Recherche avancée

Sur d’autres sites (7874)

• Cannot use pydub.AudioSegment on audio files downloaded by youtube-dl library

  26 octobre 2018, par Han M

  After I download a audio clip from a youtube link by the youtube-dl library, I cannot use pydub.AudioSegment to slice the audio file. The following is the information about my code.

  import youtube_dl
  
  options = {
  
          'format': 'bestaudio/best',
  
          'extractaudio': True,  # only keep the audio
  
          'audio-format': "wav",  # convert to wav
  
          'outtmpl': whole_path,  # name the file the ID of the video
  
          'noplaylist': True,  # only download single song, not playlist
  
          'audioquality': 1
  
  }
  
  with youtube_dl.YoutubeDL(options) as ydl:
  
      print('url is:', web_url)
  
      ydl.download([web_url])
  
  
  
  from pydub import AudioSegment
  
  audio_data = AudioSegment.from_wav(path_to_downloaded_file)

  Then, I got an error as follows first on Mac and second on Linux :

  b'avconv version 12.3, Copyright (c) 2000-2018 the Libav developers\n  built on Jul 26 2018 18:08:50 with Apple LLVM version 9.1.0 (clang-902.0.39.2)\n/Users/***/random17560390.wav: Invalid data found when processing input\n'
  
  
  
  b'ffmpeg version 2.8.15-0ubuntu0.16.04.1 Copyright (c) 2000-2018 the FFmpeg developers\n  built with gcc 5.4.0 (Ubuntu 5.4.0-6ubuntu1~16.04.10) 20160609\n  configuration: --prefix=/usr --e
  
  xtra-version=0ubuntu0.16.04.1 --build-suffix=-ffmpeg --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --cc=cc --cxx=g++ --enable-gpl --enable-
  
  shared --disable-stripping --disable-decoder=libopenjpeg --disable-decoder=libschroedinger --enable-avresample --enable-avisynth --enable-gnutls --enable-ladspa --enable-libass --enable-li
  
  bbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libgme --enable-libgsm --enable-libmod
  
  plug --enable-libmp3lame --enable-libopenjpeg --enable-libopus --enable-libpulse --enable-librtmp --enable-libschroedinger --enable-libshine --enable-libsnappy --enable-libsoxr --enable-li
  
  bspeex --enable-libssh --enable-libtheora --enable-libtwolame --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx265 --enable-libxvid --enable-libzvbi --e
  
  nable-openal --enable-opengl --enable-x11grab --enable-libdc1394 --enable-libiec61883 --enable-libzmq --enable-frei0r --enable-libx264 --enable-libopencv\n  libavutil      54. 31.100 / 54.
  
   31.100\n  libavcodec     56. 60.100 / 56. 60.100\n  libavformat    56. 40.101 / 56. 40.101\n  libavdevice    56.  4.100 / 56.  4.100\n  libavfilter     5. 40.101 /  5. 40.101\n  libavresa
  
  mple   2.  1.  0 /  2.  1.  0\n  libswscale      3.  1.101 /  3.  1.101\n  libswresample   1.  2.101 /  1.  2.101\n  libpostproc    53.  3.100 / 53.  3.100\n[wav @ 0x20854c0] invalid start
  
   code [0][0][0][28] in RIFF header\n/home/users/test_download/random10485395.wav: Invalid data found when processing input\n'

  I think the problem might be about the encoding on the downloaded files by youtube-dl. I can use pydub.AudioSegment on other normal .wav files. After checking the details of the downloaded file by youtube-dl, I found the following feature :

  Stream #0:0(und): Audio: aac (LC) (mp4a / 0x6134706D), 44100 Hz, stereo, fltp, 125 kb/s (default)

  It is different from other normal .wav files, which have the following :

  Stream #0:0: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 44100 Hz, 2 channels, s16, 1411 kb/s

• Music bot returning errors at core.py and bot.py

  13 septembre 2021, par Loganox

  The code for the music bot I'm trying to write in python keeps returning an error with the following code :

  


  @bot.command(name='play_song', help='To play song')
async def play(ctx,url):
    
    #if not ctx.message.author.name=="Rohan Krishna" :
    #     await ctx.send('NOT AUTHORISED!')
    #     return
    if 1==1:#try :
        server = ctx.message.guild
        voice_channel = server.voice_client
        print("try 1 was a success")
        
        async with ctx.typing():
            filename = await YTDLSource.from_url(url, loop=bot.loop)
            #voice_channel.play(discord.FFmpegPCMAudio(executable="ffmpeg.exe", source=filename)) #ISSUE HERE
            print("try 2 was a success")
            voice = get(ctx.bot.voice_clients, guild=ctx.guild)
    
            #voice.play(discord.FFmpegPCMAudio('test.mp3'), after=your_check)
            #voice.source = discord.PCMVolumeTransformer(voice.source)
            #voice.source.volume = 0.5
        await ctx.send('**Now playing:** {}'.format(filename))
    if 1==0:#except:
        await ctx.send("The bot is not connected to a voice channel.")
        print("try 3 was unfortunately a success")


  


  It normally has a try/except function but I replaced it with if true and false statements to force it to run instead of breaking to determine the exact line of it breaking. The issue supposedly lies in

  


  voice_channel.play(discord.FFmpegPCMAudio(executable="ffmpeg.exe", source=filename)) #ISSUE HERE


  


  The error I get returned is :

  


