The main goal of developing stereo systems is enriching the sound reproduction quality and enchancing the perception of sound spatialization. However, all available stereo systems achieve surround sound only by perceptual principles instead of actual physical reproduction. As a result, the ideal listener area narrows to one location, the so-called "sweet spot".
In this thesis I present a state-of-the-art technique called sound field reproduction, which is able to synthesize the sound field of a primary source with numerous loudspeakers. In the first part of the thesis I present the principal theory of sound field reproduction. In order to synthesize the sound field of a primary source, the proper driving of the loudspeakers is necessary. The conventional method of calculating the driving functions called Wave Field Synthesis is presented.
The classic method only ensures correct synthesis under ideal circumstances, which is not feasible in practice. The diffraction effects caused by finite source distribution are being discussed. In microscopy, similar effects are caused by the optical system. A new method is being presented on the analogy of optics and sound field reproduction. The diffraction effects are being reduced based on the deblurring algorithms used in image processing.
Finally, the simulation of sound field reproduction is being presented, which was implemented in MATLAB environment. The new solution used for reduction of diffraction waves is being demonstrated. In the last chapter I compare the non-ideal sound fields synthesized by different methods to the ideal wave field. The results and limitations of the new technique that offers more precise synthesis is being presented.