Abstract
The definition of masking is extended to include both complete and partial masking. The masking of one sound by another is the increase in the SPL of the masked sound needed to reach a given loudness level or to reach threshold. Plotting old and new data in accordance with this definition shows that complete and partial masking obey the same basic rules. There is no discontinuity in the masking functions when the criterion level is raised from threshold to suprathreshold levels. Three fundamental aspects of masking are dealt with. (1) The asymmetry of masking whereby low frequencies mask high frequencies better than high frequencies mask low. This asymmetry holds mainly at threshold and at low criterion loudness levels against strong maskers. At higher loudness levels or against weak maskers, the masking is nearly symmetrical, and may be reversed at the highest loudness levels. (2) Masking increases in nearly direct proportion to the masker level provided the amount of masking exceeds about 15 dB and provided the signal frequency lies within the frequency limits of the masker. (3) At low criterion levels masking does not change with bandwidth once the critical band is exceeded; at high criterion levels it increases with bandwidth. Within the critical band, masking at threshold increases about 1.5 dB per doubling of bandwidth; it is not known how masking at suprathreshold levels changes within the critical band.

This publication has 14 references indexed in Scilit: