An adjoint-based approach for sound field synthesis using non-smooth arrays
Recently for wave field synthesis specialized analytical driving functions for sharp-edged secondary source distributions were introduced. In practice, a spatial discretization of these driving functions is required. In addition to the potential occurrence of well-known diffraction and spatial aliasing artifacts, sampling of the sharp edge induces specific errors in the synthesized sound field. To further analyze this artifact we present an adjoint-based approach for determining driving functions for sharp-edged secondary source distributions in the time domain as it promises insights that are not straightforwardly found with optimization in the frequency-domain. For further analysis, we aim at a consistent comparison with analytical and typical numerical optimization schemes.