PWM option in Zebra 2 OSC...I don't get it

[mabian]

OK, getting to the roots now, expanding the context to the general oscillator behaviour.

I took your preset, set the exact oscillator settings for OSC1 and OSC2, making sure everything is the same.
So, no inversion, no PWM. Osc reset set to on (for both of course).
I would expect the sound of one of the oscillator being just doubled. That's what happens, but the sound is not "fixed", it's like there's a light phase drift between them. And global - voice drift is off (if on, the drift is much more noticeable, but I think that makes perfect sense)

I'm using Melda Production MOScillatorScope to check, but it's not only a visual thing, I can actually hear that the waveform is not static.

Does Zebra have some kind of magic to make sound more live in this regard?
I'm now going to try with another multi oscillator synth...

EDIT (sigh): I had another Zebra playing at the same time. Now with just one instance the waveform is perfectly static... sorry!

Thanks,
Mario

[Howard]

"more tests" - that's the spirit, Mario!
Yes, sorry, I tend to use very recent u-he internal betas. Just resave any preset that complains.

[mabian]

Yes thanks, I'm enjoying a lot this journey, also discovering, even before messing with the oscillator effects, how many possibilities are in zebra to create awesome sound by just combining the oscillator basics with the fundamental techniques like layering, careful design of shapes and envelopes and ring + sync + detune + voice multiplier and so on. A lot of ground to be covered with them alone. And.. The hz additional filters are astonishingly better!
- Mario

[mabian]

Finally, I managed to get back to this (thanks to Zebra Legacy which made me discover TDZ again) and I think I have the final recap by gathering all the info collected in the previous posts:

with PWM on, the inverted waveform is shifted by "-phase" (phase 0-100 range covers, as documented, the range 0-180°)

Various cases:

- With phase 0, the inverted copy NULLs the original waveform (it's added with 0° = no shift).

- With phase 50 (90°) the inverted copy is shifted -90°, so the original and its inverted copy are 180° apart, which, in the case of a sawtooth waveform, brings indeed to a square wave.

- With phase 100 (180°) the inverted copy NULLs again the original waveform because the waveforms are -180° vs 180° = 360° apart -> no shift.

I tested upward and downward sawtooth and sine oscillators with a simulation made on LibreOffice Calc charts (yes I'm crazy I know) that allowed checking various phase values.
Everything seems to match perfectly.

Thanks again for the help.
- Mario