A redetermination of quiet thresholds as a function of stimulus duration

Abstract

The quiet threshold of audibility has been measured for five subjects at frequencies from 125 to 10,000 c/s, for tone durations from 1 msec to 5 sec and for continuous tones. The measurements were made in free field conditions, and a series of preliminary experiments were performed to determine what factors can influence accuracy in this type of measurement and to establish an optimum procedure. The results obtained in the main series of measurements have a probable error of not more than 1 dB over most of the range and show that for durations less than 1 sec, and frequencies up to 1000 c/s, the “integration” theory of the threshold provides a fairly good description of the measured curves. Outside this range, however, the results show two considerable departures from this theory. Firstly, for frequencies above l000 c/s the threshold function below 1 sec is found to be frequency dependent, the steepness of the curve decreasing as the frequency is increased. Secondly, the threshold for tones longer than 1 sec is found to rise with increasing duration, the rise being frequency dependent and greatest for continuous tones at high frequencies. At 10,000 c/s the threshold for continuous tones is about 11·0 dB higher than the threshold of a 1 sec tone of the same frequency. The rise in threshold for long durations and the frequency dependence throughout the range are not predicted by existing theories of the threshold.