TranceAddict Forums (www.tranceaddict.com/forums)
- DJ Booth
-- Sound Levels - Normalizing

Sound Levels - Normalizing

How come when I normalize a track in Soundforge there are still louder and softer parts in it when i play it. Shouldnt it take the average of all peaks and level them?

when you normalize, what happens is the peak value is adjusted to it's highest level before cliping. (usually -0.5db for peak level) If you want to raise the gain in certain parts of the recording you can insert volume envelopes and manually adjust it to your needs or just select that small section than normalize that one part. (warning: it may sound weird going from a normalized section to a section that's not normalized at the same time)

AHhh...depends on your normalization algorithm. I think what you have in mind is closer to a compressor than a basic normalizer.

so are you saying that normalising gets the peak value of the entire recording, finds out how much that can be boosted, and boosts everything by the same amount? Or are there algorithms that use rolling averages to that, for example, every 5 minute section gets boosted so the peak of that section is at max?

quote:

Originally posted by Tranc3 AHhh...depends on your normalization algorithm. I think what you have in mind is closer to a compressor than a basic normalizer.

well, that is most likely true... anyways im not a sound engineer btw.

AFter valid education, I've said it before and I'll say it again.....

NEVER EVER NORMALIZE!!!!!!!

compress instead...it's "healthier" shall we say, for the music.

TRUST ME

quote:

Originally posted by SgtFoo AFter valid education, I've said it before and I'll say it again..... NEVER EVER NORMALIZE!!!!!!! compress instead...it's "healthier" shall we say, for the music. TRUST ME

can u go into a little detail about why not use it.....i would like to learn....just a quick lesson

For some odd reason, most (as in close to ALL) normalizing algorithms are far too imperfect in their sound quality. The idea behind the process is to take the averages of your music's amplitudes and boost everything in the file by another averaged number. In music mastering terms, this is inacceptable, especially considering the result is often a tainted-quality sound file, with troubles such as tinny-sound, added aliasing or noise (like gaussian blur in graphics terms).
Normalizing often changes dynamics in amplitudes, which could throw off a mix ever so slightly.

Compression has been used and proven over the years because is works with the dynamics of your music based on ratios, so the effect of it can be either transparent and un-noticeable, to the most annoying "pumping-and-breathing" effect that compression can do. Good compression can bring up the quieter parts, and relatively soften the louder parts (reduce the dynamic range) leaving you with a fuller sounding volume and so you can boost the entire output gain a few more dB to make the track sound louder. Normalizing doesn't quite do this at all. Another thing is that most (pretty much all) professional-grade compressor gear and software compression algorithms these days are virtually noise-free, leave you no side-effects, and are much cleaner with your music.


"the more you know"

thanx bro

from what i understand, normalizing peak value doenst really do much but make everything louderby making the peak value as high as it goes. normalizing rms will change the dynamics of your music meaning...less distance between the louds and softs so that everything can be louder. compressoin lets you take a certain threashold and then compress it to bring up the rest of the music

sgtfoo that was a great post. thanks

only problem with compression is that its so hard to get right. i hear alot of music thats overly compressed and it sounds really bad. its always an ongoing problem for me when i complete mixes. i always just try to get my gains right and then normalize the peak value so i dont have to mess with so much.

quote:

Originally posted by DjJade only problem with compression is that its so hard to get right. i hear alot of music thats overly compressed and it sounds really bad. its always an ongoing problem for me when i complete mixes. i always just try to get my gains right and then normalize the peak value so i dont have to mess with so much.

Once you start working with it, you learn pretty quickly, or at least I did. If your original recording was already good, then a weak compressor setting will smooth it out.

haha exactly...the hard part for me is just keeping my levels right : P

quote:

Originally posted by DjJade haha exactly...the hard part for me is just keeping my levels right : P

yep... i'm in the same boat. i still need to work on my levels before compressing will really work for me.

ok compress you say...
the only compression i know is from dat/wav to mp3

how?

What sort of things should i listen for when compressing different parts of my track? any tips....

you should look for squared off looking areas of the waves... it should sound like clipping but cleaner.... sorta like the highs or the punch is just not dynamic at all like someone took the volume knob to the loud parts and kept punching it down when it hits. it is hard to say depending on your parameters...

*i think* attack time makes affects how fast the compression initiates once the waves hit the threashold whereas the release time controls how smoothly it trails off. so you want it all to be smooth so that its not so abrupt that its easy to hear yet not soo slow that it sounds so muddy. the threashold lets you pick which level to to start compressing and the ratio to me is like... the slope of compression which would be sorta like the speed at the compression attacks.

Powered by: vBulletin
Copyright © 2000-2021, Jelsoft Enterprises Ltd.