Dear folks, Sonar X3 user here and trying to learn good technique.
When I record, I am trying to ensure that signals hover around -18db and that peaks don't exceed much more beyond -6db. When it comes to mixing tracks, I like to start with my faders at the -6db mark in order to be able to move them slightly up and down in smallish increments.
 
My question is this. When I have my fader at the -6db mark, the output of the recorded track is obviously not outputting at -18db but at -24db. So should I apply gain to the track whilst the fader is at -6db mark to ensure the output of that track is at -18db?
During my mixing I don't usually push the fader up anymore than 2 or 3 db.
 
So, is this an acceptable technique to apply gain if you start with your faders at less than 0?

The important thing about recording is avoiding extremes.  If your bass guitar in a typical rock song is peaking at -24 dB, you are going to have to add volume to it to make it work in a mix.  If you record your cowbell at -.01 dB I hope you'll end up bring the fader down (all the way down, but that is just me).  You want a nice, strong signal you can work with, but one not peaking the meters (unless you are working in analog).  A hardware compressor going in can help, not only to reduce peaks but to firm up the quieter parts.
 
As to mixing, think about it this way - you have a bunch of tracks going out through a single stereo track.  If every fader is at 0 dB (unity - no gain or decrease) you are most likely going to drive your output buss pretty hard.  That is one reason why we record softer than 0 dB, to give the system (or in the old days, analog board), some breathing room.  You stumbled, just like I did, into a workable habit of lowering your channel faders on everything before mixing.  I found -6dB a good starting place too.
 
But there is nothing wrong w/ bringing up tracks more than 2-3 dBs.  I'll often have some tracks +3 dB or more, a rise of 9 dB.  No harm, no foul.  Other tracks are way down there - 24 dB.  Don't be a slave to the numbers - they just represent reality but aren't it.  The numbers I've given above are just best practices as I use them, not gospel.  Use all such guidelines as, well, guidelines, not laws.  Instead, do your tweaks by ear.  What you find pleasing is your style, and it will be different (slightly or perhaps more) than anyone else.  Start from the averages others provide and find your own.  Then follow your ears.
 
@

i always set my individual faders at zero.
 
i set my INPUT (either from my card, or other device) in such a way as to get a very clean input level,
where i'm seeing RMS of anywhere from -12 to -22.
 
if you are tracking individual input at -18, and your faders are set to 0, you still have tons of headroom, and when you go to mix a bunch of tracks, where you THEN set your faders in relation to the output at the master bus, is all that really matters.
 
point is, at EVERY stage of inputting level to track, you have to consider what you will do to that track in the future...
 
if you never applied any fx or eq, you could run with it a bit hotter...
but you want HEADROOM to be able to further process later.
 
 
as soon as you set up a send for a reverb, you are pushing more gain into that channel that is feeding the master bus, and after 40 tracks, that gain starts to add up to distortion very easily if you are not careful.
 
it is easy to raise level to be heard,
it is hard to turn down signals that were captured TOO HOT

bottom line for me:
don't have tracks peaking higher than about -10.
 
yes, -10.
 
-12 is even better.
 
try it, and see.
 
then when you sum all those individual tracks to the master bus, set them all starting at about -6,
your headroom should be sitting nice and pretty.
 

good article:
 
 
http://www.soundonsound.c...icles/level-headed.htm
 

Thank you everyone. Very helpful. Gives me a bit of confidence. I wanted to begin the mixing process with my channel faders at around -6db position but wanted nevertheless for them to be outputting at around -18db. I don't know why I thought this important. I thought perhaps that mixing with too many tracks at very low levels (-24db for example) would result in a final stereo mix that would require too much of the limiter to bring the finished mix to an RMS of around -12db. I thought I should try to use the final limiting stage to have around -3db maximum gain reduction.
I know that plugins and eq would affect the output. But something was telling me that output levels of all tracks should be calibrated so that better balance and plugin decisions could be made. Hence adding gain to the tracks so that they output -18db with the fader position at -6db

One thing to watch with the audio recorded at too low a level is that it won't push a compressor or saturation plugin in the same way as a louder track. If you think of an audio track, once the audio has been recorded, as being a mixer channel playing back a track from a multi-track tape deck the signal level that hits any processing on the track entirely depends on how loud that audio is at the start of the channel strip. The track fader sets the volume from the track output so can't alter that.

Which is where either normalising all tracks (to a sensible level, not 0dB) or the input gain (which can also cut) control on each track comes in. The track input gain is very useful to standardise how things like compressors respond.