Real-time synthesis and transformation of ambient sounds

Ambient sounds, those tones and noises that occur in our environment, are a field neglected in many ways, one of them being sound synthesis. Almost all sound synthesis techniques aim at the creation of artificial music or speech, their suitability for the synthesis of ambient sound has not been subject to thorough research. Such a short evaluation is one focus of this article, the second one is the description of a concept named high-level controls. Sound transformation has become a necessary feature for interactive applications, a requirement that returns some of the lost popularity of parametric sound synthesis techniques from sampling, which has developed to the standard sound production method due to its ease of usage. Unfortunately, the possibilities for transformation are inherently connected to the employed technique, they are dependent on the technical details of the particular method. Contradictory, interactive applications, no matter if virtual environments or performance installations, change according to some kind of story and thus exhibit semantic transformation. The logical conclusion is to translate the technical properties of sound synthesis into abstract, semantic properties nto high-level controls. This article investigates methods to achieve such a translation and discusses its advantages and disadvantages, highlighted on some prototypical implementations in Native Instrument Reaktor.