Fundamental frequency measures in the comparison of habitual and disguised voices

Purpose  This study aimed to investigate the stability of measures of fundamental frequency of voice by comparing samples of habitual and disguised speech.

Methods  This observational, analytical, cross-sectional study analyzed a voice bank of 40 people of both sexes (20 men and 20 women), aged 19 to 58 years. They read a passage in both habitual and disguised speech. The types of disguises were analyzed by three expert voice judges. The study used the PRAAT^® acoustic analysis software to extract the mean f₀, median f₀, minimum f₀, maximum f₀, and base f₀ (f_b) from the complete reading in both situations, using pre-determined procedures. The data were tabulated and statistically analyzed.

Results  The f_b, maximum f₀, and minimum f₀ values were similar in habitual and disguised speech among speakers of both sexes, whereas the mean f₀ and median f₀ differed between the two situations.

Conclusion  In general, the measures of f_b, minimum f₀, and maximum f₀ were more stable in disguised speech, which may suggest future usefulness in intraindividual comparisons. The speakers’ various types of adjustments to disguise their voices did not significantly alter the measures of f_b, maximum f₀, and minimum f₀ in the comparison between habitual and disguised speech.

Keywords:
Voice; Voice Quality; Speech Acoustics; Speaker Recognition Interface; Acoustics