Normalize vs Gain

  • Thread starter Thread starter NYMorningstar
  • Start date Start date
N

NYMorningstar

Recording Modus Operandi
I'm new but I think I'm getting better at this, doh! Last night I was working on a song that had a weak vocal track recorded digital at maybe -12db. I imported the track to wave lab, chose "select all" in the edit menu and then normalized the track to -0.5db. Here I usually quit and send the track back to the mixer but I listened to it and there were still weak spots. I then said what the heck and started choosing(highlighting) the weak sections and brought them up to -0.5db. Hey! That was working great and I got the whole track pretty much up to snuff. I even repaired a part where the voice crackled by finding a matching section on the track and using cut and paste. Woa! One section was real weak so I selected it and this time I chose "change gain" as opposed to normalize. Well that worked and I took the section up to -2.0db right where I wanted it. I thought to myself, hmmm what's the difference? Well I don't know! What is the difference?
 
Well simply said:

If you select a part of audio and normalise it to -0.5dB, it will put the peak of that selection at -0.5dB

If you select the same part of audio and change the gain it will change the gain by the amount you said. Just look out for clipping.
 
Is it any better to use one over the other? Am I thinking correcting in assuming now that they both perform the same function, raising and lowering volume?
 
Yes, though normailzing will seldom be used to lower the volume. It's used to reach the max possible
 
Normalizing is increasing of the gain relatively to 0 dB ( loudest )level. Program search for loudest part of selected wave, measure its distance from peak level (0 dB), and if this is, say 4.3 dB, gain of whole selection is raised for 4.3 dB.
Manual level editing is totally ok, and better than just compression of vocal track.
If you raise gain of selection so that it excedees peak level, Wavelab, automatically compress this part dynamically, so no clipping ( digital distortion ) occurs.
 
Downside Studio said:
Yes, though normailzing will seldom be used to lower the volume. It's used to reach the max possible
Click to expand...
...and raise the noise floor in the process.

"Normalization" is not commonly used by pros...


Bruce
 
i'd advise you to ride the gain rather than normalising.

changing the gain in spots where the vocal is weak is the same as raising the fader on your mixer. you can raise it just a little, or you can raise it alot so that you keep but somewhat limit the dynamics of the recording

you are definitely going to need to get good at playing with the gain to get an even leveled performance. i'm learning this same lesson myself, and just posted a thread about it in the cakewalk forum.
 
"Normalization" is not commonly used by pros...
Click to expand...

Bruce, I'm not sure what you mean with this remark.

I do on little budget recordings the mastering myself. I have the waves NPP (L1,Q10,C1 etc) and the C4 to work with. Now the L1 (which is used last) trows the whole signal after limiting up to the -0.3db level(digital)

This is the same as normalizing. If pro's don't normalize how do they get rid of the headroom.
 
BTW with normalizing I don't mean hit the normalize button, but the process of getting rid of headroom.
 
The NORMALIZING feature - which is simply to mulitply all amplitude by the amount from the current peak to 0dbFS (or a selected peak) - is not common because is raises the noise floor by the same amount and does nothing to assist in terms of apparent loudness.

In addition, bringing the signal up to peak like that eats away at available headroom that is better used by more common functions such as Limiting, Compression, and EQ.

All normalizing is, is amplitude-multiplication - there are no other benefits.

It sounds like you're confusing it with Limiting, which is far more useful and affects apparent level.


Bruce
 
Bruce,

I don't think I'm confusion things. Limiting and compression are used for altering dynamics for some reason (more punch, tighter sound etc). Normalizing is indeed used for reducing headroom.

If have made a mix with peak level at -4dB, for mastering going through final compression, the peak level will drop because of the compression.

I guess nobody will put a track on cd that's for example 5dB below digital max. The signal will be brought up to max level. This can be done manually by adding gain or done automaticly. both I consider nomalizing.

