Recording level for backing layers?

I typically try to keep my incoming volume the same (meaning mic settings, etc.) and use the track fader to control the volume during mixing. Acoustic instruments will obviously come in quieter or louder depending on dynamics, but I set the mic at the same level.

My take comes from working in analog studios for years. Not only are old habits hard to break, I also tend to follow the 'if it ain't broke, don't fix it' rule. So maybe folks who cut their teeth in the digital age will have completely different advice - I kinda hope they do, because I'll learn new tricks! Anyway... my analog perspective, which I use in the digital world:

1. You want the best signal:noise ratio you can get. The noise (like tape hiss) isn't going anywhere. So you want the strongest signal you can get cleanly. When doing a level check, I'll push as hard as I can on the instrument - what's the maximum output you can squeeze out? I want input levels set so that's right at the point of clipping.

2. You need some headroom. So from that starting point I'll back off just a touch. That ensures that no matter how loud you're going to get, you'll capture all of the sound.

Then you record.

Sure, the classical guitar or harp or whatever will be captured at a lower volume than the amplified guitar or Hammond B-3. But that's what the mixing process is for - adjusting the balance. I figure I want to catch ALL the sound when it happens, and worry about how it matches up with the other stuff later. If you set your levels too low, you can't go add them back in later on.

Like the tailors say, you can always make the pants shorter. Going the other way is a heck of a lot harder. I'd rather have the original signals as hot as I can get them without clipping... because that gives you more options when you mix.

Never having worked in a studio...
My understanding is that it's exactly the same for digital, for very similar reasons: you want to make the most of the dynamic capacity of the recording. So you record at the highest level you can while avoiding clipping. The differences are that you don't have tape compression to help you avoid clipping and that it's easier to change the levels in the mix.

I've never done analog, so this digital thing doesn't take the place of old knowledge for me. It is all I know. With that in mind, I didn't know that there was such a thing as "tape compression". I guess that's why so many tracks in the digital world need some sort of compression going on?

As far as volume and recording levels go, on most tracks, I do aim for a certain spot in the faders' meters and watch to make sure any heavy handed-ness does not cause clips. I thought the line in the fader meter that I was supposed to sort of hover around was much lower than it actually is. It took quite a few iterations of tracking/bouncing/testing to get a good feel for where my volume needed to me. Same goes for other aspects as well.

That was for the main tracks, though. The back of the mix tracks were something I was quite unsure of. I've read arguments to both ends. Someone complained somewhere recently that new music sucked because everyone thought they had to crank each track up to the limit when recording. This was actually someone's article, not some mindless forum rant.

I guess the bottom line on the volume thing for me would then be to test the track at lower volumes to make sure it is indeed what I think I want to hear after it's all said and done. If the dynamics are taking up too much space, track less of them compress them out. (?)

Back to compression, I'm still quite new to this. I don't know the difference between with and without compression tones until it gets strong. It's better than it was, but still more ear training required. Many of the online tutorials get lost on me when they try to show the difference ith or without compression. I just sort of sit, smile and bob my head in polite approval to my laptop monitor. :roll: Where it leaves me, though, is a sort of an inability to fine tune the compressor plugin. I can use a pre-set, but "tweaks" don't mean anything for me yet. I could be making something worse for all I know. I tend to leave compression at the problem solving level and move on.

The main reason to keep the recording level at the highest possible level on the meter is to minimize noise , whether it be background noise or electronic in nature . The level that everything is in relationship to the rest of the recording is then adjusted to suit . Maintain a good signal-to-noise ratio and your recording is quieter with things you DON'T want heard . This is especially important in quieter passages . In this digital age we have the capability to do things easily that were once much harder to accomplish : noise suppression , pitch correction , tempo corrections , etc. However the easiest thing to do is capture the best recording where those things aren't necessary in the first place .

Cool thread, Roy.
Thanks to all who have posted.

I get max'ing the level with a loud instrument: piano, electric guitar, sax.

Where I get a little lost is with elements that generate a low volume/signal.
An extreme example would be something like a shaker. Surely you wouldn't juice the recording volume to try to approach clipping?

A less extreme example is a bass guitar signal. Run DI through my audio interface I get a signal maybe a 1/4 of my electric guitar signal. If I goose it I get a ton of noise introduced...

How do you handle elements like that?

