PhD Thesis Proposal Defence


"Evaluation of Wavetable Generation Methods for Musical Instrument
Matching"

By Mr. Cheuk-Wai Wun


Abstract:

Musical instrument matching is a classic problem in digital sound
synthesis. The goal of instrument matching is to find the best set
of parameters to synthesize an acoustic instrument tone. Multiple
wavetable synthesis reproduces an instrument tone by adding several
enveloped static waveforms. One successful approach to deciding the
basis spectra (i.e., the spectra of the wavetables) involves
selecting snapshots of the original spectrum. Sophisticated
selection strategies such as a genetic algorithm (GA) are required
to find excellent solutions. An alterative to spectral snapshot
selection is to generate the wavetables, which has received little
attention. Previous generative methods do not compare with GA
selection in effectiveness, but actually there is a largely
unexplored world of possibilities for artificial basis spectra
generation. This proposal describes several generative methods for
wavetable matching using numerical and statistical techniques.
Preliminary matching results for a range of instruments show that
local search finds near-optimal solutions almost all the time, and
fuel the speculation that some generated basis spectra are even
better than the best selected spectral snapshots.


Date:     		Monday, 25 October 2004

Time:                   3:30p.m.-5:30p.m.

Venue:                  Room 2407
			lifts 17-18

Committee Members:      Prof. Andrew Horner (Supervisor)
			Prof. Brian Mak (Chairperson)
			Prof. Lydia Ayers
			Prof. C.Y. Tsui (ELEC)


**** ALL are Welcome ****