A new way to account for binaural detection as a function of interaural noise correlation

The purpose of this communication is to describe a new method for predicting masking level differences (MLDs) as a function of the interaural correlation of the masking noise. The general idea is to decompose a binaural noise masker having an arbitrary value of interaural correlation into two parts: an No constituent and an N pi constituent. It is shown that the amount of masking produced by a noise having an arbitrary interaural correlation is equal to the addition of the masking effects produced by the No and N pi constituents that compose the masker. The new approach was used to generate predictions of MLDs for N rho S pi and N rho So conditions. Those predictions accounted for 98% of the variance in the classical data of Robinson and Jeffress [J. Acoust. Soc. Am. 35, 1947-1952 (1963)]. Finally, it is shown that the "additivity of masking" method is mathematically equivalent to assuming that binaural detection consists of a stage of errorless binaural processing preceded by a corrupted representation of the stimulus.