Mark R. Sydorenko
Jont B. Allen
Acoust. Res. Dept., AT&T Bell Labs., Murray Hill, NJ 07974
Several studies in the literature have reported that the masked threshold in the presence of a tone masker versus a narrow band of noise of equal intensity can differ by as much as 20 dB. This difference is known to be important in music coders, and will probably turn out to be important in future speech coders. This presentation describes results that show that the difference in masking ability of tone versus noise maskers is due to the difference in acoustic intensity when a probe is added to these maskers. In addition, a series of psychoacoustic experimental results are reported that suggest a fundamental relationship between the just-noticeable difference in intensity (jnd[sub I]) for a sub-critical bandwidth masker and the masked threshold of a tone. This result is a quantitative demonstration of the relationship between jnds and masking originally hypothesized by Miller [G. A. Miller, J. Acoust. Soc. Am. 60, 115--142 (1947)]. It was concluded that (a) the difference in masking ability of tones versus sub-critical bandwidth noise depends on the phase relationship between the probe and masker and (b) that the masked threshold for a tone in a two-interval forced choice (2IFC) procedure can be predicted from the jnd[sub I] of the masking stimulus.