Sketching sonic interactions by imitation-driven sound synthesis

Sketching is at the core of every design activity. In visual design, pencil and paper are the preferred tools to produce sketches for their simplicity and immediacy. Analogue tools for sonic sketching do not exist yet, although voice and gesture are embodied abilities commonly exploited to communicate sound concepts. The EU project SkAT-VG aims to support vocal sketching with computeraided technologies that can be easily accessed, understood and controlled through vocal and gestural imitations. This imitation-driven sound synthesis approach is meant to overcome the ephemerality and timbral limitations of human voice and gesture, allowing to produce more re- fined sonic sketches and to think about sound in a more designerly way. This paper presents two main outcomes of the project: The Sound Design Toolkit, a palette of basic sound synthesis models grounded on ecological perception and physical description of sound-producing phenomena, and SkAT-Studio, a visual framework based on sound design workflows organized in stages of input, analysis, mapping, synthesis, and output. The integration of these two software packages provides an environment in which sound designers can go from concepts, through exploration and mocking-up, to prototyping in sonic interaction design, taking advantage of all the possibilities offered by vocal and gestural imitations in every step of the process.