Abstract
An artificial reverberator is proposed in this paper to synthesize room responses. The method employs the virtual source representation and the comb-nested allpass filters to generate the early reflection and late reverberation, respectively, of room responses. The virtual source method is based on a simple representation of sound field with a distribution of discrete simple sources on the boundary. The complex strengths of the virtual sources are then calculated by solving a frequency domain least-square problem. Parameters such as room geometry, size, and wall absorption are naturally incorporated into the synthesis process. The filtering property of human hearing is also exploited in a nonuniform sampling procedure to further simplify the computation. Apart from the early reflection, a comb-allpass filter network is adopted to simulate late reverberations. Optimal parameters of the comb-allpass filter network are obtained using the genetic algorithm (GA). The energy decay curve (EDC) is chosen as the objective function in the GA procedure. Numerical simulations are carried out for a rectangular room and a concert hall model to verify the proposed technique. Subjective listening experiments demonstrate that the present technique is capable of conferring remarkable realism of room responses.
