Keeping the recorded volume even

Forge Farm

Member
Oct 6, 2004
33
0
6
48
Aveiro, Portugal
Hi all.

I'm recording a death metal band demo and I need your help.

The problems I got are mainly with drums and vocals. I've set the gains of the several mics (headroom about 10dB, no compressor / limiter used) and they seem to be fine during sound check and during the beginning of the songs. But, as the drummer starts to get in the groove, he starts to hit the crap out of the drums, and the mics start clipping.

Also, when we are punching in the mistakes, as it's only a little piece of music, the drummer doesn't hit as hard as before. So, when I listen to the track with the punch in, I have noticeable fluctuations in volume.

Finally, if I record the same track the next day, the drummer may not be with the same mood and doesn't hit so hard, or perhaps starts hitting even harder.

The same happens with the vocals. These guys simply aren't consistent.

I think the solution is on the use of compression / limiting during recording, but I'm afraid of overcompressing and ruining the mix.

Could please someone advise the best way to solve these problems?

Thank you very much.

Nuno Seco
Forge Farm Studios
 
These guys simply aren't consistent.

That's your problem. If they are not constant there is no way to make them constant after recording. Compression is not a real solution. It equals only the level, but much more important: the sound changes with the level.
Natural voices produce much more overtones when they are louder.
As a drastic example: you cannot make a quiet voice sound like a loud (screaming) voice by raising its level.

Trigger the drums. And I assume you have to quantize it, too.

Let the singer sing the whole verse or chorus. Or make as many takes until it is good and edit the parts.
Make them to feel good in the studio: no stress, enough beer, a good loud headphone mix etc.

but I'm afraid of overcompressing and ruining the mix.

Don't be afraid: a real Metal Album _is_ overcompressed these days ...
 
when you do the punch in/outs and you have differences in the tone, show it to
them so that they can actually hear the difference.

what i do with singers who aren't consistence in their performance, i go with him
over the lyrics and trying to get the desired emotions of of them.
sometimes it helps.
"you wanna be fuckin evil? then show me some ..."


Try hitting the drummer with a 2x4 when he hits too softly.
and yes, throwing (heavy) stuff at the drummer works very nicely too :D
 
Telling them that it's not working if they don't perform better is also the task of a producer/engineer.
You would be surprised to find out how fast they learn sometimes...
Hey, They ARE metal, they can take it! :headbang:
They'll consider you more pro too, if you tell em to perform better. :worship:

Cheers and good luck,

Tony
 
That's the thing - they need to be consistent. Triggering up the drummer though would help you alot in terms of the kicks. I would also suggest doing what Andy does and using an application like Drumagog or Aptrigga and mixing in some samples with the hits, to give the feel of them being more consistent. If you can't rely on the humans.. always rely on the computer.
 
Thank you all for your help. I will look into triggering the drums and using the 2x4 :D .

Anyway, do you think I should use some compression/limiter during recording, at least to prevent clipping?

Nuno Seco
Forge Farm Studios
 
You still gain accuracy at the least significant bit by recording as hot as you can. There is certainly no harm in using a limiter on the front end just to prevent clipping. Just set it so it only activates at -1dB and you won't lose anything except the transients that would clip. Especially if the drummer is that inconsistant.
 
24bit represents a dynamic of 144dB. The best AD-converters have a dynamic range of around 110dB (18-19bit). More than 120dB is physically impossible as long as you don't record in liquid helium. 24bit are oversized, 20bit would be enough, even to handle the round error during processing. It's just to have 3 full bytes.
Best microfones have around 70dB.

You have enough reserve in dynamic. There is no need to reache 0dBfs. Its no harm, of course to use limiter. But it is not necessary, nor will it improve you sound. That's why we have 24bit now.

Btw. the LSB is used even by lower levels, because audio data are signed integer numbers with twos complement. So the other way round: the LSB is not set on the highest value. A set LSB means a negative value...
 
If you put a little compression during recording on the
vocals might help the singer too.

I used to split the vocal signal before going into the A/D and record the vocals dry,
and giving the signal which goes to the headphones a little compression and reverb.
Can help to make them feel more comfortable.
 
ThomasT said:
24bit represents a dynamic of 144dB. The best AD-converters have a dynamic range of around 110dB (18-19bit). More than 120dB is physically impossible as long as you don't record in liquid helium. 24bit are oversized, 20bit would be enough, even to handle the round error during processing. It's just to have 3 full bytes.
Best microfones have around 70dB.

You have enough reserve in dynamic. There is no need to reache 0dBfs. Its no harm, of course to use limiter. But it is not necessary, nor will it improve you sound. That's why we have 24bit now.

Btw. the LSB is used even by lower levels, because audio data are signed integer numbers with twos complement. So the other way round: the LSB is not set on the highest value. A set LSB means a negative value...

I wish I understood one bit of what you just said there. I think I might go and take a physics lesson...
 
I wish I understood one bit of what you just said there. I think I might go and take a physics lesson...

That's not a proplem as long as your sound is good and you know how to tweak the poties.
It will become a problem for you when the marketing of expensive gear succeded and you buy expensive 192kHz need 4times of the HD space just without any physical improvement. (But there may be a psychological one...)

But: electronic and physic are the basics of our work/hobby. It helps to understand what really happens.
Without the law of induction there would be no guitar pickup...
 
Hopkins-WitchfinderGeneral said:
Thats the science part, which is part of the technique... but I always considered audio production to be an art rather than a science.

I believe science is an art!!

I'm doing a Bsc in Music Tech specialising is Audio Systems and for everything I learn about theoretical methods I learn how to apply them to a studio situation.

The best part though is when I load up my custom made trigger-sample programs or have conversations with engineers twice my age who sit there not knowing why hell they're desk does this or how a compander works.

To be good at what you do you must know your tools, and to an engineer/producer that means a knowledge of acoustics and electronics as well as computer programming and audio dsp.

Some people can go through life just using something, I personally have to know why and how.