first rack, opinions please

Here it is. I have a few of the modules already (braids, kinks, the LFO and BRST, the power supply and uJack) and have already had a blast experimenting with what little I can.

Overall, I am interested in generative patches, particularly ambient and noise.
The Mutable Instruments modules are pretty awesome in general; played a lot with the ones available in VCV.
This rack has mostly been inspired by Comparative Irrelevance over on YT. His three module series has helped me learn a lot and shown me that less = more.
Some background: I have some experience with mostly virtual synths through Reason; took some courses on MIDI and synthesizers in college, so I have a general idea of what it is all about.
I'm an electronics major and computer guy, definitely something that makes modular appealing to me, along with watching a machine make music with little to no human interaction. Random patterns and all that sort of thing have always fascinated me.

Looking at this, I'm not seeing a lot that can generate elaborate stochastic structures. As a synth, it's an excellent build, but I'm thinking it'll be rather limited for complex generative work, particularly something self-regulating. Have you perhaps considered a second cab to tandem with this one that can handle that level of complexity? Yes, less CAN = more...but for what you're envisioning, it seems a little TOO 'less'.

Consider something that adds some complex stochastic CV pattern generation, like a Catalyst Audio Time's Arrow, then tandem that with some methods for altering clocking behavior that includes both a stochastic element (such as a Euclidian sequencer) and logic functions that complicate potential trigger/gate patterns, which would work nicely with the Pamela's. Lastly, consider some sort of modulator 'cascade', in which you have some low-speed modulators that can continuously vary some downstream LFO or envelope sources, or a function generator such as a Maths, and maybe tandem this with a quantizer + shift register to 'grab' voltage curves for quasi-random arpeggiations. So that way, you'd have a top cab for 'voice' and the lower for 'control'. With something like that, you could also add some sound processing and also subject that to generative control.

Thanks for the feedback.
The size is temporary; I figured I would start small, get a feel for things and not get too overwhelmed. As of now, I have just an 84HP panel. I do intend to expand to something a little larger, or add another 84HP panel. When I do, it will likely consist of more control/modulation modules, perhaps another voice or two, as you suggested.
I'll have to check out Time's Arrow.
Maths is definitely a module I am considering, pairing it with various logic as you suggested sounds pretty interesting.
Are there modules with quantizers and shift registers? I know there are plenty of logic functions out there. I am definitely not against making some myself as well; would be a really good use for the random IC's I have floating around in drawers.

Well, lessee...the Toppobrillo Quantimator does both quantizing and has an onboard shift register. Also, Mark Verbos's Random Sampling combines a lot of Buchla-esque random source capabilities with a shift register. Those would be an idea. Also, there's several quantizing modules that can also output chords based on incoming CVs, with one huge (and very complex) example being the Instruo Sinfonion, and those could be interesting in a generative situation. Using some separate modules might not be a bad idea, either, and would let you pair something quirky like Erica's microtonal quantizer with a couple of Elby analog shift registers for some real strangeness.

Another useful module would be a quantizer; several of these can extract gate/triggers from voltage curves and output a number of different control signals that could then be fed into logic to create even more complex rhythmic behavior. Joranalogue's Dual Window Comparator is especially neat, since it has several different ways of obtaining gates/triggers beyond the usual single crossing-point activation behavior. As for logic, ARC's Artificial Neural Network looks like it has a lot of 'abuse potential' beyond the typical Boolean gates. A couple of stochastic 'skippers' could be fun, too, and Ladik's are pretty cost-effective/

Plus, don't neglect some of the random sequential devices, most notably a Turing Machine. In its full-blown expanded version, it's a generative fun-house! And as for interesting random sourcing, Nonlinearcircuits' 'Sloth' variations offer a lot of slow-motion change across, in theory, several hours. One of those hooked into an LFO cascade as an extra modulation-variance source could get pretty entertaining.

