Spectral Characteristics of Speech at the Ear Implications for Amplification in Children Research Article
Research Article  |   June 01, 2003
Spectral Characteristics of Speech at the Ear
 
Author Affiliations & Notes
  • Andrea L. Pittman, PhD
    Boys Town National Research Hospital Omaha, NE
  • Patricia G. Stelmachowicz
    Boys Town National Research Hospital Omaha, NE
  • Dawna E. Lewis
    Boys Town National Research Hospital Omaha, NE
  • Brenda M. Hoover
    Boys Town National Research Hospital Omaha, NE
  • Contact author: Andrea Pittman, PhD, Boys Town National Research Hospital, 555 North 30th Street, Omaha, NE 68131. E-mail: pittmana@boystown.org
Article Information
Hearing Disorders / Speech, Voice & Prosody / Hearing / Research Articles
Research Article   |   June 01, 2003
Spectral Characteristics of Speech at the Ear
Journal of Speech, Language, and Hearing Research, June 2003, Vol. 46, 649-657. doi:10.1044/1092-4388(2003/051)
History: Received June 17, 2002 , Accepted January 22, 2003
 
Journal of Speech, Language, and Hearing Research, June 2003, Vol. 46, 649-657. doi:10.1044/1092-4388(2003/051)
History: Received June 17, 2002; Accepted January 22, 2003
Web of Science® Times Cited: 16

This study examined the long- and short-term spectral characteristics of speech simultaneously recorded at the ear and at a reference microphone position (30 cm at 0° azimuth). Twenty adults and 26 children (2–4 years of age) with normal hearing were asked to produce 9 short sentences in a quiet environment. Long-term average speech spectra (LTASS) were calculated for the concatenated sentences, and short-term spectra were calculated for selected phonemes within the sentences (/m/, /n/, /s/, /∫/, /f/, /a/, /u/, and /i/). Relative to the reference microphone position, the LTASS at the ear showed higher amplitudes for frequencies below 1 kHz and lower amplitudes for frequencies above 2 kHz for both groups. At both microphone positions, the short-term spectra of the children's phonemes revealed reduced amplitudes for /s/ and /∫/ and for vowel energy above 2 kHz relative to the adults' phonemes. The results of this study suggest that, for listeners with hearing loss (a) the talker's own voice through a hearing instrument would contain lower overall energy at frequencies above 2 kHz relative to speech originating in front of the talker, (b) a child's own speech would contain even lower energy above 2 kHz because of adult-child differences in overall amplitude, and (c) frequency regions important to normal speech development (e.g., high-frequency energy in the phonemes /s/ and /∫/) may not be amplified sufficiently by many hearing instruments.

Acknowledgments
This work was supported by Grant R01 DC04300 from the National Institutes of Health. We gratefully acknowledge the contributions of Maureen Higgins and Mary Pat Moeller on previous versions of this paper.
Order a Subscription
Pay Per View
Entire Journal of Speech, Language, and Hearing Research content & archive
24-hour access
This Article
24-hour access