You just opened a project and you want to make sure all tracks do not exceed -12dB.
 
What do you do or use?
 
Do you reduce the gain?
 
Do you limit?
 
Do you compress/limit?
 
Just wondering, thanks.

I start by getting proper levels on the way in during tracking. Then, I adjust the Gain to make sure the highest peaks do not cause any Pro Channel modules to turn red. Then if any Pro Channel modules increase gain too much, I insert an FX chain to adjust the output to the next module so as to make sure it does not turn red.

The first thing I do is get a recording with levels close as possible to where I want them to be. I used to record much hotter but now I shoot for something comfortably over the noisefloor but with room for gain staging. Once recorded, I use channel gain/trim to a target average. This will vary by track type but I always do this before any compression or eq. Once I am satisfied, I (if necesary) do a high pass rolloff and then leveling via compressor  if necessary.  For me, this approach allows me to be gentler with compressor.

I don't have a problem with clipping or anything.
 
I just want to get each track to behave at about -12dB.
 
Normally when I fire up a fresh new project, I am at about -6dB or somewhere in that area depending on instrument etc.

As was said above, get it right going in and you have zero problems come mix time.
 
This is fine when you're recording with microphones, but a lot of "recording" these days is done either by recording midi into a vsti or composing in the track/staff views.
 
You have to take each case as it comes. If you're happy with the dynamic variations you've applied to your midi sequencing and the levels are too hot, go into the vsti and reduce the level there (the alternative is to use the Gain control at the top of each track (in console view) or via the same control in the Track Header in Track View.
 
Some synths put out a HUGE volume which clips immediately if left at the default - like BFD for example.
I always start by going into the BFD mixer and reduce every fader by at least 9dB

I only record with mics...
 
Old school prog/rock band.
 
No keys...yet.

Now... as everyone should know by now, I'm still learning (particularly general production skills) but I was under the impression that there is absolutely no need to record at -12dB when dealing with a DAW. I make sure I get as hot of a signal as I can before clipping then turn it down ever so slightly just to avoid any momentary overages. Once it's tracked digitally (without clipping) it should not matter... right? I get a powerful signal and if anything is being overloaded in a track I can just turn down the channel gain.
 
I will however export at -12dB if I am sending raw tracks to someone else to mix and I will make sure my tracks are turned down to -12dB before beginning my own mixes but as far as the actually capture to disk I thought "as loud as possible without clipping" was the way to go.
 
If I am mistaken on this I would very much like to hear the arguments AGAINST this method.
 
Cheers and thanks.

You can track hot if you want, but it gives you less headroom to play with when you start adding compressors, eq, etc. many pros will throw in a gain plugin or waves q1 with no bands enabled and drop the gain by about 5 or 6 decibels on each track and only then do they start adding plugins. Also, if you push your preamps too hard to get those hot signals, they may end up making more noise, since the noise they make is not linear, but concentrated towards the end of their gain limit.

If I want no track to exceed -12dB then I'd argue the correct thing to do is pull the volume fader of that track down to -12dB.
 
When recording into that track, I'd aim for it to be as hot as I can get it with no risk of clipping. How to do that depends entirely on what instrument is being recorded. If it comes in far too quiet, then I'll push the track gain up to compensate.
 
The ideal for me is that with faders at 0, the meters tend to be hitting the orange zone. That's kind of what it's there for, after all - to say "this signal is warm, but not too warm". Then I pull down faders to control what is sent to the buses. Not much else to it.

Kylotan
 The ideal for me is that with faders at 0, the meters tend to be hitting the orange zone. That's kind of what it's there for, after all - to say "this signal is warm, but not too warm". Then I pull down faders to control what is sent to the buses. Not much else to it.

Good. I've been operating like this for a while since I learned about input into the DAW a couple years back but have seen a couple posts alluding to recording at lower levels lately which has been making me question it. I have never had any headroom problems using this method (because I can always turn things down). I'm guessing this idea of tracking at -12dB is a remnant of old school analog recording wisdom where you had to be very careful about headroom and noise floor when actually tracking.
 
I think what happens is people forget there is a lonely little Gain knob at the top of the channel strips that can be turned down if the audio is slamming into the effects too hard.
 
Also... when using those "Ride the Input Trim Until X Happens" that is on sims like TH2 when you acheive that desirable level without changing any levels in Sonar the resulting signal is far hotter than -12dB. That to me is an indication that a hotter input is desirable in a DAW.
 
Cheers.