The neat thing about where Eurorack's at at present is that there are SO MANY choices and possibilities, it ultimately makes creating most any sort of device a piece of cake...provided you're willing to wade through the gazillions of possible modules and combinations of modules. But what we have now is raw power on a scale I'd never envisioned when starting off in electronic media way back at the end of the 1970s. Pretty cool!

Awesome, definitely going to look into more of those.
It's interesting, I had learned quite a bit about digital logic, and had never really applied it to much outside of controlling stuff. I'm thrilled to see that it can be applied here for all kinds of really interesting sound generation.

I had seen the Turing Machine mentioned several times; it is another on my "to get" list.

That's actually really true; like I said, it's really crazy to see concepts that I had learned about used in these really cool ways. I like that Eurorack has really become more than just VCO, VCA, VCF, envelopes and all that. If it's changed that much since the 70's, I can only imagine what it will be like in another 40 years.

Right...consider this: you have an LFO cascade with irregular but very long periods. You then set up a comparator to view this, then send the LFO curve on to a filter cutoff, perhaps with some attenuation. The comparator's gate output then goes to the start/stop of a sequencer, so that the sequencer only runs when the LFO cascade's output voltage exceeds X amount. Then you also send the comparator on to a XOR gate, which has a clock on its other input. When both signals are present, the gate sends nothing. But when only the clock is there, it passes and then goes on to some other device, such as a Euclidean sequencer's clock input, so that when one sequencer is running, the rhythmic pattern generation of the Euclidean stops, and when it's stopped, the rhythms pick back up again. This is what's neat about having all of these digital control subprocesses going on behind the scenes; they allow for an incredible amount of complex control staging that can be just as generatively controlled as pitch CV processes are in a generative compositional system. In short: it'll drive ya NUTZ...but that's a good thing!

The big changes in recent years, which partly (I think) got kicked off by Don Buchla via the early 200e modules, is the whole 'digital in analog's clothing' paradigm, where you have analog control over a module which, behind the panel, is based around some sort of DSP setup, either via a dedicated DSP, or a cheap micro such as an Arduino or Raspberry Pi. The ability to seamlessly merge that sort of digital generation architecture in with all of the analog stuff is blowing things wide open right now.

That does sound pretty interesting. I can see how it can get out of hand quickly, though I cannot imagine that is a bad thing.
When you say chained LFOs, do you mean one LFO modulating another's frequency/rate? If so, that is actually a really interesting concept and I can see how that could create chaotic and interesting waveforms. Sampling it with a quantizer would make it all the more interesting.
I assume that S&H's can be worked into this in a similar fashion as a quantizer, no?

Yea, merging digital and analog is definitely really awesome. It really brings the best of both worlds and blending them together really makes for some neat results. Interesting you mention Arduino, that is something I had gotten into in the past; I'm not superb coder, but I can get by sort of. Using it for synthesis would be really fun.
There's a company called Bleep Labs that makes standalone synths and devices, a few of them use Arduino now that I think of it.
I might try implementing some of their ideas and designs into a Eurorack format.

Right...let's say you have three separate LFOs, plus...oh, a Nonlinearcircuits Sloth and an adder. So...LFO 1 and the Sloth go into an adder so that the LFO's output is constantly subject to slow alterations. Adder feeds LFO 2's rate CV, then that LFO feeds into LFO 3's rate CV. There is pretty much NO WAY that LFO 3 can output anything steady and/or repeating with that sort of a control chain feeding it. Now, go nuts and feed that LFO into a quantizer that's set for something like a minor hexachord with A1 as the root. Then clock that quantizer with...ah, let's get stupid with a comparator that pulls its gates off of a triangle wave from LFO 2. So, each time that fires, it locks in a new voltage on the quantizer. Feed said quantizer into a shift register, clock that critter with another comparator from a triangle from LFO 1. Now, feed those shift register outs to different VCOs set with very different waveforms, mix this down, feed to a VCF of considerable weirdness with an ADSR gated from YET ANOTHER quantizer, this time off of LFO 3's ramp. Then...

