Sounds in noise: Behavioral and neural studies of illusory continuity and discontinuity

ability to parse an auditory scene into meaningful components varies greatly between individuals; some are able to parse out and write down competing musical pieces while others struggle to understand each word whenever they have to converse in a noisy environment. Using a simple discrimination task, healthy, normally-heari ng adult participants were asked to judge whether a pure tone (with or without amplitude modulation) was continuous or contained a gap. One quarter of the participants consistently heard a gap when none was present, if the tone was accompanied by a higher-frequency noise burst with a lower edge beginning one octave away from the tone (that did not have any energy overlapping the tone). This novel form of informational masking (perceptual interference between components with non-overlapping sound energy) was named 'illusory auditory discontinuity’. The phenomenon appears to reflect natural differences in auditory processing rather than differences in decision-making strategies because: (1) susceptibility to illusory discontinuity correlates with individual differences in auditory streaming (measured using a classical ABA sequential paradigm); and (2) electroencephalographic responses elicited by tones overlaid by short noise bursts (when these sounds are not the focus of attention) are significantly correlated with the occurrence of illusory auditory discontinuity in both an early event-related potential (ERP) component (40-66 ms), and a later ERP component (270-350 ms) after noise onset. Participants prone to illusory discontinuity also tended not to perceive the ‘auditory continuity illusion’ (in which a tone is heard continuing under a burst of noise centered on the tone frequency that completely masks it) at short noise durations, but reliably perceived the auditory continuity illusion at longer noise durations. These results suggest that a number of attributes describing how individuals differentially parse complex auditory scenes are related to individual differences in two potentially independent attributes of neural processing, reflected here by EEG waveform differences at ~50 msec and ~300 msec after noise onset. Neural correlates of the auditory continuity illusion were also investigated by adjusting masker loudness, so that when listeners were given physically identical stimuli, they correctly detected the gap in a target tone on some trials, while on other trials they reported the tone as continuous (experiencing illusory continuity). High er power of low-frequency EEG activity (in the delta-theta range, <6 Hz) was observed prior to the onset of tones that were subsequently judged as discontinuous, with no other consistent EEG differences found after the onset of tones. These data suggest that the occurrence of the continuity illusion may depend on the brain state that exists immediately before a trial begins.