Auditory Masking

Measurement

In the research literature on auditory masking, the sound to be detected is referred to as the signal, and the interfering sound is the masker. Audibility of the signal in the presence of the masker is measured through many experimental trials. A typical procedure involves presenting on each trial an instance of the signal-plus-masker and masker alone in random order and having the listener judge which of the two sounds contained the signal. This procedure ensures that any bias the [Page 175]listener might have to overreport or underreport the presence of a signal will not influence the results. After the listener has received a period of practice in the task, the level of the signal is found that produces a criterion level of percent correct detection performance. This level is taken as the masked threshold for the signal. The difference in decibel (dB) between the masked threshold and the threshold for the signal in quiet is then taken as a measure of the amount of masking.

Effects of Factors

The types of signals and maskers that have been studied most extensively in research on auditory masking are speech, pure-tones, and bands of noise. The amount of masking produced by any one of these sounds depends on a complex interplay of factors involving the sound's intensity and its spectral, temporal, and spatial properties in relation to the signal. Years of research in audition have mapped out in great detail the major effects of these factors and their interactions.

As might be expected, increasing the intensity of the masker generally increases the amount of masking regardless of the type of signal or masker one might choose. The rate of increase, however, is generally greatest when the masker spectrum falls just below the frequency of the signal: an effect referred to as the upward spread of masking. If the level of the masker is fixed, the amount of masking will also decrease rapidly as the masker is moved away from the signal in frequency. The rate at which masking decreases provides a measure of the frequency selectivity of the auditory system, the degree to which the ear is capable of resolving individual details in the spectrum of sounds.

Sounds also have masking effects that extend over time. Forward masking occurs when the masker precedes the signal in time; backward masking occurs when the masker follows the signal in time. Little backward masking is obtained for practiced listeners unless the signal is quite brief and the masker follows the signal closely in time. Forward masking, on the other hand, is obtained for much longer duration signals and extends over a much longer period following the offset of the masker. The time course of forward masking provides one measure of auditory temporal resolution, the ability of the ear to follow rapid fluctuations in the amplitude of sound over time. Forward masking is also commonly used as a means to study complex processes occurring in the cochlea that are known to affect auditory frequency selectivity.

...