Normalize vs Gain

  • Thread starter Thread starter NYMorningstar
  • Start date Start date
I am kinda following the ideas of Downside studio. When you limit to control your peaks, you still have to add gain to get the recording up to approx. 0dB, or at least the level as loud as possible (in the consideration you want to song to sound as loud as possible). At that moment, you do the exact same as normalizing, aren't you?

Am I an idiot when I say I sometimes normalize a track after I used compression to take away the peaks who are to high?
 
Thank you Downside, Blue Bear and crosstudio. I believe I understand the dynamics of this particular process now.
 
nave said:
oops....sorry I edited a bit....Ok I understand that....But what does that have to do with limiting?
It has to do with taming the peaks that are causing an overall lower level of the entire track (because if you boosted the level up higher, the peaks would distort!)

So if you bring the peaks down without affecting the overall "body" of the song, you can raise the levels up without distortion (at the expense of some dynamics... so it's a balancing act)

Bruce
 
Am I an idiot when I say I sometimes normalize a track after I used compression to take away the peaks who are to high?

Now students please pay attention to Blue Bear and me.

Normalizing has nothing to do with removing peaks. Thats the compression part. Normalizing has to do with peaklevel compared to max possible level on cd. (0dB)

BTW one good method to win headroom without adding compression or limiting is the FIND PEAK function of the editor. Most of the time a couple of peaks that just are a few milisecs, are a couple of dB above the rest. manually selecting them at samplepointlevel and lowering them using a change gain function could result in winning a couple of dB's.
I know the story doesn't make much sense for those of you who have never done this, but i have difficulty discribing the process in plain and simple english. Maybe somebody else could explain this feature so people actually understand what i'm trying to say?
 
Downside Studio said:



Normalizing has nothing to do with removing peaks. Thats the compression part. Normalizing has to do with peaklevel compared to max possible level on cd. (0dB)


Downside, off course I know that and hopely I didn't cause confusing with my post. I know that normalizing has nothing to do with removing peaks.

What I meant with my question: When I get some dynamic differences or peakes that are too high, I use compression (everyone knows that I gues:)). but, when my compressed signal is too quiet, I sometimes normalize it (instead of just boosting my gain on the compressor) to get it at the loudest volume possible. So why is there someting wrong with normalizing then? It doesn't make difference if I normalize it, or boost up the gain with my compressor gain button, with the exception I let the comp find the best amount of dB he has to add??
 
Well that's exactly what I'm trying to say.

Now if Blue beer would only listen.;)
 
:D :D

My point about normalizing is that you shouldn't get into the "need to normalize all my tracks" mode...

For specialized use, normalization is simply a tool... but every tool has it's appropriate use!


Bruce
 
Hahahahaha

I knew that would get you pissed

:D :D :D :D :D :D :D :D :D :D :D :D :D :D

Blue Bear, I think we understand each other just fine !
 
BrettB said:


So why is there someting wrong with normalizing then? It doesn't make difference if I normalize it, or boost up the gain with my compressor gain button, with the exception I let the comp find the best amount of dB he has to add??

There is nothing wrong with normalizing. Blue Bear originally said pros don't commonly use it and that was misconstrued. I believe what he meant (correct me if I'm wrong Blue) is you don't want to rely on that as a method to reaching peak volume because you have to trade off raising your noise floor. Using compression, limiting,EQ and recording your signal hot is preferred.

In contrast to that, Downside is talking about situations where there isn't an option. He is saying if he has to work on a mix that was recorded at a lower volume say -4db, then it is best to reduce the headroom(normalize) than not to.

I hope I have it right guys:)
 
NYMorningstar said:
There is nothing wrong with normalizing. Blue Bear originally said pros don't commonly use it and that was misconstrued. I believe what he meant (correct me if I'm wrong Blue) is you don't want to rely on that as a method to reaching peak volume because you have to trade off raising your noise floor. Using compression, limiting,EQ and recording your signal hot is preferred.
Exactly right...!

Bruce
 
In contrast to that, Downside is talking about situations where there isn't an option. He is saying if he has to work on a mix that was recorded at a lower volume say -4db, then it is best to reduce the headroom(normalize) than not to.

Exactly right...!

Downside
 
Excellent thread gentlemen!

Very enlightening stuff. Thank you all.:)
 
Just a note- 24bit native recordings will Normalize with less noticeable noise increase than 16bit. Another advantage to 24bit's lower noise floor.
 
I disagree -- the noise floor that gets increased is the noise floor of the recording, NOT the converter's low threshold. The recording's noise floor is going to be a lot higher than the low threshold of the converters, whether 16-bit OR 24-bit....

Bruce
 
BlueBear - You just looking for fights today? ;)

In my experience 24bit recordings will have lower noise ON THE RECORDING than 16bit. 24bit sound cards generally have better noise specs than the older 16bit cards and especially the venerable soundblaster (this is home wrecking after all).
Therefore you have more room to play with.

Obiviously this would not apply to all cases.
 
Blue Bear Sound said:
the noise floor that gets increased is the noise floor of the recording

Normalizing, will always sample the entire spectrum, including noise, and will boost it with everything else during the process. For example, take a sample of some clean acoustic guitar. Make sure your average is about -12 db to -9 db for the sample. Then take that sample and normalize it up to -.1 db. Look at the before and after on a spectral analyzer, and you will see where your noise floor indeed did move up. Why is it important? Because you effectively alter your signal to noise ratio, at the sametime as removing headroom. Mastering Engineers typically do not want normalized and compressed data/music, it limits them too much. I know alot of newer engineers compress the master bus, but try to keep it transparent. If your compressing your master bus more than 2:1, I would recommend fixing the mix. If your 16 bit converters are noisy that will get normalized, 24 bit noisy converters have same problem. Noisy equipment is the thorn in my side!

Peace,
Dennis
 
TexRoadkill said:
BlueBear - You just looking for fights today? ;)

In my experience 24bit recordings will have lower noise ON THE RECORDING than 16bit. 24bit sound cards generally have better noise specs than the older 16bit cards and especially the venerable soundblaster (this is home wrecking after all).
Therefore you have more room to play with.

Obiviously this would not apply to all cases.
Heh-heh.... no fight - I just disagree.... simply because the noise floor of most gear is above the noise floor of even a modest typical 16-bit or 24-bit converter, so the converter doesn't really factor into the equation....

:)

Bruce
 
I agree with Bruce, although I should make it clear that we are talking about tracking a final mix, not multitracking.
 
ummm.
I have a audio engineer friend who told me that normalizing actually loudens the quiter parts in the sound as a result. Like if you set normalization to 90%, then it finds the level that is 90% of the highest peak and bring the quieter parts up to that level and the louder parts down to that level.

is this correct- or might I be totally confused. It had made sense to me because the word 'normalize' would seem to imply such an idea.

??
 
Back
Top