Analysis of the sound attack in context of computer evaluation of the singing voice quality

Sound attack is meant as the initiation of the tone. Good attack involves the simultaneous start of exhalation along with the emission of sound. This situation is known as a soft attack. Much more common is a hard type of attack, which is normally used while speaking. It involves tightening of the vocal cords before the expiration. Unfortunately, this state is tiring for the ligaments and can cause damage to the voice. The idea was to propose a method to analyse this problem, which in the future would become a part of set of methods for computer analysis of the singing voice. This article presents a method of the extraction of attack parameters. The estimation of these parameters is carried out with the recorded audio samples. In the study sound sample of ´a´ registered several times for each pitch will be used.