Perceptual Wavetable Matching for Synthesis of Musical Instrument Tones

Recent parameter matching methods for multiple wavetable synthesis have used a simple relative spectral error formula to measure how accurately the synthetic spectrum matches an original spectrum [1]. It is supposed that the smaller the spectral error, the better the match, but this is not always true. This paper describes a modified error formula, which takes into account the masking characteristics of our auditory system, as an improved measure of the perceived quality of the matched spectrum. Selected instrument tones have been matched using both error formulae, and resynthesized. Listening test results show that wavetable matching using the perceptual error formula slightly outperforms ordinary matching, especially for instrument tones that have several masked partials.