Acoustic Voice Analysis of Normal Versus Pathological Voices Using PRAAT Software

Abstract

Background: Voice is a fundamental component of human communication, and disturbances in its quality can adversely impact social, professional, and emotional well-being. Acoustic voice analysis using PRAAT software provides an objective, non-invasive, and cost-effective method for differentiating normal and pathological voices by evaluating key parameters such as pitch, formants, and pulse count  Objective: This study aimed to compare the acoustic characteristics of normal versus pathological voices among Pakistani adults using PRAAT software, with a focus on identifying clinically relevant differences in core acoustic parameters. Methods: A cross-sectional study was conducted on 26 participants aged 18–60 years, divided equally into normal and pathological voice groups. Voice samples were recorded primarily through spontaneous speech in controlled environments and analyzed using PRAAT software to extract pitch, first formant frequency, and pulse count. Statistical analysis was performed with SPSS v27 using independent t-tests and effect size calculations, with p<0.05 considered significant. Results: Mean pitch was 202.2 Hz (SD=53.3) in the normal group and 188.9 Hz (SD=57.5) in the pathological group (p=0.55). Formant frequency averaged 536.8 Hz versus 571.8 Hz (p=0.46), while pulse counts were 5.0 versus 5.8 (p=0.24). None of the differences were statistically significant. Conclusion: Pitch, formant frequency, and pulse count did not significantly differentiate normal from pathological voices in this cohort. Broader acoustic metrics, such as jitter, shimmer, and harmonics-to-noise ratio, may provide greater diagnostic utility.