From memory all the KP style ugens in the STK framework expect a 'lowest freq' at init Voic-form is a 4-band formant synth not actually a speech synth. good vowel sounds though Check the orig STK project at http://ccrma.stanford.edu/software/stk/ index.html for more info