THE SIGNAL-DESIGN WINDOW REVISITED

A method for determining the optimum physical characteristics for audi tory warnings was developed in the 1980s. Designated the Signal Design Window, the method took. into account the noise environment, the hear ing ability of the workforce, and the effects of hearing protectors. T he method was computerised and, from the knowledge of the noise enviro nment and the hearing ability of the workforce, it was possible to des ign signals that would be neither too loud nor too quiet for workers w ith poor hearing ability (as determined from their audiograms). The or iginal model had drawbacks in that the masking calculation employed wa s symmetric and it did not take full account of the ranges of hearing selectivity in the working population. Hearing selectivity is not meas ured by a conventional audiogram. In particular, the experimental evid ence suggested that a proportion of the working population could have elevated masked thresholds when wearing conventional hearing protector s. The safety consequences of this effect are self-evident and they ha ve implications for the routine use of hearing protection in industry. This paper presents a modified signal design window procedure, which incorporates revised calculations for the auditory filter including pe rcentiles of auditory filter width based on a large sample of measurem ents of hearing selectivity. The effects of hearing protectors are sim ulated and the distribution of elevated thresholds is compared across hearing protectors with differing attenuation characteristics. The imp lications for safety policies are explored.