Validation of room acoustic simulation models
Aspöck, Lukas; Vorländer, Michael (Thesis advisor); Weinzierl, Stefan (Thesis advisor)
Berlin : Logos Verlag GmbH (2020, 2021)
Book, Dissertation / PhD Thesis
In: Aachener Beiträge zur Akustik 35
Page(s)/Article-Nr.: 1 Online-Ressource (VI, 180 Seiten) : Illustrationen, Diagramme
Dissertation, RWTH Aachen University, 2020
Already in ancient times, the goal existed to design appropriate environments to transport acoustic signals from a sound source to an audience. Achieving this goal requires an understanding of room acoustic effects and, in the best case, the ability to accurately predict them. Simulations based on the concepts of geometrical acoustics have been developed and extensively investigated for more than 50 years. However, even today, there is still insufficient knowledge about the uncertainty and validity of the acoustical simulation results.In order to offer developers and researchers a possibility to comprehensively validate the results of their room acoustic simulation, a database with eleven acoustic scenes was developed, documented and published. The scenes cover both simple environments, where individual acoustic effects such as reflection or diffraction are isolated, and complex room situations such as a concert hall. In addition to a detailed description and the definition of input data, acoustic measurements were carried out for all scenes, which serve as a reference for simulations. Based on this data, a first study is presented which was designed following the concept of three previously conducted round robin experiments on room acoustic simulation. Here, the simulation results of up to six participants, who were not informed about measurement results in advance, are compared with each other and with the results of the corresponding measurements. The evaluation of the simple scenes revealed various weaknesses of the simulation software. In case of the complex scenes, acceptable results were observed in the mid-frequency range. In a second study it was determined how much room acoustic simulations deviate from measurements if the user knows the measured results beforehand and input data of the simulation is adjusted manually or systematically. This work demonstrates which deviations from actual, measured values can be expected when simulating either unknown or known rooms. It furthermore provides a basis for the revision of room acoustic simulation software and for the improvement of the determination and selection of correct input data.