zenaxe
Fractal Fanatic
I have to agree with one thing in particular: with every firmware update the modeling is getting more and more real and with such radical changes to a given amp model's tone throughout the firmwares this is a little confusing. We tend to say for years that the Axe amp modeling is spot on - many people agreed with it, tried it next to the real thing and swore that it's just the same, etc. and now FW18 is even more realistic. So what's the truth, then?
One problem here is that there are so many lay people arguing this topic who have no idea what it even means to measure the accuracy of a system like this or the subtleties involved in modeling a specific circuit much less providing a general framework for it.
As a result, people seem to think there is some underlying "it is accurate" origin point that all the other models flow. This is a terrible assumption and generally all further discussion goes off the rails because there is this binary it is either 100% accurate 100% of the time for 100% of the inputs for 100% of the modeled characteristics (remember we are modeling more than one thing non-linear distortion effects, frequency response, transient response, etc) or it is "not accurate" sort of view.
In fact, the framework and modeling may be accurate to the point of no meaningful differences with respect to certain circuit configurations or behaviors/characteristics and partially or noticeably off in other respects/effects/configurations at any time during the system's evolution. Addressing these inaccuracies in no way diminishes the existing accuracies, yet people seem to think it somehow does.
Some models, setting configurations, or listener preferences might be wholly covered by these essentially fully realized operating points in the current system; whereas others will benefit noticeably from incoming changes.
In some cases the measurement device (for the end user, my ears and fingers) will not be able to discern any difference because even though it HAS improved from a circuit behavior accuracy perspective from the way I dial, listen, and play, nothing big has happened. Witness the V18 (game changer or meh posts).
The only way to know FOR SURE is to measure things in a non-objective electrical sense. And this may not always be an easy thing to do. But, that's WHAT CLIFF DOES. That's why I dig his design process and like rolling with these changes and feel this unit is really like owning a zillion great amps but still look forward to it being improved. I have faith in this process. That's why generic distortion simulations and filter stages with feedback loops are not what I'm looking for in a modeler. So, when I have to redial, I believe what I am getting is more accurate and worth the effort. In some cases it is subtle or a no-op other times it feels like a leap.
Most people do not get this. I have worked in circuit simulation in some capacity or another for 20+ years so it is easy to accept the 'it is accurate AND ALSO will continue to improve and in some cases you may or may not need/notice that improvement' thing. If you are looking for a system that has 200+ totally indistinguishable dead bang across every single operating point under all listening conditions for all players amp models, I think you might need to have your OCD checked or start saving because all those amps are going to cost a ton.
Last edited: