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

[mabian]

Hello,

I am trying to get back to Zebra 2 basics because I want to learn something more about sound design.
I'm a bit scratching my head about the PWM option in OSC modules.

This is what the manual says:

"PWM
Adds an inverted copy of the waveform. Applied to a sawtooth, phase modulation cre-
ates an effect very similar to classic PWM (only the phase of the original wave is shifted
around, while that of the inverted copy remains fixed). See PHASE below."

But then, if I don't enable phase modulation, shouldn't I just get complete silence when enabling PWM (because the inverted waveform should null the original one)?
I am instead getting sound, a sort of square wave (using an oscilloscope plugin to check).
For sure I'm missing something, but what is it?

Thanks,
Mario

[Viktor [TUC]]

Hi Mario,

look at the "Phase/PW" knob on the right side of the osc panel. At 100, the inverted waveforms overlap in that way that produces silence. At lower values, the phase positions between original and inverted waveform are not identical so they produce non-silent results.

Viktor
u-he team

[Urs]

For convenience, at a phase setting of 0 the inverted waveform is shifted by half a wavelength.

Code: Select all

*
 *
  *
   *
    *
     *

+

Code: Select all

  *
 *| 
* |   
  |  *
  | *
   *

=

Code: Select all

***
  |
  |
  |
  |
   ***

[EvilDragon]

ASCII to the rescue!

[mabian]

Thank you for the replies, it's much clearer now!

A second question though then... how does this inverted waveform phase interact with the "reset" button?
Is its phase always reset, so retriggered from its start at each note? Or does it have a half wavelength constant offset (excluding the effect of the "phase" button)?

- Mario

[EvilDragon]

When Reset is enabled, you get the phases set up exactly how Urs showed on every new note.

[Howard]

Note: The phase of the inverted copy remains fixed at 180° while the original wave reacts to the Phase setting (and modulation).

[mabian]

Note: The phase of the inverted copy remains fixed at 180° while the original wave reacts to the Phase setting (and modulation).

Thanks Howard, maybe I'm a bit dumb but still don't understand what happens with the OSC "reset" setting to this inverted copy when a new note is played...

1) Does it always retrigger at 180° no matter if "reset" is on or off?
2) Does it retrigger at 180* when "reset" is on and run free when "reset" is off, regardless of note retriggers?
3) Other?

- Mario

[EvilDragon]

It's always at 180° he just said

[mabian]

It's always at 180° he just said

He just said it mentioning the phase knob, not the "reset" button... but I'll try checking it myself more in detail.

Thanks,
Mario

[Howard]

... how does this inverted waveform phase interact with the "reset" button? Is its phase always reset, so retriggered from its start at each note? Or does it have a half wavelength constant offset (excluding the effect of the "phase" button)?

The Reset button makes no tonal difference. Its status doesn't affect the phase offset between the original and inverted copy.

[mabian]

Wow I must say I'm amazed of having three (major) members of U-he replying to this thread, thanks!

I'm researching thoroughly with oscilloscope plugins and similar, finding what follows - mostly in agreement to what replied so far, not completely though. The reset button is not considered at all here, to make clean examination of the "phase" parameter.

At phase "100" (as mentioned by viktor) the inverted waveform nulls completely with the original wave form; so the inverted and the original are completely in phase, good.
Also, at 0, there's a strong level reduction, like the inverted and the original are almost in phase. Good again, we can speculate that phase 0 to 100 relate to full 360° turn around. This is also confirmed by the fact that when moving the phase knob end to end the change is continuous and 0 seems adjoining to 100.

Then, to have 180° offset between the inverted and the original, the phase value must be set to 50 (default value of the initialize patch) and not 0; in the example of Urs, the sawtooth becomes a square when the PWM button is activated (a sawtooth plus an inverted sawtooth shifted of 180° is a ideally a square indeed). But this happens at phase value 50.

Finally, bringing back the reset button, I found, as mentioned by Howard, that it doesn't change anything; the phase offset between the inverted and the original stays the same, the original waveform phase is simply reset to the Phase parameter value at every note retrigger, the inverted keeps its offset untouched.

So, just two little details aren't in my opinion completely right on what discussed so far:

1) The phase knob goes full circle (360°) from 0 to 100, not only 180° as mentioned in the manual:
"Phase/PW
Adjusts the oscillator phase (from 0° to 180°, but you can modulate it up to 360°). Rapid phase modulation normally affects pitch, but if the PWM switch (see above) is on, the effect is similar to pulse width modulation"

2) The 180° offset between PWM inverted waveform and original is achieved at 50.

Not wanting to be pedantic, just willing to fully understand how it works.
Am I missing something?

Thanks,
Mario

[exmatproton]

... how does this inverted waveform phase interact with the "reset" button? Is its phase always reset, so retriggered from its start at each note? Or does it have a half wavelength constant offset (excluding the effect of the "phase" button)?

The Reset button makes no tonal difference. Its status doesn't affect the phase offset between the original and inverted copy.

It does when using more then 1 osc ofcourse

[Howard]

The Phase knob really does have a range of 180°. However, I suspect that the phase of the inverted copy does not remain static (as stated in the manual and above), but is shifted in the opposite direction. [EDIT: Urs confirmed this]. Very interesting, thanks for the discussion, Mario!

BTW for proof that the range is normally 180°, see attachment. OSC2 is an inverted saw: Adjust OSC1 Phase.

[mabian]

Hi, getting back to this after a lot of sidetracking.

@Howard, even though I have version 9709 installed (VST2, x64, Windows) I cannot load your preset because Zebra says my plugin version is too old... as far as I can see, version 9079 is 2.9.1, the latest official release, but maybe you're already testing with a bleeding edge futuristic version?

Anyway, I'll do some more tests in the meanwhile, I'd really like to fully get how this all works.

EDIT: nevermind, the warning message does not prevent the patch from being loaded.

- Mario