MULTI-SLOPE ENERGY DECAY CURVES GENERATED BY MOVING AVERAGE APPROACH

Sound energy decay is typically represented by an energy decay curve (EDC), which depicts the decrease in sound energy over time after the sound source in a room is turned off. In spaces with single-slope decay, the EDC appears as a straight line. Such curves can be analyzed using linear regression to calculate parameters such as reverberation time. However, due to factors like room geometry, absorption, and other effects of sound transmission in enclosed spaces, an EDC may exhibit multi-slope decay. When linear regression is applied to such a curve, the results depend heavily on the range over which the curve is approximated by a straight line.This paper analyzes EDCs generated using the moving average approach, based on impulse responses measured in two rooms: a reverberation chamber and a classroom. The focus is on multi-slope EDCs and a comparison between moving average EDCs and those generated using the Schroeder backward integration method. The moving average EDCs exhibit a greater dynamic range than the Schroeder-integrated EDCs, revealing the final part of the reverberation decay, which is obscured in the latter due to the cumulative summation of background noise.