Fractal Talk

AxeFx III MK II Turbo, A/D Sensitivity of 8.0% has self noise that is equivalent to my JCM800 and 6505, lower than 8.0% is noisier than the real amps.

Why should you care?
Using real boost pedals before the modeler is much quieter than the equivalent digital boost after the A/D conversion, and real boosts need more headroom before clipping the A/D converter unless you already have very hot pickups.
 
Last edited:
Cool he is using new measuring techniques
I wonder if this is next gen behind the scenes
Starting to trickle down
 
Cool he is using new measuring techniques
I wonder if this is next gen behind the scenes
Starting to trickle down
Yeah - really sounds intriguing:

Updated various default values (and hidden parameters) in Amp block based on new analysis methods. For many amps this improves the feel and makes the amp sound more “open”. Existing presets are automatically updated.

I wonder if this helps with what @MirrorProfiles was describing in the low mids & overall darker tonality than the amps they model. FWIW, that was also my take on the Fractal vs amps back when I had my last unit.
 
I wonder if this is next gen behind the scenes
Starting to trickle down
W/o a doubt it has been so for quite a while, is my guess.

Cliff is having such extraordinary epiphanies for the next gen that he has chosen to share some crumbs to pretend he's still working on the III, so to speak ;)

:rollsafe
 
Last edited:
Typically from what I've seen in Fractal-verse, I speculate the new generation of Fractal modelers will be a continuation of current gen and will start off from where the current AxeFxIII is. There will be an overlap of several months like the move from 2 to 3 was because development and refining the algorithms never stops. I also speculate a massive UI overhaul which work on has probably already began.

I love the Amp block improvements, especially when I don't hear of feel the difference anymore when directly comparing with my real amps which is already the case.
Fractals current gen is so good that in my opinion when their next gen arrives it will be a tough call for people to move to it unless it really is a game changer in terms of UI, Routing and Functionality. After all, it is just guitar gear, what more can you imagine you need?

You know what would be cool?
A 'Realistic/Visual' window in the Editor where I can open each block/channel and interact with the individual models like they were the real hardware similar to how VST plugins strive to look like. Yes it's only cosmetics but it does a lot to the perception of the modeler.
Expanding that thought, if Fractal only change the visual side of things like the Editor and UI they can easily put a 'IV' on it and everyone will buy it and be happy.
 
Typically from what I've seen in Fractal-verse, I speculate the new generation of Fractal modelers will be a continuation of current gen and will start off from where the current AxeFxIII is.
I was joking just a little in my previous message, but I think Cliff's original epiphanies are probably "in higher resolution" -- to be pared down even to fit the III that soon may become a Game Boy too... ;) In the next gen he can unleash the original ideas/algorithms in fuller detail, so to speak. The new IRs are an example of my point. Even w/o actual new epiphanies, all these small boosts in dozens of algorithms will probably offer something we want (it's not about what we need ;) ).
And possibly some solid state amps, etc., etc. to pull us over. Otherwise he would not give us the current already amazing "crumbs," IMO. He can't wait to really show off! ;)
 
Last edited:
Typically from what I've seen in Fractal-verse, I speculate the new generation of Fractal modelers will be a continuation of current gen and will start off from where the current AxeFxIII is. There will be an overlap of several months like the move from 2 to 3 was because development and refining the algorithms never stops. I also speculate a massive UI overhaul which work on has probably already began.

I love the Amp block improvements, especially when I don't hear of feel the difference anymore when directly comparing with my real amps which is already the case.
Fractals current gen is so good that in my opinion when their next gen arrives it will be a tough call for people to move to it unless it really is a game changer in terms of UI, Routing and Functionality. After all, it is just guitar gear, what more can you imagine you need?

You know what would be cool?
A 'Realistic/Visual' window in the Editor where I can open each block/channel and interact with the individual models like they were the real hardware similar to how VST plugins strive to look like. Yes it's only cosmetics but it does a lot to the perception of the modeler.
Expanding that thought, if Fractal only change the visual side of things like the Editor and UI they can easily put a 'IV' on it and everyone will buy it and be happy.
When plugins and hardware look better they sound better, its a fact
 
Typically from what I've seen in Fractal-verse, I speculate the new generation of Fractal modelers will be a continuation of current gen and will start off from where the current AxeFxIII is. There will be an overlap of several months like the move from 2 to 3 was because development and refining the algorithms never stops. I also speculate a massive UI overhaul which work on has probably already began.

I love the Amp block improvements, especially when I don't hear of feel the difference anymore when directly comparing with my real amps which is already the case.
Fractals current gen is so good that in my opinion when their next gen arrives it will be a tough call for people to move to it unless it really is a game changer in terms of UI, Routing and Functionality. After all, it is just guitar gear, what more can you imagine you need?

You know what would be cool?
A 'Realistic/Visual' window in the Editor where I can open each block/channel and interact with the individual models like they were the real hardware similar to how VST plugins strive to look like. Yes it's only cosmetics but it does a lot to the perception of the modeler.
Expanding that thought, if Fractal only change the visual side of things like the Editor and UI they can easily put a 'IV' on it and everyone will buy it and be happy.
I'm not buying it really. When you compare a real amplifier under load to any of these platforms, the amp modellers are nearly always too polite, not snarly enough, not offensive enough, and not aggressive enough. They are very good, and Fractal is certainly at the top of the pack, but they're still not as good as a real amplifier.
 
