Author(s): Meng Fanyu, Georgiou Fotis, Stienen Jonas, Hornikx Maarten, Vorländer Michael
Urban environmental noise consists of various components, among which a significant contribution is caused by moving vehicles such as cars and trains. To auralize urban acoustic scenarios, appropriate synthesis of moving sound sources and sound propagation modelling are critical to achieve plausibility. This work introduces a concept of auralizing pass-by vehicles with synthesized signals of moving sound sources based on measurements as well as measured and simulated impulse responses (IRs) in urban environments. First of all, two virtual environments are considered in this work: a street without buildings and a street canyon. Further scenario information is set up, such as initial location, source speed and moving orientation. Second, the spatial locations and signals of the moving sources are extracted from array recordings of pass-by vehicles by using beamforming. Moreover, IRs in line with the virtual scenarios are generated at discrete locations on the moving trajectories by measurements and simulations. IRs in the first scenario are computed with the pseudospectral time-domain method (PSTD) and image source method. In the second scenario, IRs are obtained from measurements and simulations with a geometrical acoustics software. Finally, auralizations are implemented by convolving the source signals, and cross-fade windows are used to deliver smooth transitions between impulse responses. Source dynamics and resulting Doppler-shifts are then applied by resampling the audio stream according to the relative movements of the direct source-receiver-paths.
Name: Mr Fanyu Meng