Abstract
Voice quality was investigated in 36 young males with a mean age of 10 years. This group represented subjects with vocal nodules and “breathy-tense” voices (N = 12); healthy vocal folds and “trained-breathy” voices (N = 12); and, healthy vocal folds and “normal” voices (N = 12). Each subject was required to produce three prolonged, isolated vowels (/a;/, /i:/, /u:/) which were recorded and analysed by the Fast Fourier Transform and the Filter Bank computer. This provided data on five variables: fundamental frequency (fo), fundamental frequency variability, spectral noise, second formant (F2) amplitude variability, and relative amplitudes. It was found that the nodular voices were differentiated from the normal voices by a higher fo and greater fo variability, and from the simulated breathy voices by higher fo, greater fo variability, less spectral noise, and less F2 amplitude variability. The nodular and simulated breathy voices did not appear to share any acoustic feature. These findings were discussed in relation to the speech pathologist's use of subjective perceptual judgements in the management of clients with voice quality disorders. Finally, the need to develop a terminology which would provide objective reference points for the speech pathologist was argued.