Transcribed features heard in repeated noise

When noise is repeated, listeners report hearing repeating percepts within the noise. The acoustic cues that contribute to these percepts are not known, and they may vary from listener to listener. To test this, we asked musically trained listeners to transcribe the repeating perceptions using musical notation, including pitch where possible. They completed three transcriptions of each of five noises (and a control sound with embedded features). Within listeners, pitch chromas were reported for the same noises at approximately the same time more than for different noises, implying that their pitched percepts were somewhat reliable. However, between listeners, there were not significantly more coincidences for pairs of the same noises than pairs of different noises, in line with listeners having more idiosyncratic perceptions.