• We would like to remind our members that this is a privately owned, run and supported forum. You are here at the invitation and discretion of the owners. As such, rules and standards of conduct will be applied that help keep this forum functioning as the owners desire. These include, but are not limited to, removing content and even access to the forum.

    Please give yourself a refresher on the forum rules you agreed to follow when you signed up.

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

ModelGuitarist

Inspired
Interesting - guess we'll have a whopper of FW friday coming up.
That would be a dream, but I doubt this would be implemented any time soon.

There's really no details if it is for the III (I would imagine) or if it's only available through the MkII, or even a future unit.
 

artzeal

Experienced
I wonder if this new feature will also be usable to match drive pedals, I'd personally be more interested in profiling those rather than amps
Give me your tired Fuzzface, your Klon, Your huddled masses of distortion pedals yearning to breathe free, The wretched refuse of your teeming effects collection might yet be reborn to see the gilded stage...
 

ModelGuitarist

Inspired
I wonder if this new feature will also be usable to match drive pedals, I'd personally be more interested in profiling those rather than amps
From what I understand, this only takes preexisting models in the FAS unit, and refines them to the input and output filters of a particular amp of similar design.
So you could have YOUR '65 Deluxe, YOUR Matchless Chieftain, or (if you can get your hands on it) Bonamassa's Twin Reverb, as opposed to the stock FAS models. But it has to be based on something already modeled, not an amp unknown to the FAS unit. That's my understanding but I don't know shit TBH.

Sorta like a Kemper, but better because it's a Fractal process. :D
 

DLC86

Fractal Fanatic
From what I understand, this only takes preexisting models in the FAS unit, and refines them to the input and output filters of a particular amp of similar design.
So you could have YOUR '65 Deluxe, YOUR Matchless Chieftain, or (if you can get your hands on it) Bonamassa's Twin Reverb, as opposed to the stock FAS models. But it has to be based on something already modeled, not an amp unknown to the FAS unit. That's my understanding but I don't know shit TBH.

Sorta like a Kemper, but better because it's a Fractal process. :D
From what I understood instead, it doesn't have to be necessarily the same exact amp as the modeled amp, you just select the model that most closely resembles it. And the variety of amp models in the axe fx covers pretty much every existing topology.

A quote from the patent:

"The user can invoke a software utility or “Wizard” in the digital modeling system 112 that guides the user through the process by instructing the user how to connect the digital modeling system to the real guitar amplifier 100 and then instructing the user to select an amp model that most closely resembles the real guitar amplifier 100"

Furthermore his previous patent talked about matching the non-linearitiy too, so the actual implementation could be a mix of both patents.

We'll see (hopefully soon) 🤷‍♂️
 

ModelGuitarist

Inspired
From what I understood instead, it doesn't have to be necessarily the same exact amp as the modeled amp, you just select the model that most closely resembles it. And the variety of amp models in the axe fx covers pretty much every existing topology.

Much better said. Yes.
 

boyce89976

Inspired
After the new release of Mooer and the next one of Neural Dsp (Quad Cortex showed at Namm and expected for September 2020), it now seems that it is also possible for other manufacturers to be able to profile an amplifier, exactly as the Kemper does since its debut, even if In the past it seemed that there was a strong copyright on "profiling", impossible to overcome.

If ever Axe Fx could ever do it, it would be amazing.

What this community think about?
🙂
I had a Kemper for 2 years. Axe III is better in EVERY CONCEIVABLE WAY. Seriously, profiling is NOT better technology. In fact, there are too many disadvantages of profiling to list, such as having to wade through hundreds, if not thousands of profiles just to find a few you like. And, you have to find a profile you like at each gain level you would use, for each amp you use.

Not to mention the fact that Kemper has said there are only one or two very complex profiles in the Kemper that all profiles are based on... Not what I would consider "leading edge tech" that Fractal should copy.

Axe III models the amp with component-level detail, meaning the model reacts just like the amp it is based on. Plus, you get Fractal's experience and upgrades with controls the real amp doesn't have... expanding the capability of the amp if you'd like. What could be better than that?
 

