In a recent video test of the new Happy Nerding Fun VCF I happened on some vocal sounding “whoa wow” sounds. Jump to about 7:05: http://youtu.be/X114GA1280s
There is a new module called the RB1Robotto that is coming out (soon?) by soundmachines that is designed for vocal singing! I really want one! Here is an early test byt the manufacturer, ‘vocal’ stuff starts at about 1:30 http://youtu.be/lpkUkJ1fwIc
Certain vowels can be realized by running one oscillator through a filter, then using another oscillator (usually higher pitched) to FM the filter via CV inputs.
The modulation waveform is often best being a square, but PWM can greatly affect the outcome and is desirable to have control over.
Filter freq, resonance and modulating osc freq and waveform are tweaked until the desired vocal.
With this above, I’ve been able to make a human sounding A, E or I.
I do know that Kevin’s tips will work on a Voyager. Years ago I did a atonal piece called “An Aibo Libido” on the Voyager - I was just tweaking the panel while recording and I got a bunch of formant sounds from that. Unfortunately that was before the OS had MIDI CC control of the panel, otherwise I would’ve captured tweaking via MIDI.
Cool that you got that happening.
Different filters can vary a lot.
The only further advice I could give you’ve probably already tried: vary the modulation depth to the VCF.
Maybe using an already hardsync’d waveform (for either audio or modulation) could make some interesting variations?