Chord Analysis Using Ensemble Constraints

Abstract

Many applications in music information retrieval require the analysis of the harmonic structure of a music piece. In Western music, the harmonic structure can be often be well illustrated by the chord structure and sequence. This chapter presents a technique of disambiguation for chord recognition based on a priori knowledge of probabilities of voicings of the chord in a specific musical medium. The main motivating example is guitar chord recognition, where the physical layout and structure of the instrument, along with human physical and temporal constraints, make certain chord voicings and chord sequences more likely than others, and make some impossible. Pitch classes are extracted, and chords are then recognized using pattern recognition techniques. The chord information is then analyzed using an array of voicing vectors indicating likelihood for chord voicings based on constraints of the instrument. Chord sequence analysis is used to reinforce accuracy of individual chord estimations. The specific notes of the chord are then inferred by combining the chord information and the best estimated voicing of the chord.