A Comparative Normative Study Between Multidimensional Voice Program, Praat, and TF32

Purpose The instrumental assessment of voice quality is influenced by several factors, including the type of recording equipment used to capture the acoustic signal and the algorithm used by the acoustic analysis software program (AASP) to analyze the corresponding waveform. In this study, various acoustic measures were compared across 3 AASPs commonly used in research, education, and clinical practice: Multidimensional Voice Program (MDVP) by Computerized Speech Lab, Praat, and TF32. Method Sustained vowel phonations for the corner vowels /ɑ/, /æ/, /i/, and /u/ were captured and analyzed for 50 healthy subjects. Correlations and inferential and descriptive statistics are reported for measures of mean fundamental frequency (mean F0), standard deviation of F0 (SD F0), harmonic-to-noise ratio, jitter (i.e., short-term frequency perturbation), and shimmer (i.e., short-term amplitude perturbation). Results Results indicate statistically significant differences between the 3 AASPs for most of the acoustic variables of interest, with MDVP consistently yielding higher values than Praat and TF32 for SD F0, jitter, and shimmer. For harmonic-to-noise ratio, MDVP consistently yielded lower values than the comparison programs. No significant program differences were identified for mean F0. Conclusion Results demonstrate that, although moderate to high correlations were found for values obtained between the various programs, the reported numerical values vary greatly. It is not advisable that voice data reports be combined or compared across AASPs.