FIG. 2.

(Color online) Mean SRTs measured in experiment 1, obtained by musicians (left panel) and non-musicians (right panel), for a monotonized target voice against speech-modulated buzzes (squares) or against two interfering voices (circles), presented either diotically (i.e., same input on both ears, filled symbols) or dichotically (i.e., different input on opposite ears, empty symbols). Lower thresholds indicate greater intelligibility. Here and later, error bars are +/− one standard error of the mean over listeners, but may not be visible given the scale of SRTs. Here, asterisks indicate a significant effect of ΔF0.