The loudness of impulse and road-traffic sounds in the presence of maskingbackground noise

Research with pure tones presented in background noise (BN) has shown that for signal levels up to about 10-15 dB above masked threshold, the loudness growth is considerably higher than that of the same tones in quiet (recrui tment), whereas for signal levels higher than about 30 dB above masked thre shold, the loudness growth for pure tones in BN and purr tones in quiet is about the same. To establish to which extent previously determined relation ships between sound level and annoyance caused by impulse and road-traffic sounds had been affected by loudness recruitment, the research on the loudn ess growth of pure tones in BN was extended to that of wideband sounds in B N. The wideband sounds included were produced by a pistol shot (P), a deton ating hand-grenade (H), and a passing passenger car (C). For the conditions in which BN was presented at a low A-weighted level of 30 dB, as had been the case in the previous annoyance studies, the growth in loudness of P, H, and C was almost equal to that of the sounds presented in quiet. (In these low-level BN conditions, the masked thresholds exceeded the thresholds in quiet by no more than about 5-15 dB at the maximum.) On the assumption that annoyance is at least partly determined by loudness, it was therefore conc luded that the relationships between sound level and annoyance caused by im pulse and road-traffic sounds presented against low-level BN, had not been affected by loudness recruitment. For additional conditions in which BN was presented at an A-weighted level of 55 dB, the loudness of low level P was moderately, and that of H and C was considerably reduced. Overall, this lo udness reduction disappeared at signal levels of about 25-35 dB above maske d threshold, which is consistent with the results for pure tones, as report ed in the literature. After corrections to the spectral levels, the results of the present experiment could be fairly well predicted by a slightly mod ified version of Zwicker's loudness model formalized in ISO 532B: The discr epancies between predictions and experimental results were smaller than 5-1 0 dB.