Detail relevance table for wave-based simulations
Explanation
When running the wave-based solver for the most accurate sound simulation, the air in the room is cut into small 3D triangles into which the sound waves are propagated.
The speed of the wave propagation algorithm is directly linked to the smallest 3D triangle in the whole model, which means that 1 unnecessary piece of detail or 1 small air gap between objects can have drastic consequences on the runtime of wave-based simulations. If the smallest length is 5 times smaller than the 2nd-to-smallest length, then running the simulation will take 5 times longer.
Both edges of objects (especially in almost-flat objects) and gaps between objects can affect the runtime, so the smallest gap between two objects and/or the smallest edge are what we refer to as the level of detail.
You can see more how it impacts the computation time on our Estimation compute page.
Table
Herebelow is a table containing an optimal level of detail according to the desired transition frequency, so that you can ensure that too detailed objects/gaps are not included in the model (especially when running a simulation with a higher transition frequency):
| Transition frequency | Optimal level of details |
|---|---|
| 177Hz* | 0.4m |
| 355Hz* | 0.2m |
| 710Hz | 0.1m |
| 1420Hz | 0.05m |
| 2840Hz | 0.025m |
*These transition frequencies most often do not contain a lot of 3D triangles due to their size and their runtime is likely to be low to start with. Multiplying the simulation runtime by 5 could simply end up being 5 minutes instead of 1 minute, but it becomes more and more problematic as the size of the space - and the simulation runtime - increases.
How to spot it?
When checking your model in either the web app or the Treble's SketchUp plugin, you will receive relevant feedback regarding the presence of such issues, as detailed here.