Multimedia 1:
This is an example of ideal time-frequency segregation applied to a speech signal masked by Gaussian noise (energetic masker) or by a 2-talker speech masker (informational masker). There are four pairs of stimuli. The first stimulus of each pair is glimpsed (i.e., consisting only of the target-dominated time-frequency units after ITFS processing). The second stimulus of each pair is the unprocessed target plus masker at the same target-to-masker ratio as was used for the preceeding glimpsed stimulus (i.e., the target information is the same only the masker has been added back in the masker-dominated T-F units). The following is the order of presentation of the pairs of sounds: 1) TMR = 0 dB for Gassian noise masker; 2) TMR = 0 dB for 2-talker speech masker; 3) TMR = -10 dB for Gaussian noise masker; 4) TMR = -10 dB for 2-talker speech masker. Again, within each pair the stimulus is presented glimpsed then nonglimpsed (unprocessed).