Guitar Amp Sim for Amp Players?

[guitarzan]

I don't know. The more amp sims are based on tests and detailed physical modeling, the stiffer and less fun they have become, IMO. I am suggesting going the other direction — if you can hear it and feel it, put it in the sim in the most pleasing and playable way possible — puting dynamics and feel first ahead of detailed authenticity. Right now the focus is on ever finer detail, while simultaneously missing the most dynamic and fun to play aspects. Screw tests, go by ear and response to playing dynamics ("feel").

Maybe having a dynamic fun sim to track, and a detailed modeler for re-amping would work for some — I am really just looking for a convenient way to play guitar that is still fun and inspirational, personally.

[Mats Eriksson]

You got a point there. As even mr Christopher Kemper made an opinion about, regarding modelling/profiling and trying mimick every aspect of the guitar sounds from a tube amp:

"So how does 'profiling' differ from digital modelling, and what are its advantages?

CK: "Modelling, in the technical sense, is bringing the physics of the real world into a virtual world, mostly by defining formulas for the real world and letting them calculate on a real-time computer. I don't know how other companies model their stuff, but I tend to listen to a circuit rather than study it —because by treating models on a theoretical basis, one tends to overlook some very important side-effects that can later be heard clearly, so you end up listening to it anyway. I did our models of the distortion pedals just by listening, and matching my model just by ear. In contrast, profiling is an automated approach for reaching a result that is probably too complex and multi-dimensional to achieve by ear, or by capturing the behaviour of individual components in isolation."

Whatever there is, there's always something that they/you/me have forgotten to model properly, or even at all. Quoted from:

https://www.soundonsound.com/reviews/ke ... ier?page=2

[guitarzan]

Excactly, Mats, except ideally something more lightweight and versatile than a what a profiled amp is typically, from what I understand. A single amp sim that could cover a lot of ground, from clean to quite high gain from the front panel, with the emphasis on dynamic response ("feel") first. Quickly dialing up something inspiring for live playing (or "tracking") would be the intended use, then for recording there would always still be the ultra realistic sounding but somewhat dynamically stiff and CPU taxing physical modelled stuff for re-amping if desired.

[JCJR]

I don't know. The more amp sims are based on tests and detailed physical modeling, the stiffer and less fun they have become, IMO. I am suggesting going the other direction — if you can hear it and feel it, put it in the sim in the most pleasing and playable way possible — puting dynamics and feel first ahead of detailed authenticity. Right now the focus is on ever finer detail, while simultaneously missing the most dynamic and fun to play aspects. Screw tests, go by ear and response to playing dynamics ("feel").

I don't necessarily disagree. OTOH-- Before home computers were affordable I was a mediocre audio hardware nerd. Would repair solid state and tube amps among other typical gadgets, and would design and build various gadgets. I had looked at enough schematics and scope traces that I thought I knew basically what they were doing.

Then later, but still before VST was invented I decided to try "crude" quasi-static modeling of guitar amp. When I would fix tube guitar amps I had never paid much attention to the craziness that happens after they start clipping-- Duh, its expected to get weird when it clips. I mainly just made sure that for instance a repaired 40 watt amp can play "fairly clean" 40 watts without exploding on the bench.

So it was eye-opening when I started studying the weirdness when you seriously crank past clipping. I couldn't guess how such a simple circuit was doing such strange things. The simple ideas how I thought it ought to behave would never make oscilloscope traces like that.

I never tried much on modeling since those decades ago. Its interesting but never had the time.

Am just saying sure, you could develop a theory about how the tubes, transformer and speaker dynamically interact. Then write something that would work kinda-sorta how you think it is theoretically behaving, with lots of knobs for break points and various time constants. And maybe some of the settings could be tweaked to sound good. Maybe some of that would be "inventing something that never really existed in a real amp". Which is not all bad.

But if you get a WRONG idea about how it ought to be working and implement the WRONG idea, then you are "emulating something that never existed" and then wonder why it doesn't sound like that classic black face super reverb turned to 10. OTOH maybe you will like it BETTER than the super reverb turned to 10. Stranger things have happened.

If I was to try to develop some theories about how the power supply sag interacts with the tubes, transformer, and speakers, then at some point I would want to test some real amps to find out if they behave anywhere near what I've imagined the behavior would be. I'd want to find out how much RMS power the Rectifier Super Reverb can deliver long-term before beginning to sag. I'd want to know ballpark figures of how fast it will sag and how fast it can recover. Etc. Sure you could make knobs for all the settings and tweak by ear but it would be nice to know in the ballpark of where the knobs ought to be set to behave like real amps. And nice to know if the knobs which one imagines would even begin to explain what the real amp actually does in the real world.

[guitarzan]

I year ya, JCJR, but developing proper theories is way beyond my pay grade. I think I might have some chance at throwing something together in energyXT using open source and a friends synthedit stuff that gets closer in feel than what's out there now. My real hope is that a discussion and attempt will get some ideas flowing and spark the interests of someone who can do it right... with like, code and stuff

EDIT: sounds like you might have the tube tech that would also be needed to do it right... because, like you said, even if the sim were to use bog standard audio processing modules the envelopes and waveforms would still need to be tweaked into a feasibility realistic shape. In your experience have you seen evidence of slew rate wave shaping or my whole "power amp behaving as a VCA — Envelope Generator" idea? I think it can be made to work anyway, but I would love to know what is really going on with an o-scope.

[guitarzan]

Depending on the nature of the speaker, you might even get some "power sag" effects due to speaker coil resistance change because of voice coil heating.

Different speakers would have different heating and cooling time constants depending on design.

