Sami A. Khaiyat
Dept. of Architecture, Texas A&M Univ., P.O. Box 2844, Bryan, TX 77805-2844
Lester L. Boyer
Texas A&M Univ., College Station, TX 77843
Two different methods, the calculation of the articulation index (AI) and the rapid speech transmission index (RASTI) measurement using the speech transmission meter, were utilized to obtain the speech intelligibility in spaces with concave sound reflecting surfaces. Many factors were considered such as room size, size of curvature, position of the sound source, and background noise level. The base cases, spaces without curvatures, showed highly correlated results indicating no significant differences between calculated AI and measured RASTI. With curvatures and the sound source at 4 ft. away from the center position of the front wall it was found that the effect of room size is significantly different between the two methods under the 2- and 16-ft. radius curvatures only; also, the background noise level showed almost the same effect on both calculated AI and RASTI. Under low background noise levels, spaces with curvatures showed that the effect of change in room size for both methods is identical. The center location for the sound source showed better speech intelligibility under all different testing conditions. Finally, a modification factor is developed and applied to the calculated AI so that reliable estimates of speech intelligibility in spaces with curvatures may be obtained.