Auditory Model of Intensity Discrimination and Vowel Formant Discrimination: Effect of Signal Frequency

An auditory model, excitation-pattern model, was used in the present study to interpret the effects of signal frequency on intensity discrimination of tonal signals and vowel formant frequency discrimination. Excitation patterns were computed for standard and intensity-changed tones and for standard and formant-shifted vowels. Simulation models based on these excitation patterns were used to predict thresholds of intensity discrimination and thresholds of vowel formant discrimination for F1 and F2 frequency of four American English vowels. Predicted thresholds showed high correlations and low root-mean-square errors with thresholds measured from human listeners for an auditory metric calculating the excitation level near the target formant. Results indicated that normal-hearing listeners needed a relatively constant change in the internal representation of signals, e.g. excitation patterns, near the signal/formant frequency to discriminate changes in tonal intensity and vowel formant frequency, independent of signal frequency. In addition, the excitation-pattern model may be used more broadly to explain discrimination of acoustic signals in a variety of listening conditions.