dynamic equilibrium

I don’t know if the Upright Go is going to be helpful or not. My back pain hasn’t really been better since getting it, though it’s probably too much to ask for instant results. It’s caught me slouching several times and maybe correcting that is good. But it doesn’t detect when I have my butt too far forward in a chair and am letting my lower back curve in a way that I will pay for later, and it doesn’t detect how I’m holding my shoulders or head. I think it would need at least three more linked sensors to make sure I’m not hunched and scrunched. But really I should treat it as a reminder not to just make sure my upper back is pointed upward, but check my entire posture.

It definitely likes to go off when I’m leaning to use the modular, reminding me that the ergonomics aren’t particularly great. I don’t think I can entirely fix that with this desk setup and this case. Maybe a vertical case, but I’m loath to switch due to both my spouse’s art on this one and the cost.

Since I’m done working on the new album, I’ve jumped into the gear studies. First on the list is Synchrodyne. I’m trying to cover it systematically and thoroughly, since it’s no ordinary filter (or VCO or PLL or wavefolder) and it’s chaotic and unstable and there’s a lot to cover. So it’s going to take several sessions. I’m taking notes about:

  • basic behavioral aspects, isolated as much as possible from other factors.
  • usage and patching which work particularly well, or particularly badly.
  • more things to try that I haven’t gotten around to yet.

The first session mainly concentrated on how the VCA and folder responds with the filter wide open, on the relationship/differences between the four main filter outputs, and on the cutoff frequency tracking behavior (ignoring the Influence input for now).

That last bit is where the filter diverges most from other filters. In an analog filter, the cutoff frequency is determined by a combination of resistance and capacitance. The majority of filter designs use a variable resistor, because those parts are common and cheap. Variable capacitance is trickier — there are bulky mechanical versions if you have the space, but not voltage control, and fully electronic versions are prone to noise and unreliability. So another design switches rapidly between different capacitance values to achieve different overall results. Normally this is handled simply with a high-rate clock and you don’t have to think about it. You do get artifacts (clock whine and aliasing and imperfect filtering) when the rate is low but otherwise it behaves like other filters.

The unique thing about Synchrodyne, its genius and its curse, is that the clock is an ordinary analog audio rate VCO, with its frequency multiplied and/or divided by a PLL. (And then the VCO is normalled into the filter input, to act as a twisted kind of full voice module…) PLLs can be very touchy things when they’re not designed for a limited range and very specific behavior, and this one is allowed to be fully weird and unstable. The relationship of the VCO rate, multipliers/divider, Track Speed knob and Damping knob all influence whether the filter cutoff is stable, wobbly, or completely chaotic, as well as how it reacts to changes in frequency. And speaking of “influence,” there’s an Influence CV input that I don’t even know what it does which also has an effect that I need to study, and a separate clock input for the PLL which I think is combined with the internal VCO rather than overriding it like the manual says.

Add to that the odd character that all switched-capacitor filters have, the VCA (with lots of gain) and wavefolding built in before the filter input, and the post-filter single-stage folding on two of the outputs and there’s a lot to examine here. And of course, every filter, even the “normal” ones, will react differently to different kinds of input signals or modulation.

What I found last night is some general rules for the (in)stability and tracking… and a pure gold sweet spot where the cutoff wobbles in tension with the input signal which works fantastically to add motion to drones, along with emphasis and distortion of some harmonics… in other words NO I’M NOT GOING TO SELL IT.

Around the same time, Klavis has been releasing a few more short videos including their Grainity filter and honestly, I have found each one less interesting than the one before. The texture and motion in the original Superbooth videos is here in Synchrodyne and more varied and interesting as far as I can tell — albeit trickier to dial in and control the exact rate.

I am still interested in Serge GTO and/or Rossum Morpheus, but I’d have to find some other thing to move out of their way if I go for either of those.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.