16000,hop_length=10,n_mels=128,window=25)


  


  Errors

  


  C:\Users\Shreya Gupta\IS-Project\generate_from_file.py:142: FutureWarning: Pass y=[-6.5381321e+37  1.5676004e-38  7.2428216e-37 ...  2.6283893e-37
  2.5346798e-36  1.9253581e-37] as keyword args. From version 0.10 passing these as positional arguments will result in an error
  mel_spec = librosa.feature.melspectrogram(audio,sr=16000,hop_length=10,n_mels=128,window=25)
C:\ProgramData\Anaconda3\lib\site-packages\librosa\core\spectrum.py:2563: RuntimeWarning: overflow encountered in absolute
  np.abs(
C:\ProgramData\Anaconda3\lib\site-packages\librosa\core\spectrum.py:2563: RuntimeWarning: overflow encountered in square
  np.abs().....
TypeError: array([nan, nan, nan, ..., nan, nan, nan], dtype=float32) has type numpy.ndarray, but expected one of: int, long, float


  


  Could someone please help, thank you !