Welcome,
I think the rule of thumb here is that you want your mixer faders to be as close to unity (0) as possible.
This takes some experience, planning ahead, and maybe a little logical thinking.
An example you ask?
You probably don't need that hi-hat track recorded at maximum. It cuts through just fine at a much lower level. This allows you to keep your mixer fader closer to 0 where you can make fine adjustments.
Things that you know that will be low in the mix can be recorded lower from the start.

What you've been reading is correct. The stronger the recorded signal, without clipping (never over 0dB), the cleaner the audio quality will be because you increase the gap between the noise floor and your desired audio signal (called dynamic range). The weaker your recorded signal, the greater the chance you have of introducing unwanted noise into your mixes. There will always be "track-volume adjustments" to do after the fact. It is time consuming to mix and stereo imaging via panning is an important step in mixing so individual sounds don't become lost. It also helps to bring desired sounds forward.

The sum of all your tracks pinned at unity gain (0dB) will clip your stereo output if your stereo output fader is also at unity gain. Adjust your channel faders to get a good sounding blend and add the appropriate dynamic effect processors with the appropriate settings on your channels (including your stereo output) to keep the overall mix from clipping. Try sending all the vocals to one Aux Bus and instruments to another. If the mix is good, but your output is too hot, drop the faders on the Aux Buses so you can keep the overall blend and still have control over the volume levels as a whole without having to drop every fader separately. Groups and Buses are a huge help in streamlining mixing (for me). Everyone does it differently and the end result is what matters.

It's an art like painting... the recorded tracks and effects are your paint, the channel faders are your brushes, and your ears become your eyes. You can have the best of all of these, but painting a beautiful picture takes time and practice. Anyone can paint... some better than others. Keep practicing.

I'm going to have to disagree. The OP is clearly speaking of digital recording (RME FF800).

What you have been reading is perhaps appropriate for back in the day when there was only analog gear that could take a few overs. In the digital realm this is a very very bad thing (with very few exceptions which I won't address here). Digital is all 1's and 0's. You can't get louder than that.

The rule of thumb in today's digital recording, and this has been true for quite a while now, is to have your peaks in a range between -20dB and -6dB. You do not want to get close to 0dB because it SIMPLY ISN'T NECESSARY. In fact, if you find you did record something too hot then pull the fader on the mixer of your DAW down. That's why it moves in two directions. An alternative is to down to -6dB. As you are finding, if there is zero headroom you cannot have a very good mix. I in fact mix with peaks closer to -12dB or less. The digital noise floor is so much less than any analog setup that it is easier for glaring errors in recording technique to show up. You can't blame it on a badly aligned tape head etc.

Now, if your front end is an analog mixer there is a rule of thumb for recording too. Your setup with THE LEAST noise involves setting all the faders at unity or 0dB whatever it is marked. That means the sticks are neither amplifying or attenuating. Now, all the level of the stick is controlled at the trim rotary at the top of the stick. Keep your main 2-bus also at unity. But, this isn't the situation you have described in your original post and the peaks are STILL kept at -20dB to -6dB.

Now let me tell you the number one mistake made running a mixer: always pushing faders up. The secret is to more often than not cut the signal on individual sticks until a mix sounds great and then push the main 2-bus louder if you need more overall volume.

Man I'm glad you posted that, I thought I was losing my mind with everyone talking about trying to get close to zero!!! I try to track with peaks around -12 both to avoid the danger of clipping the odd loud hit, and because I've been told that sometimes it's difficult to know for sure whether there's a small amount of clipping occuring even if things in the AD and DAW say they never hit 0, it can just result in an over-all harshness to the sound. Not too sure about that second bit, but I do know I can always raise the volume later, but I cannot undo a clip!

Wow thanks for such good information and discussion!

TheJackAttack - I think you've help add some clarity around most of what I learned when I "first" started, and stuff that has confounded me ever since I got my own gear. I picked up a lot of those kind of basic principles when analog was still very much in play. Re-adjusting the way I think about levels within the digital realm, and watching my Input monitor levels so signals fall between -12 and -6 would certainly be different for me. But it sounds like perhaps that's my problem!

I think the best thing for me to do is just immediately put this into practice, and also adjust my personal monitoring practices (i.e. turn up those headphones! lol) and see how things go!

Thanks again - I'll continue reading through other threads too. This forum is really pretty awesome - I hope to contribute as I can as well!

ps - and yes, for the benefit of this thread, my setup is as digital as you can get. I have no external hardware around mixing at all. Just a couple rack effects units to bus things through.

AToE, post: 360639 wrote: I try to track with peaks around -12 both to avoid the danger of clipping the odd loud hit

An analog compressor/ limiter would remedy that danger.

I prefer a strong signal when I record digitally and find it easier to drop the fader over trying to increase the amplitude after the fact. I hate putting "gain inserts" on the channels because the recording was too weak (obviously operator error during the recording process). If your final outcome is clear and not full of distortion or noise (unless that's what you're looking for) then it's done right.

I never try to get my signal right at 0, but I do like it strong (just below yellow on the meter... -6dB). I still think a weak signal is bad and I would definitely regard a -6dB or -12dB signal as a strong signal.

As The JackAttack said, going over 0dB on digital recordings is a no-no. Done it before and it sounded absolutely terrible. Especially on a clipped vocal transient.

Yes, a comp or limiter would save me from digital distortion, but I can guarantee that in the conditions I do tracking in (I'm mobile, so unfortunately my monitors aren't always with me or in the best position) I wouldn't make as good of settings on that gear as I would during mixdown. There's no loss for me to just record at -11 and then make sure I have my dynamics control exactly how I want it later. Then I don't have to worry about it. Plus, I'd need an awful lot of comps/limiters to cover all my tracks, and I can't afford to buy 16 of the good ones, so they'd all have to be cheap! Better to just avoid that whole mess for me IMO.

If I have my peaks running between -16 and -8 or so I seem to be pretty happy with levels during mixing, I don't find myself pushing faders up much, just pulling them down generally. My monitors go plenty louder than I could ever want anyways, so if my whole mix is "low" it doesn't really matter. It's a personal rule of mine to not worry about overall volume, just to keep it from clipping. I'll leave worrying about the level to the mastering engineer! :)

EDIT: and I missed that you thought -12 was a plenty hot signal anyways, so I guess I wasn't really arguing with you at all!

No problem. I didn't mention what I thought was a strong signal in my original post so it was all up to interpretation anyway. :biggrin:

This place is great.

I agree with the posts above that say you want to aim for a digital tracking level of -6 to -12 dBFS. Aiming for 0dB may have made sense in the days of 16 bit recording, but not with 24 bits. Signal to noise ratio to burn.

Let me make a comment about this -

TheJackAttack, post: 360636 wrote: ...Now, if your front end is an analog mixer there is a rule of thumb for recording too. Your setup with THE LEAST noise involves setting all the faders at unity or 0dB whatever it is marked. That means the sticks are neither amplifying or attenuating. Now, all the level of the stick is controlled at the trim rotary at the top of the stick. Keep your main 2-bus also at unity. But, this isn't the situation you have described in your original post and the peaks are STILL kept at -20dB to -6dB....

If you are using mixer preamps in your signal chain, I suggest adjusting the trim pots to keep the pre fader level fairly low - below 0dB (usually the pfl level is related to either dBV or dBu). This may mean you will have to push the main faders above 0dB to get the final signal to the level you want, but it is still better to add a little gain there and keep the preamps in the most linear range and give them as much headroom as possible. (This was a tip from Remy years ago that I found very useful.) The end result might be about the same as what John suggests, but the focus is a little different.