next up previous
Next: Background Up: Organ Sound Synthesis by Previous: Organ Sound Synthesis by

Introduction

Current sound-generation techniques for organs generally involve some combination of (1) playback of pre-recorded samples or wave tables, (2) sampling playback manipulation or interpolation to adjust pitch, (3) physical models describing pipe output [1]. The first method, which requires organ tuning, recording, and editing of samples sets is expensive, time consuming, data intensive, and limits playback to only recorded pitches. The second method, when combined with the first, allows any pitch to be produced, but only at exact harmonic structure as the original sample. The third method, though inexpensive and appropriate for real time work, does not take into account slight variations in pipe construction across the range, and so again results in a static harmonic structure. Hence, no current method is inexpensive, easily available, and accurate to organ acoustics.

This paper presents a method to address these shortcomings by proposing a relatively simple, inexpensive approach to model a harmonic structure across an entire range. This method requires only a limited number of recordings, which need not be of high quality or in tune. A set of continuous functions is produced, which will be able to generate unique outputs at any frequency.


next up previous
Next: Background Up: Organ Sound Synthesis by Previous: Organ Sound Synthesis by
mjibson 2009-01-06