100% phase-driven oscillator?

While the V/Oct input and (exp) FM input on many VCOs are essentially identical, my guess in this instance would be that the V/Oct input is treated as a ‘pitch/notes’ input and therefore does not need to work at audio rate, whereas the FM input does. Having the V/OCt input not have to sample at audio rate might save a bit of CPU.

1 Like

Well you’ve opened up a whole new rabbit hole here :smile:. The phase modulation is what I was actually looking into, but linear FM is a whole different ball game for me and I haven’t really experimented much with it. It would be great if you can post a video sometime, but I completely understand about the “long backlog of interesting stuff” :grinning:. It’s always exciting to learn something new.

Yes, that is what I was assuming, but did not want to guess. I had looked at the code to see if I could identify the issue, but couldn’t see anything. I imagine an experienced plugin developer (or someone familiar with DSP) would spot the design issue in the code relatively quickly.

1 Like

With my Scannibal modules you can build a phase driven oscillator with up to 16 segments. With 1 segment you can do sawtooth. With 2 you can do square and triangle. With more you can invent a variety of shapes, with a somewhat limited ability to give each segment its own S- or J-shaped curvature.

It wasn’t intended to be an oscillator, but don’t let my intentions limit you.

1 Like

Ah yes, I know these modules, they’re neat! I hadn’t thought of this use. “Don’t let [the designer’s] intentions limit you” must surely be the motto of this entire modular synthesis endeavor

1 Like

Oh and: Your FUNC is probably my #1 most used module in the past few months. It’s JUST SO USEFUL!

1 Like

getting a lot of mileage out of the @docB phased oscillators this evening

2 Likes

dumb question - how is this any different from shaping a VCO with any kind of lookup-table / mutli-segment thing / bezier mapper / shape master / wave shaper? and with most of these won’t you get a ton of aliasing if you try it at audio rates?

well, without having thoroughly tested, I suspect the docB SPL and PHO are cheaper to run than for instance Shape Master, which certainly looks like a CPU monster. They’re also smaller.

Both of these have particular control features that I like; the PHO takes sixteen CV channels of additive partial levels, which I can mix to suit any occasion. This gets more fun when a few of them are driving one another in series, producing highly varied, but totally determined, oscillations from very little input.

As for the aliasing, I don’t know - I don’t use these at audio rate, only as LFOs

It’s really not… SM is very efficient for what does. Although given it has 8 channels it will use more CPU than one instance each of SPL and PHO for sure.

For sure, right now I’m running one ShapeMaster scanning one set of splines at about 1.4% - 1.9% CPU, and two instances of docB PhO at 0.2%-0.4% each