Slow-Played Speech with and without Time Restoration: A Phonemic Analysis

Abstract

Intelligibility of speech divided by a frequency factor of two was explored in 12 young adults with normal hearing. Two conditions investigated were (1) 50% frequency division with 50% time expansion (FDTE), i.e., frequency halved and time doubled, and (2) 50% frequency division with 50% time-compression (FDTC), i.e., frequency halved and time restored. The latter condition was obtained by use of the time-compressor developed by Fairbanks el al. in 1953 and modified by Zemlin in 1970. Test stimuli were CVC words recorded by a male speaker and presented monaurally under phones. The response format was of the closed-set type. Test words were constructed so as to emphasize ten consonants in the initial word position, the same ten in final position, and ten vowels always in the medial portion. Results showed: (1) the initial consonants and the vowels were significantly more intelligible in the FDTE condition than in FDTC; (2) intelligibility of final consonants did not differ significantly in these conditions; (3) in the FDTE condition, the initial consonants were as intelligible as the final ones and both were significantly more intelligible than the vowels; (4) in the FDTC condition, the initial consonants did not differ significantly from the vowels and both were significantly less intelligible than the final consonants.