Digital Signal Processing Applications in Monophonic Music Analysis

Abstract

This paper presents an approach to monophonic music analysis using guided, continuous, overlapped fast Fourier transformation. This method is based on the physiological activity of the ear which can be modeled as a bank of contiguous bandpass filters. The analysis involves pitch detection, determination of note duration and identification of instrument limited to flute, clarinet and trumpet.