Wavetable PWM oscillator on the TI OS v2

[Timo]

So how does this work, then?

Does it duplicate the waveform, and then variably adjust the phase of the duplicate against the original (a bit like a chorus effect)?

...Like something that was discussed a few years ago (pseudo PWM):

http://www.infekted.org/virus/showth...ht=pulse+width (pictures and audio demos of which are further down the page).

Or is it different to that technique?

Any demos to showcase the effect? The only bit I've heard of the TI's PWM wavetables was from the January 2007 TIOSv2 youtube demo by Ben Crosland at NAMM last year, which was isolated and a little quiet, and I couldn't really discern it properly.

Cheers.

[bishopkris40]

I know there is a description somewhere.

Will try to find it for you.

If anyone has the TI software then please post it

[Timo]

Just found a description from the V2 addendum:

Wavetable PWM Oscillator

The WavePWM oscillator takes two instances of the same wavetable, and phase-shifts them against each other to achieve an effect reminiscent of the traditional pulsewidth modulation of a pulse wave oscillator.

Index
This determines the playback position within the currently selected Wavetable. Each of the 128 available values represents either a particular wave or the interpolation of the two nearest waves. Modulating the Index of a wavetable with an LFO or Envelope will often result in a highly-dynamic variation in the timbre of a sound, impossible via any other means.
This parameter can be modulated via the Mod Matrix and the LFO’s – please select Wavetable 1/2 Index from the list of available destinations

Please note that Table 0 (Sine) contains only a sine wave, and as such, the Index parameter will have no effect on it.

Wavetable
Selects the current Wavetable – each being a unique collection of different waves from which all manner of different timbres may be achieved.

PulseWidth
At zero position, all the even-numbered harmonics are cancelled out, creating a hollow sound similar to a 50% pulse wave -
then the value reaches 127, the whole wave is almost entirely cancelled out, resulting in a much thinner sound.

This parameter can be modulated via the Mod Matrix and the LFO’s – please select Osc1/2 Pulse Width from the list of available destinations

Detune
As with traditional PWM, much of the fun comes from modulating the pulsewidth by an LFO, which creates the characteristic warmth of detuned oscillators - the Detune parameter creates this effect automatically, with no further modulation assignments.

Interpolation
Use this parameter to determine the how smoothly the different waves in the current Wavetable are blended into each other as the Wavetable Index of the respective oscillator is swept.
Hint: If you do not make use of either the PulseWidth or Detune parameters, switch back to Wavetable mode to save on DSP and increase polyphony!

Sounds quite similar to the one described on my other thread. But by their description of the so-called PulseWidth...

"PulseWidth - At zero position, all the even-numbered harmonics are cancelled out, creating a hollow sound similar to a 50% pulse wave - then the value reaches 127, the whole wave is almost entirely cancelled out, resulting in a much thinner sound."

... isn't the same. Whereas on my audio example of Korg's implementation of waveform modulation the two saw waves rocked between a standard saw wave and one an octave higher, at no point would they ever cancel each other out.

It appears that Access' do the same, but they mirror the duplicate wavetable on the x-axis (flip it upside down) before phase-shifting it with the original wavetable. That way the waveforms WILL cancel each other out depending on the phase, like true PWM.

It would sound markedly different to something like the original audio example of Korg's implementation of the waveform modulation (which sounds really nice and brass-like), but I guess the options on the Access implementation are much more flexible - being indiscriminate in the waveforms that the effect can be applied to (Korg's was limited to the saw waveform only. - Of that I can think of [and it's a bit of a headfuck] I don't think any other waveform apart from a sawtooth can effectively be pitched up an octave by duplication and phase-shifting while still retaining the identical sonic characteristic of the original?).

Would be very interested to hear what the Access PWM implementation sounds like anyway.

By the way, what are the waveforms like on the wavetables? Are they anything like the LFO waveforms?

Cheers.

[bishopkris40]

some of them are very strange, but listening to them alone I get more of a bell like sound coming from most. However I have been shown that some wonderfull things can be done with them. I just have to wait until I can program patches better before I can use them to their potential.

As far as how they look compared to the LFO's nobody has pic's yet, but I know they're just one cycle samples. They all have daft names that don't describe the shape so hard to tell. Maybe a quick look on a scope will help me describe them, will see if I can get time.