Singing voice analysis on the basis of acoustic parameters

Voice plays a fundamental role in human relations. In addition to its communicative function in everyday life, the voice also acts as an instrument or a working tool for singers, teachers or actors. It is said that singing is an extension of speech, but performing it correctly is a complex task that requires hard work and training. This paper draws attention to the problem of insufficient training in voice emission and voice control of singers in amateur choirs, which can cause strain and disorders of the phonatory system. Tool that can assess the quality of a person's singing on the basis of acoustic parameters may prove useful. In order to determine parameters that could help evaluate the correctness of singing, a study was conducted on a group of 10 choir members and one professional singer. The study consisted of recording the singers' voice during singing and speech. The subjects performed simple vocal exercises consisting mainly in upward and downward sound modulation. In this study, portions of the recordings were analysed to determine parameters like Maximal Phonation Time (MPT), Singing Power Ratio (SPR) or signal integral. The values of obtained parameters for the choristers were compared with the results of the professional singer, which allowed to select those parameters that may be helpful in the evaluation of the singing voice. The parameters for which the connection between their value and singing correctness has been shown create a vector of features that can be used to assess the correctness of classical singing. The paper also describes further research plan.