Need help with gain staging when recording guitars helix -> interface -> REAPER

[Stiman]

I know this topic is discussed often, but I was curious if I could get some specific advice on my current setup. See the attached image for my setup. My issue is, when tracking guitars, the tracks is fairly quiet, peaking at about -15db. I know that ideally, my guitars would peak at about 6db (I think?).

- Any suggestions on getting the gain staging right?
- If I'm getting into the yellow on the Scarlett, then I'm assuming the gain staging up to that point is correct, right?
- Is there a way to adjust the recorded gain inside REAPER? Is that what I should be trying to do? In the past I would "normalize" the guitar takes to get them up to about -6db, is this a no no?
- Once I normalize the individual files, they sound good, although maybe lacking a little bit of clarity. Would fixing the gain staging fix the clarity issue?

 

[Winspear]

All sounds good - absolutely no need to run things that hot. It can be a bit trickier to say and gauge with 'peaky' sounds like a guitar DI as opposed to say, a sustained distortion or synth, but around -15dB average is about perfect. Do feel free to have the actual peaks of the DI hitting higher than this though, but it is completely normal to have the waveform itself appear miniscule, etc. On a mix where there are a whole bunch of elements like metal, -15dB is great, and you should find if you run everything around there (this also results in turning down the default output volume of a lot of amp sims, drum and synth plugins), your group busses come in with perhaps just a tiny bit of clipping for compression to shave off, and the master channel comes in similarly. If you run everything around 6dB you'll find you are having to turn down faders all over the place to stop outputs from clipping.

 

[Winspear]

P.S. How did you make that diagram, it's aesthetic af

 

[Stiman]

but around -15dB average is about perfect

Ok, interesting. You're right, I'm often dropping the faders everywhere. EZDrummer comes in pretty hot and I guess I tended to want to normalize my guitar wav files to bring them up to a suitable level for the drums.

P.S. How did you make that diagram

It's this site: https://excalidraw.com/
I use it a lot at work for making diagrams and wireframes

 

[Stiman]

All sounds good - absolutely no need to run things that hot. It can be a bit trickier to say and gauge with 'peaky' sounds like a guitar DI as opposed to say, a sustained distortion or synth, but around -15dB average is about perfect. Do feel free to have the actual peaks of the DI hitting higher than this though, but it is completely normal to have the waveform itself appear miniscule, etc. On a mix where there are a whole bunch of elements like metal, -15dB is great, and you should find if you run everything around there (this also results in turning down the default output volume of a lot of amp sims, drum and synth plugins), your group busses come in with perhaps just a tiny bit of clipping for compression to shave off, and the master channel comes in similarly. If you run everything around 6dB you'll find you are having to turn down faders all over the place to stop outputs from clipping.

Out of curiosity, do you use an always-on master bus chain (EQ/limiter/etc...) to bring the levels up while tracking and mixing?

 

[crushingpetal]

EZDrummer comes in pretty hot and I guess I tended to want to normalize my guitar wav files to bring them up to a suitable level for the drums.

Yes. I use a trim plugin to get ezd/sd down to a better level while tracking/mixing.

You also don't want to hit your output bus too hard---leave headroom for mastering.

 

[afterzaar]

Yes. I use a trim plugin to get ezd/sd down to a better level while tracking/mixing.

You also don't want to hit your output bus too hard---leave headroom for mastering.

What plugin are you using for this?

 

[Stiman]

I know that EZDrummer 3 has an overall volume slider, I guess that would probably do the trick. Curious to hear what others say though.

 

[drgamble]

One of the first things I do with Superior Drummer is turn the instrument down 5-6 db using the instrument volume. This especially helps if I am using any of the presets in Superior Drummer. Once this is done a guitar coming in at -15 db will be more than loud enough for tracking.

 

[crushingpetal]

Yes. I use a trim plugin to get ezd/sd down to a better level while tracking/mixing.

You also don't want to hit your output bus too hard---leave headroom for mastering.

Just any plugin that acts like a trim pot on a mixer. I don't know reaper, but it must have one.

The reason you'd want to do this, rather than on a fader, is that a fader has more resolution (able to make finer adjustments) when it's close to 0.

 

[Drew]

P.S. How did you make that diagram, it's aesthetic af

In addition to all @Winspear 's other excellent points, I'll ask this, as well.

Out of curiosity, do you use an always-on master bus chain (EQ/limiter/etc...) to bring the levels up while tracking and mixing?

Why do you need to bring levels up while tracking and mixing? Why not just turn up your monitoring chain a hair? You don't need to be monitoring a signal that internally is running close to 0db, you just need to be monitoring a clean signal that isn't clipping anywhere you're not intentionally clipping it, and then you can control the in-the-room volume as you see fit independently.

You seem to have already figured this out, but the problem with trying to get all your tracks up over -6db, is that you're just going to be turning them down in the mix anyway, or your master bus is going to be clipping like crazy.

EDIT - probably more explicitly helpful color - remember that "gain staging" isn't "getting my tracks to an ideal peak of -6db," but "making sure every gain stage in my signal chain is operating within its optimal range, to sound as good as it possibly can." The -6db thing isn't an "ideal" so much as a cap - most "prosumer" level stuff (i.e - not a Neve board) sounds good through most of its range, but transient response sn't as good as you begin to push it, and beyond that digital clipping is extremely undesirable, so -6db is a great maximum because you're not really risking clipping your transients in the preamps, OR in analog-to-digital converion. But, there's nothing that says you have to reach that cap, just not go over, and the main reason engineers used to attempt to track as hot as they could safely get away with was driven by a concern completely alien to digital; tape's self-noise. Tape's self noise is something like -60-70db, so regardless of where your signal is peaking there's going to be steady noise that far down. If your peaks are at -0db, you'll have (lets take the midpoint and say tape noise is -65db) 65db of usable headroom. Peak at -5db, 60. Peak at -20db, 45db. Digital meanwhile has more usable headroom than any of us will likely ever need - 24 bit has a theoretical headroom of something like 144db, which is insane. You can easily give up 20db here and still have plenty of room.

