"safety" compressions for vocal recording?

  • Thread starter Thread starter Mo-Kay
  • Start date Start date
M

Mo-Kay

Dragon Soul Productions
Hey everybody...

I've read that it's best to record rap-vocals with just a little "safety" compression to avoid clipping if sudden peaks should occur...

2 questions:

-can I do this with a plug-in in Cubase...and how.
-what settings should I use for this purpose?

thanks

Mo-Kay
 
You definitely will have to have a hardware compressor for this application. If a peak makes it to your audio card's A-to-D you're screwed at that point and no software app. can help you then. Just remember, once digital clipping makes it to your hard disk, there isn't much you can do about fixing it. :)
As far as which levels to use, I haven't had a lot of experience there but I would guess and say that the ratio would need to be something like 2:1.

-JV
 
hmmm that's not gonna work I guess... since my pre-amp is also my soundcard (1820m)...

Well, I'll just keep doing it the way I did then, and watch them levels ;)

thnx for the help.
 
One thing you could also look at doing is getting something like a Presonus TubePre or some other type of preamp that has a tube front end. The only reason I mention Presonus is because that's what I use. The tube will act like an extremely (and I do mean extremely) soft compressor that will make your peaks more manageable. It also works really well for a DI box on bass guitar.

-JV
 
silentman said:
You definitely will have to have a hardware compressor for this application. If a peak makes it to your audio card's A-to-D you're screwed at that point and no software app. can help you then. Just remember, once digital clipping makes it to your hard disk, there isn't much you can do about fixing it. :)
As far as which levels to use, I haven't had a lot of experience there but I would guess and say that the ratio would need to be something like 2:1.

-JV
Click to expand...

I don't think a ratio of 2:1 is going to be much help in controlling an unruly or un-trained vocalist though. What he probably needs is a limiter or extreme ratios like 10:1 or higher if the compressor will do it.
 
You're definately right about the ratio's if the vocalist is very unstable. I've just gotten into the habit of using light compression and watching the levels closely. It works better for me, but I am usually looking for a large dynamic range with my vocals.

-JV
 
I've been looking into that tubepre preamp, and it seems like pretty cool thing.

Do you think it would be better to record through that, and then into my 1820m instead of directly into the 1820m?
Cause the 1820m pres seem to be rather good.

Warming things up is always good, though :)


letmeknow!
 
The pre's in your sound card probably are very good and pretty transparent. They are going to give you a better represention of the sound than a tube pre, but the tube pre gives you the warmth that you talked about. It does color the sound slightly, but in a good way. Along with this comes the slight compression that I talked about earlier.

-JV
 
ok then...maybe I'll look into buying one...

sadly I won't be recording vocals for a few months...since I'm going to live with my uncle in sept. and oct. :(... oh well meanwhile I'll just make some new stuff with my computer and keyboard :)
 
If you're recording at 24-bit, you really don't need to record with very high levels. Just have your singer go through the song (especially focus on the "loudest" parts) and set your gain so it peaks around -6. This will leave you plenty of "safety" room if the singer gets really into it during the take. Since you're recording 24-bit, you'll still have plenty of resolution to work with.
 
I'd even go further down... Imagine that you have additional 48dB of resolution compared to 16bit. If you even peak at -48, you (theoretically) still have the quality of old 16-bit recordings. There were some you could listen to these days... :D

Your noise will nevertheless get worse, the lower you go down, this is the real lower limit. I tend to think that -10 to -12 peaking would be ok, especially if you consider the fact that SW-meters often are a little slow and don't catch the fast transients like plosives from untrained singers...

aXel

P.S.: take a look at the first takes' wave images. There you should see whether your 'safety margin' is enough and where you really peak...
 
ok people...thnx for the computertech input... I WILL be recording at 24bit...48khz I guess...

I still think I'm gonna go the tubepre route though...

You see, you can always test levels... but with hip-hop...it's not always an option to "look at the first take"...since that first take is THE take with some people.


Thanks for the comments everybody :)
 
Mo-Kay said:
but with hip-hop...it's not always an option to "look at the first take"...since that first take is THE take with some people.


Thanks for the comments everybody :)
Click to expand...


You should be using a limiter then.
 
or do the "fake" mic trick - put a mic for the vocalist - connect a wire to it so it looks real, the put a second (live one) one back a bit (which will be connected...) and let the performer know to keep back from the fake mic a bit... then open the gain on the live one to capture the vocals and let the performer go wild on the fake one... should help suppress pops, clicks, and any other closeness/presence issues... only thing is you need a decent vocal area to pull this off - quiet, reasonably dead... could even have the front mic be live and just record both tracks so you have a safety track in case...
 
You must log in or register to reply here.
Back
Top