Virtual microphones and virtual loudspeakers for multichannel audio

Multichannel audio is the fastest growing sector in the home audio market. We present a novel method for synthesizing multiple virtual microphone signals from a single microphone reference signal. Our method uses linear prediction to generate filters that can be used to synthesize microphone signals at various locations within the hall, while preserving the correct acoustical characteristics of the venue. We also present a method for rendering multichannel audio over two loudspeakers. Adaptive algorithms are used to create virtual loudspeakers at various locations around the listener by simulating the characteristics of the human ear.