Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
×
Music Media

Aligning Audio Levels for Bulk CD-to-MP3 Conversion? 9

cmaxx asks: "Has anyone got a tool or method for aligning the recording (playback) levels of their CD's when converting to mp3? It's winding me up that some CD's were recorded quietly (for dynamic range reasons perhaps) while other were recorded normally or loudly, or with too little bass/top-end/whatever. It means that playing my mp3 collection randomly means I'm forever having to change the volume or equaliser settings to get the most out of the music. I know it's not a straightforward problem, but I'm curious to know, in this age of napster, if someone's cracked it - and if I can too." Interesting thought. At the very least some form of database is necessary to store the audio levels for each song.
This discussion has been archived. No new comments can be posted.

Aligning Audio Levels for Bulk CD-to-MP3 Conversion?

Comments Filter:
  • Something like:
    sox -v `sox foo.wav -e stat -v` foo.wav -t wav - copy | <favorite mp3encoder>
    (Those are back-quotes.)
  • I know that AudioCatalyst has a built-in option to do this - it is defined as "if the peak volume of this track is less than x% or more than y%, normalize it to z%."

    It seems to do a pretty good job. Unfortunately it only does it on a track-by-track basis... hardly ideal, say, for classical music. If someone knows of a good way to do it over the course of an entire album, I'd like to know too!

    --
    Make Money on the 'Net [geocities.com]

  • Another WinAMP plug-in goes by the name of Audiostocker...I use it exclusively. (Just 'cause I'm testing win2k...erm...yeah. :)..) Check out Audiostocker here. [tsx.org]
  • Still, doesn't that look familiar?

    I guess MS had to remove all the good parts from sh to make it seem like they weren't copying ... and of course they changed $ to %%, and changed '*' not to match periods, and removed the semicolon-at-end-of-line, etc.

    Okay, maybe they did rewrite it...
  • I wrote up a batch script in SoundForge to run my WAVs through an EQ and a Normalize them... You can have it apply any number of effects with certain settings to a batch of files, which is pretty nice... I just queue up a bunch of them and leave, and I come back with a bunch of nice sounding WAVs... I even do this on a lot of the music I make. (http://www.mp3.com/atomly/)
  • Sounds like you need your waves normalised, like this guy here says [slashdot.org] - normalising is where you boost the volume of a wave so the loudest point of the wave is at the maximum amplitude for a wave - normalising all your waves will leave you with the loudest points of all your waves as loud as each other. To do this you'll have to go from CD->WAV, normalise, and then convert WAV->MP3.

    If you don't have any music software that can do batch processing, you'll want a command line util that can normalise waves. You can then batch normalise all of the waves using the batch processing abilities of your OS - assuming Dos/Win95 and a program called norm.exe you'll want

    for %a in (*.wav) do norm.exe %a
    which will run norm.exe with each wav in turn as a parameter. You'll then need to batch convert your wavs into mp3s.


    tangent - art and creation are a higher purpose

  • normalising all your waves will leave you with the loudest points of all your waves as loud as each other.

    There's a further problem. As cmaxx says, some CDs have a greater dynamic range than others. Normalising a classical CD at the same apparent volume as a pop CD is likely to leave the classical piece effectively much quieter, since there may be just one short part of it that reaches the highest volume level.

    Depending on the source, music may also be more or less compressed (in the audio sense; with quieter parts louder and vice versa); in general, more compression will make the music sound louder for the same apparent volume.

    Making louder music quieter to match music recorded with a larger dynamic range loses quality. Heavy compression of all music is unacceptable to all but the people who run popular radio stations. So the only solution I can think of is, as suggested, putting volume level metadata either in a database, in the filename, or somewhere in an extension field MP3 doesn't already use, if there is such a thing.


    --
    This comment was brought to you by And Clover.
  • by The Iconoclast ( 24795 ) on Thursday March 30, 2000 @05:13PM (#1160480)
    sox should be able to do renormalization in unixland. No need for fancy Winders Software.

    cdripper | sox -normalization_flags | mp3encoder Voila!

    A wealthy eccentric who marches to the beat of a different drum. But you may call me "Noodle Noggin."
  • by kcarnold ( 99900 ) on Thursday March 30, 2000 @03:04PM (#1160481)
    There's a nice (and I do mean nice) plugin for WinAmp called RockSteady (you can grab it and a whole bunch of other plugins at WinAmp.com [winamp.com]).

    "Saves you from the varying volumes of your MP3 collection" (or something like that)

    I can't help you for Unix. (but did you notice how familiar that DOS shell script was in another comment? Bourne shell code below all that? Without the source, you'll never know...)

The moon is made of green cheese. -- John Heywood

Working...