Dynamic extremes of voice in the light of time domain parameters extracted from the amplitude features of glottal flow and its derivative.

In the first part of the study, the subjects (4 females, 4 males) produced sustained breathy, normal and pressed types of phonation. The maximum AC flow (f(ac)) and the negative peak amplitude (d(min)) of the first derivative of the glottal flow were extracted from glottal volume velocity waveforms estimated by inverse filtering. The ratio between f(ac) and d(min) (amplitude quotient, AQ), which gives an estimate of the closing time of the glottal flow, and its normalised version (NAQ) were computed. The NAQ values correlated strongly with the closing quotient of glottal flow. In the second part of the study, the aim was to investigate the dynamic extremes of the human voice in terms of F(0), SPL, f(ac), d(min), AQ and NAQ as well as the intraoral pressure (p) used for subglottal pressure estimation. Eleven normal subjects (5 females, 6 males) participated in this experiment. The subjects started with the production of soft phonation (approximately 55 dB at 40 cm) and increased their vocal intensity at 5-dB steps up to the level of 105 dB. At the analysis stage, the voice samples were divided into three categories: soft (<70 dB), normal (70-90 dB) and loud (>90 dB). In soft and normal phonation, intensity change was correlated with changes in the shape of the glottal flow waveform, whereas the loud voice samples showed a large increase in F(0) and p values. At the loudest extreme, d(min) approached its maximum and AQ its minimum values and NAQ started to rise. In terms of the NAQ values of stressed syllables, the type of phonation was breathy in the soft intensity category, whereas in the normal and loud intensity categories it was within the range of the pressed type of phonation of sustained phonation samples.