Predicting Speech Intelligibility in Rooms from the Modulation Transfer Function. Part 4: A Ray Tracing Computer Model

摘要

A computer model for calculating the speech transmission index (STI) in an auditorium which takes into account the geometrical and acoustical characteristics of room boundaries is described. The speaker is simulated by a point source emitting a large number of rays, e.g., 7000, and the audience by an area of spheres (radius 1 m), resulting in an STI value for each listener position. The model accounts for the influence of partly diffuse reflections, resulting in a hybrid model, i.e., a combination of ray tracing acoustics and statistical room acoustics. Model predictions agree quite well with STI values as measured for a variety of conditions.