Fundamental frequency: an analysis of age and gender

Abstract

The phenomenon of speech is very complex and every time a speaker produces it, s/he presents personal characteristics, making the speech an individualizing element. Acoustic analysis of speech helps the identification of these personal traits of the speaker, with the help of numerical results and graphics. Acoustics is one of the objects of study of Phonetics, which has two other areas: Physiologic and Perceptual.One of the fields that uses the three areas of Phonetics is Forensic Phonetics, offering knowledge and methods of linguistic analysis in contexts of criminal investigation involving speech. One of the activities of Forensics is speaker identification, and fundamental frequency (F0) is a promising phonetic parameter in this area due to the fact that most of the disguises directly affect it. Gender and age also have influence on it, because physiological differences for both sexes and ages can be seen when analyzing F0 separately, for example, high or low pitch. In this study, we analyzed F0 comparing normal and disguised voice of 50 speaker, men and women in two age groups,and investigated which disguises were most used in the different groups. We used perceptual and acoustic analyses, and provided numerical data to observe the changes in the voice, after applying the VPAS protocol (LAVER, 1980) adapted to Brazilian Portuguese (CAMARGO; MADUREIRA, 2008) to label the features/parameters of voice used in the disguises.The objective of this study was purely instructional. We wanted to learn how to deal with this new area of linguistics: forensic phonetics.Partial results have shown that most of the participants raised their F0 and that the choice of raising F0 was most frequent among men and lowering F0 among women. Concerning age related to the type of disguise, we could not find very significant difference rather than more young participants lowered their F0 than adults, and more adults had their F0 higher.