normalize vox?

  • Thread starter Thread starter paresh
  • Start date Start date
P

paresh

Member
I never seem to get enough level in recording vocals. I don't want to get any closer to the mic so I've been normalizing lately. The noise floor doesn't bother me after it's normalized. I know compression helps somewhat but a little bit goes a long way. Is there a better option? Thank you
 
What do you mean by "level"? Personally I try to record everything around -18dB, with peaks no higher than around -14 to -12dB.

To make your vocals more prominent in your track, you can:
  • heavily compress / limit your signal, which will allow you to turn it up without peaking at transients,
  • set up your instruments and vocals on two separate busses, and do sidechain compression on the instrument buss triggered from the vocal buss,
  • figure out which frequencies you want the vocals to punch through on, and either duck those frequencies on your instruments or set up sidechain EQ to automatically duck that frequency when the vocals come in (such as TDR Nova),
  • make spatial decisions such as hard-panning rhythm guitars, etc. so that the vocals sit in the center of the mix and backing instruments are off to the side,
  • play with reverb to make some instruments sound farther "back" / gently washed out and the vocals sound right up front,
  • double / triple your parts to make them sound fuller, and add delay and reverb,
  • saturate your vocals to make them sound more aggressive
I also do a lot of volume automation - manually adjusting the volume of the vocal track syllable-by-syllable to make sure the volume is consistent.
 
Normalizing is nothing more than finding the highest peak, looking at the difference between that and 0dBFS, and adding that much gain. If your highest peak is -12.3, then normalizing will add 12.3dB of gain to EVERYTHING, peaks and noise floor. If your noise floor is low enough that it won't change the relative level. It's the same as moving the fader up 12dB.

You have two options. #1 is to get closer or sing louder to get more signal without raising the noise floor. #2 is to raise the input gain on your mic preamp which is fine as long as you don't hit peaks higher than 0. In digital audio, there is no +1dB. It's a ceiling.
 
Great ideas, thanks! I do some ducking already. By level, I meant a small waveform + too quiet in relation to other tracks.
 
You can always edit the section that is too low and raise just that, making sure you don't wonk it. If you have a good wave editor, just do some surgical improvements to the wave itself.
 
Forgive me, but if it’s too quiet, then turn it up, or turn everything else down? Have I missed something? In cubase, visually i like to see the waveform, so i often stick in a compressor on 1:1 and then use the makeup gain feature. Then the channel fader can be pulled back, giving me the possibility of oushing it if I need to?
 
Since this problem is unlikely with a capacitor microphone I am going to assume OP is using a dynamic? (everyone and their aunt seems to favour the SM7b these days!). Therefore try an 'inline preamp'. I found the FetHead to be fine for its money but there are lower noise examples.
I am also assuming the existing AI pre is at its gain/noise floor limit? Of course, putting more gain in the mic circuit will worsen ambient noise pickup and increase "room effect" but what can yer do?!

Another point to consider is that this is very often the biggest problem with live bands. Guitars etc can be as loud as you like but vocals are always limited by feedback. So many bands struggle to understand the simple fact that everyone else must TURN THE ****ER DOWN!

Dave.
 
Thanks again for all the advice. I do use a dynamic mic + fethead. I sold my ribbons just for this reason. I think part of it is that I don't have a strong voice & need to compensate afterward.
 
paresh said:
Thanks again for all the advice. I do use a dynamic mic + fethead. I sold my ribbons just for this reason. I think part of it is that I don't have a strong voice & need to compensate afterward.
Click to expand...
But it is all relative! I don't know what backing instruments you have but if say electric guitar is slamming -2dBFS? Don't let IT!
IMHO you need to bring all your other tracks to at or even below your voice. Remember! People were making very decent recordings of singers and bands with a 2 track tape machine and very basic mixer.

One problem peeps get with everything at -18 to -20dBFS is that monitoring is not loud enough. Many AIs have too little gain in their headphone amps...enough POWER at 0dBFS but no way to crank the low stuff. Solution? an external headphone amplifier.

Dave.
 
Turn the other tracks down and your headphones up. All your levels should be well away from 0 dBFS throughout the recording and mixing process. The time to get the final level is when you're mastering.
 
paresh said:
I don't want to get any closer to the mic
Click to expand...
Why Not?

paresh said:
so I've been normalizing lately. The noise floor doesn't bother me after it's normalized. I know compression helps somewhat but a little bit goes a long way. Is there a better option? Thank you
Click to expand...
Track louder and with more compression on a parallel track.
 
How close is "close"? You don't need to be 2 inches from a cardioid mic, remember that sound gets lower with distance. A few inches can do a lot, and to be honest, most people like the proximity effect as it gives a nice full sound. If you're 24 inches away, then you have plenty of room to play with.

You say you don't have a strong voice, but you can learn to belt it. Too many people think singing should be about the same volume as when you're talking on your telephone.
 
TalismanRich said:
How close is "close"? You don't need to be 2 inches from a cardioid mic, remember that sound gets lower with distance. A few inches can do a lot, and to be honest, most people like the proximity effect as it gives a nice full sound. If you're 24 inches away, then you have plenty of room to play with.

You say you don't have a strong voice, but you can learn to belt it. Too many people think singing should be about the same volume as when you're talking on your telephone.
Click to expand...
In any case Paresh you have been fluking about with normalizing, compression? A bit of LF cut is a p of p to implement and once you have found the frequency of the voice/mic combination you've got it.

As for "belting it"? *****k yes! People made a living as singers because THEY COULD BE HEARD. Big bands in the 30s and 40's had primitive PA but almost no pubs or small halls did. Just yesterday watched a doccy about Bob Dylan. Not a nice voice but he was heard!

#It's a LOUD voice but tho' it's not exactly flat she'll need a little more than that to earn a living wage. PLEASE Mrs Parkinson, on my KNEES! Mrs Parkinson, don't put.....#

I now remember it is "Mrs Worthington" Probably Freudian because I am besotted with Katherine Parkinson!
Dave.
 
Last edited:
paresh said:
I don't have a strong voice & need to compensate afterward.
Click to expand...
Plenty of popular artists don't have what we'd call a "strong" voice - you could listen to Elliott Smith, for example, to get some ideas of how to work with thin, soft, or whispery vocals in your mix
 
I think something else I've had some success with on "softer" vocal performances is parallel compression - in a DAW, duplicate your lead vocal track to a fresh track, and then absolutely smash it with a compressor. Then, set the track level all the way down, and then slowly roll it up until you don't really hear your lead vocal track getting "louder" so much as "fuller" - all the transients will still come through clearly from your lead, but your parallel compressed track will add a lot of body behind it.
 
You must log in or register to reply here.
Back
Top