Fig. 4
From: Ideal-observer model of human sound localization of sources with unknown spectrum
The average polar error rate (a, c) as function of ripple density in case of phase 0 and phase \(\pi\) (40dB ripple depth); (b, d) as function of ripple depth for phase 0 and \(\pi\) (1 ripple/oct); (e) as function of ripple phase (1 ripple/oct, 40dB ripple depth). The experimental results are shown in black (the whiskers mark the minimal and maximal error rate of the 6 subjects); the different model predictions considering the different features are shown in color.
