Effect of the Face Wears on Acoustic Parameters of Speech

Wearing a mask and maintaining a social distance, washing hands, and using hand sanitizers have all become conventional, and people are gradually adopting them as part of their daily routine. If crimes involving voice are committed while wearing the mask or handkerchief, there is cause for alarm. Because mask quality influences speech samples, this issue has piqued the interest of speech researchers all around the world. Because people were at home or hiding, the number of offences reported to the police station was substantially reduced. Medical industry, Government sectors etc. which fell under the emergency services, there the crimes like bribery, extortion where only voice of a person did a job, unexpectedly appeared. As a result, the number of offences involving voice did not decrease. The temporal and acoustic properties of voice samples obtained from people wearing tight masks (cotton masks), handkerchiefs, good grade N95 masks, and without any face covering are assessed in this paper. The results show the phenomenal spectral effect for the instances where tight cotton mask and handkerchief was used as a face cover as there was an articulatory constriction due to the face wear. In the case of N95 mask there is enough space for normal articulatory functions of the vocal tract, hence no spectral change was observed. In all instances fundamental frequency F0 remained untouched. The harmonics to noise ratio, effect on energy deterioration observed is also reported in this paper.