The CNFB Method

Only the Amp block on Axe Fx III at the moment.

I'm thinking we may see it the Drives (at least some) but that is pure speculation and Cliff has not commented on any specific applications.
True.

I'm trying hard not to read too much into the recent drive pedal comparisons.

Which didn't in fact indicate whether they were current shipping firmware, or CNFB experiments ;)
 
Last edited:
Only the Amp block on Axe Fx III at the moment.

I'm thinking we may see it the Drives (at least some) but that is pure speculation and Cliff has not commented on any specific applications.

Given that CNFB is also applicable to diodes/non-linear solid state devices, etc., I think it's quite reasonable indeed to speculate that will eventually be the case. I would also think that the method will be applied/ported to the preamp modeling in a forthcoming release so the entire amp uses the new techniques.
 
I've finally perfected the "Chase Nonlinear Feedback" (CNFB) method for the modeling of nonlinear networks. And it works amazingly well. Has the accuracy of high-order integration methods with less computational burden.
...
Could be a revolution in nonlinear network modeling.
patent applied ?
 
Only the Amp block on Axe Fx III at the moment.

I'm thinking we may see it the Drives (at least some) but that is pure speculation and Cliff has not commented on any specific applications.

Also very much speculation, but I doubt that the same kind of gains can be made in minimising processing "real-estate" while maximising model accuracy, so won't apply quite so much for Drive modelling. Compared to a tube amp and its interaction with a speaker circuit, the important bits of an overdrive pedal's characteristics must be a fair bit simpler. So I agree we'll probably see it, but not sure it can be that far up the priority list. We already have a lot of Drive models, and their behaviour is already all kinds of awesome.

Liam
 
Also very much speculation, but I doubt that the same kind of gains can be made in minimising processing "real-estate" while maximising model accuracy, so won't apply quite so much for Drive modelling. Compared to a tube amp and its interaction with a speaker circuit, the important bits of an overdrive pedal's characteristics must be a fair bit simpler. So I agree we'll probably see it, but not sure it can be that far up the priority list. We already have a lot of Drive models, and their behaviour is already all kinds of awesome.

Liam
You may be right that that's not that much to gain from applying this tech too drives blocks, but I hope you're not. Drive blocks do use quite significant amounts of cpu.
 
Also very much speculation, but I doubt that the same kind of gains can be made in minimising processing "real-estate" while maximising model accuracy, so won't apply quite so much for Drive modelling. Compared to a tube amp and its interaction with a speaker circuit, the important bits of an overdrive pedal's characteristics must be a fair bit simpler. So I agree we'll probably see it, but not sure it can be that far up the priority list. We already have a lot of Drive models, and their behaviour is already all kinds of awesome.

Liam
However, I believe the work Cliff did was only the power tubes, not all the various other bits.

If this can replace the existing diode modeling with 60% of the current requirement, that's a big win.

As @Dave Merrill mentioned, the Drives can be CPU hogs and I think the diode-based are the most intense.

And again, this is my interpretation/speculation ;)
 
Back
Top Bottom