A complex speaker edge and asymmetric resonator panel in which the acoustic vibration damping capacity of the speaker edge varies longitudinally around the speaker edge. The resonator panel has an aspect ratio of approximately 1.3:1 or more, and is composed of top and bottom panels held in spaced apart relationship by a plurality of longitudinally extending ribs extending therebetween. The ribs extend at an angle of from approximately 5 to 35 degrees to the longitudinal axis of the panel. The angle is acoustically matched to the complex speaker edge to improve the accuracy with which said acoustic vibration is reproduced. The effectiveness of the differential damping capacity of the edge in improving the quality of sound output from a speaker assembly is determined by observing the average magnitude of the excursions of the sound level pressure versus frequency curves for comparable complex and single speaker edges, particularly in the 200 to 10,000 Hertz range. The speaker edge and the angle of the ribs are acoustically matched by iteratively adjusting the edge and/or the angle in response to the quality of the sound that is perceived by a trained human ear.