is there any way of normalizing (by which I understand giving all tracks the same perceived volume/loudness) several sound files in a directory using Audacity? I think this is possible file-for-file but I mean a batch of files at once.
Before posting I searched for previous posts regarding this and found ‘Normalizing across multiple tracks’ but didn’t feel that the post adressed my problem. Because a problem it is, especially late at night when there are neighbors to consider. You listen to five or six songs and then one comes on that has you rushing head-over-heels to mute it.
Well, no biggie but an annoyance to be sure. And if there is no way using Audacity for this, does anyone know of software that will do it?
What’s wrong with turning the volume down on the loudspeakers? Or listening on headphones? In Audacity, the playback volume and record volume should be independent of each other. Or are you recording via “Stereo Mix”?
Take a look at [u]ReplayGain[/u]. It analyzes your files for perceived loudness and tries to match the levels. There are variations such as MP3Gain and WAVgain, depending on your file format and your player, or player software. Apple has something similar called Sound Check if you are using iTunes or an iPod. (I use ReplayGain with WinAmp, and it works great!)
Normalizing works on the peaks, which don’t correlate well with perveived loudnness. Average or RMS is better, and the algorithm used by ReplayGain is better yet because it takes into account the ear’s different sensitivity at different frequencies and the way that frequency-sensitivity changes with loudness. But nothing is perfect, and nothing is as good as the human ear. Two different people might not agree which file is louder with different styles of music, or if one song starts-out quietly and finishes loudly, etc…
I assume “Normalizng as a group” adjusts everything in the group by the same amount. So for example, the loud songs on an album would remain relatively loud, and the quiet songs relatively quiet, as originally intended. There is an album gain option in ReplayGain in case you want to adjust the album as a whole.
Most songs are already normalized (maximized), including most quiet sounding songs. That means that you cannot boost the quiet songs without clipping (distorting). So, you will find that ReplayGain (or any other linear solution) will end-up reducing the volume for most songs in order to match the volumes. There is an option in ReplayGain to prevent clipping (for tracks that need to be boosted), and if you choose that setting it will go as far as it can go without clipping.
DVDdoug answered along identical lines just as I was answering (thanks).
If you or anyone wants to request rms or Replay Gain normalisation in Audacity, please do so.
Just to add - you can do a crude fake of rms normalization in Audacity by getting the rms level from Analyze > Contrast then normalizing to that level (to use Contrast on a stereo track you will have to use the Track Drop-Down Menu to “Split Stereo Track” then select only one of the channels).
If you or anyone wants to request rms or Replay Gain normalisation in Audacity, please do so.
Something based on ReplayGain or EBU R128 would be awesome! I don’t know of an audio editor that has it (or a plug-in), so you alkways have to use a stand-alone program. (Of course MP3gain does have an advantage as a stand-alone program when your original is an MP3, since it doesn’t decode/re-encode the MP3 and it only changes the gain header without affecting sound quality.)
