Automatic tuning of the op-1 synthesizer using a multi-objective genetic algorithm

Calibrating a sound synthesizer to replicate or approximate a given target sound is a complex and time consuming task for musicians and sound designers. In the case of the OP1, a commercial synthesizer developed by Teenage Engineering, the difficulty is multiple. The OP-1 contains several synthesis engines, effects and low frequency oscillators, which make the parameters search space very large and discontinuous. Furthermore, interactions between parameters are common and the OP-1 is not fully deterministic. We address the problem of automatically calibrating the parameters of the OP-1 to approximate a given target sound. We propose and evaluate a solution to this problem using a multi-objective Non-dominated-Sorting-Genetic-Algorithm-II. We show that our approach makes it possible to handle the problem complexity, and returns a small set of presets that best approximate the target sound while covering the Pareto front of this multi-objective optimization problem.