Towards a Perceptual Framework for Synthetic Performance

I am  developing a computational framework that acts as a real-time front end analysis system to a musical synthetic performer that can perceive pitch.  The work includes the development of a constant Q transform, known as the Geometrically Spaced Transform, as a tool that can be used to analyze audio samples that contain music.