I have finally reposted my doctoral dissertation, this time in HTML format as well as PDF. The title is Perceptual Scheduling in Real-time Music and Audio Applications. I propose an algorithm for improving computational performance of expensive synthesis techniques, such as additive synthesis and resonance modeling that preserves audio quality, and measured both the improved CPU performance and the perceptual quality as measured by expert listeners in controlled experiments.
I think this actually a good time to review and reflect upon this work. Five years have passed since I graduated from UC Berkeley with my PhD. I probably have the only doctoral dissertation in Computer Science that includes James Brown as a citation. While I enjoyed working on the dissertation, including the formal experiments, the work I do now developing music software (and then using for my own composition and performance) is really a better match for who I am.
As discussed in an earlier post, I have had a sometimes challenging relationship with academic science. I have the technical and analytical “chops”, but I am too much of a creator and a romantic to find personal meaning and reward in rigorous experiments and analysis of data. I love the aesthetic appeal of science and mathematics, and especially look for unusal and serendipitous connections rather standard incremental results. Simply put, I am an artist, not a scientist, even when I'm working on software engineering projects.