The limiting function only works for getting the mix to the max if the signal is brought in at high level, eg above max. Peaks will be reduced to max.
If the peak is below max (eg -5dB) limiting will never get it up in level. it only pushes downwards.
 
Ok... but normalizing is just amplitude-multiplication -- you do all your other processing then end it all by normalizing, you are raising your noise floor along with the rest of the signal - not a good thing.

If you had used limiting as the final step, the gain increases to 0dbFS above the threshold your set, but doesn't affect the signal below the threshold. So your noise floor doesn't get touched........


Downside Studio said:
If have made a mix with peak level at -4dB, for mastering going through final compression, the peak level will drop because of the compression.
Click to expand...
And in this case I would use the compresor's make-up gain feature to restore the original level.



Bruce
 
And in this case I would use the compresor's make-up gain feature to restore the original level.
Click to expand...
which is raising the noiselevel too.

If you had used limiting as the final step, the gain increases to 0dbFS above the threshold your set, but doesn't affect the signal below the threshold. So your noise floor doesn't get touched........
Click to expand...
which results in a peaklevel below max level, which results in people having to set their stereo louder, which results in raising the noisefloor.


I understand your point, and there is no way to normalize without raising the noisefloor. This doesn't mean pro's don't do some sort of normalizing, especially in these days when loudness os a big concern for most bands. It would be not done to have the peaklevel of the cd a couple of dB's below zeromax.
 
I have never Normalized -- I've never needed to... my mixes are hot enough going to 2-track, then I use some careful limiting to draw out some of the apparent loudness... using this procedure, if I were to then use NORMALIZE, the multiplication factor would be 0....

I guess it's all a function of the tools we have at our immediate disposal!

Cheers, DS!

Bruce
 
Blue Bear Sound said:
Originally posted by Downside Studio
If have made a mix with peak level at -4dB, for mastering going through final compression, the peak level will drop because of the compression.

And in this case I would use the compresor's make-up gain feature to restore the original level.



Bruce
Click to expand...

This final compression you guys are talking about. How and why do you do it?
 
Wait a minute, you're talking about limiting the 2-track mix while mixing down. I'm talking about mixing down without anything between the mixer and the recording medium. All alternations are done afterwards. That way if I missed something I always have a non-processed master.

Now suppose a mastering studio gets a mix to be mastered, and its peak level is at -4dB. You're saying whatever they do, the peaklevel would never get above this point. Unless your adding some gain somewhere along the pad to raise it up to the max level, which is basicly the same as normalizing.

I personally would be non-happy susrpised if I got a mastered version back from a masteringstudio and the peak level is significant below digitalmax.
 
Downside... actually, I use a Masterlink, so I get the unprocessed mixdown, AND a processed one, as needed!


NYM.... I use limiting to draw out some additional punch if the song needs it -- So far, I haven't ever compressed a 2-track mix... (although, technically, limiting is specialized compression, but anyways, you get what I mean...)



Bruce
 
Hey Bruce.....This may sound kind of stupid...BUT...what the heck is "apparent loudness" and "apparent level?" Isn't something just "loud" or "not loud?" I'm just interested in understanding what it is exactly that you are talking about when you use those phrases...:) I understand that an engineers job, when mixing is to make something sound loud w/o actually raising the volume...I guess this is what your talking about, right? Maybe I'm just confused about how you see the concept of "apperent loudness" applying to limiting when limiting, as I think I understand it, does actually physically raise volume...I dunno...maybe I misunderstood something you wrote....then again..I haven't used limiting yet, so I really don't know what i'm talking about, and that's why I'm asking :)


thanks


- nave
 
Last edited:
Apparent Loudness: the subjective impression of the intensity or magnitude of a sound.

A flute and a saxophone can play at the same volume level, but the saxphone will ALWAYS sound louder than the flute, due to the weight our ears give to certain frequency ranges over others.

Bruce
 
oops....sorry I edited a bit....Ok I understand that....But what does that have to do with limiting?
 
You must log in or register to reply here.
Back
Top