(Sorry for the caps lock but I’m super excited!)

This is so coool… deep gratitude émilie! please take your time =) our marbles won’t be rolling away!

To all the people who are “voting” for the lenght to be superlocked aswell, try magine it the other way arround and all the nice things you could do. Couldn’t you easily do beautiful ployrhythms and still have way more control than before the superlock?
For example: You have a 4 step melodie going that you like, now you just double(or half) the lenght of the rhythm - or make
The lenght 6, let it evolve or change it back to 5 or 3 or whatever. Or take a 3 step pattern, and play with the lenght and/or varations of the melodie - make it 3/5 or 3/7.
I could go on forever, the kaleidoscopic, hypnotic possibilities seem endless and to me, this looks even more exciting that way. (If you’re not looking for that kinda ride, i guess disabling superlock while changing the lenght would also do the trick?)

Seriously almost can not believe that you gave this idea a shot. This is amazing! Was this ever mentioned somewhere before? Just whatever way it is gonna be, if its gonna be, it will be super, just as super as if its not gonna be! Have a nice :wink:


I could test my code on a real module, it works!

Official post here:

Why I didn’t implement the locking of the LENGTH? Simply because implementing it would hint at another feature that would be impossible to achieve: keeping a different length for X and t; but keeping their DEJA VU loop evolving in sync with the DEJA VU knob.

So let’s keep things consistent: the t and X buttons control the scope of the DEJA VU knob, as represented by the dotted lines on the front panel, and that’s it.

As an aside… The module already has many polyrhythmic possibilities as is. I’m not even talking about the division/multiplication t mode. Many people don’t seem to understand this, but a loop length of 4 doesn’t mean that the global behavior of the module will repeats itself every 4th clock tick. For example, with a length of 4, this pattern can be realized:

t1: X---X---X---X---X---X---X---X---

X1: 1---2---3---4---1---2---3---4---
X2: 12341234123412341234123412341234
X3: -123-412-341-234-123-412-341-234

it makes marble much more intuitive.

Bravo @psy_dive! , et un Énorme merci à Émilie pour ce “tweak”.

1 Like

Finaly :):):):)::):slight_smile: make so we can do 4/3 super lock for techno groove :wink:

I think I’m going a bit mad here, as I’m sure i had this working before !

I’m trying to record in a sequence into marbles, so that it can later replay it jumbled up.

I have my midi -> cv … trig into X clock, cv into spread. O button lit. deja vu at 12oclock, and X not lit.
(spread 12 o’clock, 16 steps, bias 0 , steps ‘chromatic’)

(so idea is each press of the key, will sample cv. (and only when i press a key)

so i play some notes, and indeed i can see X1 changing as i press keys - so far so good

I then press X deja vu to ‘freeze bufffer’, and remove X clock.
then move deja vu to 5 o’clock.

what i would now expect (and thought i had working before) , is for it to play a jumbled up set of notes id played.

but all i get is the same note being played over an over.
as if there is nothing in the buffer…

note: if i don’t use x clock, and play some notes then it does seem to record into the recording buffer.
(but its not what i want, since the cv sampled is now dependent on main clock, and not just when it has new notes i.e. key trig)

as i said, i thought i had this working before … am i misunderstanding something about x clock?

is there a better way to do this ‘recording of sequence’ for jumping?

note: on latest firmware with super-lock , but not engaged.

ok, so looks like this was an issue I had previously see here
@pichenettes, supplied a fix, here , which was ‘tentative’ (… I totally forgot it was tentative, as ive been happily using) , but I guess this didn’t make it into the ‘superlock’ firmware version.

so im not completely mad, this did work before on my marbles :slight_smile:

i guess i could roll-back…
wondering if there is a workaround?
(i dont think i can patch t2 into x clock, as then x1,2,3 are all sync to t2, rather than t1,2,3)

This change did make it into the ‘superlock’ firmware. I’ll have a look at your problem next time I sneak into my office.

1 Like

I was wondering if there is any way to have separate panel control settings for spread, bias and steps for the respective x1, x2 and x3 outputs. It seems like if “green mode” is selected these are global. I see there are also yellow and red modes but they don’t seem to quite provide the same level of control.

My use case is I want to have say x1 used for a bass line and x3 used for a lead.

I guess I could always attenuate the x1 output but then this will mess up any quantization. I would need to either run this through another quantizer or something that can give precise voltages.

No, this is not possible.

Ok thanks. I am thinking of getting a quantizer like maybe o_c and then I can just use stepped random voltages for the x outputs and then I can attenuate and quantize externally while still utilizing all the great features of Marbles. Does this seem like a sensible plan? I’m new to eurorack so not quite sure if my ideas are sensible. :slight_smile:

You can probably achieve a similar result with a sample and hold like kinks or stages, etc…

Send a marbles sequence that you are going to use for a lead line into a buffered multiple.