Driven by a clean solid state amp (which is not typically real picky about output load impedance except that power = rms volts ^ 2 / impedance) voice coil heating might be mostly clean power compression.

But a tube/transformer coupling might do funnier interactive things as temp modulates speaker impedance?

The sound of speaker IR's is just too accurate to ignore, so I would want to involve them somehow in any amp sim — but they do need to have a more dynamic feel. I always think of multi-band compression as the answer for better speaker sim dynamic response, but there does need to be more to it than that.

Again, parallel processing would probably be a simple way add a lot of apparent complexity. It could be parallel paired IR's each with their own complimentary multi-band compression settings plus a hint of dfx skidder type randomization before, then maybe summed and a bit of light clipping afterward, or vice versa (clipping first...), whatever works in actual use. There was a vst, I can't remember what it was called, that produced the coolest buzz/rattle distortion that I thought was perfect for scuffing up glassy IR tone in an almost mechanical sounding way — like a hint of speaker buzz when used sparingly.

Maybe parallel IR and algo speaker sim...

There are a bunch of different things going on in an amp that get called sag, but I think it could all generally be simulated by differing degrees of envelope follower in parallel gain stages (so some gain stages have more envelope and others have less or even none, so less overtly "effect" sounding) at different stages of the sim... close enough for r&r, etc.

[Mats Eriksson]

Yes, too many people confuses SAG with compression. Which is easy to mix up, and most people can't and won't tell the difference between them. But tube compression, which can occur at the preamp as well, should not be used to "mimick" voltage sag. It's a different ballgame.

I think Blue Cat Audio were one of the few ones that had discovered the liability of speaker IR's, and - while impressive at first - at prolonged listening binges, they get fatigued and have had second thoughts. I thought this was mentioned in the other Anderton comparison thread. There need to be other things available to more closely track down what happens between speaker and the "flyback" to the power amp transistor, tube or whatever. I want something like ... if you happen to accidentally mismatch the Ohm, resistance. Pluggin a 4 ohm output into a 8 ohm speaker, or 8 ohm power amp output to a 4 or 16 ohm speaker... but without frying any of them later on... the one way works, but not the other. Say 8 ohm output into a 16 ohm speaker will not be as bad as the other way around.

[JCJR]

There is so much compression on most recorded studio or live music, tis doubtful one could find many "pure undiluted" examples of power supply sag on youtube videos or commercial recordings. The quoted examples of Neil Young, Cortez and the other song. I listened to them and PROBABLY the live location recording guy had a safety compressor on the mic, with additional compression on both mix and mastering (and re-mastering) so how the hell one could tell between the amp giving it up versus multiple levels of compressor boxes I don't know.

Just saying one would need to go to the source, amp right there in the room. Make special un-compressed, un-EQ'd, dry as a bone anechoic analysis recordings, in order to evaluate. With such as Earthworks ruler-flat, omni measurement mics capable of insane high SPL without excessive distortion. Maybe there are plenty of such analysis reference recordings available, never looked for such.

Edit: I mean, have heard loud live amps sagging and giving it up. Just never tried to quantify it. And anechoic clean recordings could be real useful even if you can hear it live in yer very own living room, because usually it happens at such insane volumes you can't guarantee what part is the amp giving it up versus what part are yer poor ears giving it up.

[guitarzan]

Bah... let's leave all that to the spice model guys. I'm going to go by ear and response feel, that's what's missing IMO, and that's the only way to get there. I'm starting to see the "project" as a dedicated Tracker amp sim — great feel, quick to adjust and find a ballpark, close enough for rock and roll tone — let it rip, then worry later about virtual re-amping with one of the "sounds great but feels dead" amp clones if desired.

[Mats Eriksson]

[JCJR]

I was just thinking the other day that an old black face super reverb might be a good demo of "nearly pure" sag, especially the ones loaded with jbl d110 speakers.

They didn't have master volume and had tube rectifier power supply. With stock speakers you could get them to rip purt good if cranked loud enough to make yer teeth hurt..

But the ultra efficient d110's, 40 watt amp, 10 watts per speaker. Insanely loud cranked all the way with the speakers still operating in their clean power range.

I foolishly sold mine long ago. Was trying to refresh my memory maybe find YouTube video of a Les Paul plugged into a super reverb cranked to 10.

I found a few examples of folks playing the amp loud but every example either had extreme camera mic auto-level compression, or "mastering compression" on the more polished videos. So I couldn't find a good enough example to refresh my memory.

Maybe it would be less masochistic to look for evidence of sag in such as old fender champs. But might not be the same in a little mini amp. Different circuits etc.

[guitarzan]

I have what amounts to a tweed champ with a Bassman tone stack crammed into a "pedal". Tube pre, tube power amp, output transformer, and a silent speaker load (looks like just a small speaker filled with some kind of silicone rubber looking stuff). No tube rectifier, but it does sag (I think that is it's intended function?). It even has a headphone jack, so there's some pretty good isolated sag reference tone right there in a handy package.

The input jack needs replacing, and by now I'm sure the caps need to be replaced since I haven't had it powered up in a couple years and it is probably at least 35-40 years old. The last real guitar shop around here closed down about two years ago, but I'm sure there must be an amp tech within 50 miles. I really want to get that thing going again. High voltage makes me nervous, though, so I would prefer to take it to a pro for repair.

[The Noodlist]

I have what amounts to a tweed champ with a Bassman tone stack crammed into a "pedal". Tube pre, tube power amp, output transformer, and a silent speaker load (looks like just a small speaker filled with some kind of silicone rubber looking stuff).

Any pictures? Sounds cool.