As explained in “Audacity Effects: Use of Decibels (Log Scale),” sound is representation better in decibel scale, especially if you want to show volume (amplitude) changes.1 Although Audacity’s default view is in linear scale, most of the effects use decibels as parameters, so it is a good thing to know how to quickly find the magnitude of a certain track or certain region in decibels (dbs).
Since I will refer to this method of finding maximum amplitude from time to time, I decided to give it a separate treatment so that it could be easily referred to.
When you use the “amplify” effect, Audacity does some calculation and show the amount of amplification needed to bring the peak amplitude to 0 db and show this amount on the dialog. We can use this fact as a trick to find peak amplitudes. For example, if the “amplify” effect shows that we need to amplify a signal by 10 db in order to bring its peak amplitude to 0 db, it means the peak amplitude is –10 db (the negation of the amount shown).
Let us look at a couple of examples.
Say we have track as shown below, and we want to find out what the peak amplitude is.
We can get that by easily select the whole track and then bring up the dialog for the “amplify” effect. Here it shows that we need to amplify the track by 3.7 db in order to bring the peak amplitude to 0 db. This in turn tells us that the peak amplitude of the signal is at –3.7 db (note the minus).
And we can use this trick not just for the whole track but for a chosen region as well. I often like to see the amplitude of the noise I am getting before I make my adjustment. Below is a simple signal with noise.
To find the peak amplitude of the noise region, I simply selected it and then bring up the “amplify” effect.
It shows that I need to amplify by 34.8 db to bring the peak amplitude to 0 db. As explained before, this means the peak amplitude of the noise is -34.8 db.2 Of course, in actual adjustment, I do not have to adjust by this exact amount as I don’t always adjust the noise level to 0.
- On the other hand, if you want to examine the sine wave of signals, it is better to use linear scale. ↩
- I find that sometimes the value shown does not correspond exactly with what is shown in the waveform. I am not sure whether it is the graph or the value found that is off, but the value is still a good approximation. ↩