Has the Axe-FX II Achieved 99% amp modeling accuracy?

All that tube mojo is so overrated I fn guarantee you I can make the KPA and AxeFx BOTH sound exactly like shit haha

I have a maxim, proven countless times over the past few decades, that says "Anything with more than three knobs has more bad settings than good settings. The challenge is to find the good ones."
 
I have a maxim, proven countless times over the past few decades, that says "Anything with more than three knobs has more bad settings than good settings. The challenge is to find the good ones."
That's because who built that "anything" has set all the other hidden knobs for you ;)
 
Ergo the need to tweak to get them to sound the same. You did see that above, correct?


Yes it is based on the assumption, from anecdotal data, that person after person that has taken blind tests of the KPA with the Amp that was profiled, couldn't tell the difference. Yes, have to take that assumption that that constitutes accurate, and if so, then if one can make the Axe model match the KPA profile, we have a transitive connection.

And, frankly, for me, That is all I care about, does it walk, talk, look, smell, taste etc like a duck, then it's a duck for all practical purposes. When I play it, does it sound and feel like a tube amp? When I record it, does it as well? When people hear me play it, or hear the recording, can they tell the difference? That's All I care about. And they both do

Your level of satisfaction is something different, as regards this. Fine, soldier on. BUT, what criteria YOU hold as the pinnacle of accuracy, holds true, and there is no other way around this, for you. My criteria holds true for me. anything else is just you taking a "holier than thou' attitude to the whole concept and discussion.

And, for what it is worth, your not entirely correct on "KPA is only 100% on target at a single operating point and extrapolates the amps behavior as soon as a knob is moved and there is no guarantee the KPA is 100% accurate since it is just a different digital emulation device." For certain aspects, like gain and volume, there is a sweep that isn't extrapolated, as that was determined as part of the profiling process. EQ, and complete interaction of all controls, yes your correct.

Just so I understand.
You profiled an amp, and then you tried to match the Kemper?
So it's a match of a match?
Error propagation doesn't do justice to the AXE FX there imo. Shouldn't the axe try to match the amp itself?

I'm asking this because I've seldom heard a profile that exactly matches the target amp, which makes me doubtful about buying one, considering that it's purpose is to generate an exact replica of the and set up in a particular way, and it doesn't generate such an exact replica.

Also, the way I see it from your post, you said the Kemper also exactly matches the gain structure, EQ, and control interaction right (non extrapolated sweep)?

Considering that amplifiers have different gain stages, and that to obtain an accurate simulation of the gain structure, you would need to simulate each gain stage separately. How does the kemper manage to get a 100% accurate reading of of the gain structure, if it doesn't analyze each gain stage separately? I mean, that's why SPICE software is used to simulate circuits as detailed as possible, and not just a transfer function that represents the whole circuit.
Component by component simulation will always be more accurate.

How is the interaction between components modeled, if it doesn't model the amp's particular components and their place in the circuit?

How is the EQ modeled with 100% accuracy? I mean, first if all, the Kemper would need to know the center frequency for each knob. Secondly, it would need to know the position of the tone stack. And lastly, because there is no such thing as infinite impedance, moving the tone knobs usually results in changes in the gain structure, which is dependant on everything I have mentioned.

My questions are purely educational, believe it or nor.
Ans my deepest apologies if anyone has felt somehow offended by what I've said.
 
Last edited:
Well, I´m concerned about the initial a/d conversion (and the d/a conversion)

If I plug in the ax8 in the effects-loop of my JVM (HJS)-Marshall. And if I turn all effects off, and sets the gain at unity I still hear and feel a pretty big differense in sound when I switch the loop on and off. I have tried sticking a cable in the loop doing the same test without any difference in sound and feel.

I guess that tells me something. If it can´t send a distorted signal to the return-jack without coloring it how can it capture the input-signal good enough before this major process. That would probably take even more accuracy than being used in an effects-loop.

On the other hand I will probably not stop using the AX8 straight into PA systems anyway because of it´s great possibillitys and it´s great sound and feel.


It depends on what you're playing. This thing is pretty astounding for clean tones, but once you start getting into breakup and high gain territory there's still room for improvement. This is just opinion of course.

It's much debated, but the high and low end on distorted tone still suffers, in my opinion, from a digital sound in top end, and a whompy low end. Also, the way the breakup sounds is still too smooth and undefined. It's gotten a lot better since the last few firmware iterations,to a point where many don't seem to be bothered by it. But I've yet to hear something as saturated sounding as, say "Silvera" by Gojira, "Psychosocial" by Slipknot, or "Vicarious" by Tool. Sure, the tone might be spot on, but the sputtering crackle just isn't there.

That said, running my Axe through the Power Station into a real cab DOES improve these qualities. Still, when I jam with guys they always say there's "something" about the top and and low end that "isn't quite right" even though the rig does sound exceptionally defined.

