In this Master's Thesis, frequency domain adaptive FIR filters that forms dynamic sound zones in spatially confined regions, when filtered playback signals are reproduced, are made. Based on the filtered signals, dynamic models have been proposed which models the acoustic pressure in dynamic spatial points with the assumptions of far-field and anechoic conditions.

By solving an optimization problem iteratively, the filters that minimizes the difference between the dynamic modeled pressure and a dynamic reference pressure are calculated. The filters are thereby adaptable to the dynamic sound zones.

Two dynamic models are proposed. One which includes the pressure contributions from other playback signals and one which models the pressure independent of other playback signals. The two methods had equal reproduction error and almost equal array effort.

The proposed adaptive filters require low memory and the updating procedure can be heavily parallelized.