Acoustic analysis of the normal voice in nonsmoking adults

Acoustic analysis was made of 154 voices of healthy non-smokers adults. A sustained vowel /a/ was recorded on a mini-disc. Acoustic analysis was carried out with Dr. Speech Science software. Fundamental frequency (F0), standard deviation of the fundamental frequency (SD F0), jitter PPQ (%), shimmer APQ (%), harmonic-to-noise ratio (HNR), and normalized noise energy (NNE) were estimated. Mean F0 was 201 Hz in women and 129 Hz in men. F0 increased with age in men. In women, changes in F0 with age were not statistically significant. Age, sex and F0 did not significantly influence jitter, shimmer or HNR. An acceptable normal threshold is 0.5% jitter. Three percent shimmer excludes 13% of healthy persons. Our HNR values were higher than those reported elsewhere (95th percentile: 18 dB). An NNE threshold of -10 dB is unacceptable because it excludes 38% of healthy persons. NNE differed with sex (mean difference: 3-5 dB). Methodology, sample characteristics and the vowel /a/ may account for some of the differences in results.