This research component addresses issues of perceptual significance of normal and pathologic speech and voice qualities. The qualities are based on the following acoustic or physiological variables: (1) degrees of oral- nasal coupling, (2) open- and speed-quotients of glottal source waveform, (3) short-term perturbation of fundamental frequency, resonance frequency and amplitude, and (5) widening and narrowing of the epilarynx/pharynx. In particular, it aims at (1) delineating physiologic and acoustic variables important for perceived vocal qualities using sustained vowels and connected speech, both synthetically and naturally produced, and (2) investigating auditory capabilities to discriminate vocal qualities. Auditory capabilities will be investigated for three groups of listeners, i.e., speech pathologists/scientists, naive listeners, and trained singers. Evaluating perceptual and acoustical analysis methodologies is a secondary purpose of the investigation. Data generated from these synthesized stimuli, because they are free from many extraneous within and between- subject variables inherent in naturally produced voices, will not only clarify critical physiologic conditions responsible for associated vocal qualities but also delineate listeners' capabilities to detect and differentiate various degrees of perceived qualities. Use of the different psychophysical methods allows comparative evaluation of quality judgment procedures as they relate to auditory capabilities and clinical practices. The synthesized speech (both sustained vowels and connected speech) and limited amount of naturally produced voice samples will be submitted to comprehensive acoustical analysis programs which include spectral correlation/eigenvector analysis and dynamical/fractal analysis. The purpose of the acoustical analysis is to furture delineate acoustic correlates of perceived voice qualities and to verify and evaluate these analysis techniques.
Showing the most recent 10 out of 151 publications
