Observing mismatch fields from human auditory discrimination of heavily damped frequency component adjustments

The aim of this study was to test the brain's ability to discriminate frequency shifts of heavily damped components of a complex sound. Since the auditory system is tonotopically organized, heavily damped transients should generate more activity than lightly damped partials. The increased activity increases the categorical resolution'. This work attempts to test this by detecting the Mismatch Field' from human subjects during a defined hearing task. Since the hearing system evolved in a complex environment, complex synthetic audio samples are constructed similar to the sound of a guitar. The sounds presented to the subject consisted of repetitive 'standard' tones, with occasional 'deviant' epochs. Standards differ from deviants in that one of the anharmonic frequency components of the heavily damped partials is transposed. The partials representing the more sustained vibrations of the string are identical for both standard and deviant. The randomized samples were presented to the subject's right ear with an average random inter-stimulus interval of 1 second over a duration of 17 minutes. The subject was then submitted to an active discrimination task where they actively attempted to count the deviants. The magnetic field responses from the brain were then detected using whole head magnetometer. The grand averaged samples were then band-pass and localized. Although the subjects could not actively discriminate the sound samples, mismatches were observed originating from the auditory cortex. These results suggest the auditory system is able to distinguish at a sensory level, shifts in the heavily damped spectral structure of complex sounds even though some cannot do so attentively. This could be due to the greater analysis given by the human brain to determining the pitch center rather then the sound timbre, yet some trained musicians have the ability to distinguish these subtle differences.