And hope this isn't a thread hijack. :)

I didn't know that there was such a thing as "tape compression". I guess that's why so many tracks in the digital world need some sort of compression going on?
Two different types of compression here - digital audio compression to make files smaller and dynamic compression which is about the volume ('squashing' loud and quiet together). Tape compression comes from tapes being a bit resistant to clipping, they sort of round off more softly than a digital recording which hits max signal as a hard barrier.
Where I get a little lost is with elements that generate a low volume/signal.
An extreme example would be something like a shaker. Surely you wouldn't juice the recording volume to try to approach clipping?
It's not so much about trying to juice the volume as making sure you're recording at the maximum level you can. If you've got things cranked so much you're capturing background noise that's not exactly what you want either, but you'll be turning it down in the mix. What you're aiming to capture is dynamics. Like Diceman says it's about getting the best signal without noise. If you're doing that shaker hard or not at all then you don't need to worry so much, if you're doing hard, soft then you want them separated and still to be able to pick up soft without any noise in the background.
So, yes the bass guitar example is a good one, there's no point turning up the recording level once you're turning up the nosie too. What you do is try and increase the signal and cut down the noise (exactly how is an exercise left to the reader...).
Someone complained somewhere recently that new music sucked because everyone thought they had to crank each track up to the limit when recording.
Yes, this is a fairly well known issue, look up 'loudness wars' on wikipedia. And it's actually about compression and how we listen to music. If you go back to 70s/80s recordings tended to exploit the maximum dynamic range, quiet passages were quiet and loud passages were loud. But that means volume varies over the track and it also means that if you apply some compression (making the quiet bits louder essentially) the average volume of the track goes up at the cost of dynamics. This has happened gradually in pop music over the last few decades. It makes songs more noticeable on the radio and easier to listen to on headphones in a noisy environment, but you're losing fidelity (and some amount of artistic vision).
Example: being a bit of a completest I got the Led Zeppelin Mothership CD when it came out, partially because I was interested the tracks were remastered. So I loaded up some tracks against the same ones on the previous individual album versions, the new ones were noticeably 'flatter', the volume much more level all the way through.

Again, caveat, I don't do any studio work, I just went through a phase of reading up about all this.

Cool thread, Roy.
Thanks to all who have posted.

I get max'ing the level with a loud instrument: piano, electric guitar, sax.

Where I get a little lost is with elements that generate a low volume/signal.
An extreme example would be something like a shaker. Surely you wouldn't juice the recording volume to try to approach clipping?

A less extreme example is a bass guitar signal. Run DI through my audio interface I get a signal maybe a 1/4 of my electric guitar signal. If I goose it I get a ton of noise introduced...

How do you handle elements like that?

And hope this isn't a thread hijack. :)
First you do everything you can to get rid of the noise caused by the bass/audio interface . Moving to a different location with the instrument possibly can lessen the noise . Computers , monitors , refrigerator compressors , neon lights , etc. all create electonic noise which can be picked up by magnetic pickups . Getting a power conditioner to clean up the electrical supply is an option . Recording shakers would require a noise free environment to begin with and near-field miking to capture the low output .

I've seen articles that say shoot for -12db and some that say -8db and others that say -6db, so I aim somewhere in that range. I've found that shooting for that range makes me do the least amount of work getting an initial mix. When I recorded everything close to max, I had to turn everything down, because when you sum up all the tracks all hitting max, it's going to clip for sure. I usually can lay down a guitar or two, bass and drums all around -8db and without touching the faders, my master is hitting right about where I want it.

I tend to leave a little headoom digitally for that same reason. Everything together triggers the meters past max. There is a happy zone where you still have head room yet minimize noise. In my old 4 track days, yes the tape compression is nice! With tape red lining your meters is a good thing. Red lining a digital recording sounds like poo. If your signal to a tape is too low ..... hssssssssssssssssssssss.

What's the concensus (or your personal preference) in the digital recording world as far as what volume levels should be when recording secondary instruments?

Record them exactly the same way as primary instruments. You may decide to mix them so that they ARE primary instruments.

Here's a link to a song I did guitar on. The intro "hook" was a guitar part I did to go under other instruments. It got bumped front and center. You can hear the same guitar in the background on the chorus doing the ersatz Fender Rhodes arpeggios ( with the "hook" dubbed in also).

http://soundclick.com/share.cfm?id=7931488