Setting the input gain on audio interfaces

Gear
#21

They probably correlate, but, well, if you want to be clipping at the preamp level, you should be damned sure you know what it’s doing to the signal. There are a lot of high end mic preamps that people intentionally push a little for the slight saturation you get out of them. They usually cost more per channel than even a fairly high end 8-input interface today will set you back, and there’s a good reason for that.

I don’t think there are too many situations where you have to worry about clipping at the preamp level but not at the converter level, in the “normal” home recording range. But I’d also generally make sure that, if you have any sort of preamp clip indicator, it’s staying out of the red, and very likely out of the yellow if you have that level of visibility, unless you’re absolutely sure that intentionally overdriving it a little is adding something very musical you want in the recording. And you definitely DON’T want to clip at the analog-to-digital conversion point, even if something else is later reducing the signal down to -12-24 when it hits the DAW channel.

There’s a lot more to gain staging than “don’t accidentally overdrive something along the way even if your finished signal isn’t clipping” but that’s the single highest-yield step in the process, making sure nothing is clipping earlier in the chain than the track VU meter you see at the end. :+1:

#22

I’ve had good results when the input your DAW sees is -24 to -12, can allow peeks to breach that but best to keep it in the middle of that range for the most part.

Adjust your hardware gain to meet that range. @Drew they do correlate, but I’m I’m sure there is some hardware or the other out there that does not, in my case it’s been RME always and it’s been reliable. I don’t need to look at the hardware monitor levels.

It’s a headroom game, and for my workflow this is the truest signal setup for my needs.

Sure there are some things you might want to overdrive at the source, it’s a flavour thing and that’s cool too if you know what your going for.

1 Like
#23

This pretty much means that nothing in your chain is likely adding and then reducing gain before you hit the DAW, then, which IMO means this is pretty safe. But, I guess the main reason I’m just warranting a little caution is if you’re tracking based on DAW volume, and you have anything active on the channel strip while tracking, that can get into trouble - if you’re monitoring output on the channel bt the channel has a few VSTs running and you’ve got like a tube screamer model and then an amp model, then the interplay of their outputs and gain stages means you absolutely could be be inducing clipping, including at the interface, that you’re just not seeing or able to hear clearly downstream. In that scenario I’d want to be monitoring the DI signal to make sure THAT’s not clipping. Otherwise, even if the modeled output is falling at -12db or so, you could be lopping off the peaks of your signal in ways that may not be overly obvious through a Marshall amp model or something, but will be robbing it of some of it’s impact and “punch” and will cause it to get a bit indistinct in the mix.

This probably isn’t a problem if you’re micing an amp, and monitoring the mic’d signal as it hits the DAW with no FX processing on the tack and the track fader at unity. And if I remember correctly that’s how you record, micing up a real amp. But if someone IS slamming the front of their interface, I wouldn’t want to have them come away from this thread thinking “cool, cool, I just gotta slide down the track slider until it falls to -12db and I’ll be golden here,” and expect that to really work.

#24

If you’re slamming the front of your interface, the DAW will reflect that as well. Though you make a fair point, for example if I was using a UAD and post processing between the interface input and DAW input, it’s not 1 to 1. But we are digressing from my point. In the DAW -24 to -12 is where I keep the signal regardless.

#25

As long as the DAW is monitoring input, not output, and there’s nothing else in the signal chain, that’s true. But it definitely doesn’t hurt to double check since clipping on input isn’t something you can fix downstream.

#26

sorry if this has been mentioned but I would always recommend a dedicated DI if you can afford it. Many interfaces (even with instrument inputs) are not that great and will clip with higher output pickups.

The countryman type 85 is a classic for a reason.

Other than that, aim for about -6db for your peaks (loud strum etc.) and you should be golden :slight_smile:

#27

Yeah, one of my guitars has Black Winters and will clip if I look at it funny with the input gain minimized. I’m probably just going to replace them, but I’ve been wondering what was wrong.

1 Like
#28

Yeah, just a weird thing with many interfaces (it even happens on my relatively expensive audient unit) so I always just use a D.I.

nothing wrong with your pickups :slight_smile:

#29

Little update that I have been using recently:

Provided the input DI is not clipping, you can always fix things after the fact by “normalizing” the DI waveform before it gets re-amped by the various plugins. Different DAWs will probably have different options and methods to do it, but in Reaper I found that normalizing by setting an overall target LUFS value (a measure of perceived loudness) gives fairly consistent results.

3 Likes
#30

Yeah - one of the ways digital recording is transformative that’s maybe a little underappreciated is that in STARK contrast to analog, where every gain stage also adds self-noise, you have access to transparent, silent, noise-free gain control. That plus the much higher useable headroom (off memory the tape noise floor is something like -50 to -60db, while 24-bit .wav files have a dynamic range all the way down to like -144db) means that you really don’t have any reason at all to try to record “hot.” You can peak at -30db in digital, which in the analog world would leave you dynamic range of about -20db in practice, and STILL have more than 110db of usable dynamic range from your peaks to the floor. That’s insane and in practice your limiting factor isn’t going to be equipment self-noise so much as whatever ambient noise is also captured with your recording signal.

It’s honestly pretty amazing. And, arguably, a lot of the “analog warmth” that people miss in digital is probably related to the amount of compression you just had to use on everything to get everything smoothly into the dyamic range that analog could support, plus things like maybe rolling off the highs and lows a little more than CD allowed because of the difficulty in geting a vinyl record to reporoduce that smoothly without the needle jumping out on bigger bass notes. But, the evolution of recording technology in the last 20 years even that I’ve been doing it, is mind-blowing.

3 Likes
#31

Absolutely, in my mind one merely needs to capture the signal appropriately (no clipping and minimal background noise), and then everything can be repeatedly changed to one’s heart’s content.