Musical expression generation reflecting player's image by soft-computing method

This paper generates musical expression, the deviation of performance from tempo marks and/or dynamical marking in a score, based on player's image for a piece of music. The generation procedure of musical expression consists of two procedures; image estimation and derivation of parameters values of musical expressions. In the image estimation procedure an adjective, an image word, is mapped in a Kansei space. In the parameter values derivation procedure, parameter values of musical expression, a tempo, a volume and a length of a note, are obtained in the form of mapping from a Kansei space to a parameters space by the fuzzy inference method. The validity and the usefulness of the proposed musical expression generation system are confirmed by subject experiments. From the experimental results it is found that musical expressions reflecting image are generated rather well by the proposed system.