The effects of microphone placement and azimuth on peak sound pressure level of type A impulses

Abstract

Single, repeatable type A impulses were generated by colliding steel spheres in an acoustically reverberant room, and measured as A‐weighted sound levels by a sound level meter having four exponential time averaging modes with time constants ranging from 2 μs to 1 s. Measurements using peak hold (2‐μs time constant) were then obtained on KEMAR at four azimuths and four microphone placements at each azimuth. Comparative measurements showed peak hold to give the most accurate estimation of A‐weighted sound level, as verified electrically. The remaining three underestimated this value. Maximum underestimation of 34.8 dB was demonstrated using the slow time constant. KEMAR measurements demonstrated peak A‐weighted sound levels to be sensitive to both microphone placement and azimuth. Measurements made in the reference field and at external microphone placements tended to underestimate peak A‐weighted sound level at the tympanic membrane (TM). Two conclusions were reached. Time constant of the exponential time averaging must be shorter than the duration of the signal. Peak A‐weighted sound level cannot be predicted by measurements made at sites other than at the plane of the TM.