Extraction of Fundamental Frequency of Human Voice by Autocorrelation Technique

Abstract

Human beings are blessed with a wonderful tool – voice or speaking ability that assist to communicate, talk, sing, and express emotions. Like a fingerprint, one’s voice is unique and distinct, and it is different from all others.  It can act as an identifier. The human voice has many components created through a myriad of muscle movements. It is specifically a part of human sound production through the vocal folds (vocal cords) as the primary sound source. It is composed of a multitude of different characteristics, making each voice different; namely, pitch, tone, and rate. [1]. Vocal folds structure and size varies from person to person making the voice as unique nature. Genetics, gender as well as the size and shape of the rest of that person's body, especially the vocal tract, and the manner in which the speech sounds are habitually formed and articulated are the governing factors to describe its uniqueness. Voice characteristics can be correlated with various electrical parameters like intensity, pitch, short time spectrum, format frequencies and bandwidth, spectral correlation etc. This paper is intended for discussion and analysis the   methods of extraction of characteristics of human voice, especially the pitch frequency.

Keywords: Vocal folds, Phonation, Kymographic parameters, Pitch, Short time spectrum, Average Magnitude Difference Function.