Ooooh... Charts and Graphs Redux

d/dx of the integral of f(t)dt from 0 to t (that's hard to type out) is just f(x) isn't it? What's the benefit of doing it that way over just using f(x), does it sort of "normalize" the data?
You use the anti-derivate OF THE FUNCTION to do the nonlinear processing. Then take the derivative OF THE SIGNAL. This suppresses aliasing products. Of course you need to figure out the anti-derivative of the function which can end up being a much more complex function which then defeats any gains you get by lowering the oversample rate.

It also means that you are using a waveshaper in the first place. In our case the functions are essentially dynamic so this wouldn't work.
 
Looking at the frequency response graph in post #131 my first thought was that it is intentional. It looks very similar to what Audyssey and Yamaha YPAO room correction does when you engage Audyssey Dynamic EQ or Yamaha YPAO Volume. Both of those products cut/roll off the high end some and boost the low end. In this case product B appears to be boosting at the lower end of the guitar range and cutting/rolling off the high end.

Is this an effort and design decision on their part to provide a little more of a fuller sound and remove some of those high frequencies that most will cut anyway? Maybe they feel that an exact tone print of an amp model is not the most welcome tone that their target audience wants to hear.

I noticed this in Guitarjon's most recent comparison of the Peavey 5150-Helix-Axe3 comparison where I thought that the actual amp and the Axe3 were for the most part identical but the Helix version had something added down low.
That was an excellent video. The tones were close in the end, but most interesting to me was when Jon at Sonic Drive Studio revealed the settings at the end. The Fractal amp settings were very close to the real but the Helix was much more manipulated. For the Helix, the presence was at 0, compared to the real on 5. I always found the same. Jon is a pro.
 
Last edited by a moderator:
You use the anti-derivate OF THE FUNCTION to do the nonlinear processing. Then take the derivative OF THE SIGNAL. This suppresses aliasing products. Of course you need to figure out the anti-derivative of the function which can end up being a much more complex function which then defeats any gains you get by lowering the oversample rate.

It also means that you are using a waveshaper in the first place. In our case the functions are essentially dynamic so this wouldn't work.
This is absolute gold. I love learning from the man behind the machine.
 
AxeFxIII does a great job with aliasing. Is it the same with other products (FM series)?
As many dudes on YouTube claim they have the same quality I guess they have the same near perfect aliasing reduction?
Is there some room for improvement? Helix did an update with « less aliasing » and I found it was a little bit better but still those harsh highs…
Also the guys of ISP tech were claiming their product had zero aliasing (theta pro or something like that)… but had zero success apparently…
BTW my AxeFxIII mkII sounds good.
 
Yeah I recall just enough Calculus 1 and 2 to follow the conversation. I always appreciated calculus as a beautiful breakthrough in mathematically describing complex systems, a triumph of human thought.

Just don't ask me to do it. I'd get the wrong answers every time.

FAS is probably the most compelling and fascinating use of math and I'm glad to kinda understand there is a real method behind identifying and solving the problems that were barriers to this technology.
 
Here's another interesting one. Frequency response comparison. 5150. All knobs at noon except MV at 9:00, Presence and Depth at zero to minimize influence of speaker impedance since Product B's speaker impedance isn't user adjustable.

Green trace is amp. Purple is Axe-Fx. Blue is Product B.

View attachment 89082
That's fascinating - the III's right on top of the real amp's curve.

Any idea if blue line mystery product's deviation in the 100Hz range is a function of the model itself being off, vs the taper of the "pots" not matching the taper used by the real 5150?

I realize that figuring this stuff out by twisting knobs and re-measuring is stupidly time consuming, but maybe you've already collected the data. :)
 
Back
Top Bottom