Recherche avancée

Sur d’autres sites (4950)

• Unable to use Multithread for librosa melspectrogram

  15 avril 2019, par Raven Cheuk

  I have over 1000 audio files (it’s just a initial development, in the future, there will be even more audio files), and would like to convert them to melspectrogram.

  Since my workstation has a Intel® Xeon® Processor E5-2698 v3, which has 32 threads, I would like to use multithread to do my job.

  My code

  import os
  
  import librosa
  
  from librosa.display import specshow
  
  from natsort import natsorted
  
  import numpy as np
  
  import sys 
  
  # Libraries for multi thread
  
  from multiprocessing.dummy import Pool as ThreadPool
  
  import subprocess
  
  pool = ThreadPool(20) 
  
  
  
  songlist = os.listdir('../opensmile/devset_2015/')
  
  songlist= natsorted(songlist)
  
  
  
  def get_spectrogram(song):
  
      print("start")
  
      y, sr = librosa.load('../opensmile/devset_2015/' + song)
  
  
  
      ## Add some function to cut y
  
      y_list = y
  
      ##
  
  
  
      for i, y_i in enumerate([y_list]): # can remove for loop if no audio is cut
  
          S = librosa.feature.melspectrogram(y=y, sr=sr, n_mels=128,fmax=8000)
  
          try:
  
              np.save('./Test/' + song + '/' + str(i), S)
  
          except:
  
              os.makedirs('./Test/' + song)
  
              np.save('./Test/' + song + '/' + str(i), S)
  
          print("done saving")
  
  
  
  pool.map(get_spectrogram, songlist)

  My Problem

  However, my script freezes after finished the first conversion.

  To debug what’s going on, I commented out S = librosa.feature.melspectrogram(y=y, sr=sr, n_mels=128,fmax=8000) and replace it by S=0.
  Then the multi-thread code works fine.

  What’s wrong with the librosa.feature.melspectrogram function ? Does it not support multi-thread ? Or is it a problem of ffmpeg ? (When using librosa, it asks me to install ffmpeg before.)

• How can i use ffmpeg to add voice note to another mp3 at specific time , let everything play without reducing the original mp3

  29 juin 2020, par Adike maduabuchi

  I have used some code which does not give me what I want, I used PHP shell to execute the code

  


               $var = 'ffmpeg -y -i ' . $voice_note_tag . ' -i ' . $original_music_path . '  -filter_complex "[0]asplit[a][b]; [a]atrim=duration=' . $time_come_vioce . ',volume=\'1-max(0*(t-1),0)\':eval=frame[pre];  [b]atrim=start=' . $time_come_vioce . ',asetpts=PTS-STARTPTS[song];  [song][1]amix=inputs=2:duration=first:dropout_transition=2[post]; [pre][post]concat=n=2:v=0:a=1[mixed]"  -map "[mixed]" ' . $output . ' 2>&1';
                            shell_exec($var);


  


• Returns wrong frame while seeking in FFMPEG Player

  5 février 2015, par user543

  I have implemented an audio player which supports all the formats using FFMPEG. But while seeking the audio, the application is getting the wrong frame at that target duration. Here is my code for the seek functionality.

           int64_t seekTime = av_rescale_q(seekValue * AV_TIME_BASE,
  
                      AV_TIME_BASE_Q,
  
                      fmt_ctx->streams[seekStreamIndex]->time_base);
  
  
  
              int64_t seekStreamDuration =
  
                      fmt_ctx->streams[seekStreamIndex]->duration;
  
  
  
              int flags = AVSEEK_FLAG_BACKWARD;
  
              if (seekTime > 0 && seekTime < seekStreamDuration)
  
                  flags |= AVSEEK_FLAG_ANY;
  
  
  
              int ret = av_seek_frame(fmt_ctx, seekStreamIndex, seek_target,
  
                      flags);
  
  
  
              if (ret < 0)
  
                  ret = av_seek_frame(fmt_ctx, seekStreamIndex, seekTime,
  
                          flags);
  
  
  
            avcodec_flush_buffers(dec_ctx);

  It is working fine for most of the songs.
  But some of the mp3 songs are getting the wrong timespan.

  For instance : If the song total duration is 2 minutes, if I play the song without seeking it works fine. But while playing, if we seek the song to some position, this will make the song ends with 2 mins and 10 seconds.

  I am getting this issue with only mp3 songs. I am using FFMPEG 2.1. The code for the seek functionality is working fine on FFMPEG 0.11.1.

  Please provide any information about this issue.