Adventures in Accuracy

FractalAudio said:
Spent an hour or so comparing a couple drive pedals. The comparisons are done at two excitation levels: high and low with low being 40dB lower. All knobs at noon.

Here we have the Axe-Fx vs. a Klon:

View attachment 130096
Blue and orange are the Klon.

And here we have another modeler vs. the Klon:
Click to expand...
The day I fired up a drive model in one of those famous modelers I owned at the time, in front of my hand-wired AC15, was the day I also decided to get rid of it. Yes, I could get good sounds out of it, but nothing like in real-world settings or pedal-amp(Vox AC15HW) combinations.

Traditionally, I never favored a Tube Screamer in front of a Vox-type amp, yet the modeler's TS surpassed other typical contenders/drive models when paired with my Vox.

I should also add, my experience is from 5 years ago, I don't know the current state of that modeler, maybe/hopefully it's more authentic these days.
 
I’ve always wanted to try a nobles, excited this is imbound!
I love comparison posts like this, always enjoy reading through them!
 
iaresee said:
Addendum: I dig that you test with a low energy pulse as well.

I love showing people how well the Fractal responds to volume changes on my guitar. More so: I love being able to ride the volume control on my guitars to change what I'm getting out of my Fractal tech and enjoying hearing it change the same way the IRL amps and pedals change when I do this. I've been running with my Telecaster exclusively for the past few months and doing more volume and tone knob manipulation on simpler presets to vary my gain and tone. It's truly wonderful.
Click to expand...

Yep you guys are making great stuff. Absolutely at the top when it comes to modeling. I used my volume in the old days too but more so now. It’s fun just sticking on one preset and 1 amp channel and working the guitar controls.
 
If you are a new customer going into the guitar modelling world….

I don’t see why you would choose anyone other than Fractal Audio Systems!

Look at all these testimonials everywhere in the forum and with the technical graphs to prove their modelling is second to none!

Along with unparalleled technical support here in the forums (from years long members) along with the official support….

You would be…. unwise….to choose other than fractal

-Fractal fanboy

An early thanksgiving thank you @FractalAudio
 
Last edited:
So to understand this how much variation can hardware pedals have?

I assume with the graphs for the other modeller it is comparing it to the hardware that Fractal has and not the hardware it was modeled on? Isn't this a bit like when people say the Fractal model doesn't sound like my amp and the response is - but it isn't modeled on your amp?

Admittedly the Klon comparison looks really out of whack so either the modelling isn't accurate is the or the there is an issue with the hardware itself.
 

Aivoryuk said:
I assume with the graphs for the other modeller it is comparing it to the hardware that Fractal has and not the hardware it was modeled on? Isn't this a bit like when people say the Fractal model doesn't sound like my amp and the response is - but it isn't modeled on your amp?
Click to expand...

Here’s how I’m interpreting the charts and the post:
  • There shouldn’t be a 10 dB difference in output between the original and the other brand's model, just as Klon wouldn’t accept the same difference between two of their units. 10 dB is very audible and is a big difference going into the front of an amp when pushing it.
  • There might be some minor variation between two Klon units and that is expected, just as there is in amps, but it shouldn’t be that extreme, so there’s a major fail in the output and upper end reproduction in the modeling of the other brand. The curve should be following the Klon’s output much more closely if it’s supposed to sound and act like a Klon, and the FX3 is doing that.
  • The upper-end falloff at 5K and above is at the threshold for the guitar’s high frequency range and should be preserved because that’s where the “sparkle” and presence live in what is typically used as a clean boost.
  • The ODR-1 traces for the FX3 are based on code that isn’t finished/fine-tuned and it’s already more accurate.
Again, there might be a little variation between the original units but the quality control should be better than that. A difference of 1 dB is hard to hear, especially in a band situation, but 5 or 10 dB…? That’s audible.
PS- Remember that the Fractal models are based on the hardware and the schematic, and that Cliff will adjust/clean the original hardware to bring it up to the specs defined by the manufacturer in the schematic so that the Fractal version is based on the closest possible “perfect” example. That’s one reason why we see a lot of variation between units other people have and the model in the Fractal, the assembly line substituted “good enough” or “that ought’a work” and Cliff wants exact. Maybe the other brand isn’t going to that point in their modeling and maybe they should.
 
Last edited:
Lol! I kept looking at the first graph on my phone n scratching my head thinking - "what's he talking about, there's only 2 lines here - where's the FAS/Klon comparison?!". Now, a day later I enlarge the image and see that the lines are right on top of one another! 🤣
 
Last edited:
Greg Ferguson said:
PS- Remember that the Fractal models are based on the hardware and the schematic, and that Cliff will adjust/clean the original hardware to bring it up to the specs defined by the manufacturer in the schematic so that the Fractal version is based on the closest possible “perfect” example.
Click to expand...
Does this imply that Cliff will swap out components in the amp if he finds they don’t match the schematic?
 
Brian Angiel said:
Does this imply that Cliff will swap out components in the amp if he finds they don’t match the schematic?
Click to expand...
I'm sure he's had to change some physical components at times because he buys amps from various places and the odds are not good that all of them were working perfectly, but it doesn't make sense that it's necessary to do it on every amp that's modeled. For instance, when he was modeling the Carol Ann amps he was working closely with Alan Phillips and probably had perfect amps to compare to. Conversely, the other day he said he accidentally blew up two transformers when modeling amps so I expect he replaced those. :) So it's not implying that every amp modeled has had physical components changed.

When he has the schematic he can set the values for the modeled components to what the schematic says they should be instead of what they might measure as, then begin measuring the output of the different stages in the target amp and comparing those to the output of the modeler's stages. His methodology probably changes depending on the particular amp and whether he has the schematic and whether he wants the amp modeled so badly that he'll do it without the schematic. He's been doing this stuff a long time so he's probably done about everything that was required to get an amp into modeling condition.

"Amp block" in the Wiki talks about how he models and issues he's encountered, especially the changes that occur in manufacturing.
 
Here's another:

This is a Deluxe Reverb vs. the Axe-Fx:
deluxe_reverb_vs_axefx.PNG

And here's the other modeler vs. the amp:
deluxe_reverb_vs_brandF.PNG

Green and violet traces are the amp.

On the Axe-Fx I set the Speaker Impedance to LB-2 UK. Here's same test but using the Deluxe Reverb impedance curve:
The amp is still using the load box. You can see the impedance has a noticeable effect on the response but not that dramatic.
deluxe_reverb_vs_axefx_2.PNG
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom