Analysis of Intelligibility of Tissue-Conducted Speech Signals
Herbert J. Oyer · The Journal of the Acoustical Society of America · 1961
Previous research has revealed that acoustic energy resulting from vocal signals differs in sound pressure level as a function of (1) the anatomical sites at which signal pickup takes place, (2) the composition of the vocal signals, and (3) the fidelity of the recording systems. It has been shown experimentally also that vocal signals picked up out of the external auditory canals of speakers are significantly more intelligible at −15 db and −18 db S/N ratios than are signals simultaneously recorded at the lips. Phonetic analysis of the signals recorded at ear and lips points toward those sounds which cause the ear-recorded signals to be superior. They are the vowels e, ɛ, i. The present research is directed toward a determination of the relative intelligibility of words as they were recorded from six subjects representing two specific body types. Recordings were made at the following anatomical locations: (1) frontal bone of skull, 1 in. above nasalis, (2) mental protuberance of the mandible, (3) immediately below superior thyroid notch of the laryngeal prominence, and (4) at the angle of the mouth. Results of intelligibility testing point toward the use of anatomical sites in addition to the lips for recording of voiced signals. (This research was sponsored by an All-University research grant No. 2771, Michigan State University.)