Predicting speech intelligibility in university classrooms using geometrical acoustic simulations

In university classrooms a suitable acoustic condition is necessary to enhance the productivity of students and the vocal comfort of lecturers. With this purpose international standards recently introduced new requirements on the quality of verbal communication. The goal of this work is to propose a procedure for predicting speech intelligibility in large learning spaces through geometrical acoustic simulations. The performance of the present approach is investigated using analytical prediction models as well as the measurements results in six university classrooms surveyed.