Loudness of Quasi-Continuous Impulse Noise

The loudness of quasi continuous impulse noise has been studied as a function of impulse duration and repetition rate, and the spectral shape and relative level of steady background noise. The measurements were performed on 1-second trains of white-noise bursts with exponentially decaying envelopes. The repetition rates were 5 to 100 Hz. The impulse durations were adjusted so that the duty factor of the impulse train (product of the repetition rate and the effective duration of the impulses) varied from 0.025 to 0.1. Steady white background noise was added at peak-to-background ratios of 10, 20 and 30 dB. The loudness levels of the stimuli relative to a steady white noise reference stimulus at a presentation level of 60.5 dB L_{rm eq} were measured in ten normal-hearing subjects using an adaptive testing procedure. Under most conditions, the loudness levels of the quasi continuous impulse noise stimuli were best predicted by the equivalent sound pressure level, L_ {rm eq}, indicating that the "equal energy" relation was followed. At the 50 and 100-Hz repetition rates, the loudness level of the impulse noise test stimuli at the 20 and 30-dB peak-to-background ratios exceeded that of the reference stimulus due to the "roughness effect". The loudness levels of the stimuli at the lower repetition rates, on the other hand, was significantly less than the equal energy prediction at the shortest durations, which suggests that auditory temporal integration depends on the duty factor of the impulse train. The data of experiment 1 were fit by a simple model involving the duty factor of the impulse train and the difference between the L _{rm eq} of the impulsive component and the L_{rm eq} of the steady background noise component of the test stimuli. In the second experiment, the spectral shape of the background noise did not significantly affect the point of subjective equality. However, there was a significant interaction between background noise spectrum and peak -to-background noise ratio which was suggestive of a masking process.