Score analysis/resynthesis environment of the “intelligent music workstation”

Abstract In this work we describe the analysis/resynthesis environment of the “Intelligent Music Workstation”. It consists of three software modules, a) ScoreSegmenter, which is able to decompose a score into a set of basic music objects and a number of transformation relationships among various occurrences of basic music objects within the score; it tries to recognize the main theme of the piece, if any exists; then it finds every point where the main theme occurs, considering also any possible transformations of it; both entire and partial instances are considered; this automatic process is controlled through many configuration parameters which can be set by the user, b) ModelSynth, which synthesizes a Petri Net (PN) model representing the generative structures of the original score; it starts from information extracted by ScoreSegmenter, such as the number of voices within the score, the presence of loops, the application of music transformation (transposition, retrogradation, mirror inversion, ...), e...