Fig. 3.

Grand-average (25 adults) fronto-central (FCz) ERPs elicited by standard (thin dotted lines) and deviant (thin solid lines) sounds for horn (top panels), speech (middle panels), and frog croak (bottom panels). The deviant-minus-standard difference signal is shown in thick lines with the significant MMN responses marked by grey shading. The deviant feature(s) are indicated in each panel: pitch (P), resonance (R), both pitch and resonance (PR), and their direction (+/−). The panels in the three left-most columns are for larger-source deviants ( −) and the three right-most are for smaller-source deviants (+). The vertical dashed lines indicate the perceptual onset of the sounds.