Real-time processing of audio signals in microphone and loudspeaker arrays often requires the implementation of fractional delays. The proposed methods are particularly relevant in applications that use windowing and frequency transforms. The short-time spectrum can be modified to obtain the equivalent delay. Various approaches to the problem are discussed and evaluated in the context of beamforming precision resulting from time-alignment errors.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.