Perception of tones versus tone streams under random multi-tone masking and its representation in auditory cortex

When repeated tones are presented among random masker tones, their perceptual organization is often variable. In some trials, they cannot be segregated from the masker tones. In other trials, often with a delay, they form coherent streams, segregated from the masker tones. A strong awareness-related negativity is evoked by each segregated tone of the stream in auditory cortex. A similar negative-going response has been observed when subtle, isolated tones are detected versus missed, raising the question of whether target tones that do not form a stream under informational masking are still perceived or not. Here a paradigm is introduced to probe this question while activity in auditory cortex was recorded with magnetoencephalography. Human listeners (13 female, 8 male) were instructed to indicate when they detected an isochronous stream of tones under multi-tone masking while listening to the stimulus. Thereafter, they indicated if a cue tone, presented subsequently to each trial, was comprised in the target and masker combination. Results confirmed that target tones could still be perceived in the subset of trials where they did not form a coherent stream, but the detectability in the cued frequency report was higher for tones forming a stream. A prominent awareness-related negativity was only observed when the tones formed a stream, suggesting that this response reflects the perceptual organization of the tones into streams and beyond the pure perception of the tones themselves. We speculate that the perceptual organization into a stream further enhances the detectability of the tones comprised in the stream. Significance Statement Tone sequences may form streams that segregate from competing aspects of the auditory scene, a phenomenon considered comparable to visual object grouping. In multi-tone masking, hearing out streams of repeated tones seemingly enhances tone detectability by a degree that makes it difficult to judge whether these tones could at all be perceived if they had not formed a coherent stream. We present a paradigm that allows for probing both, the perception of target-tone frequency and the formation of coherent streams in the same trials. Applying this setup in magnetoencephalography showed that the perceptual organization into a stream is more strongly represented in auditory cortex than the ability to correctly detect a tone frequency within a random tone masker.