Simulating reverse log pot

[mark2]

I'm trying to make my pot most responsive at the very top, i.e. most pot travel gives values between 0.75 to 1.

So my first (admittedly naive) assumption is to do the nth root. e.g. POT1 raised to (1/4). Is this possible or should I be trying to approximate it with some form of EXP(LOG(X)) ?
Or other ideas?

[frank]

While this is targeted at FXCore it has POT curves towards the back and the equations for the different curves so this may help: https://www.experimentalnoize.com/manua ... s/an-7.pdf

[mark2]

A little unconventional, but maybe another idea is to use SKP conditionals to divide it into 3 sections. Something along these lines

Code: Select all

rdax pot0, 1
skp zro, ENDPOT0 ; retain zero

sof 1, -0.33   ; bottom third of travel gets 3/4 of values
skp neg, QUICKTRAVEL

SLOWTRAVEL:
; TODO: scale/offset 0 to 0.33 to 0.75 to 1
skp ENDPOT0, gez

QUICKTRAVEL:
; TODO: scale/offset -0.33 to 0 into 0 to 0.75

ENDPOT0:

resulting in this:

Is there a saner way to achieve this goal, or if not which method do you think is preferable? This latter is probably more instructions. For the EXP(LOG) method on the other hand I'll need to think through getting around those 1/16 and 16x factors which (tbd) may require a lot of instructions anyway.

[mark2]

While this is targeted at FXCore it has POT curves towards the back and the equations for the different curves so this may help: https://www.experimentalnoize.com/manua ... s/an-7.pdf

Brilliant, thank you. I'll give that a go instead. Disregard my last post about the linear skipping unless that's a good idea for any reason. Assuming the linear approximation sounds fine and a true curve isn't needed - which I'll guess is the case.

[mark2]

In case anyone comes across this with a similar question, or in case I did something stupid that should be pointed out, here's where am at now. I haven't tested it extensively but at first blush it seems okay.

Code: Select all

sof 0, 0.999
rdax pot2, -1.0     ; (1-POT2)
wrax tmp, 1
mulx tmp            ; (1-POT2)^2
mulx tmp            ; (1-POT2)^3
wrax tmp, 1
sof 0, 0.999
rdax tmp, -1        ; 1-(1-POT2)^3