Speaker Variations and Vocal Disguise

Abstract

The process of recognizing the speaker based on parameters like pitch, loudness and other acoustic attributes is called speaker recognition. Speaker recognition is considered to be challenging. The voice changing apps facilitating vocal disguise have further made this process even difficult. The voice changing apps can induce variations, some of these variations may be predominantly different, even though these apps can disguise a person’s voice, certain parameters may stay real to the original/habitual samples. The current study was carried out with the aim of determining intra and inter-speaker differences in vocal disguise in six adult speakers. The habitual voice of these individuals was recorded and three variations were induced using a voice-changing app. As the sample size was limited, descriptive analysis was carried out for all the six participants. The first three formant frequencies were determined and the intra and inter-speaker differences were determined. The intra speaker differences were more when compared to inter speaker differences.