ACOUSTIC ANALYSIS OF SPEECH SIGNALS (BONE-CONDUCTED SOUNDS) PICKED UP AT THE HEAD

• Published in: Nippon Jibi Inkoka Gakkai Kaiho Vol. 79; no. 9; pp. 963 - 972
• Main Authors: OHYAMA, MASARU, MIYOSHI, YASURO, SHOJI, KUNIO, YAMAMOTO, SHINPEI, TANIGUCHI, CHIEKO, WATANABE, SHINGO, INOMOTO, KENJI
• Format: Journal Article
• Language: English
• Published: The Oto-Rhino-Laryngological Society of Japan, Inc 1976
• ISSN: 0030-6622; 1883-0854
• DOI: 10.3950/jibiinkoka.79.963

The bone-conducted vowels were picked up by the vibration transducer of the accelerationtype applying to three anatomical sites on the heads (forehead, upper jaw and mastoid region)of the healthy adults. A comparative studies on the speech signals from these three regions of the head were done acoustically. The results were summarized as follows; i) for the pickup of speech signals from the body surface, the vibration transducer of small acceleration type (Rion Co. PV-06)was the most satisfactory in point of a signal-to-noise ratio, intelligibility of the speech and practical use. 2) the relative intensities of the speech signals were decreased in regular order i, u, e, o and a. 3) the spectral characteristics of these vowels were highly similar to those which were recorded from the micropone at the position in the distance of 30 cm from the lips, although there was an inclination of the increased low-frequency components with the decrease of high-frequency in the spectrum. 4) from the results of the articulation score, intelligibility of the vowels was much better in recorded sounds from the upper jaw and forehead compared with that from the mastoid region.