Sound annoyance evaluation of single pass-by vehicle noise.
Traffic noise monitoring usually relies on acoustic metrics, such as A-weighted equivalent continuous sound level. With the current development of noise radars, the noise emitted by a single vehicle can be tracked from a fixed position close to the road. However, equivalent sound level may not fully explain the sound annoyance generated by each single vehicle. Such sound can be considered short and time-varying. In this study, fifty naive participants were asked to rate the perceived annoyance when listening to 3 s pass-by vehicle sounds excerpts. The sounds were played from a loudspeaker at four fixed percentile loudness levels. The audio excerpts were chosen to span over typical range of values in terms of roughness and fluctuation strength. The correlation of the subjective annoyance ratings with several psychoacoustic metrics of the literature was investigated. The results suggest that A-weighted equivalent continuous sound level as well as Zwicker’s model of sound annoyance may not be the best indicators of the actual annoyance perceived by listeners for this type of sounds. The averaged instantaneous loudness, computed with either Zwicker’s model or Moore-Glasberg’s model, provided the best correlation with the subjective annoyance ratings.
Traffic noise monitoring usually relies on acoustic metrics, such as A-weighted equivalent continuous sound level. With the current development of noise radars, the noise emitted by a single vehicle can be tracked from a fixed position close to the road. However, equivalent sound level may not fully explain the sound annoyance generated by each single vehicle. Such sound can be considered short and time-varying. In this study, fifty naive participants were asked to rate the perceived annoyance when listening to 3 s pass-by vehicle sounds excerpts. The sounds were played from a loudspeaker at four fixed percentile loudness levels. The audio excerpts were chosen to span over typical range of values in terms of roughness and fluctuation strength. The correlation of the subjective annoyance ratings with several psychoacoustic metrics of the literature was investigated. The results suggest that A-weighted equivalent continuous sound level as well as Zwicker’s model of sound annoyance may not be the best indicators of the actual annoyance perceived by listeners for this type of sounds. The averaged instantaneous loudness, computed with either Zwicker’s model or Moore-Glasberg’s model, provided the best correlation with the subjective annoyance ratings.