Audio Signal Processing and Coding

5.5 NONSIMULTANEOUS MASKING

As shown in Figure 5.9, masking phenomena extend in time beyond the window of simultaneous stimulus presentation. In other words, for a masker of finite duration, non simultaneous (also sometimes denoted “temporal”) masking occurs both prior to masker onset as well as after masker removal. The skirts on both regions are schematically represented in Figure 5.9. Essentially, absolute audibility thresholds for masked sounds are artificially increased prior to, during, and following the occurrence of a masking signal. Whereas significant premasking tends to last only about 1–2 ms, postmasking will extend anywhere from 50 to 300 ms, depending upon the strength and duration of the masker [Zwic90]. Below, we consider key nonsimultaneous masking properties that should be embedded in audio codec perceptual models.

Of the two nonsimultaneous masking modes, forward masking is better understood. For masker and probe of the same frequency, experimental studies have shown that the amount of forward (post-) masking depends in a predictable way on stimulus frequency [Jest82], masker intensity [Jest82], probe delay after masker cessation [Jest82], and masker duration [Moor96]. Forward masking also exhibits frequency-dependent behavior similar to simultaneous masking that can be observed when the masker and probe frequency relationship is varied [Moor78]. Although backward (pre) masking has also been the subject of many studies, it is not well understood [Moor96]. As ...