Loudspeaker nonlinearities-analysis and compensation

Loudspeakers produce linear and nonlinear distortion, especially at low frequencies, which seemingly can be modeled by a Volterra filter. If parameter adaptation of this Volterra model can be successfully performed, one can reduce these distortions by prefiltering the loudspeaker input signal with an appropriate nonlinear inverse filter. Some experiences in the accommodation of a Volterra model to a real world system are shown. The concept of linearization, the restrictions, and the implementation on a digital signal processing (DSP) system are elucidated. Measurements of the loudspeaker output signal with and without compensation are presented.<
>