Formant feature differences between whispered and voiced sustained vowels

摘要

Previous work on whispered speech has indicated that there is some shift in formant frequencies in whispered vowels when compared with voiced vowels. The purpose of this investigation was to determine the relative differences between whispered and voiced formants. The experiment was performed with five Serbian vowels /i/, /e/, /a/, /o/ and /u/. A total of 10 speakers, 5 males and 5 females, sustained their production of each vowel in a normal (voiced) and whispered manner. Acoustic and articulatory analysis indicated the following: (i) the first and second formant frequencies of whispered vowels /i/, /e/, /a/ and /o/ show shifts toward higher frequencies, (ii) all formants of whispered /u/ shifts toward lower frequencies, and (iii) the formant bandwidths are expanded and are nearly constant throughout the all formants and whispered vowels. [References: 20]