  Traceback (most recent call last):
  File "C:\Users\heyin\anaconda3\lib\site-packages\discord\ext\commands\core.py", line 85, in wrapped
    ret = await coro(*args, **kwargs)
  File "app2.py", line 75, in play
    voice.play(discord.FFmpegPCMAudio('test.mp3'), after=your_check)
  File "C:\Users\heyin\anaconda3\lib\site-packages\discord\player.py", line 225, in __init__
    super().__init__(source, executable=executable, args=args, **subprocess_kwargs)
  File "C:\Users\heyin\anaconda3\lib\site-packages\discord\player.py", line 138, in __init__
    self._process = self._spawn_process(args, **kwargs)
  File "C:\Users\heyin\anaconda3\lib\site-packages\discord\player.py", line 147, in _spawn_process
    raise ClientException(executable + ' was not found.') from None
discord.errors.ClientException: ffmpeg was not found.

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "C:\Users\heyin\anaconda3\lib\site-packages\discord\ext\commands\bot.py", line 939, in invoke
    await ctx.command.invoke(ctx)
  File "C:\Users\heyin\anaconda3\lib\site-packages\discord\ext\commands\core.py", line 863, in invoke
    await injected(*ctx.args, **ctx.kwargs)
  File "C:\Users\heyin\anaconda3\lib\site-packages\discord\ext\commands\core.py", line 94, in wrapped
    raise CommandInvokeError(exc) from exc
discord.ext.commands.errors.CommandInvokeError: Command raised an exception: ClientException: ffmpeg was not found.


  


  I added the print statements to determine which lines were working so those can obviously be ignored. It seems to be working fine except for this, other commands in the bot are working appropriately.

  


• Batch generating spectrograms with ffmpeg or SoX : how to customize their appearance, and is my code so far correct ?

  4 mai 2024, par Mark I

  I have a .flac music library, and for organizational and quality assessment reasons I want to have spectrograms of every song in it. After finding out that the program I usually use to view them, Spek, doesn't offer a batch export option, I have found ways to batch generate spectrograms with both SoX and fmpeg. Below you can find the two scripts I've got, which take as an input a folder which contains flac files or subfolders containing flac files, and output their spectrograms (with the same names as their corresponding flacs) to an output folder with the same structure as the input folder :

  


  SoX :

  


  @echo off
setlocal enabledelayedexpansion

REM Root folder containing flac files
set "root_folder=%~1"

REM Output folder for spectrogram images
set "output_folder=0. Spectrograms"

REM Create the output folder if it doesn't exist
mkdir "%output_folder%" 2>nul

REM Loop through all flac files recursively
for /r "%root_folder%" %%F in (*.flac) do (
    REM Get the directory of the flac file
    set "directory=%%~dpF"

    REM Get the relative path of the directory from the root folder
    set "relative_path=!directory:%root_folder%\=!"

    REM Create the corresponding subfolder in the output folder
    mkdir "%output_folder%\!relative_path!" 2>nul

    REM Get the filename without extension
    set "filename=%%~nF"

    REM Get the filename with full path
    set "filename_with_path=%%F"

    REM Generate spectrogram image using sox
    sox "%%F" -n spectrogram -x 640 -y 360 -t "!filename_with_path!" -o "%output_folder%\!relative_path!\!filename!.png"
)



  


  It's slow but it works as intended, this is what the output looks like :

  


  ffmpeg :

  


  @echo off
SETLOCAL ENABLEDELAYEDEXPANSION

:: The root directory of the input files
SET "rootdir=%~1"

:: The output directory where the spectrograms will be saved
SET "outputdir=%rootdir%\0. Spectrograms"

:: Function to create spectrograms recursively
CALL :processFolder "%rootdir%"

GOTO :EOF

:processFolder
FOR /D %%D IN ("%~1\*") DO (
    SET "subdir=%%D"
    SET "spectrodir=!subdir:%rootdir%=%outputdir%!"
    IF NOT EXIST "!spectrodir!" MKDIR "!spectrodir!"
    CALL :processFolder "%%D"
)

FOR %%F IN ("%~1\*.flac") DO (
    SET "filename=%%~nxF"
    SET "spectroname=!filename:.flac=.png!"
    SET "spectropath=!spectrodir!\!spectroname!"
    ffmpeg -i "%%F" -lavfi showspectrumpic=s=1280x720 "!spectropath!"
)
GOTO :EOF


  


  It's much faster, and this is what it outputs :

  


  I added them to the Windows Send To folder, and use both by right clicking the folder containing the subfolders with the flacs, then Send To and then the script I want to use.

  


  I have 3 questions :

  


  

  1. Is my code as optimized as it could be ? Is everything correct ? And is it "folder structure agnostic" ? When outputting the spectrograms will the scripts mimic the structure of the input folder in all cases, or are there flaws in the code that would prevent them from being able to do it in some cases (i.e. if the .flac files are 3 layers under the input folder or something like that) ?

  


  2. How do I modify the function calls at each of these scripts (whichever approach I stick with, I'm leaning towards ffmpeg cause it's much faster and appears to be better supported) to have the spectrogram image include some info about the file, customize its font, x axis time format, and other things ? I want it look like what Spek outputs as much as possible. Can't find much in the documentation of SoX and ffmpeg, but Spek uses ffmpeg as its backend right ?

  


  3. Is it recommended to show both channels on the spectrograms, like the SoX script does, or is one sufficient for my purposes ?