Creative Sound Modeling with Signed Distance Fields

Abstract

Wavetable synthesisers often use 2D or 3D surface representations to visualise the internal table of values that produce the final audio waveform. This representation can be used to extend the expressiveness of synthesizers, by allowing more intuitive creation and manipulation of the 3D forms. In this work we explore the use of Signed Distance Fields (SDF) for representing and visualising 3D wavetables. SDF representations allow intuitive interactive sculpting and animation of 3D forms. We demonstrate different methods for transforming these forms into sound in real-time, by intersecting arbitrary 2D surfaces with the 3D forms and interpreting the results as sound, in both time and frequency domains. This extends the concept of wavetable synthesizers, and as SDFs lend themselves to real-time, immersive display and interaction through motion and gesture, the technique can be used to construct novel interfaces for musical expression.