Department of Computer Science, University of Chicago

Title: Department of Computer Science, University of Chicago

Research Question: How can we decompose steady-state instrument data into excitation and filter components for creating source-filter models?

Methodology: The researchers used a source-filter decomposition method on steady-state instrument data from various recordings. They analyzed the frequency domain data and built models for each series, representing the data as a source-filter system. The source part was considered a harmonic excitation system with frequency-invariant magnitudes, while the filter part was responsible for any spectral inhomogeneities.

Results: The method was applied to the SHARC database of steady-state instrument data to create source-filter models for a large number of acoustical instruments. This work has several applications, including improving wavetable and physical modeling synthesis, creating "hybrid" instrument timbres, and enhancing pitch shifting.

Implications: This research provides a new approach to decompose steady-state instrument data into excitation and filter components. The source-filter models created can be used for various applications, such as improving the quality of digital instrument synthesis and creating more realistic and varied timbres.

Link to Article: https://arxiv.org/abs/0103005v1 Authors: arXiv ID: 0103005v1