Multitone Masking of Selected Speech Material.
AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OHIO SCHOOL OF ENGINEERING
Pagination or Media Count:
A continous electronic noise whose amplitude statistics and frequency spectrum approximate the amplitude statistics and frequency spectrum of continous speech was generated by combining four frequency modulated sine waves and filtering the combination signal. The masking effect of the multitone noise was compared to the masking effect of a white noise source for the vowel rhyme test material. Within the range of SN ratios studied, the multitone noise creates probabilities of error comparable to the white noise source with approximately 2.5 dB higher SN ratios. The results clearly indicate that the multitone noise is more effective in masking the vowel sounds than white noise. Author