A solution method for active suppression of reflections in anechoic chambers

A solution method to improve an anechoic chamber at low frequencies with the use of active noise control is presented. The approach uses the Kirchhoff-Helmholtz integral to compute the reflected sound field resulting from the primary sources together with an algorithm to compute the filter coefficients of a controller driving secondary sources on the walls of the enclosure using reference signals as inputs, which are measured on a contour enclosing the primary sources. A causal frequency domain method with conjugate gradient iterations is derived to determine the controller. The method is sufficiently efficient to allow computation of a causal time-domain controller with hundreds of secondary sources and hundreds of reference sensors in two- (2D) or three-dimensional configurations using a fully coupled multiple-input multiple-output system. The paper shows the results of a simulation with 200 secondary sources, 200 reference sensors, and 225 performance sensors based on a 2D finite element simulation. The method is verified in real-time in an experiment with the objective to suppress the reflections from the walls in a smaller 2D setup. Measurements with verification microphones show that the reverberation time is effectively reduced in the real-time experiment.