Abstract
An attempt is made to increase the intelligibility of speech sampled in time (interrupted speech) by filling in the blank spaces with repetitions of the speech samples (reiterated speech). The experiment involves the evaluation of these processes of speech distortion in quiet and in noise with two duty cycles (speech-time fractions), with various rates of interruption of the speech signal, and with various buildup and decay times in the switching operation. The results indicate that (1) there is little difference between the intelligibility of reiterated speech and interrupted speech in quiet, (2) reiterated speech is more intelligible than interrupted speech in noise if signal measurements are based on instantaneous amplitudes of the speech samples, and (3) buildup and decay has little effect on the intelligibility of interrupted speech when the entire buildup phase is confined to the “on” time of the switching period.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
