Effects of face protector devices on acoustic parameters of voice

Objectives. The SARS-CoV-2 pandemic required the use of personal protective equipment (PPE) in medical and social contexts to reduce exposure and prevent pathogen transmission. This study aims to analyse possible changes in voice and speech parameters with and without PPE. Methods. Speech samples using different types of PPE were obtained. Recordings were then analysed using PRAAT software (version 6.1.42). Statistical analysis was conducted using ANOVA in Jamovi software. A post-hoc test was performed to compare PPE-related results. Results. Statistically significant differences were found in Cepstral Peak of Prominence-Smoothed, Harmonic to Noise Ratio (HNR), slope of Long-Term Average Spectrum (LTAS), tilt of trendline through LTAS, shimmer parameters, HNR mean and standard deviation of vowels, vowels and consonants formants. HNR values increased whereas shimmer parameters and formant values reduced using PPE [PPE combined>filtering face piece (FFP)> surgical masks>no PPE]. Conclusions. Our data show improvement in many parameters of voice and speech quality and modification of speech articulation when using masks, particularly in case of combined PPE. The most relevant changes were found with a combination of face shield and FFP2 masks. This may be due to unconscious improvements in speech articulation and increased demand on vocal folds to achieve better speech intelligibility.