Cherry Audio Forums

I would like something : I searched in the Forum, I tried to understand the modules explanations filtered with "Glide" and "LFO", but I didn't found an answer.

let's take the random output of an LFO, send it to the input of a glide module with constant response option. Let's feed the glide output into an oscilloscope, or anything voltage controlled.
Let F(LFO) be the LFO frequency and T(Glide) be the Glide time

Here is what I would like: a way coupling F(LFO) and T(Glide) with the formula T(Glide) (ms) = 1000 / F(LFO) ( Hz), and therefore
If F(LFO) = 2 Hz, T(Glide) = 500 ms,
if F(LFO) = 1 Hz, T(Glide) = 1000 ms,
if F(LFO) = 0.8 Hz, T(Glide) = 1250 ms,
if F(LFO) = 0.2 Hz, T(Glide) = 5000 ms, and so on, to obtain a variable sawtooth :



You'll tell me: make your patch, adjust your frequency and your glide time ! Yes, but I would like to vary the two, in order to increase the randomness of the stuff... But they are not voltage controlled.
Does anyone knows a module that does this ? My knowledge of English language is too weak to fully understand what certain third-party modules do.

Thanks in advance to help me.

If I understand correctly then, what you are looking for is a glide module with voltage control of glide time.

An alternative approach is to use a VCF. This won't give you exactly the same results but may do the job...

ColinP wrote: ↑Sat Sep 09, 2023 3:30 pm If I understand correctly then, what you are looking for is a glide module with voltage control of glide time.
An alternative approach is to use a VCF. This won't give you exactly the same results but may do the job...

Thank you very much for the suggestion, ColinP, I hadn't even thought of that! And yet, I am a retired neurologist, and I know the consequences of too intense filtering on electroencephalogram signals, in particular the damping of the curves and the blunting of the spikes !
Due to the limitation of the filter slope to 24 dB/Oct Max, if the LFO frequency drops below 3 Hz, even with a cutoff frequency of 0.000 Hz, and even when combining low-pass and high-pass in series, the time-constant is too large and the curve reaches the target value too early. I want to be able to go down to very slow frequencies, with variation times of a minute at least, therefore much more than the 5 seconds of the CA Glide module. I would have to look into the Module designer, but it doesn't work with me: impossible to compile anything. I must be too old now!

But maybe Cherry Audio could implement a Glide Amount Control :

Before : And after :

You might look into one of Benard's Sloth LFO's, which can give you really long periods. I've also got something that can do that in the works, but it's a month or so out from release. I suspect that Dome's Big Rat might be able to do it, as well.

Follow-up: that module is now out, and is temporarily on sale. It's called Quadrature VLFO.

Reid

Jean d'Oran wrote: ↑Sat Sep 09, 2023 11:17 pm if the LFO frequency drops below 3 Hz, even with a cutoff frequency of 0.000 Hz, and even when combining low-pass and high-pass in series, the time-constant is too large and the curve reaches the target value too early. I want to be able to go down to very slow frequencies, with variation times of a minute at least, therefore much more than the 5 seconds of the CA Glide module.

Yeah, if you want slews of a minute or more it gets tricky.

You can however extend the slew time by patching multiple low-pass filters in series but it's hardly elegant. In the patch below the LFO is running at 1 Hz.



The Ladder Filter uses a different filter architecture and the effective cutoff can get lower. But there's a lot of instability and phase lag. Still that might not be a problem.

In the patch below the LFO is running at 0.25 Hz.



I've not tried running the Ladder Filter in series. Not sure if that would work but I suspect it'll get pretty unstable. Still if all you want is a slowly evolving random wave then it might be worth investigating. Unfortunately I'm short on time.

Jean d'Oran wrote: ↑Sat Sep 09, 2023 11:17 pm I would have to look into the Module designer, but it doesn't work with me: impossible to compile anything. I must be too old now!

You could start a thread in the Module Designer part of this forum and I'm sure someone will walk through any VMD issues with you.

A slew limiter is pretty easy to code so it might be a good starting project once you get past the building issue.

We recently discussed how the SmoothValue class might work and that's a similar application.

Well, the closest I could get using only Cherry Audio modules is this:
The idea behind it:
- use performance knob to control "Frequency" on LFO and "Amount" on Glide module at the same time,
- use min, max and curve settings to adjust CV mapping
- remotely control performance knob using CV

This is not a 100% solution, but you probably can get arbitrarily close by tuning the min, max and curve settings accordingly.

To actually glide to a given random value and generate a new one once it is reached, you will need some form of value comparison. If you own Comparator and Random Seed you could easily set this up (see this thread) and use Remote Control on the Glide Amount for CV control over the rate/speed. Since Glide Amount is the limiting factor here, you cannot get slower than 0.2 Hz (or 5 seconds glide time).

Jean d'Oran wrote: ↑Sat Sep 09, 2023 11:17 pm But maybe Cherry Audio could implement a Glide Amount Control :

Out of curiosity, I used Remote Control to add CV control to Glide. It doesn't increase the max glide time beyond 5 sec, no matter where I set the initial amount. The limitation seems built into Glide's code.


Andrew Macaulay has a slew processor that maxes out at 12 sec. At least that's better than 5

https://andrewmacaulaymodules.com/help-slew-processor/

Thank you, thank you very much to each of you for your help, your suggestions, your advices, your contributions, all these things that I had no idea, that enrich me and allow me to move forward!
I found something: you'll need :
- An LFO Module
- two Glide Modules
- a DC Source Module
- A Sample and Hold Module
- an Amplifier Module
- two Formula Modules.

Send in the input X of a formula module the random output of the LFO and in its input Y an output of the DC Source.
The formula will be: ((5+X)/2) + Y
Thus the S&H values will be limited between 0 and 5, with a second control by the DC.

Send in the input of the first Glide module the output of this Formula module, with the Amount of the Glide at 5000 ms, Response Con in order to dampen the changes in value.

Send to the CV Freq of the LFO, with the Potentiometer at -100%
- the exit of the Glide
- the other DC Source.
Set the LFO frequency to 1 Hz

So the frequency of the LFO will change at each new cycle, randomly between 0.03 and 1.0 Hz. By putting the DC which enters the formula at 5 V, a cycle may last 12 minutes
By adjusting the value of the second DC amount, we can adjust the frequency of the LFO and lower it very, very, very low, since with the two DC amount at 5 V, a cycle lasts approximately 50 minutes, with the LFO at 1 Hz.
So I'm afraid of what can happen if we set the LFO frequency to 0.002 Hz!

Good
It's a bit sad this LFO with constant amplitude, isn't it?
Send the square wave of the LFO Module as a trigger for a Sample & Hold module, (remember to switch to “External Source!)

Send its output into the X of a second Formula module, whose formula will be (X+5.1)*1.5

Send the output to the input of a second Glide module, with the Glide Amount at 5000 ms, and Response Con so as to absorb value changes.

Send the output of the Glide Module into the CV amount of an Amplifier module, connect either the sine or the triangle of the LFO to the input (and why not both?)
and modulate wath you want, randomly and smoothly with the output of the amplifier !



Okay ! now that this is done, I'm gonna explore what you suggested, that I did not know :
Big and Smol rates, Sloth LFO, comparator and random speed, slew processor, Remote control !
Smothing values fear me !

When I can get the module I mentioned out, it'll go down to a three-hour period.

It's out, and it does. Quadrature VLFO.

Reid