Automated extraction of motivic patterns and application to the analysis of Debussy's Syrinx

Abstract

A methodology for automated extraction of repeated patterns in discrete time series data is presented, dedicated to the discovery of musical motives in symbolic music representations. The basic principle of the approach consists in a search for closed patterns in a multi-dimensional parametric space, comprising various features related to melodic and rhythmic aspects, which can be organized into note-based and interval-based descriptions. The pattern description is further reduced through a lossless pruning of the sequence description. This requires in particular a detailed estimation of the specificity relations between patterns. For instance, a pattern is more specific than its suffix, and a melodic-rhythmic pattern is more specific than its rhythmic component. A notion of cyclic pattern is introduced, enabling an adapted filtering of a different form of combinatorial redundancy caused by successive repetitions of patterns. The use of cyclic patterns implies a necessary chronological scanning of the musical sequence. The resulting algorithm offers compact motivic analyses of simple monodies. As an illustration of the analytic capabilities of the computational system, a complete analysis of Debussy's Syrinx is presented. © 2009 Springer-Verlag.