I guess it's just a matter of modeling being too shiny and perfect. Don't know if we'll ever see it actually "get there". Sorry if that sounds like sour grapes.
 
Well, I´m concerned about the initial a/d conversion (and the d/a conversion)

If I plug in the ax8 in the effects-loop of my JVM (HJS)-Marshall. And if I turn all effects off, and sets the gain at unity I still hear and feel a pretty big differense in sound when I switch the loop on and off. I have tried sticking a cable in the loop doing the same test without any difference in sound and feel.

I guess that tells me something. If it can´t send a distorted signal to the return-jack without coloring it how can it capture the input-signal good enough before this major process. That would probably take even more accuracy than being used in an effects-loop.

On the other hand I will probably not stop using the AX8 straight into PA systems anyway because of it´s great possibillitys and it´s great sound and feel.
How do you set it to unity gain? And are you sure the global eq is flat?
 
You can shoot an IR through an Axe-Fx to measure its color. I might actually do that out of interest today.

I would shortly like to discuss the topic and once again emphasize that we rarely get an apples to apples comparison. Just about every instance of "it's not there yet" is comparing a 4x12 in the room vs a PA system. That is not what the Axe-Fx is modeling. The accuracy is so close that if every band member can tell the difference blindfolded it's most likely not an apples to apples comparison.
 
Well, I´m concerned about the initial a/d conversion (and the d/a conversion)

If I plug in the ax8 in the effects-loop of my JVM (HJS)-Marshall. And if I turn all effects off, and sets the gain at unity I still hear and feel a pretty big differense in sound when I switch the loop on and off. I have tried sticking a cable in the loop doing the same test without any difference in sound and feel.

I guess that tells me something. If it can´t send a distorted signal to the return-jack without coloring it how can it capture the input-signal good enough before this major process. That would probably take even more accuracy than being used in an effects-loop.

On the other hand I will probably not stop using the AX8 straight into PA systems anyway because of it´s great possibillitys and it´s great sound and feel.
It actually tells you that the output/input impedance doesn't play nice rather than the converters. And that for all purposes can be just the amps issue.
When crawl down that rabbit hole every thing makes that proverbial difference that we have to decide if it matters.
 
You can shoot an IR through an Axe-Fx to measure its color. I might actually do that out of interest today.

I would shortly like to discuss the topic and once again emphasize that we rarely get an apples to apples comparison. Just about every instance of "it's not there yet" is comparing a 4x12 in the room vs a PA system. That is not what the Axe-Fx is modeling. The accuracy is so close that if every band member can tell the difference blindfolded it's most likely not an apples to apples comparison.
Yes please!!!
 
I don't think it's an impedance missmatch. The marshall loop sounds transparant with
Strymon, rocktron, and boss units. Those units have an analog direct-signal-chain.
I have allways had a hard time with a/d converted loop effects. Speaking of impedances.
All the above mentioned units, including the ax8 have high input imp and low output imp.

I'm pretty shure that the conversion makes a difference in some way.
 
I set it up so the volume is the same when the loop is in or out.
Yes, I mean what settings and in/out do you use? Consider that there are several ways to have the same volume but those can lead to a different sound in some circumstances. The best way to set it right is to measure the output of the ax8 (all shunt preset) with an external VU meter and compare it real-time with a reference signal. Furthermore, it's better to use fx loop (input 2) on the ax8 since I don't think input 1 can handle a line level signal.
Also check output level in global settings
 
Doesn't surprise me at all that a device with no analog path sounds different in the loop of an amp. The fx send on the amp will have harmonic content well beyond the bandwidth of the converters. You can't hear that but it will change the response of the power amp as those harmonics mix in the power amp and create intermodulation products. They also affect the behavior of the PI and power supply.

The AX-8 is first-and-foremost a standalone processor. When used in the loop of an amp and you want the most transparent sound you should use a mixer along with it. The FX-8 is specifically designed for use with an amp and has a cleaner analog path. No where do we advertise the AX-8 as being intended as a processor in the loop of an amp. I'm sure it will work okay but that's not its intended function. The FX-8 can actually operate with an analog dry path and a digital wet path on the Post I/O but I don't recall if we ever implemented that in the software (due to the fear of support issues w/ people not understanding they have to set the mix to 100%, yada-yada).
 
Yes, I mean what settings and in/out do you use? Consider that there are several ways to have the same volume but those can lead to a different sound in some circumstances. The best way to set it right is to measure the output of the ax8 (all shunt preset) with an external VU meter and compare it real-time with a reference signal. Furthermore, it's better to use fx loop (input 2) on the ax8 since I don't think input 1 can handle a line level signal.
Also check output level in global settings

I maybe should have mentioned that that I allways check the level-matchings before connecting stuff to other equipment. So there was no clipping going on, and there was no extrem low levels going on. I even turned of the noise gate to exclude the possibility of the gate coloring the signal in any way.
 
Back
Top Bottom