And, unlike the analog world, you can add or remove gain completely transparently in the digital domain - analog, any gain stage adds noise that will be compounded by subsequent stages. Digital, if you boost the output gain on a plugin to feed a better signal into the next stage, unless that output control models gain stage noise to replicate some analog piece of gear, this make-up gain is transparent and clean. you can stack hundreds of gain stages adding and then removing 5db one after another, and have no ill effect to your audio, whereas analogue it would rapidly degrade into noise.

So, tl;dr - worry about getting levels that work best for each stage of your signal, don't worry about where your peaks are beyond not wanting to exceed -6db or so. That'll make a much bigger impact on the quality of your recording, than making sure everything's hot enough to hit an arbitrary target (that you're just going to need to turn down in the mix anyway)

 

[Drew]

One of the first things I do with Superior Drummer is turn the instrument down 5-6 db using the instrument volume. This especially helps if I am using any of the presets in Superior Drummer. Once this is done a guitar coming in at -15 db will be more than loud enough for tracking.

Same. Sometimes more. For the same reasons as above - leave it peaking at 0db, and you're going to clip your master as soon as you add literally anything else to the mix.

 

[Stiman]

I'll ask this, as well.

https://excalidraw.com/

Why do you need to bring levels up while tracking and mixing? Why not just turn up your monitoring chain a hair? You don't need to be monitoring a signal that internally is running close to 0db, you just need to be monitoring a clean signal that isn't clipping anywhere you're not intentionally clipping it, and then you can control the in-the-room volume as you see fit independently.

I guess this is a misunderstanding on my part. I understand I can bring the volume up on my monitors/headphones, but I thought some suggest tracking/mixing into somewhat of a mastering chain. I know this is controversial, but I don't know any better so I'm trying out what I read on the internet until someone tells me something better or I learn for myself of a better way.
So then, do you suggest I don't do this? I can understand the limiter and the argument for just raising my volume on my monitors, but what about additive EQ, compression, subtractive EQ, saturation? Do you suggest having these turned off and just turn them on periodically to check the mix or bounce progress takes? Or not at all until the mastering stage?

You seem to have already figured this out, but the problem with trying to get all your tracks up over -6db, is that you're just going to be turning them down in the mix anyway, or your master bus is going to be clipping like crazy.

Yeah, this all makes more sense now that I've read this advice, and applied it. I guess I was concerned with tracking the loudest takes possible to avoid the noise floor and the commonly experienced "weak" sound.

Same. Sometimes more.

You quoted @drgamble here, but I'll ask you, does "using the instrument volume" mean you literally turn down the fader that SD is on, or internally inside the VST?

 

[Drew]

https://excalidraw.com/



I guess this is a misunderstanding on my part. I understand I can bring the volume up on my monitors/headphones, but I thought some suggest tracking/mixing into somewhat of a mastering chain. I know this is controversial, but I don't know any better so I'm trying out what I read on the internet until someone tells me something better or I learn for myself of a better way.
So then, do you suggest I don't do this? I can understand the limiter and the argument for just raising my volume on my monitors, but what about additive EQ, compression, subtractive EQ, saturation? Do you suggest having these turned off and just turn them on periodically to check the mix or bounce progress takes? Or not at all until the mastering stage?


Yeah, this all makes more sense now that I've read this advice, and applied it. I guess I was concerned with tracking the loudest takes possible to avoid the noise floor and the commonly experienced "weak" sound.


You quoted @drgamble here, but I'll ask you, does "using the instrument volume" mean you literally turn down the fader that SD is on, or internally inside the VST?

So, I just made a LONG edit, which should help.

Mixing into a mastering chain - controversal subject. Some people swear by it. Others don't. I personally have my reservations - if you need to slap a compressor on your master bus to get dynamics under control in your mix, there's probably something in their that could be better served by an individual compressor on that particular track. That said, while I track with an empty master bus, at some point along the way in the mixing process I'll probably drop some sort of a volume maximizer/limiter on the master bus, mostly leave it bypassed, but flip it on every once in a while just to make sure it's not doing anything unexpected, and then if I'm sharing a rough mix it's a good way to at least be able to share something relatively CD-level. Stuff like additive or subtractive EQ, though, I mean, never say never... but you almost alway want to get that right at the track level, because doing so buys you a lot more potential headroom to work with at the master bus level.

"weak sound" - paradoxally, tracking too hot, and subtly smashing your transients while doing so, is a GREAT way to get a weak sound. When i doubt, track lower than you think you need to. And the digital noise floor, well, I address this in my edit, but it's EXTREMELY low. Your mic pre/DI in noise floor is the one to worry about, and as long as you're nowhere near that you're probably ok.

Superior - honestly, I don't remember. Usually at some point along the way I stop using a Superior preset while demoing, and flip over to the raw tracks that I then process myself within Reaper, which largely takes care of the issue. But, I'd say it's really one and the same, just turning the SD output down, vs just turning the track down, because digital gain staging is 100% transparent and creates no additional noise, and the difference between the two should be nonexistent.

 

[Stiman]

Excellent information @Drew, thank you so much. This clears up a lot of things that I didn't fully understand.