The setting for AxeFX (around 17.5dBu=0dBFS) is +18dB. Adding 18dB of gain (think of it like 17.5-18) puts you pretty close to -0.79dBu, doesn't it?
What's the issue exactly? IMO it's mostly a solved problem these days. The plugin could just as easily calibrate for the most common audio interfaces with gain at 0 and have optimal levels.
That's a good way of putting it... I wish all plugins that have gain-dependent behavior had a simple dropdown that allows you to select from a list of common audio interfaces (with the instruction that the interface should be set to minimum gain)
So if I was to run a 1vp sine wave into the preamp on my apollo solo instead of using the hi-z input, what dbfs level should I be targeting?
Why would you use a sine wave, guitar doesn’t produce sine waves, it has transient peaks which affect levels a lot, and in the case of audio interfaces it’s those transients that will cause clipping.
Yeah, if this is the only amp sim plugin you’ll use it’s an easy way to do it. The real fun is mixing and matching plugins and I think most plugin users are quite used to getting things set up right. It’s worth exploring as even using a Fractal interface will require some adjustment for other plugins (my spreadsheet has some common values for other companies).
In your case, I’d just set your gain to minimum and then set ICONS to 12.87. Your signal will be pretty hot already and it’s easily recallable. No need to measure.
You use a sine wave to calibrate because it’s constant voltage and repeatable by anyone anywhere. A guitar is not constant so it’s not good for this purpose. Your guitar voltage is proportional to your calibration signal so it scales accordingly. Imagine you were somehow able to make your guitar isolate at one frequency indefinitely, every time without fail. Every guitar is different, every 1k 1V sine wave should be identical.
Exactly. I think people get way too hung up about trying to maximize signal-to-noise, when its functionally simpler and more consistent to run at minimum gain and just calibrate max input levels across devices (especially when the input gain on interfaces adds noise anyway).
totally agree. Optimising SNR is table stakes and often interface design is doing this so you don’t have to do anything (much like Fractal have). There some outliers and you obviously need to be aware of things but it’s pretty easily solved and needn’t be a stumbling block.
Most interfaces produce way too much noise and don’t have nearly enough headroom for a lot of humbuckers/playing styles.
With very few exceptions, I couldn’t even use most interfaces, including Fractal, with gain set to anything but minimum, and even then half of them clip.
Are you using pedals? I can just about clip 13dBu but I don’t think i’ve come across a guitar that’ll clip 14dBu.
No, god forbid. I hate pedals, always have.
My Les Paul Custom with stock 498s clipped 14 dBu Duet all the time, it was unusable.
Some Music Man guitars have built-in preamps that can be trimmed internally for up to +20 dB.
Even without that boost a regular crunchlab is too hot for 14 dBu.
I suspect that the documentation you’re referring to states RMS voltage, while clipping is caused by peak to peak transient spikes.
Tbf, a lot of those transients are going to get clipped by whatever you have first in your chain. That may be an amp, or it may be a pedal or buffer.I suspect that the documentation you’re referring to states RMS voltage, while clipping is caused by peak to peak transient spikes.
It’s also very unclear how that is even measured. It depends a lot on the guitar - where exactly the pickup is placed. On pickup height as well, of course - how close it is to strings. New strings will give higher output.
And the elephant in the room is technique, of course, how exactly you’re playing and what. I believe that the maximum output is produced by very fast string motion perpendicular to the pickup, so palm muting with some pick angle/hand position may produce more voltage than another with the same guitar and pickup.
And manufacturers, they just don’t care about instrument inputs.
Also, when you say that one device clips more than the other one, how do you know? Are indicators working the same way?
Sometimes there is no clipping indication for a fast transient, while you’ll see it in the recorded waveform, and probably hear that something’s off, but too late…
Well, digital clipping is always bad, I don’t think there’s any context where it’s beneficial. And we’re talking about interface specs, exactly where this digital clipping occurs.
That’s another puzzling thing I hear so often and can’t understand where it even comes from.
Is Gibson 498T an edge case? It can absolutely produce more than 14 dBu.