Analysis of Speech Sounds In the Praat Program

Abstract

The article identifies the possibility of using the Praat program for comprehensive and complete analysis of speech sounds. It also indicates the range of functions of the software package and the perspective of using the program for comparing sounds.

Praat is a program for speech analysis and synthesis written by Paul Boersma and David Veninko at the Department of Phonetics, University of Amsterdam.

Praat is a very flexible tool for speech analysis. It offers a wide range of standard and non-standard procedures.  The functions of the program that were used in the work: spectrograms, pitch analysis, formant analysis, intensity analysis, jitter, shimmer, voice breaks. There are at least three ways to get sound in the program: sound recording, read sound from the disc, create sound from formulas. In this article, the first method was used and the vowel sounds "a" and "e", as well as consonants "n" and "p" were considered.

With the help of automated algorithms, a number of important data were quickly obtained, such as the average value of the formant on the selected interval, the maximum, minimum and average value of pitch, maximum, minimum and average intensity, the overall signal report (jitter, shimmer, voice breaks). Also, this program can create high quality images that can be exported and inserted into scientific articles or personal research. Due to this fact, the images of not only spectrograms and waveforms, but also the images of formants and intensity was presented in the work.

Based on the obtained results, the selected sounds were compared. As a result, a conclusion was made about the feasi bility and effectiveness of using the software package Praat in the analysis of speech sounds.

It was also noted the ease of use of this software package, as well as the importance of using the program in research in various fields of human activity.