One buffered output you use for the lead line. Then take other(s) buffered outputs into sample input of s&h module. You will need another module to generate the timing (gate/trigger) of your bassline.

This method works best if you are looking to generate slower musically related parts like ambient pads or basslines.

1 Like

But won’t I have the same issue potentially where the range of cv is still too high for a bass line? I can s&h the bass line at a slower clock rate than the lead but won’t I still be able to get some very high cv values?
I would like to constrain the cv for the bass to say one octave while the lead can move over two or three octaves.

I had a thought about Marbles yesterday. It might be a bit esoteric though.

Unless I’m mistaken, Steps isn’t probabilistic – at any given setting all notes in the scale are equally likely, but as Steps moves from noon to fully CCW, some notes are completely removed. I usually find myself putting Steps somewhere in the 2 to 5 o’clock range, because equal weighting of the entire scale rarely feels right.

One possibility is to modulate Steps. If it changes over time in the noon to full CCW range, on average the “stronger” notes of the scale (root, fifth etc.) will be chosen more frequently than the more dissonant notes, but they all have a chance. Sequencing Steps might be smart too – with stricter note choices on stronger beats or key portions of a phrase.

So then I started thinking about a feedback process. If there was a mode where the Y output reflected the “rarity” of the note at x2, several things could be done with that information. It could be slewed and fed back to Steps, to temporarily suppress subsequent rare notes. It could be fed to Rate, to encourage quickly moving on so that it acts as a passing note. It could be patched to Spread or Bias for possibly interesting results. Or it could affect timbre, level or envelope parameters to make the note less prominent (or more, I’m not judging :slight_smile: )

Does that make sense?

Is it possible to add a quick ‘external sequence capture’ mode? It can be done manually but personally I’d love a quick button combo. Idea being that it would allow for ‘sequence pass thru’ of CV/gate while recording a sequence, and then once you jump out of the mode - panel settings resume except both t/X are locked so your sequence is captured. Gate from the ext sequencer would go into the clock input.

Enabling it would trigger the following settings:
Rate: 12 o clock
T mode: orange
T Bias: 12 o clock
Jitter: 7 o clock
Spread: 7 o clock
X Bias: 12 o clock
Steps: 12 o clock
*using the T button as a “gate rest“ input would be an incredible addition here.

My use case centers around sampling a sequence and then mutating it while locked, and then returning to the base sequence occasionally. I do understand I could just continually play the sequence in the ext processing mode, but I would like to ‘save’ the sequence into Marbles and then use my sequencer to complement Marbles ideally.

1 Like

HI , was wondering if is possible to switch J. Output voltage range using one of the inputs with a firmware change and button combinations etc? I would love to randomise that. I am still struggling with this module and find it very difficult to understand. Its mostly suck and see for and i am unsure how to use it like the Turing machine, where you can allow a little bit of randomness into a fixed x pattern? Also changing octaves is difficult for me. Any ideas please. All the best.

No and no. There’s already enough in the module!


with respect i was only asking not demanding, as i said i struggle with the functions of this module though i do love it, i would have appreciated some advice on how i might have gone about some of those things without firmware changes perhaps. i guess that’s too much at this time with the abrupt no’s? Sorry for any stressfulness i may have caused.

I don’t think added and added layers of complexities, to the point that it becomes impossible to know which knob or CV input does what is a solution to any problem. This has been a solution for some people between 2014 and 2018 maybe, and see what kind of monsters of complexity it engendered!

Using the largest voltage range (+/- 5V) and then modulating the actual range with SPREAD could be a solution to get different ranges. You could even use the Y input sent through an attenuator to randomize that.

Enable DEJA VU for the X section, and set the DEJA VU knob slightly before 12 o’clock.

A dedicated module would be a better solution.

1 Like

Hi, thank you very much for the reply. I got a great patch going yesterday and love to make generative music, its the main reason i put a modular together in the first place, so marbles is very important to me.
“Monsters of complexity”, sounds like a super villain or a terrible nightmare!

I also have an Ultra random Analog, doepfer a-149-1 and a Turing Machine, yours is the most fully featured and my current favourite.

Anyway some of the best moments in the piece came when i switched the voltage ranges and that’s when i wondered if it was possible to somehow CV it. Never thought it as a failing of your module.
Thanks for the suggestion about using y and spread will try it later. I have yet to get octave switching working accurately with my gear, it maybe loading effects, who knows. I got a t43 from and a caltrans recently. Also have 2 buffered mults.

I will persevere, its easy to get disheartened in the middle of a creative process, when it gets too technical and ruins the moment. All the best and thank you so much for all your modules and your generosity of ideas.

The (truly) great thing about Emilie’s modules (along with 4ms, TiNRS, etc) is that they are open source, clearly coded and easy to follow. This allows anyone to modify their own modules to enable them to do what they want! It’s a great way to learn coding on embedded stuff…