The concept of "strongest signal" is somewhat redundant when recording in a 24bit environment.
 
It was certainly valid back in the bad old days of 16bit but these is SO much headroom in a 24 bit system that it's simply not worth the risk of having anything clipping by recording even close to 0dB.
 
You can easily prove this to yourself by recording something at -12db or thereabouts and crank up the recording afterwards and listen to the "noise floor". It will be non-existent (assuming you follow proper recording practices to start with).
 
Gain staging becomes much easier following good recording technique

I've started to become very conservative with levels, both in recording and setting up initial gain staging. When recording, I typically aim for an average of -18 to -12, regardless of the source. Then those random peaks are still nowhere near clipping. 
 
For gain staging, and this is something I recently picked up from David Glenn, I aim for -15 for kick, snare, bass and lead vox, and everything else for -18. That typically gets a pretty good starting point and leaves plenty of headroom. As I stack effects, I try to compensate the plugin output so that I'm not increasing volume, at least not by a lot. It helps determine if the plugin settings are actually making things better or if they're just making things louder.
 
Track count plays a significant role in this of course, but with this type of setup I usually end up hitting the mix buss at about -6db, leaving plenty of headroom for mastering or mix buss processing.

what great thread responses!
 
i have continued to experiment with lower levels and find the results especially satisfying. always good to pay attention.
 
since i have multiple simultaneous tracks for bass and guitar (through different preamps, cab sims, etc...) i have been sending the channels to a stereo bus and adding VSTs there. lower overhead and chances for unintended gain.
 
i have also been checking out some mastering plugins for the master bus.
some really great, FREE saturation, brickwall limiters and mastering EQ plugins out there.
 
quick and thorough education!
thanks all!
fred
 
 

I highly recommend checking out Graham Cochrane's tutorials at Recording Revolution and David Glenn's tutorials. I've totally stolen a lot of their tips in regards to gain staging (and mixing in general) and have found their advice to be very helpful. 
 
That isn't to say that plenty of people are able to mix great music without following any of those guidelines, but for me personally, they've really made a huge difference. I'd much rather turn my monitors up and not worry about meter levels, fader creep, and clipping.  I've found that by keeping everything low, I'm free to ride automation or add grit/distortion to my heart's content without ever having to worry about clipping.
 
I also do a lot of aux/bus routing. Much easier to control levels or process instruments from the aux/bus level then fighting with a bunch of individual tracks.

I guess the reason I blast my inputs as loud as I can before clipping is because I can. It's just me and I have plenty of time to to redo tracks or reset levels if something clips. Usually though once I set up a track/channel and jam out through it a bit I never go over from then on.
 
Now if I were in a LIVE situation and/or recording others and/or was pressed for time/unable to sit there and tweedle around so much then I'd definitely bump the inputs down more than need be just for extra clip protection. Things generally tend to get louder after sound level checks so a 3db or more buffer would be how I'd go just to be safe.
 
And of course I use my hardware as my first reference/line of defense against clipping. On the mixer if I'm using one or using the input trim on the interface if I'm going DI. Clipping outside the box (before/at the AD converters) is something nothing can be done about after the fact.
 
That said... I guess if there is some kind of sound quality benefit to turning down a bit I'm open to it. I've just always tried to get any out of the box signals as strong as I can because it seemed like the right thing to do and I HATE having compensate for low levels after the fact (I used to get that a lot before I knew what I was doing).
 
Most of what I do is DI guit and bass work anyway though or using line in from outboard gear. Then I add emulators and whatnot. So maybe this stuff doesn't affect me as much and if I did more mic/room recording the benefits would become apparent.
 
Really I've mostly cobbled together my methods by simply experimenting with things over and over again while occasionally adding little techniques I've read here and there or been told by my buddies here (thanks doods).
 
I think I'm getting close to maybe, sorta, kinda getting some decent/consistent results. I'm actually FINALLY about to start screwing around with some old band projects I've been putting off because I didn't want to make them teh suxxors due to incompetence.
 
Now I believe I can pull it off properly... all thanks to this place and a MOUNTAIN of reading, watching, studying and experimenting.
 
Of course... it could still all end up teh suxxors but it's time for that proverbial pot to get pooped in or remove my scrawny buttocks from it.
 
;-)

I used to record as close to 0 as I could get, but no more. I think the historical idea of that was to get as loud of the "good" sound (e.g. singer's voice) to mask noise caused by the equipment. This is still true to a degree, but recording equipment has come a long, long way, especially the ability to record at higher bit rates.
 
I record at 24 bits at about -15. I do gain staging, and bring the levels up in the mix as I pass through buses. BTW I am told that many effects are at their best when the signal is at about -6 or so going into them. My goal it to have the master bus outputting at about -3 or a little under when all sliders are zeroed. 

streckfus
I've started to become very conservative with levels, both in recording and setting up initial gain staging. When recording, I typically aim for an average of -18 to -12, regardless of the source. Then those random peaks are still nowhere near clipping. 
 
For gain staging, and this is something I recently picked up from David Glenn, I aim for -15 for kick, snare, bass and lead vox, and everything else for -18. That typically gets a pretty good starting point and leaves plenty of headroom. As I stack effects, I try to compensate the plugin output so that I'm not increasing volume, at least not by a lot. It helps determine if the plugin settings are actually making things better or if they're just making things louder.
 
Track count plays a significant role in this of course, but with this type of setup I usually end up hitting the mix buss at about -6db, leaving plenty of headroom for mastering or mix buss processing.

I do something very similiar. When I master an album (for my and my wife's consumption, plus any friends or relatives that will lisiten to our music!) I aim for -16LUFS for each song. When I play the CD on my car's radio, the volume of the CD is much lower than commercial music, but I'm not into the Loudness War, and just want to have the volume consistant. The -16LUFS number comes from Dave Katz.
 
Walt

LUFS?
made me Google-Look...
 
http://www.tcelectronic.com/loudness/loudness-explained/
 
good stuff.

Here's a link to a good overview of gain staging. Hope it helps you.
 
https://www.gearslutz.com/board/so-much-gear-so-little-time/463010-reason-most-itb-mixes-dona-t-sound-good-analog-mixes-restored.html

Google is your friend