Fig. 2. Experiment 1 (passive hearing): classification of filtered impulses by three bats (A–C) and their means ± s.e.m. (D). The performance is plotted as percentage of highpass classification versus the filter slope. Trials with presentation of white-noise background are shown with black symbols, and trials with pink-noise background are shown with red symbols. The two training/control conditions (±6 dB/octave and white noise) are indicated with open symbols. The grey bars show the difference in highpass classification between pink- and white-noise background. Their mean difference across all six filter slopes (indicated by M) is shown as a black bar. The perceptual classification boundaries per background are indicated by vertical lines in the lower part of each panel (black, white-noise background; red, pink-noise background). The shift in the perceptual classification boundary from white- to pink-noise background is given above the vertical lines in units of dB/octave. Only the mean perceptual classification boundary could be tested for significance. ^*P<0.05, ^**P<0.01.