Assessment of the human ability to reproduce known sounds via a synthesis procedure

One method used to measure the properties of tinnitus perception is allow listeners to “synthesize” their tinnitus by summing sounds of various center frequencies, intensities, and bandwidths. The values of the parameters can be set by the listener, and the acoustic parameters of the resultant sound are thought to describe the characteristics of an individual’s tinnitus. However, the validity of this method is unknown. Here, we present data establishing the abilities of normal-hearing listeners to reproduce the acoustics of known stimuli including pure tones, complex tones, and noises containing various bandwidths and frequencies. Of particular interest is the validity of this method in estimating acoustic properties of high-frequency sounds >8kHz. Results will be interpreted in the context of the limits of auditory perception and in the context of data collected on listeners reporting tinnitus of tonal and noisy qualities.