Optimizing the signal-to-noise ratio in speech rooms using passsive acoustics.

Abstract

Adults with normal hearing require a roughly 0‐dB signal‐to‐noise ratio for good speech intelligibility in classrooms and lecture halls. However, significantly higher values may be needed to compensate for neurological immaturity, sensorineural and conductive hearing losses, language proficiency, and excessive reverberation. ANSI 12.60 addresses ways to lower the noise interference due to background levels and reverberation time. However, it is also possible to increase the signal, by reflecting or diffusing early reflection. While speech power is delivered in the vowels which are predominately in the 250–500‐Hz frequency range, speech intelligibility is delivered in the consonants, which occur in the 2–4‐kHz frequency range. Therefore, effective core learning designs can incorporate scattering surfaces, rather than surfaces that absorb in the 2–4‐kHz region, on the front wall, lower side walls, and central ceiling areas, to increase the speech signal. The decay time can be controlled with broadband absorption on the perimeter of the ceiling and upper wall surfaces. A computer model analysis of various speech environments will be presented.