I'm not buying it really. When you compare a real amplifier under load to any of these platforms, the amp modellers are nearly always too polite, not snarly enough, not offensive enough, and not aggressive enough. They are very good, and Fractal is certainly at the top of the pack, but they're still not as good as a real amplifier.
Could very well be, I mostly compare at very low master volume settings so the idiosyncrasies don't get to show up.
Can I interest you in a nice lufs based null test? on second thought, forget about it.
 
I'm not buying it really. When you compare a real amplifier under load to any of these platforms, the amp modellers are nearly always too polite, not snarly enough, not offensive enough, and not aggressive enough. They are very good, and Fractal is certainly at the top of the pack, but they're still not as good as a real amplifier.
I think this sort of stuff is always very case by case.

When I pipe my Mark V and the AM4 model through my Fryette PS-100 into real cabs at a reasonable loud volume, I can get them to sound and feel pretty darn identical.

But it takes work to do so. The right speaker impedance curve, figuring out what settings match your real amp. Especially difficult with the Mark V when the Fractal amp model's GEQ does not work like the real amp.

I think Fractal gets a lot of the nuances right, like I did not expect the master volume to have the same effect to the feel of the amp as on the real amp, and most models default to a much higher setting that on the real amp would be equal very loud band levels.

Output system always matters, and makes things more complex. My Fryette PS-100 will always be a fairly tight, ballsy, big poweramp and it doesn't really sag or loosen up. I feel like I have more control over the feel with the options my real Mark V offers - tube vs diode rectifier, power scaling, variac mode...but luckily most of the time I like a big, ballsy, tight sound so it's not a real issue.

I've even gotten tones out of Fractal that I liked more than the real amp. Or made unusual pairings that I liked as their own thing, like the Class A 30W Brilliant model -> Fryette PS-100 -> 2x12 V30 cab. It has a cool "crisp" to it that is fun for low to medium gain tones.

If you just approach it as "do I like playing this, does this sound badass" thing in a vacuum, then it's easy to say hell yes for Fractal. I feel it takes me less effort to get kickass tones out of the AM4 than other modelers. That's always a good thing.
 
When I pipe my Mark V and the AM4 model through my Fryette PS-100 into real cabs at a reasonable loud volume, I can get them to sound and feel pretty darn identical.
Right... but connect your Mark V to a V30 load .... and then record it under a mic, or better yet by tapping the power-output of the amplifier with a high quality DI box like the Neve, and then try and get THAT sound purely on desktop without any tube poweramp involved... much bigger struggle.

And I'd argue that is the primary use case for modelling.
 
Right... but connect your Mark V to a V30 load .... and then record it under a mic, or better yet by tapping the power-output of the amplifier with a high quality DI box like the Neve, and then try and get THAT sound purely on desktop without any tube poweramp involved... much bigger struggle.

And I'd argue that is the primary use case for modelling.
Couldn't you just use the line-out from the amp? Or is the Neve more about you liking how that DI box sounds?

I wonder if you could use e.g the Fractal mic preamp emulation to get something that tickles your fancy.
 
Couldn't you just use the line-out from the amp? Or is the Neve more about you liking how that DI box sounds?

I wonder if you could use e.g the Fractal mic preamp emulation to get something that tickles your fancy.
The Neve is transparent. Doesn't colour the tone. I can trust with the Neve that when I'm connecting the THRU to my real cab for the load, that the DI signal is getting a true cabless version of the amplifier in full.

I don't trust the line outs on most amps, and if I used the FX loop send as my signal, then I'm missing the gloriousness of the poweramp.

Using the Neve isn't an aestethic choice really. It is a technical one. It ensures that I'm getting the true reference signal.
 
The Neve is transparent. Doesn't colour the tone. I can trust with the Neve that when I'm connecting the THRU to my real cab for the load, that the DI signal is getting a true cabless version of the amplifier in full.

I don't trust the line outs on most amps, and if I used the FX loop send as my signal, then I'm missing the gloriousness of the poweramp.

Using the Neve isn't an aestethic choice really. It is a technical one. It ensures that I'm getting the true reference signal.
According to the Mark V schematic (which might not be entirely accurate), the Mark V 90 slave out is a very simple circuit so it should be about as transparent as it gets too.

Would be interested if you can spot a difference between the two.

Though I understand the Neve is more convenient for you when messing with other amps.
 
Using the Neve isn't an aestethic choice really. It is a technical one. It ensures that I'm getting the true reference signal.
A 25k volume box in parallel with the cab/load into a buffer then into a passive DI box will be the most 'invisible' to the amp, have the flatest response, maintain best impedance ratios and isolate grounds.
 
I think there are still things that can be done , I suspect w next gen super computer CPU you might see even higher resolution ,less aliasing etc lower latency ,even more immediate attack

A new poly pitch / and synth blocks

New Dyna cabs 2.0 which has already started

NAM block w as good or better conversion than anything else currently available ,

Who knows what else but I think there will be improvement to the core modelling , cabs
I don’t think it will just be GUI
 
Hopefully next gen one will have full polyphonic synth guitar stuff and perhaps some weird glitch and unique pedals etc.. Although you can get a lot of weird and glitch and unique things as is in Axe-FX 3 but dedicated things that's made specifically for it would be nice too.
 
Back
Top