The rediscovery of diffuse reflection in room acoustics prediction

Abstract

Around 1986 ‘‘hybrid methods’’ appeared where specular ray tracing was used to speed up image source validation and were implemented in a range of software (sometimes in the form of specular cone tracing). Unfortunately, these methods neglected diffuse reflection and room acoustics prediction suffered. Consequently, most software that started out with such methods have today either changed their algorithms to incorporate diffuse reflection or use them only to predict the early part of the echogram. A possible reason for the popularity of these methods was the more detailed point-receiver echogram, but already in 1980, Kuttruff published an Acustica paper clearly indicating the importance of handling diffuse reflection. Nevertheless, software exist today, and are presented in journals and doctoral theses, that neglect diffuse reflection. One reason could be that diffuse reflection does not always need to be taken into account since when rooms act geometrically mixing or have an even absorption distribution a specular-only prediction may suffice. However, if none of these conditions are met the RT may be severely overestimated while, on the other hand, Sabine/Eyring formulas will severely underestimate the RT. This paper presents examples where neglecting diffuse reflection leads to large prediction errors, especially regarding the RT.