Is arrived the time for Axe-Fx III to profile an amp?

Status
Not open for further replies.
Yup, people are too hung up on recreating rather than creating. They scrutinize audio clips and say "the model has a little less bass than the real amp". So turn the bass up.

Or "I really need a model of that Death Monger Seal Clubber XT500 amp, I can't bring teh brutalz without it." For chrissakes it's probably just a tweaked SLO 100 like every modern metal amp ever made. Or I spend months modeling it and "eh, it's okay but I still prefer the Recto models".
OK you need to name the next FAS amp model "Death Monger Seal Clubber XT500".

That's comic gold right there.
 
I've had my eye on a Mesa DC-5 Dual Caliber and found a guy that had two of them. He was kind enough to let me take both of them back to my studio and choose which one I wanted. They are COMPLETELY different. One has a lead channel like a fire-breathing dragon and a very Fendery clean channel. The other one has half the amount of gain on the lead channel and the clean channel is spitty and harsh. It just goes to show how different amps of the same type can be, . . .
This, this, this! To me, says more than everything else posted here.
 
I think the only thing that could satisfy the "I want to sound like 'X'" users would be the ability to capture the sound of a recording; for example, as an extreme, the tone Andres Segovia's 1945 Albeniz/Granados album using a PRS electric.
 
I have something even better. Waiting for my patent approval.
Is it this one? Filed in Feb., published in Sept. :)

"Amplifier matching in a digital amplifier modeling system"

Looks like I don't have enough posts to qualify for URLs yet, but for others, paste this into your browser (inside the quotes) for a deeper dive:
"patents.google.com/patent/US20200279546A1/"
 
To follow up.. Looking through the patent application -- and trust me, I am nothing close to an electrical engineer. As I understand it in a simplified form, this proposed process involves capturing a low-level input filter and high-level output filter from a physical amplifier using various exponential sine wave sweeps. Then those captured filters get applied to the input/output filters of a digital amp model to produce corrected input/output filters. Now you have a corrected amp model that responds like your specific physical amp.

All that extra space in the MkII would be a great place to store these corrected filter sets.
 
Last edited:
To follow up.. Looking through the patent application -- and trust me, I am nothing close to an electrical engineer. As I understand it in a simplified form, this proposed process involves capturing a low-level input filter and high-level output filter from a physical amplifier using various exponential sine wave sweeps. Then those captured filters get applied to the input/output filters of a digital amp model to produce corrected input/output filters. Now you have a corrected amp model that responds like your specific physical amp.

All that extra space in the MkII would be a great place to store these corrected filter sets.
This initially sounds like Tone Matching, which has existed for quite a while already... But checking the link, it seems to be a "corrective" technique to be applied to existing amp models to make sound them more accurate.

https://patents.google.com/patent/US20200279546A1/
 
This initially sounds like Tone Matching, which has existed for quite a while already... But checking the link, it seems to be a "corrective" technique to be applied to existing amp models to make sound them more accurate.

https://patents.google.com/patent/US20200279546A1/

My understanding of the process described in the patent, that it is very similar to the description of the MIMIC Technology in the white paper.

https://www.fractalaudio.com/downlo...actal-Audio-Systems-MIMIC-(tm)-Technology.pdf
 
Last edited:
That’s why many of us also enjoy synthesizers, so we can chase those 80s synth tones from 30 years ago lol

I don’t know about “new” sounds though.... last time someone tried that we all had to suffer through Dubstep for a few Years
No shit. Nothing beats the sound of an old Oberheim.
 
My understanding of the process described in the patent, that it is very similar to the description of the MIMIC Technology in the white paper.
Yea, it does look similar to the section in the whitepaper on pg.8 "Frequency Response Deviation":

"A complex series of tones is injected into the amplifier being simulated and the output is measured. These measurements then form data sets that are stored with each model and correct the various filters."​
Interesting. With MIMIC, the analysis was done by Fractal and included in the firmware. Now this capture process would be end-user accessible, there's use of a "wizard"/guide in the Axe interface described in the patent. So perhaps this is an improvement to the process? Or just a refinement that allows non-Cliff laymen to run this analysis easily at home?

I'm looking forward to seeing what comes of it.
 
What is the difference between this and Kemper's profiling process?
Only Cliff can give a definitive answer.

What I've understood from a quick read is, it seems, Cliff's amp matching is able to determine the frequency response of the amp both pre and post distortion separately, and the way it uses the harmonics generated by the amp's clipping to "whiten" the test signal is genius (I don't know if that's a Cliff's invention or a common technique).
That may be one of the differences with the Kemper.

Another difference might be not having an automatic model selection but adapting the existing model (selected by the user) parameters so that the clip shape matches the real amp. Even though in a paragraph the patent talks about Volterra kernels which I think is a different approach (the same one used for example by Acustica Nebula plugin), so I'm not sure about that.

Anyway, I think what the Kemper does is just selecting a topology and matching the amount of gain and the post eq
 
Last edited:
I re-read Kemper's patent and it seems I was wrong here:
What I've understood from a quick read is, it seems, Cliff's amp matching is able to determine the frequency response of the amp both pre and post distortion separately, and the way it uses the harmonics generated by the amp's clipping to "whiten" the test signal is genius (I don't know if that's a Cliff's invention or a common technique).
That may be one of the differences with the Kemper.

Even the kemper is able to determine pre and post filtering separately and with a similar process apparently:

IMG_20201010_114532.jpg

So the difference might lie on how the characteristics of the non-linearitiy are detected and reproduced.
And if the axe fx still uses its amp models to generate these "non-linearities", the following is certainly a difference with the kemper:

IMG_20201010_115419.jpg

We know for sure that real amps have "memory" due to sag, recovery times, etc.. so they (just like the axe fx) don't always deliver the same output value for the same input value.


PS: here's Kemper's patent
 
Last edited:
I know we're way off topic from now, but to address the OP - we already have everything in the Axe to "profile" amps, it just requires us to manually select the underlying model before tone matching. Since there is component based modelling in the Axe-Fx the results are, IMO, far more accurate than a "profile" as we can set our initial conditions much closer to begin with.

I outline a few different strategies in this video.


Thanks Leon! Really good and informative video.
I have a question. Can you use the line out of a THD Hotplate for the mic + DI method of capturing irs?
 
Status
Not open for further replies.
Back
Top Bottom