Normalizing

[rsp2rsp2]

I am about to try “normalizing” for the first time. Anyone have any comments or warnings before hand?

I did notice that despite all my tracks being initially recorded at the same level, that after I tried a test run with normalizing, a few tracks got jacked up WAY high (and I thought about using the “blue-bar” to manually decrease volume back down on these tracks).

I have read on the internet that normalizing should not always be done so easily (without thought & care first).

Comments?

[tacman7]

I guess I use compression instead of normalizing to bring up the gain on a track.

[P.P.T.]

I use the track fader to bring up the gain.
I suppose it is possible that the recording is too low, in which case normalization could be useful.

In general there is no need to normalize. All ir does is raise the volume and bring the peaks to a determined level.

[Bigduggieface]

Normalizing has exactly the same effect on your audio track as raising the fader.

There is no inherent benefit, other than signal gain.

A better mode of thought would be to turn the loud elements of your mix down rather than boost the quiet ones.

Therefore normalization is most often a waste of time and can even become counterproductive.

[sparqee]

Almost all faders (both hardware & software) are scaled. They are not linear in the effect they have on volume. The highest resolution of most faders is found in the area around 0db. It is therefore beneficial to have the average amplitude of your track signal in that general range (i.e. so that fader moves are as accurate as possible).



I also find it helpful to have a visual wave form displaying with peaks high enough that I can judge sections of a sound by it's shape (e.g. being able to visually distinguish a region of sibilance).

[Audio~Geek]

Compression is different from Normalizing. Compression reduces gain by a certain amount (make up gain turns it up again, its the same as turning up the fader) Normalizing analyzes the peak or rms level of a sound clip, then allows you to bring the peaks up to an absolute level. Both processes will increase noise level. Compression changes dynamics, normalizing doesn't.

I usually normalize my dialog for the podcast, in addition to compression. Usually only to -4db peak
I do this so the compressor has a more consistent level to work with. The track fader is post inserts (not that it makes much difference) and doing a lot of automation for a podcast just isn't practical.