Sustainability Characterization of Normal Voice Features with Voice Obtained After Wearing Different Masks Using Praat Software: A Comparative Study

ABSTRACT

The basic frequency, formants, pitch, and strength of an individual's voice are compared to a disguised voice using various masks in this study. The greatest obstacle arises when scenarios involving disguised speech samples are encountered for assessment and recognition. Such audio samples of anonymous callers are widespread in crimes such as abduction, blackmail, fake extortion, and many others, where the speaker makes a conscious effort to mask their identity due to fear of being discovered. Because the use of face masks during pandemics has resulted in voice-altering strategies that may mislead the identification of a person based on voice, the study was conducted on thirty male and female subjects from the Punjab region wearing various types of masks. As a result, the goal of this study is to compare the durability of features present in normal voice with sound from the same subject wearing various types of COVID-19 protective masks. The voices were analyzed using Praat software, and a comparison study was conducted. The results of spectrographic study of mono and stereo sounds of both male and female voices revealed considerable alterations in their overall spectrum. The entire study yielded significant results that may be useful to a forensic scientist studying instances involving voice identification and analysis. © 2023 American Institute of Physics Inc.. All rights reserved.