Do you save your patches?

Oh, boy...the horror that are "dope sheets". These work nicely for prepatched stuff, but pretty much NOT for modular, mainly because you'd have to draw the patch layout on top of the panel diagram with the knob settings. The end result tends to look like refrigerator art created by a profoundly disturbed individual.

Instead, I worked out something decades ago that seems to work better. Not as all-in-1 as a dope sheet, but hey...

First up, your knob settings. Only list the controls that have a specific setting or which change, and then do so as a list...not a diagram. Anything that doesn't change or have a specific setting is assumed to be at "zero" or some similar setting. Once the control settings are in place, then the patch itself, and for that I just use a shorthand that looks something like...

VCO1 out -> SVF1 in

SVF1 out -> DRF in (these are from a piece using my Digisound, fyi)

DRF out -> MIXER1 in

EG1 out -> SVF1 CV

EG2 out -> MIXER1 pan CV

...and so on. While this looks sort of kludgy, believe me, it's not; I've used this notation method for some 25 years for recording _specific_ modular patches for works where I need to replicate the patches at a later date, such as for live performance. But you'll notice that I said "specific" patches that require replication...most of the time, I just tear into things and deliberately DO NOT make any notes on patches. That way, I keep exploring instead of laying back on a pile of in-reserve patches, and since I know what instruments in my studio do which things better (although some new exploration will be needed after this huge refit), there's not too much blundering around.

That is, unless the project is created around simply blundering around. That method does yield results sometimes...

Precisely. It's sort of like working with any other tool...once you know what it does, what it's for, how it's used, and you're comfortable with that...well, there you are.

F'rinstance, the VCOs on my Digisound 80. They have no octave switches...just "coarse" and "fine" tune knobs that sweep from tens of seconds to ultrasound. But after using them for some time, I knew that if I wanted LFO-type behavior, I set the knobs to X, Y and Z...and if I needed bass, then...well, you get the idea. It's not that you want to precisely log the settings and such; part of the beauty (or annoyance) of modular synths is that tiny adjustments can result in HUGE sonic differences, and by avoiding this sort of "precision", you open up other possibilities for "accidents" and, thereby, your music evolves.

This is a huge part of why I try to NOT use memory settings on synths that have usable controls. In some cases you have no choice; not saving a patch to memory on, say, a Kawai K5 would be a massive mistake. And in a few others, like I describe above, it's necessary to save something that'll be used repeatedly for live performance. But...well, take the example of my CS-80. I think I've opened its "memory hole" lid more times to show how utterly MAD the "patch memory" is than to store settings. Plus, in that case, you have ZERO guarantees that what you've saved is precisely what you've programmed, since it's never possible to replicate the panel settings on those teensy-weensy memory panels...which is sort of a plus in that it opens up the accident potential yet again, but it's still loads easier to start grabbing "live" tabs and sliders on the fly and having at the sound directly. But again, I know what all those funny-colored thingummys are and what they do...and they're THERE to be screwed around with, unlike the aforementioned K5, which pretty much just has buttons and an alpha-dial.

Also, getting "stuck" on a modular patch isn't quite the same problem as the sort of sonic painting-yourself-into-a-corner that you can run into on single-entry-method synths. An example from decades ago that comes to mind is from my synthesis proficiency project back in my first undergrad year in 1981...

We had to submit a tape that had some specific requirements for sound sources, and one of these was to have a tone cluster of three sine waves. But the dirty trick there is that, between the ARP 2600 and the EML 400/401, we DID NOT HAVE three oscillators that outputted sines. Just two. Hmmm...BUT...while using a resonant filter was "off limits" for reshaping, the prof said NOTHING about NONresonant ones...and hey, presto, there's a patchable slew limiter on the 2600, and I can get a spare triangle wave, soooooooo... Long story short, I definitely passed the proficiency, AND got extra points for the explanation of the patch and how I was able to get the three pitches in only one pass, without overdubbing, on a problem that was solved by the rest of the class by overdubbing that last pitch onto the other two.

But this is what you get when you know your tools. A lot of the class didn't know that a slew limiter can be used as a lowpass-ish "wave smoother" on triangle waves to back the few harmonics that are there down to a much lower (and sine-like) level. I read the textbook, though...but also, I'd spent a lot of time in the EM studio with the 2600 (and the EML, later on) getting used to what it did and how, and I'd gotten to see (on the studio's scope) how the harmonic reduction worked with the slew limiter, and that time was far more of a key to solving the problem than just poking at the textbook. Had there NOT been a slew limiter that I could patch into...well, that would've been a problem, yes, and then I'd had to have fired up the 4-track. When you have no way to alter the signal paths in your synth, getting stuck is far more likely than if you have ZERO signal paths prepatched.

So, sure...saving patches CAN be useful. But it's not quite as useful as simply sitting down and getting to know your tools.