Calibrating Input Level for Plugins

[2dor]

Absolutely. I purposefully kept it generalised to make it more accessible and easy "for the masses". Optimising SNR is for sure a better way to go, but if people discover they have to do some work to achieve it, then no one would bother. And for a lot of people, they're already recording sufficiently above the noise floor.

Really you want to satisfy both - If noise is presenting an issue, then it would be crazy not to increase the gain. If the noise is sufficiently below the background noise of the pickups, then the next step would be finding the easiest way to achieve a known input headroom. Many interfaces are designed to be pretty optimal at 0 gain - thats why Rabbea said "why would you add more gain?". Humbuckers on a 12dBu input basically have no room to add more level without clipping, so in those cases all you need to do is calibrate. The Behringer is quite an outlier (and was used in the video to prove the point) because its VERY noisy, and also produces a weak signal at 0 gain.

I'm a bit concerned that some dumbasses will selectively reference the clickbait aspects of the video, and ignore the good advice on calibrating and optimising SNR. The video is totally correct, but people are going to selectively draw their own narrative from it.

Pfff - on a +13 dBu input I can get very close to clipping the input with my Suhr Aldrich pickup in the bridge and a 550K pot; no boost pedal or anything else after it. Raising the input knob will mess it up for good.

Anyway, at the end of the day, folks on the internet will always do whatever the hell they want

 

[MirrorProfiles]

Pfff - on a +13 dBu input I can get very close to clipping the input with my Suhr Aldrich pickup in the bridge and a 550K pot; no boost pedal or anything else after it. Raising the input knob will mess it up for good.

Anyway, at the end of the day, folks on the internet will always do whatever the hell they want

Ha yeah, the internet is full of people cherry picking things they've read and repeating them without always understanding why. As long as the information is out there, anyone who wants to learn and understand can.

 

[MirrorProfiles]

it's just that the message delivery is a bit douchy / "mine is bigger than yours".

I didn’t really have an issue with the videos presentation at first, but when I think about it, saying “everyone is wrong” is a bit misleading and is potentially divisive. It’s understandable why Rabbea’s advice was “keep the gain at 0” - he wasn’t advocating recording guitars where there is room to boost 20dB, or using noisy interfaces. His advice is absolutely solid *for his situation*. Half of the video is in agreement regarding knowing your headroom, and many situations are not going to require increasing the preamp gain. Pretty sure Cordy did some videos showing the difference of interface noise at different levels and explained why it often isn’t going to be an issue.

So now the majority of comments are “see, I knew I was right to record my DI’s really hot” and absolutely ignoring the other half of the video that’s about actually calibrating.

TBF, he does repeatedly say the video is trying to get across that optimising SNR and calibrating are 2 seperate things and you shouldn’t just do one and expect the other to be taken care of. The issue is the amount of comments that are treating it as an either/or.

 

[DLC86]

I just received (what feels like) my 131st FM3 and measured the DI signal. I have connected it to my audio interface via S/PDIF and have to attenuate the output by exactly 12.1 dB to match Helix level.
View attachment 31674

Did you select "input 1" as the spdif source? Otherwise you'd get a boost if the source was the output.
The dBu value seems quite low compared to my fm9 (17.5 dBu)

 

[yeky83]

A good video just demonstrating an aspect that gets a little overlooked. As with many things audio, so much gets repeated online by people without really understanding why (I think this video addresses some of those things that get repeated).

I purposefully avoided too much focus on optimizing SNR, because that would mean most people would have to manually measure their interface, and most guitarists simply aren't going to do that. Even experienced engineers I know get frightened at the sight of a multimeter. So, given the most popular interfaces out there have a reasonably sufficient amount of headroom AND SNR at minimum gain settings, its an easy starting point for accurate calibration.



Anyway, I think the points are well made and should help those who want to understand the topic in a little more detail.

I brought this up some time ago and you had shooed me off

TwoNotes GENOME!!!

Let’s see if I have this right. If you don’t know the input gain the plug was built around then your choice for input gain is one more thing you might be doing wrong in your pursuit to use the plug at its ‘designed to be‘ optimal, state. When coming from the real amps and cabs side of things...

thegearforum.com

Optimizing SNR doesn't take a multimeter

 

[MirrorProfiles]

I brought this up some time ago and you had shooed me off

TwoNotes GENOME!!!

Let’s see if I have this right. If you don’t know the input gain the plug was built around then your choice for input gain is one more thing you might be doing wrong in your pursuit to use the plug at its ‘designed to be‘ optimal, state. When coming from the real amps and cabs side of things...

thegearforum.com

I wouldn’t go as far as shooing you off, but I was demonstrating how it’s generally not going to be an issue for most people. The behringer is an outlier interface that is going to require a need to minimise noise, he even said he chose it out of about 20 interfaces because of this.

If you are using a noisy interface and the peaks have room to boost 20dB or more, then obviously you’d benefit from optimising SNR. No one has ever said otherwise.

The most common interfaces are not going to have that much room to boost the signal, nor will they be as noisy to begin with. So long as your interface noise is sufficiently below your pickup levels, all is fine.

Optimizing SNR doesn't take a multimeter

For the average guitarist to work out how much gain they’ve added is not always going to be easy.

