Effects of Temporal and Spectral Combinations on the Judged Noisiness of Aircraft Sounds

Abstract

Tests were performed in an anechoic chamber using 20 college students to determine the judged noisiness of stimuli varying both temporally and spectrally. The investigations were divided into three test series. The first test series used 30 stimuli with six different time patterns and five different spectra. The second test used stimuli selected from the first test but modified to include signal durations ranging from 1 to 100 sec. Duration in this case is the amount of time the stimuli were within 10 dB of the maximum level. The stimuli for the third test consisted of recordings of turboprop, turbofan, turbojet, and helicopter flyovers. The results indicated that the most accurate predictor of the judged noisiness was preceived noise level using a tone correction suggested by the FAA and an integrated measure of duration. To illustrate the improvement in noisiness predictability of EPNL over PNL, the results of Test II (duration test) are used. For this test, 75% of the data was within 4 dB of the standard for EPNL with the FAA tone and integrated duration measure as compared to 11 dB for PNL.