The geometric room model as a surround-sound auralization resource

Abstract

A large church near the WPI campus was chosen as an experimental site for its splendid acoustics as reported by musicians who perform there. The church was modeled using the CATT-Acoustic software, and the resulting computed acoustic parameters compared favorably with MLSSA measurements taken in the church. A brass choir was recorded in the church using a surround-sound multitrack microphone technique. The brass players were also recorded as individuals with acoustic isolation between the instruments. Each layer’s actual position in the church during the recording session was defined in the model as a CATT source with its associated directivity pattern, and positioned in the church CATT model. Impulse reponses were computed from each source location to a set of CATT receivers which duplicated the surround-sound microphone array. The isolated brass instrument recordings were then convolved with these impulse responses and auralized to a surround-sound mix. The auralized version and the actual recorded version were compared for authenticity by a panel of listeners.