Acoustic Ray Tracing uses a precomputation step to determine the acoustics of an environment in advance, typically when creating the scene in the engine editor. This data is managed using a Meta XR Acoustic Map component, which serializes the acoustic data to a custom file format. This asset is then loaded at runtime and used to define the acoustics. The runtime component consists of a lightweight update which interpolates/extrapolates the precomputed acoustic properties to the current listener and source positions. Since almost everything is precomputed, this mode can be hundreds of times faster than dynamic simulation, and produces similar quality results. The downside is that anytime the scene geometries or materials are modified, the Acoustic Map must be baked again to hear the result (Note that control zones do not require rebaking to hear results).

Using an Acoustic Map requires the following steps:

To activate the precomputation, the first step is to add a single instance (regardless of how many geometry components) of the Meta XR Acoustic Map script anywhere in the scene. The Acoustic Map does not need to be attached to the same parent as the Acoustic Geometry. Whatever the Acoustic Map is attached to should not have any Scaling applied on its transform, as this is not supported for Acoustic Map.

The Acoustic Map manages the prebaked data for the scene. The prebaked data is defined using a collection of points placed in the scene, which are visualized by yellow spheres. These points are analogous to light probes used for graphics rendering. Each one captures the acoustic properties of the environment surrounding the point. This includes reverb, early reflections, and diffraction.

Before computing the acoustic properties of the scene, the component is first set up by placing the points throughout the environment. The acoustic map points can either be placed automatically, or manually in specific locations. The automatic point placement algorithm uses the geometry which was previously tagged to determine the ideal locations to place points. The algorithm tries to place as few points as possible while still spanning all walkable areas of a minimum size. It is possible to use a combination of automatic and manual point placement. You can use the automatic placement to place most of the points, and then add to or move those points to customize the locations. The automatic point placement can be done by pressing the Map Scene button. Regardless of whether Custom Points are enabled or disabled, Map Scene can be used to automatically generate points .

The location of an acoustic map point can be edited by clicking on it in the viewport, then using the usual transform controls. Points can be added at the center of the viewport by clicking the Add Point button, or removed by clicking the Remove Point button. You can also press the backspace key to delete the selected point. Any points that have been moved, or added, will be preserved if you map or bake the acoustic map again.

Once a scene is mapped, the acoustic data can be baked by pressing the Bake Acoustics button. This will do a series of ray tracing simulations to determine the acoustics at each point. Baking the acoustic map will also automatically map the scene if it has not yet been mapped. Baking can take anywhere from seconds to minutes depending on the complexity of the scene and the power of your computer.

If baking is successful, the Status of the Acoustic Map will display READY, indicating it is ready to be used in-game. If the scene has been mapped but not yet baked, the status will display MAPPED. If the acoustic map is neither mapped or baked, it will display EMPTY for the status. An acoustic map that is not READY will have no effect on the audio (it is treated as if the acoustic map was disabled).

Now you are ready to play the game and hear acoustics while taking advantage of the prebaking system.

For large worlds consisting of many sub-scenes there are two ways to approach acoustic mapping. The simplest and most robust way is to do a multi-scene bake for the entire space using Acoustic Scene Groups. The other way which, is more detailed, is to have an acoustic maps for each sub-scene and stream them in with the sub-scenes.

Note: In either case, the acoustic geometry (which is usually far larger than the map file) will be streamed with the sub-scenes. Only the acoustic geometry relevant to the area that is loaded will be present in memory.

To bake multiple scenes, at once create a Scene Group ScriptableObject and include every sub-scene. Then, reference that Scene Group in an Acoustic Map. When you bake that map, every scene will temporarily be opened in Unity and the acoustics for the entire space will be calculated. Make sure you have a GameObject with that acoustic map loaded once (and only once) the entire time you’re in the space.

In order for reverb to properly propagate between different acoustic maps, they must share precompute points at the boundaries so that the acoustic maps can be matched up. There are a few different ways to achieve this, with the most robust solution listed first:

If neither of the above are done to ensure continuity, acoustics will still produce correct sound within each sub-scene, but reverb from a source in one sub-scene will not be audible when the listener is in another sub-scene. This does not apply to direct sound, which will always be heard across sub-scene boundaries.

In order to support games that use a moving coordinate system origin, the acoustic map can be dynamically transformed using a 3D transformation, which includes position and rotation. You can see this in the unity editor by changing the position of the object to which the acoustic map is attached. This will cause the precompute points to be moved by the same translation. If your game uses a floating coordinate origin, the acoustic map game object can be moved so that the acoustic data stays at the correct position relative to the scene geometry when the origin changes locations. Changing the acoustic map transformation does not require re-baking, as long as the geometry is in the same relative position to the precompute points. For best performance, it’s recommended that you avoid changing the transform every frame when using multiple acoustic maps as discussed above. Every time the transform is changed, it requires re-merging the acoustic maps, which uses some additional CPU resources.

Acoustic Map does not support any Scaling on its transform. If you apply scale to an Acoustic Map, it will not render accurately, and an error will be logged to the console.

By using Acoustic Geoemtry and Acoustic Map, your game can automatically generate accurate acoustic simulation for your scene. However, you may find you’d like to tweak the sound of the reverb for various parts of the game. There are two tools for controlling reverb properties: