The generation and propagation of noise is an ongoing research topic that affects everyone. Urban environments in particular suffer from high levels of noise pollution, which can negatively impact the health and well-being of citizens. It is therefore essential to understand the propagation of sound from one or more sources to a receiver. However, human perception of sound is highly subjective, making generalisation and research very difficult. A common way to gain insight into sound propagation is through computer aided simulation. The image edge model is one such simulation model that identifies all significant sound paths from one or more sources to a receiver, before calculating transfer functions for each path. It is designed for urban environments and is based on the geometric acoustics approach, while including diffraction using the Uniform Theory of Diffraction. However, specular reflection is handled by simply reflecting on each surface with a constant reflection factor. This does not accurately capture the effect of different materials, which exhibit frequency and angle dependence, on the reflected sound levels. In this thesis, the existing image edge model workflow is revised, enabling material-aware specular reflection that considers these factors. Specifically, the hypothesis that material-aware specular reflection lead to different and more plausible transfer functions is investigated. To achieve this goal, two new implementation types are created and com- pared against the old, constant reflection factor. Validation results using the Benchmark for Room Acoustical Simulation (BRAS) dataset show, that material-aware specular reflection overall significantly outperforms the old implementation type, especially for absorbing materials. This supports the hypothesis that material-aware specular reflection leads to improved transfer functions.

Melden Sie sich hier an um Einladungen zu den Kolloquium-Vorträgen per E-Mail zu erhalten.
Register here to receive the invitations to colloquium talks via e-mail.

Zoom-Meeting-ID: 954 4073 3814
Passwort: 450783