Development of an annoyance index for heavy-duty diesel engine noise using multivariate analysis

Annoyance response to vehicle noise is commonly reported by many people in society, A need to improve the sound quality of vehicles is therefore apparent, The engine is one of the most predominant sources of vehicle noise causing annoyance, Judgments of annoyance due to engine noise were made by 160 subjects in five separate listening tests, An M-S (mid-side) stereo microphone was used to record engine sounds under idling and running conditions for listening Tests A-C, and E, For listening test D, two microphones were used to record engine sounds in stereo under idling conditions, All sound stimuli were presented to the subjects through a pair of loudspeakers in an anechoic room, One of the listening tests was conducted using a paired comparisons method and the other listening tests were conducted using a sequential rating method known as the method of successive intervals, A prediction model of annoyance response was developed by the use of principal component analysis and partial least-squares regression, In this case, all original annoyance scores for all five separate tests were transformed into a common annoyance scale, The prediction model was based on three psychoacoustic descriptors: loudness, sharpness, and harmonic ratio, The model was validated internally and also externally by three new sound stimuli, The prediction of annoyance for these three sounds was found to be consistent when judged by 20 additional subjects. The model gave good predictions of annoyance judgments for 6-cylinder in-line engine noise. (C) 1997 Institute of Noise Control Engineering.