I want to control these parameters with sequenced voltage levels. So instead of controlling the pitch of an oscillator I want to control the position of the Bias, Spread and Steps etc in the same way. It doesn’t have to be quantized CV. I know you can send stepped random into these inputs. I thought sending sequencer values would be similar only not random and more controllable. Does this make more sense?

A practical example would be if I wanted the bias to go up on every other bar. Or step up a little bit on each bar and then back again.

Ah, I see, I didn’t understand why you called these V/O CVs…

Yes of course, all these parameters can be controlled by a sequencer.

When a knob is at 12 o’clock, a voltage of -2.5V will turn it fully CCW, and a voltage of 2.5V will turn it fully CW. Maybe the voltages you were sending were outside of this range, causing the knobs to be stuck at extreme settings?

1 Like

OK those voltages are extremely helpful! I’ll try to measure what I’m outputting and find the proper range and then stay within those limits.

And the knob acts as an offset: if the knob is fully CCW, a voltage of 0V won’t change anything, and a voltage of 5V will turn it fully CW. If the knob is fully CW, a voltage of -5V will be needed to turn it fully CCW.

1 Like

Ummm silly question maybe (cause who would want to stop it) but is there a more elegant solution to stopping Marbles as opposed to a dummy cable into Clock?

There is no way, the module runs continuously.


Id would be great to have an alternative firmware in which the the “steps” control spams the quantized options through the whole knob (from min/7 o’clock to max/5 o’clock). For many of us we use it exclusively in quantized mode so modulating this control could be much more precise.

3 posts were split to a new topic: Marbles: gate outputs’ shape

The explanation for X Mode 3 in the manual is incorrect.

It’s the other way round (X1 is opposite and X3 follows the panel) like the icon depicts.

1 Like

Hello! New to Marbles and having a small issue:

With the “E Button” solid orange, the clock has always been solid. Just now after updating my firmware to 1.2 I decided to try pressing and holding the E button so it blinks orange, and it appears to be having some timing issues at the most extreme settings of 5 o’clock and 7 o’clock.

My understanding is that in this mode, at 5 o’clock, for example, t1 is getting a constant division of 8 and t3 is getting a constant multiplication of 8 relative to the system clock (t2). So there is effectively a pattern that loops every 8 clock cycles. The issue seems to be that every time it wraps back around to the beginning, the clock stutters. This appears to be happening whether I am clocking Marbles internally or externally, and the problem only seems recognizable when RATE is 12:00 or to the left of 12:00.

If you need a video of the issue I’d be happy to post a link…hoping you can help me out with this as I can see myself using this setting. Thank you!

Hi, just checking in to see if my issue was confirmed as a bug or not, or perhaps I’m using the module incorrectly. Thank you!

I’ll have a look!

1 Like

Random discovery: I was looking into CV recorders/loopers modules like the Dnipro Krait or the Planar, and I realized my Marbles might actually already do this job quite well. For some reason whenever I use Marbles as a CV sampler, it was always with sequences, never to send CV controlled directly by hand. It turns out it’s a great way to just live-record up to 16 steps of CV, and then modifying your looped signal with Marbles’ slew/stepped/quantized/offset/amplitude controls. Super fun.


Emilie, I’d be interested in knowing, if you wouldn’t mind sharing, why you decided not to make Length CV-able on Marbles. I’ve always been impressed with the rationales behind your design choices and have resisted loading any alternate firmwares for that reason.

1 Like

Panel and board space. There is a certain knob-patched jack distant that I find uncomfortable.

As you can see from the answer above, there is nothing to be impressed about :rofl:


Ha ha. Well, I certainly understand that. But even so, why did you decide to CV, say, Jitter over Length? I might’ve chosen otherwise, but I do get the sense that you have a very specific concept for your modules in mind when you make these choices. Or, prioritize them given the limitations.

So we could, for instance, send a quantized output from stages into an oscillator & the input Cv of marbles and play a melody with the hand and it would record ? That’s sounds quite cool

This module was designed in 2015 so it’s very distant in my mind – but more things than you think can be explained by considerations of panel layout and symmetry.

Symmetry makes sense, the only way you add a Length CV is to put it above the Deja Vu input and then things start to get very cramped.

I think part of the reason I was wondering is that Length CV feels like it would change the functionality of the module pretty considerably, while jitter does not? At least how I use it so far.

If I understand you right, yep! Instead of playing by hand, you can just send a sequence (no need to have it quantized) and modify it on the fly with Marbles using a quantized position of STEPS. It’s a simple and super intuitive way to make a melody evolve without having to re-program it. Kind of an good middle-ground between Marble’s interesting features, yet without being completely dependent on its randomness, since you feed it your own “source” sequence.