Chapter Excerpts

Chapter 10: Designing For Speech Intelligibility
The Receiver-Listener

Herein is the most challenging component. It can be analyzed but you have no direct control over it. Inarticulate talkers or handicapped listeners or some combination of both may be insurmountable and become the critics of your design.

Speech intelligibility centers on the 2 kHz octave band and so does the peaked response at the listener’s eardrums. Note that at 2,500 Hz the highest pinnae response is 72 dB and the lowest is 58 dB or a 14 dB difference. In tests that we conducted during an intelligibility workshop at Indiana University, we found that a person with a high pinnae response in the 2,500 Hz region but with a poor frequency response, will score higher on an intelligibility word test than the person with a low pinnae response at 2,500 Hz but with near perfect frequency response curve. This information was gathered from 30 listeners that had very acceptable audiometric charts of their hearing. The pinnae response is largely due to the ear lobe configuration and the individual ear canal resonances.

Hearing sensitivity, ear configurations that allow significant differences between ears and training in focusing on speech can all provide very significant variations in speech intelligibility scores in individual listeners. For example, musicians will often ignore high harmonic distortion in a recording but immediately detect a wrong note or a pitch due to wow or flutter. Analysis of listeners is a psychoacoustic task and awareness of listener types—ministers, choir directors, organists, allows “weighting” of their testimony.