Binaural reproduction of head-worn microphone array recordings with adjustable field-of-view control

This paper investigates an approach for reproducing head-worn microphone array recordings over headphones, such that the listener is also able to augment the rendering to emphasise sounds arriving within a particular field-of-view (FoV), while also attempting to preserve the spatial properties of the captured scene. This type of processing may find application within future augmented reality contexts. The directional emphasis is realised by applying an additional direction-dependent weighting term, when conducting the magnitude least-squares fitting of the array directivities to the binaural directivities. The proposed approach is presented alongside perceptual metric analysis and evaluated via a perceptual study involving 20 listeners. The results suggest that achieving a gain within a defined FoV is attainable, but there exists a trade-off between increasing gain and negatively impacting the spatial aspects of the reproduced sound scene.