MC-202 LFO Delay

#1

Hey there i am trying to reconstruct the envelope of the Roland MC-202 LFO Delay with axoloti.
When you hit a key on a roland mc-202 with lfo delay set to long. After a while the lfo effect on different parameters gently comes in. If you set it to a short delay it goes very fast. ( dark blue signal is short delay time, light blue is long delay time)

mc202_lfo_delay_envelope.jpg891x702 315 KB

on the screenshot of my ltspice simulation of the 202 circuit to do this. you see what happens on a gate signal.
The amplifier of the lfo signal goes quickly to low voltage than it starts soft exponentially and stops soft.
It looks like the curve of limited growth formulas (google it..).
I dont need to have a perfect copy of the 202 circuit but I would like to have this soft beginning and soft stop of the envelope. Also the behaviour of stretching the envelope with increasing delay time would be nice.
I could not manage to find a solution how to do this so far. Can someone help me out a little bit?
thanks alot and all the best,
Flub

#2

Are you trying to emulate the exact behaviour of that especific LFO Delay or just wnat to add Delay to an LFO?
If the case is the later just add a VCA (or multiplier) after the LFO and control it with an AD envelope (Yo may choose one with exponential or linear or inverted exponential behavoiur to suit you taste).

#3

I tried it with AD envelope objects + an inverter. But is another curve. it is only half of the curve. Only the soft end is there but exponential attack or decay curves start straight not like in the picture above.... I am searching a possibility to get soft start and soft end of the curve. But i dont know why at the moment.

#4

The inverter will no change the exponential behaviour of the envelope. There are various objects the let's you choose the curve of the envelope, I'm dont't have my Axo near, but if you search the envelope objects you will find them...

#5

thats exactly the problem I have... i tried all of them and I can only get an envelope which is soft on the one end if you know what i mean ( it seems hard to expain to me )

#6

did you try something like this?

lfodelay.axp (2.5 KB)

#7

i think i get it now...

a sigmoid function:

should be doable as a module on axoloti...

#8

If you want a kind of inverted exponential/exponential (an "S" curve) you could use an inverted exponential envelope an pass it through a tubelike distortion/waveshaper to smooth the end of it...

#9

Hey, thank you very much for the ideas. I am bit busy now but I ll concentrate on this later to see if it is possible with the distortion.

It might take to much cpu power. actually but perhaps not...
It perhaps its just the right thing to try!

#10

I tried to understand how to translate the formula ( f(x) = x / (1 + abs(x)) ) to fixed point math but I dont really understand it.

I read about it here https://sebiik.github.io/community.axoloti.com.backup/t/coding-axoloti-objects/2606 but it is not clear to me after reading the fixed point math section. I dont know what a magnitude is in this context and also not why logarithms are important.
What I think i understood is that the comma is fixed between bit 27 and bit 26 like this
bit 31 -> 0 0 0 0 0 POINT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 <- the right being bit 0
so 1 is 0 0 0 0 1 POINT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
and the maximum number in decimal is 1 + 2 + 4 + 8 + 16 = 31.99999999999 or something
so to calculate the formula in fixed point i have to take the inlet which has the type frac32 ( what is this? ) and ...
Can you help me with this?