Spatial Audio

Continuous Ray Casting

Spatial Audio uses a raytracing engine to efficiently evaluate reflection and diffraction paths. In the previous version of Wwise, when the emitter or listener movement exceeded a predefined threshold (see the Movement Threshold Spatial Audio Setting), a number of rays were cast, and the reflection and diffraction paths were recalculated. The combination of the ray casts with the recalculation increased the load on the frame, which caused CPU peaks. In Wwise 2023.1, rays are continuously cast on each frame until a recalculation is triggered. This helps spread the load of ray casting over several frames, which prevents CPU peaks.

For this reason, the default value of the Number Of Primary Rays Spatial Audio Setting has been lowered from 100 to 35. Make sure you re-evaluate the preferred number of primary rays for your project after migrating.

Watertight Geometry

In 23.1, we have introduced new Spatial Audio features that require the Geometry of Rooms to be watertight in order to calculate containment and proximity to transition regions. This means that there can be no holes in the Geometry; every edge connects exactly two triangles. When a Room is sent to Spatial Audio, the Room's Geometry is searched for boundary edges. If any boundary edges are found, the Geometry is not watertight and an error message appears.

If the Room is a Reverb Zone, its Geometry must be watertight. If you encounter the error message, fix the Geometry by looking for holes and boundary edges. Any hole can be covered by a surface with a Transmission Loss value of 0. If there are no apparent holes, look for boundary edges caused by overlapping edges or surfaces.

Note that it is also important to have a watertight Geometry if you depend on Spatial Audio to do room containment. However, the Unity integration does not use this feature by default. In any other scenario, you can disregard the error message.