The comments on that video are predictably ignoring almost everything that he’s saying - his advice is to calibrate as suggested, and then if there is room to optimise SNR, add gain on the preamp and then subtract that from the plugins.

The typical response in the comments is just interpreting it as “cool, set gain to just below clipping”.

There’s nothing I disagree with, but users should be aware of what issues they are facing and then decide how best to tackle them.

The discussion about setting levels for an accurate gain response from the amp is being turned into a discussion about noise. They are 2 different things that each need to be considered. It’s nothing that hasn’t been discussed already and that everyone is in agreement on.

 

[James Freeman]

While everything in the video above is true, there is a flaw in his method, he assumes guitarists can and will measure noise floor and SNR with expensive plugins and adjust the peak signal to be the same as before adjustment.

We are working with concrete numbers from the manual to simplify things for guitarists, while he introduces a random factor (gain adjustment before clipping) which unless one is proficient with measuring SNR with expensive plugins the results will also be random.

Also, most consumer interfaces Instrument Inputs are in the +12 dBu range, humbuckers are close to clipping that as is.

 

[James Freeman]

I just received (what feels like) my 131st FM3 and measured the DI signal. I have connected it to my audio interface via S/PDIF and have to attenuate the output by exactly 12.1 dB to match Helix level.

Yep, I did the same a few weeks ago, the value is -12db for the SPDIF output to get Helix values (11.5 dBu).

While the input of the AxeFx may be 17.5dBu, the digital output from SPDIF does not keep the same value.

 

[MirrorProfiles]

he assumes guitarists can and will measure noise floor and SNR with expensive plugins and adjust the peak signal to be the same as before adjustment.

Guitarists seem to be having a hard time reaching the end of the video and drawing the right conclusion from what he’s saying.

 

[2dor]

Guitarists seem to be having a hard time reaching the end of the video and drawing the right conclusion from what he’s saying.

lmao Neural DSP have a very cheeky preset in their Fortin Nameless X release aimed at the average guitar player hot-head.

 

[James Freeman]

TikTok recommends creating videos between 21 and 34 seconds long

 

[Sascha Franck]

Anyway, I think the points are well made and should help those who want to understand the topic in a little more detail.

Yeah - but one of the main takeaways would be that it's not a great idea to keep your interface's input gain at zero. Which will become an issue for those of us not being able to have a clear digital readout of the input trim setting. Such as me with my Motu M2. Or such as on the Behringer he's using in the video.
Now, once I've done the required math, measurements and basic adjustments needed for ideal plugin performance, I will be able to always reproduce input gain properly, but that requires some re-measuring of your input level each time when you may deliberately or accidentally have altered your input trim pot (for which there are several reasons).

By now and mainly using HX Native (which I have made all my basic calculations/adjustments for), I'm doing it the halfassed way buy running my guitar through a chain of my usual gain plugin (suitable to get me the proper HX level with the interface gain set to zero) and a level metering plugin. I do then just strum something, look at the level meter, switch the gain plugin off, strum again to see where I'm ending up and keep adjusting the input level of the interface until the level meter shows the same. Works pretty fine but it's obviously not all that accurate. For more accuracy, I connect one out to the used input and play some steady signal through it.
Whatever, I need to do that all the time after moving my stuff (which atm happens daily as I'm on a 2 week tour). Conclusion: My next mobile interface will feature a digital input level setting readout.

 

[2dor]

Conclusion: My next mobile interface will feature a digital input level setting readout.

RME

 

[Sascha Franck]

RME

Yeah, sure. But I got the M2 for reasons, even outside of price, mainly because the form factor seemed relevant for me (and better on the M2), but by now I think it's not really that important. So yes, my next interface will likely be a Babyface.

 

[2dor]

Yeah, sure. But I got the M2 for reasons, even outside of price, mainly because the form factor seemed relevant for me (and better on the M2), but by now I think it's not really that important. So yes, my next interface will likely be a Babyface.

Does the Babyface have a Hi-Z instrument input nowadays?

 

[MirrorProfiles]

Yeah - but one of the main takeaways would be that it's not a great idea to keep your interface's input gain at zero.

*Might* not be a great idea, it really depends. If when the gain is at zero and you have a sufficiently loud signal that is well above the noisefloor floor, you’ll have an easier time calibrating from a known value than having to work out how much gain is added. It’s certainly an issue on a noisy Behringer that records a very quiet signal at minimum settings. I suspect your MOTU isn’t as noisy?

RME and UAD have digitally stepped input levels, but I also have to leave both at 0 or else I’ll clip. So while it’s handy in theory, it never actually benefits me for guitar recordings.

 

[TheTrueZoltan!]

Does the Babyface have a Hi-Z instrument input nowadays?

Yes, the Pro-FS does have a 1MOhm Hi-Z input.

 

[TheTrueZoltan!]

[…] RME and UAD have digitally stepped input levels, […]

I love stepped controls. So satisfying. Everything should have stepped controls - no matter if digital or analog.

 

[MirrorProfiles]

I love stepped controls. So satisfying. Everything should have stepped controls - no matter if digital or analog.

They're the best, especially when there's a satisfying click between settings. But there's one big downside:

and even on newer gear, stepped gain controls is very fiddly to build:

 

[TheTrueZoltan!]

[…] when there's a satisfying click between settings. […]