Influence of Vowel Type and Vocal Intensity on Formant Frequency Estimation: Implications for Acoustic Voice Assessment
Аннотация
Abstract To quantify the independent and interactive effects of vowel type and vocal intensity on first ( F 1) and second ( F 2) formant frequencies across the full vocal dynamic range, and to establish normative F 1 boundary extremes for clinical voice assessment.[Methods]Thirty-eight vocally healthy native Mandarin speakers (19M/19F; aged 19–24 years) sustained four vowels (/a/, /æ/, /i/, /u/) at four intensity levels (softest, comfortable, loudest modal, loudest falsetto) in a 4 × 4 repeated-measures design. F 1 and F2 were extracted via linear predictive coding and analyzed using non-parametric Friedman tests with Bonferroni-corrected post-hoc comparisons.[Results]Both vowel type and intensity significantly affected F 1 and F 2 ( p < .001). High vowels (/i/, /u/) yielded lower F 1 than low vowels (/a/, /æ/); F 2 was highest for /i/ and lowest for /u/. In modal register, F 1 increased with intensity for all vowels. In loudest falsetto, however, F 1 declined for /a/ and /æ/ but rose further for /i/ and /u/, while F 2 decreased significantly for /a/, /æ/, and /i/—indicating vowel space centralization. The absolute F 1 minimum was /i/ at softest intensity (median 295 Hz); the maximum was /a/ at loudest modal voice (median 782 Hz).[Conclusions]Vowel and intensity exert non-linear, interactive effects on formant frequencies, with the falsetto transition inducing divergent F 1 responses and generalized F 2 decline. Standardizing both vowel token and intensity in modal register is essential for valid clinical formant assessment. Normative F 1 extremes—from /i/-softest to /a/-loudest modal—provide a quantifiable benchmark for detecting articulatory restrictions.