Masking

Most sounds do not occur in isolation. Sounds from different sources interact such that sound from one source may interfere with the detection of sound from a target or signal source (i.e., masking occurs). A masker only provides significant masking of the signal when the masker and signal have about the same frequency. Figure 11 shows results from a psychophy-sical tuning curve masking experiment. In the middle masking contour, the signal was a brief (10-msec), 1000-Hz tone presented at a low level, about 20 dB above its unmasked (absolute) detection threshold. The level of the masker is varied until the listeners in the experiment are at threshold in their ability to discriminate the signal-plus-masker presentation from the masker-alone presentation. At the tip of the psychophysical tuning curve, when the masker and signal have the same or nearly the same frequency, a very low-level masker interferes with discrimination, indicating an effective masker. When the frequency of the masker differs from that of the signal, a greater masker level is required for masked threshold, indicating that for these larger separations in frequency the masker is not an effective masker. Similar functions are

Figure 11 Psychophysical tuning curves for three signal frequencies [250 (K), 1000 (A), and 3000 (■) Hz] showing the level of the tonal masker required for threshold detection of the tonal signal (from Yost, 2000).

obtained when the frequency of the signal is changed in different experiments.

These experiments suggest that only masker frequencies near that of the signal are critical for masking. A noise masker is a complex sound containing a continuum of frequencies. Experiments have been conducted in which the noise is filtered with a bandpass filter centered on the frequency of the sinusoidal signal and the detection of the signal is measured as a function of narrowing the bandwidth of the filter. The detection threshold remains constant until the filter bandwidth reaches a critical bandwidth, and further decreases in filter bandwidth lower the signal threshold, indicating that the signal is easier to detect. This result is consistent with the assumption that there is an internal filter centered on the signal frequency, and it is the power of the noise masker coming through the internal filter that determines signal detection threshold (Fig. 12). Several different types of noise-masking experiments can be performed to estimate the width of this internal, critical band filter. The width of the estimated critical band is proportional to signal frequency such that critical bandwidth increases with increasing frequency. The critical bandwidth is approximately 130 Hz wide when the signal is 1000 Hz, and it is 1100 Hz wide when the signal is 10,000 Hz. The similarity between the psychophysical tuning curves (Fig. 12) and neural tuning curves (Fig. 7) suggests that the internal, critical band filters are based on the tuning properties of auditory nerve fibers.

Maskers also mask signals occurring before a pulsed signal (forward masking) and after the signal (backward masking). For a fixed temporal separation between pulsed signals and maskers, there is usually more masking in forward than in backward masking.

Figure 12 (a) The maximum power comes through the internal filter for a broadband noise resulting in maximal masking. (b) Less power comes through the internal filter for a narrowband noise resulting in less masking.

When the signal is shorter than the masker and when it occurs at the very beginning or the very end of the masker, there is more masking (masking overshoot) than when the brief signal occurs in the temporal middle of the masker.

Unraveling Alzheimers Disease

Unraveling Alzheimers Disease

I leave absolutely nothing out! Everything that I learned about Alzheimer’s I share with you. This is the most comprehensive report on Alzheimer’s you will ever read. No stone is left unturned in this comprehensive report.

Get My Free Ebook


Post a comment