Tongyu Zhou, April 2022
EVERTims is an open-source framework for 3D, real-time auralization in architectural acoustics and virtual reality. Using Blender for room modeling, it is able to predict the acoustics of any given room by passing geometry and material information into its own custom raytracing client. This client simulates acoustic wave propagation, then sends this information to an auralization engine to reconstruct ambient sound for biaural listening. A full flow-graph of this process is visualized below:
This project was developed at as a collaboration between LIMSI/CNRS and the TKK/Department of Media Technology.
When following the instructions after curling the repository, make sure you make all the shell scripts executable (chmod u+x). Otherwise, following the installation instructions on the site returns a permission denied error.
After installing Blender, download the EVERTims add-on and follow the corresponding instructions.
EVERTims functions as an add-on plugin on top of Blender, and thus all file formats importable into Blender can be visualized. For best results, import closed rooms into the scene. Standard audio formats (.wav, .mp3) can be imported into the auralization engine.
(more detailed instructions are located on the webpage itself)
After installation, open a new Blender scene and navigate to the EVERTims tab.
Click on "template scene" in the import element section.
Start the auralization engine from Blender and import an audio file of your choice. Start playing this audio.
Start the raytracing client from Blender.
Click on "on the fly auralization" to see visualizations.
Minimum System Requirements
OS: Mac OSX 10.6+, Linux*, Windows*
I did not actually install and test the software on Linux or Windows, but on the github this is indeed a shell script for Linux installation. There is no Windows equivalent, so the installation would have be achieved manually by following the commands inside the shell scripts.
Graphics: OpenGL 2.1 w/ 512 MB RAM
Ram: 2 GB, 1 GB available space
Processor: 32-bit dual core 2Ghz CPU with SSE2 support
EVERTims is an open source framework and thus is available for free.
Time to Setup: Familiarity with Blender helps a lot
Beginner: No coding or graphics experience | 40 min
Intermediate: Some coding or graphics experience | 20 min
Advanced: Both coding and graphics experience | 15 min
EVERTims supports the following features:
Real-time auralization (modifications to EVERTims elements (room mesh, audio source, audio listener, etc.) automatically trigger uploads to the raytracing client
Real-time input (if you have a microphone)
Powerful rendering engine due to sitting on top of Blender
Binaural audio listening
This is a powerful software that allows users to visualize how different room materials affect the propagation of sound. Changing room materials and audio sources is easy and the re-rendering each time is fast and feels smooth.
It is a little difficult to get started from scratch without using the template. There are a couple of small confusion points (sometimes, the auralization does not match the orientation of the mesh object representing the audio source), which can be improved with better affordances.
White lines are drawn onto the scene to clearly depict the raytracing process. This is very useful and information as it demonstrates how the sound waves interact with the environment. However, with multiple audio sources in a small scene, the lines may get chaotic and it is difficult to differentiate which path belongs to which sound source.