When I select a section of my recordings (they are lectures, not music), then choose Effects Amplify, the box gives me a number that is the maximum I can choose to avoid clipping. Is this a measure of the volume of the source recording? I’m trying to figure out the best way to have all my lectures end up at the same volume level. I’m not rying to put them onto one disc like a music CD, they are all separate talks, but I would like them to sound about the same volume to a listener. Should I be doing normalization instead? Why can’t I choose the value of normalization that I want? It only has the option of setting the normalization to -3db.
I may be wrong, but as I see it, normalization simply means amplifying the sound as much as possible without loss or distortion. The -3 dB option only means that you don’t amplify that much, but leave a 3 dB bite from the top, just to avoid further distortion if you do other processes to the sound.
The Amplify effects does the same, but you have the option to choose the amplifying rate. It gives you the maximum rate as default, which equals normalization without the -3 dB thing. but you can slide the controller in both directions from that, either lowering the rate to avoid further distortion, or raising the amplification, which will cause dirtotion on the peaks. Your phrase “Choosing the value of normalization” would actually be using the amplifying effect. See normalization just as a preset amplification.
You might also want to check out the compressor effect. It kind of amplifies the weak sounds without amplifying and distorting the strong sounds.
You’ll want to use the Compressor, not Amplify or Normalize.
In order to use it correctly, do this:
On your unprocessed track, highlight a portion that is just background noise, no talking (this is the Noise Floor). Then click Effects → Amplify. Write down the default number here and click Cancel (don’t make any changes just yet).
Now, highlight the whole signal and click Effects → Compressor. Take that number you wrote down earlier and add 5 to it. This is the value you want to use for the Threshold (inverted, of course).
Now set the ratio to 10:1 and the attack time to the lowest it will go to (.1 sec). Leave the box marked “Normalize…” checked. Click OK.
If you do that to every track, they should end up roughly the same volume. And if you use Audacity 1.3.3 you can try out the Batch functionality to do them all at once. Just don’t delete the originals until you’ve listened to them and made sure they sound ok, if the noise floor is different from track to track you might need to adjust the Threshold value for each one individually.