Studying the Effect of Carrier Type on the Perception of Vocoded Stimuli via Mismatch Negativity

Vocoder processing has been long used in many studies to examine how acoustic cues affect speech understanding and auditory processing. Early behavioral studies have shown that the type of carrier (i.e., pure-tone or noise) used in vocoding process affected the intelligibility of the perceived speech, and tone-vocoded stimuli had a perceptual advantage over noise-vocoded stimuli. This work further assessed whether the auditory evoked cortical response could objectively measure the perceptual difference between the two types of vocoded stimuli using an oddball-paradigm based event-related potential (ERP) experiment. A vowel stimulus was processed by noise- and tone-vocoding processes, and the processed stimuli were presented to normal-hearing listeners in an ERP experiment. The noise-vocoded and tone-vocoded vowel stimuli served as the deviant stimuli and the non-vocoded vowel stimulus as the standard stimulus. Experimental results showed that tone-vocoded stimulus evoked a significantly larger mismatch negativity (MMN) amplitude and a significantly shorter MMN peak latency than noise-vocoded stimulus did. Results in this work suggested that compared to noise-vocoded stimulus, tone-vocoded stimulus had a larger perceptual difference relative to the reference stimulus, and this effect caused by the usage of different carrier signals could be reflected by the MMN response.