PEAK NORMALIZER
When digital recording first became popular many engineers liked the fact that you could record a signal at a low volume and then raise the volume to suit your needs after the fact. In the days of analog recording the idea was to put as much signal onto the magnetic tape as possible without going over into distortion. However engineers then realized that because of the nature of digital recording, when you record something quietly it uses fewer bits than a louder signal. Therefore you still want to get a good strong signal when you record in digital.
How Does it Work?
A peak normalizer takes the loudest peak in the audio and raises it to the predetermined level. It then raises the volume of all the other audio relative to this loudest peak. The result is identical to the original only louder. Here's something to remember, the overal volume of each track is set by using the faders. If this is the case why normalize at all you ask? Well let's say you have a plugin that you want to use that has no input volume setting. You might want to use the normalizer to change the volume of the audio to match the input sensitivity of the plugin. This may also be useful when sending a signal out of your computer to use with an outboard effects processor, or if you want to take a raw guitar signal and re-amp it. If the signal is too hot you might overdrive the outboard gear if it's too quiet you might pick up some noise.
Yeah but what does normalization actually do?
Well really all it does is raise the volume! A typical normalizer does nothing to the dynamic range. In a digital system 0db is the loudest you can go and so normalizing to 0db is probably not a good idea. This is beacuse it leaves no headroom and the second you compress or add anykind of dynamic effect you run the risk of going over 0db and into digital distortion. Many normailization plugins allow you to normalize to a certain volume, so it's a good idea to normailize to a level BELOW zero like -6db or -10db to give you some room for processing.
Things to keep in mind
Normalizing the audio is not about making things louder in the final mix where you will set the individual volumes of each track with the faders. For me it's more about getting the audio to a level where you are able to hear anything that might be wrong with the audio during tracking when it's easiest to fix! There's nothing worse than doing a mix and now that you have something as loud as you want it there's noise and problems you can't get rid of. Or even worse,you no longer have access to that musician or you have to pay the musician to come back to re-record something. When I was working more often in professional studios I would go back a few hours after a tracking session and listen to the tracks one at a time with headphones to make sure they were clean enough to use.
Louder does not always mean better, sometimes when you have recorded a track and it's already at a good level you can just leave it alone! You have to learn to use your own discretion here. If you have something that's overly quiet it's usually a good idea. Having said that I have worked with many engineers that normalize absolutely everything, and some that never normalize anything! I myself try to stay away from the extremes, so I use normalization only on things that I feel are too quiet.
If you are looking to make something more even or increace the apparent loudness of something, a compressor is probably a better tool for that job. However there is another type of normalizer known as an RMS Normalizer that raises the average volume of the track. This is basically another form of compression. RMS normalizers are more commonly used in mastering when you are trying to make all the individal songs playback at the same volume.
▾
▾