Voice source characteristics of some different modes of phonation as p
roduced by a trained voice are combined with an analysis of the same s
inger's vocal fold vibrations as recorded by a digital high-speed imag
ing technique. Considerable differences between modes are found with r
egard to the waveform and amplitude of vocal fold vibration. The resul
ts support the assumption that the physiological input parameter under
lying differences in mode of phonation is glottal adduction.