...well, you get the point. It's possible to do some extremely crazy stuff of a generative nature like this, or even with simpler patching. And yes, a S&H can work as a quantizer's front end; the important aspect of a quantizer is that, once the incoming signal is sampled and locked-in (either via detecting CV changes or via a clock), its output CV derived from the sampled source is then constrained to a specified scalar tuning. So the S&H is a quantizer part...but not the whole thing.

There's a number of Arduino-based modules here on MG, also...looking around carefully should uncover them. If you're feeling secure in your coding, some even have ways of having their firmware rewritten by the adventuresome. Dangerous...but potentially fun!

I see, its starting to make more sense. That sort of setup seems really interesting. I can see how just that control chain is capable of making completely random outputs. The sloth looks pretty interesting by the way.
That does get pretty crazy; limitless possibilities, definitely something that I like about modular.

Gonna have to dig more into those, and polish up on my C++. It isn't stellar, but I can use all this as a really good excuse to learn more of it.
In fact, that's the fun of all this; not only applying things I already know, but using it as an opportunity to learn even more.

Right...and the interesting thing is, people tend to neglect these sorts of modules because they're not quite as 'sexy' as others. True, LFOs, logic, comparators and the like don't tend to make sounds in of themselves, but what they can do in the process of using a modular as a fully-integrated instrument...that's where they get damned interesting, indeed! Many of my larger-scale test builds on here devote half or more of the instrument to these various control and modulation functions; you don't need a lot of noisemakers to...well, make noise, but if you want those noises to behave in elaborate and amazing ways, you have to deal with the 'unsexy' parts and build in accordingly massive scales. I've gone literally so far in this, myself, that I acquired many years back some salvaged analog computers. The idea with those is to modify them (they operate in the +/- 100v range, I'm planning to have that hardwire-attenuated down to +/- 10v at suitable outputs), then using all three, build a single fully-operational machine with two fully-restored 'program panels'. Even with just ten opamps, these should still be capable of rudimentary chaotic-function calculations such as Lorenz attractors, oscillator patterns such as Lotka-Volterra curves, and so forth. Technically, these can be done on a modular synth as well, but synth modules don't tend to have sufficiently open architecture to allow the user to directly get at the opamp internals on the same scale, so it's actually a bit more difficult there.

Right, I figure that most people are looking more at the oscillators and noise makers more than anything else when they first get started, and don't think about how they are going to control them. It makes sense really. But like most things, I suppose you have to get the "unsexy" things out of the way, and figure out how to make everything work. Fortunately, I already had that in mind, for the most part; just a matter of getting everything together and working.

Oh, that sounds really fascinating. I'd like to see how that turns out. What particular analog computers are you working with?
How are the architectures of modular synths restrictive in that regard?

The analog computers I have are Systron-Donner 3300s...they were intended as a teaching machine, to demonstrate the principles of analog computing in a student lab situation. And like a lot of earlier analog computers, especially the very ancient vacuum tube ones, they operate in a +/- 100 volt scheme. It wasn't until near the end of the widespread use of these machines that +/- 10 volts was finally implemented, but one of the last makers of these machines, Comdyna, even marketed an analog computer (about 15-20 years ago) designed for synthesizer applications. Damn shame it came out when it did, because back then Dieter Doepfer had just began the whole Eurorack thing, Roger Arrick was just getting the Dotcom thing going, and so on. Had Comdyna lasted until the present, they'd likely be selling those units at a nice pace!

But because the Systron-Donners I have use the higher voltages, I'm going to have them modified with some 3.5mm outputs that also have hardwired attenuation to pull the voltage down to the more useful +/- 10v, and I might also have some attenuators added to allow me to scale that down even further as needed. It's also going to be necessary to add some circuitry that will allow some common LFO-type behavior, such as CV control over the clocking, a reset gate input, and the like. But the idea is to keep the internal hardware fairly original, which means those removable program panels will have big red letters at the top to remind me "DO NOT PATCH OUT! 100 VOLTS!" so that I don't make some random, dumbass mistake and patch from that to something that can't handle it.

