MithraSound is a unique simulation tool capable of audio rendering for outdoor urban soundscapes. It integrates 3D visualization of the modeled site, and couples quantitative simulation results with an immersive listening experience. A high level of realism is achieved thanks to several advanced synthesis techniques. Both at the design stage and for collaborative and communication purposes, MithraSound is a great solution for evaluating perceptively the acoustic impact of urban projects.
MithraSound advantages
Variety of noise sources
To create realistic outdoor urban soundscapes, MithraSoundSound sources can be modelled in several ways: point or line sources, moving point sources along user defined trajectories, or background noise.
Dynamic traffic simulation
Audio rendering of transportation noise uses an integrated real-time traffic engine which models individual vehicle motion. The traffic model includes numerous parameters to precisely control the flow of vehicles on the network, including crossing priorities, traffic lights and speed limit zones.
Realistic vehicle noise synthesis
Relying on a CSTB patented approach, MithraSound generates vehicle noise in real time with excellent realism. Engine noise is synthesized based on engine speed, which continuously varies as a function of vehicle cinematic parameters (speed, acceleration, transmission, wheels, etc.) and driver behavior. Tire/road noise synthesis is determined by vehicle speed and road surface type. Overall, the vehicle synthesis allows accurate rendering of non-stationary traffic noise commonly found in urban situations with numerous accelerating and decelerating vehicles.
Advanced propagation modeling
All acoustic calculations in MithraSound use the sound mapping computation engine of MithraSIG. This engine developed by CSTB implements most of the standardized methods for outdoor noise prediction (NMPB 2008, ISO 9613-2, CNOSSOS-EU, …). It uses a fast ray tracing algorithm coupled with the calculation of all significant propagation effects between a source and a receiver (diffractions and reflections on buildings or noise barriers, ground effects, road surface absorption, …). The same propagation effects are transformed into advanced signal processing nodes to achieve accurate and artifact free audio rendering of the scene.
Calculation of dynamic noise indicators
During a simulation, MithraSound computes and records the time evolution of the instantaneous sound pressure level at the listener position, as would be obtained from a real sound-level meter. Standard noise indicators such as the maximum level LAmax or statistical levels LA10, LA50, LA90 are thus available to characterize the predicted noise exposure.
3D audio spatial rendering
For an optimal listening experience, MithraSound integrates several types of spatial audio algorithms to suit various available 3D audio systems:
- Stereo (two-loudspeaker systems)
- binaural (headphone systems)
- Ambisonics and VBAP (multi-loudspeaker systems)
A specific audio calibration procedure gives the user the possibility to guarantee calibrated levels for the listener.
MithraSound is the result of several years of research by CSTB in the field of outdoor sound propagation and auralization. The auralization techniques were initially implemented in By-OASIS, a construction site noise simulator developed for Bouygues TP. Integration of traffic simulation with vehicle engine and road noise synthesis then followed, which finally made the creation of MithraSound possible. It was used for the first time by CG59 (Conseil Général du Nord, France) for public meetings on three new road infrastructure projects.
