Some waveform and spectral features of vowel roughness

Area of Science:

• Speech Science
• Acoustic Phonetics
• Laryngology

Background:

• Vocal roughness is a common voice disorder.
• Acoustic measures like jitter and shimmer quantify voice perturbation.
• Understanding the relationship between acoustic variability and perceived roughness is crucial for diagnosis.

Purpose of the Study:

• To investigate the relationship between acoustic wave-period variation (jitter) and acoustic wave-amplitude variation (shimmer) and perceived vocal roughness.
• To compare the efficacy of jitter, shimmer, and spectral noise levels as indicators of vocal roughness.
• To explore differences in these measures across various vowel phonations.

Main Methods:

• Twenty normal-speaking males and twenty males with diagnosed laryngeal pathology produced five vowels (/u/, /i/, /v/, /a/, /ae/).
• Vocalizations were band-pass filtered to isolate the fundamental frequency.
• Jitter, shimmer, spectral noise levels, and perceived roughness ratings were analyzed.

Main Results:

• Increased jitter and shimmer were associated with increased spectral noise levels and perceived vocal roughness.
• Cyclic peak amplitude variation (shimmer) appeared to be a better index of perceived roughness than cyclic period variation (jitter) for most vowels.
• Vowel spectral noise levels showed potential as a more clinically useful indicator of roughness than waveform variability indices.

Conclusions:

• Acoustic wave variability (jitter and shimmer) is linked to vocal roughness and spectral noise.
• Spectral noise levels may offer a more clinically relevant measure of vocal roughness compared to waveform variability.
• Further research into spectral noise analysis could enhance clinical voice assessment.