Annoyance due to machine noise is critical to those that hear the sound and may determine the acceptability of the machine itself. As such, the development of an accurate model for assessment of the annoyance of time varying sound could benefit the designers and operators of many kinds of machines. Challenges in the development of an algorithm to predict the annoyance of sound are many. Besides accurately accounting for a number of subjective parameters (e.g. loudness, tone prominence, spectral balance, etc.), the algorithm has to account for time variation, growth and decay of sound within a meaningful time interval. This paper summarizes an effort to derive a generic algorithm capable of predicting the annoyance of complex sounds. The Time Resolved Annoyance (TRA) method is presented to estimate annoyance at an instant in time through a combination of various subjective parameters scaled by specific loudness. For this initial effort, the capability of the model is examined by comparison of predictions to the measured response of 32 subjects who judged the annoyance of 99 simulated aircraft flyover sounds with complex tonal and temporal characteristics.
展开▼