Casper

Inspired
The hotplate also lets you use an LED at the front to do... something. Can't remember what it's supposed to do but I never noticed any difference. I also never really got the fan to spin either :(


The manual says that the LED is part of a noise reduction system to reduce hum and hiss. I have the 16 ohm version that stays connected to a 16 ohm cab. The light glows, and the fan spins, seemingly in proportional intensity to the amp output.
 

boyce89976

Inspired
I'm guessing that is construct a virtual amp from bits of others that you like. I know you can't tell us.
Can't you do that already with Axe III? I mean, you can borrow EL34's and some Marshall power amp settings for your Matchless. What else did you have in mind?
 

Fractalist

New Member
Ok, I juste saw this morning a dude on Facebook who just made me just cancel my preorder for the Quad Cortex.
He shared a research article from the guys who actually work at Neural DSP claiming their neural learning and network system has a HUGE issue = it generate A LOT of aliasing. That’s exactly what people is pointing out on YouTube, some harshness added to capture algorithms running the profiles VS real amp.
it’s very noticeable and it’s something I don’t want, it’s ear tiring and brain just don’t like it.
So I’ll wait for the « new patent » of Cliff. Sooner or later something as good as the Kemper will come from FAS, I’m pretty confident.
 

FractalAudio

Administrator
Fractal Audio Systems
Moderator
Ok, I juste saw this morning a dude on Facebook who just made me just cancel my preorder for the Quad Cortex.
He shared a research article from the guys who actually work at Neural DSP claiming their neural learning and network system has a HUGE issue = it generate A LOT of aliasing. That’s exactly what people is pointing out on YouTube, some harshness added to capture algorithms running the profiles VS real amp.
it’s very noticeable and it’s something I don’t want, it’s ear tiring and brain just don’t like it.
So I’ll wait for the « new patent » of Cliff. Sooner or later something as good as the Kemper will come from FAS, I’m pretty confident.
If you want profiling or similar buy a Kemper. I steadfastly believe white-box modeling is superior to black-box approaches.
 

yeky83

Power User
Ok, I juste saw this morning a dude on Facebook who just made me just cancel my preorder for the Quad Cortex.
He shared a research article from the guys who actually work at Neural DSP claiming their neural learning and network system has a HUGE issue = it generate A LOT of aliasing. That’s exactly what people is pointing out on YouTube, some harshness added to capture algorithms running the profiles VS real amp.
it’s very noticeable and it’s something I don’t want, it’s ear tiring and brain just don’t like it.
So I’ll wait for the « new patent » of Cliff. Sooner or later something as good as the Kemper will come from FAS, I’m pretty confident.
To be fair, they say in the very same paper that oversampling can be used to lower aliasing, and they just hadn't done it by the time of their research paper. I dunno if I hear aliasing cus I dunno how to identify it by listening, but I don't like the IRs they use in demos, very harsh.
 

FractalAudio

Administrator
Fractal Audio Systems
Moderator
To be fair, they say in the very same paper that oversampling can be used to lower aliasing, and they just hadn't done it by the time of their research paper. I dunno if I hear aliasing cus I dunno how to identify it by listening, but I don't like the IRs they use in demos, very harsh.
Except that it barely runs in real-time without oversampling. Once you oversample it won't run in real time.
 

yeky83

Power User
Except that it barely runs in real-time without oversampling. Once you oversample it won't run in real time.
Ah, gotcha. I didn't put two and two together when you'd mentioned it.
Well... that sucks. I guess it's got Kemper-like aliasing performance even with much greater processing capability. Sounds like it's algorithmically worse in that regard, yay for "neural" processing.
 

PdotCdot

Member
“...obtaining an amplifier output filter and an amplifier input filter for the real guitar amplifier, using the amplifier output filter and a model output filter to obtain a corrective output filter, using the amplifier input filter and a model input filter to obtain a corrective input filter, and applying the corrective input filter and the corrective output filter to the digital amplifier model.”

Just a taste of this word salad would kill the mightiest of tube amps known to humankind.
 
Top Bottom