There are larger (MUCH larger) systems out there, but when you start looking at their power draws, you realize pretty quickly that they won't work nicely being just plugged into yr.basic wall outlet. And ten opamps isn't too shabby...considering that you can patch these up in a massive amount of possible configurations, and control/balance them all sorts of ways. The closest parallel I can think of would be if you had several Maths, but they were bristling with patchpoints that allowed you to totally reconfigure every sort of parameter's signal path. That's actually the big difference here; you simply can't have that level of configurability in Eurorack without the control panel looking like Swiss cheese from all of the jacks, and even if you did, it would create such a level of confusion in users that they'd likely give up on synths altogether and take up the ocarina or some such.

Ah, looked it up and found a brochure for that machine. Surprising how much it looks like a modular. I've actually not seen many like that. Is that one in particular a tube unit? Would explain the +/- 100V.
Definitely would be interested to see how it all turns out, especially how you implement the voltage shifts.

Ah, yea that would make sense then, with not being able to access all the inputs and outputs you would need to really dig into analog circuits on a lower level.
Would be fun to see it done though. Maths is definitely a module I want to learn a bit more about; it's on my "to-get" list.
Is it the Comdyna GP-6 that you are referring to? That was one of the first things that came up when I looked it up.
Might just have to find some of these old analog machines; they look really fascinating, and not just from a synthesizer point of view.

No, not tube, but one of the first transistorized analog computers. But since it comes from that switchover period, it still uses the older +/- 100v standard that was typical up until that general time. It was right around then when the first +/- 10v machines began to appear, but this certainly isn't one of those. As for it looking like a modular, that's because these were the sort of things the modular synth was modeled after; Don Buchla certainly had experience with them as a researcher with NASA, and I suspect Bob Moog saw a few while in his academic studies, pre-synth.

As for the Comdyna GP-6, that's really about the last analog computer that was being made, right up to around 2000 or thereabouts. Comdyna had another machine that they'd specially specced out for music work, though, although I'm not sure if any of them sold. I forget the model number of that one, tho. After they went out of business (so it seems), I don't think anyone is currently making analog computers anymore. And that's kind of...irritating, actually, because they do have more uses than in just an academic/research setting, but one has to actually know how to program one to get any use out of it and that's a bit of a lost art, it seems. In theory, though, something like one might be cobbled together from discrete synth modules, but the thing that makes the Systron-Donner useful is that it's patched more like a Serge, with stackable banana plugs, which opens some not-exactly-computational behavior up for (ab)use.

That definitely would make sense to design something they would be familiar with. Just put a different application.
Ah, early transistorized stuff, that odd in-between time. In general, analog computers seem pretty fascinating. Maybe I'll try to get my hands on one some day.
Wow, they made that one that late? I assume someone must have been using them then, even into the early 2000's. I suppose they do have their own niche uses.
Perhaps some intrepid hobbyist will come up with a more modern one, that works similarly to some of those.
Sounds interesting, to see what sort of potential these old machines could have for synth usage. Or really anything beyond their intended purpose.

Well, the history of analog computing sort of dead-ends around 1968, when digital computing became more common and available. It was actually in the 1980s when mathematicians were forced to rediscover this hardware, because they found that in studying and calculating chaotic processes, a digital computer with discrete logical steps wasn't able to sufficiently parse some of these new mathematical topologies. But by that point, a lot of the machines had been consigned to the scrap heap (which is where I found the Systron-Donners) as 'obsolete', and the tooling necessary to produce these on a factory scale again just wasn't available; in fact, many of the firms making them were long-gone. It's sort of like what was happening to modular synths in the early 1980s, when everyone jumped to digital in the rush to something 'better'...although synthesists rediscovered modular synthesizers in time to keep the whole thing from totally dying out, whereas analog computers had long been considered worthless, obsolete dinosaurs in computer science by the time they were pressed back into service again. That's the difference between 'art' and 'science'; sometimes the cutting edge ISN'T what you want/need!