Diva Triple VCO Question

[rsilverst]

Hi all -- I've been playing around with Diva doing a lot of tutorials, and I discovered something weird, that I don't see in the manual, and am curious if it's a "feature" or something I'm just missing in the settings.

If I use Triple VCO, in Poly mode, and set all 3 oscillators to the exact same saw wave, with all of the voice detunes zero'd, the Detune Amount zero'd, and the Voice Drift zero'd (and no LFOs or envelopes enabled), I still see fluctuations in the waveform when OSC3 is in the mix. If I turn OSC3 off, then I get the exact same waveform every time I press a note. But if I have OSC3 on, then I see fluctuations that sort of look like a small random phase change.

I attached a few images. The perfect saw is with just OSC1 and OSC2 on. And the other 2 images are 2 (of several) variations I see when I bring up OSC3.

Screenshot 2025-03-29 at 11.32.32 PM.png

Screenshot 2025-03-29 at 11.31.22 PM.png

Screenshot 2025-03-29 at 11.31.08 PM.png

I looked for anything that could be disabled, and see nothing else. I also don't see this effect with any of the other oscillator modules.

Here are my settings if you want a rough idea of what's I'm doing. Anything that looks like it's on zero has been verified to be on zero.

Screenshot 2025-03-29 at 11.35.55 PM.png

Screenshot 2025-03-29 at 11.36.14 PM.png

Feature? User error? What?

Thanks

[AmethystClassic]

Definitely a feature. Here's a quote from the excellent soundonsound review when it first released:
"I was impressed that the Detune control prevented the oscillators from perfectly soft-sync'ing together. Two oscillators, when mixed, create a rich and swirly lushness — very like a real Mini".
https://www.soundonsound.com/reviews/uhe-diva
Analog oscillators never line up perfectly. It's part of the sound.

[zerocrossing]

Definitely a feature. Here's a quote from the excellent soundonsound review when it first released:
"I was impressed that the Detune control prevented the oscillators from perfectly soft-sync'ing together. Two oscillators, when mixed, create a rich and swirly lushness — very like a real Mini".
https://www.soundonsound.com/reviews/uhe-diva
Analog oscillators never line up perfectly. It's part of the sound.

But what he's saying is that the behavior changes when oscillator 3 is in the mix, which sounds like a bug. I'm not in my studio right now, but I'll look to see if I can get the same behavior out of it.

[ere2learn]

try setting transient mode to - osc rest

Screenshot 2025-03-31 015955.png

[rsilverst]

OK - maybe this is even more weird - I created a new track with a fresh instance of the plugin, and now I cannot get it to repro the fluctuation that I was seeing. all 3 VCOs are on, and they're giving me perfect unison, with what seems to me to be the exact same settings as I had before (fwiw, @ere2learn, still using Analog for Transient Mode).

I did a little more troubleshooting though, and I DO think it may be a bug? Here's why.

Went back to the original track where I was seeing the fluctuation. It still repros. So then I switched to "osc reset" as ere2learn suggested. The fluctuation goes away. I switch back to analog -- the fluctuation does *not* come back!!

But if I then reintroduce Detune amount, and Voice drift, the fluctuation comes back -- and remains, even if I zero out those knobs -- it remains until I switch to Osc Reset again (and then the reset persists, even if I switch back to analog).

So it seems like if there's ever any detune or drift dialed in, you need to click "osc reset" to "sync up" and then if you switch back to analog it will stay synced unless you detune/drift again.

Maybe it's not a bug? But it's a "phenomenon"

[Urs]

In a freshly initialised Diva instance, each oscillator in each voice starts out rendering at a different phase. In some voices, phases of two oscillators may coincide while another is somewhere else. I don't think in a fresh instance there is any voice where all 3 oscillators line up.

I use the term "fresh" here because one can of course sync oscillator phases up, and without any detune applied at all, those phase relations may survive preset changes, playing melodies and other tweaks. Hence, the observed behaviour may differ based on what settings were chosen prior to the test settings.

Btw. even if phases seem to line up perfectly on a tiny oscilloscope, it's usually much easier to actually *hear* very small phase shifts and movement. So even if some fresh voices appear to have phase aligned oscillators, I can easily hear phase cancellation.

[rsilverst]

Btw. even if phases seem to line up perfectly on a tiny oscilloscope, it's usually much easier to actually *hear* very small phase shifts and movement. So even if some fresh voices appear to have phase aligned oscillators, I can easily hear phase cancellation.

Thanks for the response, Urs. I can definitely hear the difference, with harmonics popping in and out, sometimes even creating the impression that the played note is the octave of the actual note (was easier to share images than sound samples for the post).

Quick follow-up on this point -- I didn't seem to notice this with any of the DCO or Digital oscillator options -- is the random initial phase per-oscillator only done for the Triple VCO model (I could see that making since, since a VCO would be less stable than anything digitally controlled).

Thanks again. Loving this product!

[Urs]

The DCO has a single oscillator of which several waveforms are derived. Those are in locked phase relation.

The digital oscillator... I don't know. It would be the same as on the reference hardware, which would probably default to an aligned phase...?