Comparison of noise metrics for predicting the annoyance of aircraft overflight noise

Abstract

Two groups of 30 audiometrically screened test participants judged the relative annoyance of two comparison (variable level) signals and 30–34 standard (fixed level) signals in an adaptive paired comparison psychoacoustic study. The signal ensemble included primarily stage II and stage III aircraft overflights, as well as synthesized aircraft noise signatures. Test signals were presented for judgment as heard indoors (test 1) and outdoors (test 2), in the presence of continuous background noise, under free-field listening conditions in an anechoic chamber. For both tests, analyses of the performance of 30 noise metrics as predictors of these annoyance judgments confirmed that the more complex metrics were generally more accurate and precise predictors than the simpler methods. EPNL was slightly less accurate and precise as a predictor of the annoyance judgments than a duration-adjusted variant of Zwicker’s loudness level. [Research was supported by NASA, Langley Research Center, Hampton, Virginia.]