Echo thresholds for reflections from acoustically diffusive architectural surfaces

Abstract

When sound reflects from an irregular architectural surface, it spreads spatially and temporally. Extensive research has been devoted to prediction and measurement of diffusion, but less has focused on its perceptual effects. This paper examines the effect of temporal diffusion on echo threshold. There are several notable differences between the waveform of a reflection identical to the direct sound and one from an architectural surface. The onset and offset are damped and the energy is spread in time; hence, the reflection response has a lower peak amplitude, and is decorrelated from the direct sound. The perceptual consequences of these differences are previously undocumented. Echo threshold tests are conducted with speech and music signals, using direct sound and a simulated reflection that is either identical to the direct sound or has various degrees of diffusion. Results indicate that for a speech signal, diffuse reflections are less easily detectable as a separate auditory event than specular reflections of the same total energy. For a music signal, no differences are observed between the echo thresholds for reflections with and without temporal diffusion. Additionally, echo thresholds are found to be shorter for speech than for music, and shorter for spatialized than for diotic presentation of signals.

Effects of the Distance from a Diffusive Surface on the Objective and Perceptual Evaluation of the Sound Field in a Small Simulated Variable-Acoustics Hall

“…Especially in enclosed spaces, humans are quite sensitive to the perception of sound in all its temporal, spectral, and spatial aspects, which makes a realistic auralization in room acoustics quite challenging. These three aspects are strongly affected by the presence of diffusive surfaces [36], thus more insight into their use is needed.…”

Investigation of Auditory Objects Caused by Directional Sound Sources in Rooms

“…An energy interpretation was successful in the work of Robinson et al on localization of diffuse reflections [15] and in [25] to model primary auditory events caused by multiple loudspeakers. Consequently, a vector approach was also used in [17] to model the orientation-dependent localization of directional sound sources in a room, which we largely adopt here.…”

“…(8) The precedence threshold given in [3,15] decays by minus 1 4 dB/ms. In this paper we use the threshold raised by a factor of two to predict our experiment well:…”

“…A ring of 24 Genelec 8020 loudspeakers (15 • interloudspeaker spacing) with a radius 1.5 m was placed in the anechoic chamber at ear height of the subject (1.25 m) sitting in the center of the arrangement.…”

“…This is known to be the case whenever the level of a reflected sound exceeds the echo threshold of the preceding sound. The knowledge of the localization itself is still rough and can often only be guessed from what is known [2,3,[14][15][16]. A more specific but still coarse description of the perceived direction was given in [13,17].…”