Neural amp modeling + Individual pedal sims a good idea?

wim

The only thing that bothers me about this concept is how it flys in the face of how real amps respond to different guitars, such as a Strat single coil vs. a fat hum bucker. If we're normalizing all inputs to the same level, we shoot that in the foot.

If I still played more than one guitar like I used to, I'd be aiming the hottest one for the 0db level and leaving the weaklings to do their single-coil thing.

btw, Rob, ANALOGyAMP (shouting "analog" and "amp" like I should), takes distortion and boost pedals better than most amps, so good job accommodating higher levels there.

Rob_Jackson_Music

@FizzyLizzy27 said:

@Rob_Jackson_Music said:
The idea (according to me ) is that you add the gain in the plugin / virtual amp - not the interface.

What’s the benefit there?

Consistency for starters, in theory at least. But still not everyone agrees on this, including the plugin developers themselves... I think most people crank the input gain until they see a clip LED then back it off a little. It's totally down to the plugin design as to whether you're now hearing that amp sim as the developer intended. So I mention it in my user guide for ANALOGyAMP.

All my humble opinion of course, but audio interfaces add gain very efficiently, in a linear way. Part of the charm of guitar amps is that they don't. So if the amp plugin is modelling an amp, it kinda makes sense in my strange world is that it should start off with the kinda signal that a guitar amp would see straight from the guitar, i.e. not the guitar signal already boosted by 12-15dB.

@wim - hmm, not talking about not going over 0dB / clipping. I'm talking about not adding any gain at all to the guitar signal. That's going to be maybe as low as around -20dB.

At the end of the day, let your ears be the final judge, but it is interesting that a lot of criticism aimed at guitar amp modellers is often that they sound overly harsh or brittle. That may be because they're just being hit way harder than the designer intended.

And yeah, it's kinda frustrating that this is still not really standardised in some way, especially as the whole "high as you can without clipping" was to get a good signal to noise ratio - virtually a non-issue with today's good quality preamp 24 bit (or higher) digital interfaces.

I was going to reference the original video that kicked off this whole debate but there are so many contrary videos now, arguing back and forth, I think it's simply a case that the developer of the plugin should say what the typical baseline input level should be.

Squishy

@wim

btw, Rob, ANALOGyAMP (shouting "analog" and "amp" like I should), takes distortion and boost pedals better than most amps, so good job accommodating higher levels there.

Kombinat4 too! 😀

Rob_Jackson_Music

@wim said:
The only thing that bothers me about this concept is how it flys in the face of how real amps respond to different guitars, such as a Strat single coil vs. a fat hum bucker. If we're normalizing all inputs to the same level, we shoot that in the foot.

I'm saying your interface gain stays at zero / the same - regardless of the guitar. But yes, the amp sim will sound different if you hit it with a single coil, a humbucker or an active pickup. Just like a real amp does.

Thanks for the kind words re ANALOGyAMP btw - much appreciated.

wim

@Rob_Jackson_Music said:
@wim - hmm, not talking about not going over 0dB / clipping. I'm talking about not adding any gain at all to the guitar signal. That's going to be maybe as low as around -20dB.

Ahh! There it is. That's what I was looking for when I asked the question.

At the end of the day, let your ears be the final judge, but it is interesting that a lot of criticism aimed at guitar amp modellers is often that they sound overly harsh or brittle. That may be because they're just being hit way harder than the designer intended.

^ This. 😎

And yeah, it's kinda frustrating that this is still not really standardised in some way, especially as the whole "high as you can without clipping" was to get a good signal to noise ratio - virtually a non-issue with today's good quality preamp 24 bit (or higher) digital interfaces.

After the coffee finally kicked in I realized this is really a non issue. "Standardization" is a sill term when it comes to guitar tone. Rock 'n Roll would never have been born if some idiots didn't break the standards and abuse the hell out of their amps.

I was going to reference the original video that kicked off this whole debate but there are so many contrary videos now, arguing back and forth, I think it's simply a case that the developer of the plugin should say what the typical baseline input level should be.

Cork sniffers. I'm glad you didn't. 😎

I'm off to break some rules. Creatively. Instead of trying to computerize life.
I keep forgetting that I really don't enjoy making mechanical ... and easy ... music as much as working at it.

Rob_Jackson_Music

So this guy argues it's actually a two stage process and that everyone else is wrong including me.

https://youtu.be/gJ59h7xfvdI?si=rrkqEUWRc1a9ZQxn

Does make sense though, but it's not what I do. I guess if I worked on a higher gain shreddy type amp, I'd have been more worried about signal to noise.

wim

not gonna watch that. 🙈

FizzyLizzy27

I’ve never once thought about dB values - it’s usually a matter of setting an AUM channel low to start, having my guitar going into my interface with low gain. Then I just fiddle with volume on the software and the guitar itself til things don’t hurt my ears. Maybe I should start paying more attention for consistency sake. Would a dedicated limiter be helpful there?