Really interested in this idea; mocking stuff up in axoloti where the knob value goes though a modulo function, so for each complete rotation, 5-6 outputs step rapidly and repeatedly though the whole range of levels. I wonder if Cold Mac can be used in a similar way

4 Likes

Not quite the one knob, but the JAG facilitates the many outputs from few inputs philosophy. Combine that with a joystick, or other dual control cv source, and you can have wide ranging and varied impact on a patch.

2 Likes

Not a knob but I use a touch panel for this sort of thing.
I have one in my Bugbrand system which is four plates with 2 CVs per plate.
That can change a patch ok on it’s own but what I really like doing is using the gate out to trigger a sequencer. I tend to use sequencers as big patch preset machines rather than stepping through a selection of notes.
I’ve got some expanders for the Bugbrand one which gives me a total of 32 CV outs with a gate or trigger out per step. Using the touch panel to trigger the sequencer onto the next step with its outputs going to various points in a patch can change things up very quickly.

2 Likes

Why did I never think of this! Does the bugbrand seq give you 32 different cv’s out!?

I do this with a lot of controls on my AIRA modules. You can internally route one knob to all kinds of parameters, and also directly to one of the outputs. Go from that output into a buffered mult, and you can split it out for all kinds of fun.

Very interesting idea! Are you thinking of 5 fixed modulo functions so each output has its own phase?

Something like this? The-one-knob (or 10v cv) on x axis, output of various cv out on y.

Yes, that’s it. I was trying it with a drum patch - mapping various FM, decay, modulation functions onto the one knob, with the idea that you just twist it until you get something you like.
I’d also really like an analog way to do this, haven’t found one yet (but haven’t spent that long looking). I suppose really it’s wavefolding or wave multiplication.

2 Likes

I maybe could have worded what I wrote better!

I have a total of 32 CVs out. It’s an eight step sequencer and I have three expanders which each give me an extra CV output and gate or trigger output per step.

So for each step I have 4 CV outputs and 4 gates/triggers I can use.

I’ll take a quick picture when I am at home tonight.

1 Like

Tangential, a bit - it’s not euro - but I was quite taken with http://www.nobcontrol.com/ . It’s a little expensive, but I like that they map touching the metal controls separately to manipulating them, and the richness they deliver from a single knob.

Here it is. The hub has all the control functions. The three expander modules on the right follow the hub controls and output a CV and trigger/gate per step. I usually clock it externally with the gate out of the touchplate or joystick for example. There’s a random step mode on it which is useful if you don’t want a repeating pattern of changes.

3 Likes

Just stumbled on this thing - still thinking of ways this could interact with my Cold Mac.

3 Likes

If I understand correctly I think we can simply abuse an adc for this… If we take its unquantized bit outputs (residue?) it will give us the same sawtooth shapes. It will have fixed (binary) intervals between the outputs though, but it might be a good starting point.

I wonder if this would work, not sure how to build it

If you like the Doepfer Morph Controller, you can achieve the same behavior and more with the Toppobrillo Mixiplexer:
http://analoguehaven.com/toppobrillo/mixiplexer/

Essentially, if you put a DC offset into the All In input, turning the Position knob will output the same shapes that you see on the Doepfer.

It’s more interesting to me because you can also take four inputs and morph them down to one output, or take four inputs and selectively switch between which one is active.

Intellijel Planar is another great choice. Again, you can send in an offset and then have useful shapes on the four quadrant outputs. Alternatively, you can again take four inputs and morph through them for one output.

I’ve tackled a few of these ideas in my Euro Reakt for Reaktor package (https://www.native-instruments.com/en/reaktor-community/reaktor-user-library/entry/show/9093/):

The 8-Way Scanner is like an eight channel version of the Mixiplexer. For the 8 individual outputs, you can choose between OUTS and AMPS mode. AMPS outputs a CV based on how “open” that channel currently is (basically, the CV used to open the VCA on that channel).

The Meta Control is a single knob that controls the level of four outputs. We put this is the Unfiltered Audio products as well. There are two modes: INT and EXT. In INT mode, the MAIN knob generates a DC offset that is sent to the four outputs. The level of each output is controlled by the associated attenuverter. The easiest point of comparison are the Macro knobs on something like Native Instruments’ Massive, where one knob controls multiple parameters. In EXT mode, the MAIN knob acts as an attenuverter for the signal present at Ext. In. So, same idea, but this time you can avoid having static outputs.

Vector Mix is more like the Planar. It has four inputs arranged in quadrants. The Mix output is determined by the X and Y controls. Simultaneously, you can see that there are both regular outputs (the input multiplied by that quadrant’s VCA) and CV outputs (the VCA’s CV for that output). So, two knobs to rule them all, unless you count a joystick as one =)

These are all open-source, so feel free to open them up and see how any of the algorithms work. The Scanner algorithm was an implementation of some Reaktor 5 standard library code. I’m forgetting which instrument is was, exactly.

4 Likes

ah, so you’re the cat that’s built the Euro Reakt package! much love. I’ve been building up ensembles to work with ES-3 in the flavor of my choice. your modules have saved me so much time.

Using a combination of ableton macros and blocks, i’ve been able to achieve some interesting results. sometimes i go crazy with the mappings (click around and choose params randomly) just to see what happens. with a single turn of a knob on my ableton Push, i can adjust clocks, LFOs, sequencer direction, anything - and still have them in relation to my host. Reaktor & ES-3 have changed the way i think about patching entirely.

3 Likes

Here’s a paper on using machine learning to achieve a “high level knob”. My initial idea for the OSC module (which is still in the works) was to make a ML module which allows you to teach gestures to your synth. This paper seems to even stretch it further with active learning. Its a bit too academic for my, but will try to understand.

3 Likes

The wekinator is a relatively painless way to achieve a machine learning gestural controller.

http://www.wekinator.org/

Link to free Kadenze class on the subject at the top of the page.

Wekinator is used in Laetitia Sonami’s Spring Spyre.

2 Likes

Went through the course and its great! Its one of the inspirations for the OSC module!

2 Likes

I’ve been using CV Toolkit with an ES-8 a lot over the past few weeks. It uses a bus for a lot of functions, which I’m finding does a really good job of changing lots of things at once given some input (knob or trigger or string of CV). It’s such an effective method, it makes me want a series of busses in my case.

I’ve been experimenting with this today - using one knob (well, two knobs coarse & fine) on an arduino midi controller, sending 9 x changing values to an eventide H9.
There’s no intelligence at all - the single value is mapped to the 9 controls as triangle waves with different frequencies - that’s why you’ll see some controls going up and some down.
It’s definitely a way to make drastic changes quickly - and the tweaking with the fine knob works well.
In the clip I think my elderly iPad is struggling to keep up with the changes. It’s kind of repeatable - you learn what the knob changes do.

4 Likes