Question about "normalise"

[er1c]

Hello all,

I have a question. So when i work on a project, i use to normalise every audio track with AS Normalise before work them with plugins and volume.

But it is good to normalise all tracks or should i prefer let them to their volume and do mixing like that ?

Thanks by advance

[allstaar]

You shouldn't normalize while editing and mixing.

You'll be throwing away headroom you may well need when it comes to eqing and other processing.

[ejwells]

Never normalize, period.

MAYBE normalize a stereo wav mix (I never do that either), but NEVER EVER normalize individual tracks.

[er1c]

Ok Thanks for advices

[murph16]

Never normalize, period.

Huh?



Um..if that was true, there wouldn't be a normalize option in Audio Suite. And you won't be losing headroom if you normalize correctly in a way that makes sense.

Use normalize if you didn't first attain your optimal record level when you tracked. Or, use it to "even out" different elements that exist on one track to make mixing easier. I would say use it AFTER editing and BEFORE mixing, although sometimes it makes editing easier, too.

If you normalize a track to say 0.0, then yes, you will be limiting your headroom and your mixing options will decrease, along with your dynamic range. But if you have for example, a track that has different regions on it, and you want the levels to be more uniform, normalizing is a perfectly fine process to organize your levels.

For instance, I often produce radio spots and I use just two stereo tracks for my music drops. One one track, I might have three or four different regions of music. Rather than automate or draw volume changes to compenasate for varying levels of those regions, I will instead normalize the lower music regions to match the others. I usually don't normalize a track to any peak level higher than -3 or -2. In doing that, I retain headroom and any volume changes or automation moves I make after that are minimal and easier.

There ARE uses for it, but it's not something you want to do for every track, just to do it.
If you are recording at decent levels to begin with, you shouldn't need normalizing.

[ejwells]

[/QUOTE] If you are recording at decent levels to begin with, you shouldn't need normalizing.

[/QUOTE]

I guess that's my point.

I've recorded 400 songs for 30 or 40 clients (not to mention my own junk), and I've never used normalize.

[albee1952]

I agree with not using it very often. I sometimes use it to boost a track that was recorded too low or to even out a weak part of a track(say when the singer sang a phrase or word softer than the rest of the track). Lately, I have been happier using the AS gain plugin as the idea of boosting 5db just makes more sense to me than increasing to 85% or whatever.

[Spiritwalkerpro]

Humm.. Why do you suggest we do not normalise? Is there a problem with the math in the process of changing the levels? Or is it something else? I agree that it is best to get your levels where you prefer to work, first time. I always try to work within the -14 to -18 range as to not overload my pres and to try and limit the levels going to my master. I never normalise to 0, if I have to use the normalise function I always do it at 75.5% to try and keep my levels somewhat consistant. That being said if I am trying to fix a vocal comp with parts that are recorded at different levels, I experiment with the normalisation settings to get the levels to match. Is this better than riding the levels when mixing? I don't know, it seems to work. If the math is flawed in the normilisation process, maybe it is better to ride levels? The math parts of these questions are way out of league, anybody with some simple answers to these questions?

Norm

[el Daniel]

You shouldn't normalize or apply any AS plugin with level changes involved (gain, eq, compression) when the audio is going to be processed later because each time you apply an AS process you are adding quantization errors or a generation of dither if it is turned on. You don't get this when mixing because of the mix bus resolution, and you just have to apply dither once in the master fader when bouncing.

If a track was recorded at low level and need to be arise more than what the fader allows, you can insert the trim plugin or use the output gain in the compressor.

Also, you don´t get to match levels when normalizing, only get the same peak level value that is not agree with how human hearing hears.

[allstaar]

For me, normalization means bringing an audio file's peak value up to 0dBfs. (this assumption stems from working with early samplers and software recording)

I call everything else "Gain change"

This is why I suggest you don't "normalize".





Just to clarify.