PresetGen addresses the target preset generation problem – given a sound and a synthesizer, finding a preset that best approximate the target sound – in the case of real-world synthesizer OP- 1. The OP-1 consists of several synthesis blocks, and it is not fully deterministic. We propose and evaluate a solution to preset generation using a multi-objective Non-dominated-Sorting-Genetic- Algorithm-II. Our approach makes it possible to handle the problem complexity and returns a small set of presets that best approximate the target sound by covering the Pareto front of this multi-objective optimization problem. Moreover, we present an empirical evaluation experiment that compares performance of three human sound designers to that of PresetGen, and shows that PresetGen is human competitive.

Members: Kıvanç Tatar, Matthieu Macret, , Philippe Pasquier.

Papers & Posters: