Reflection Probes vs. Screen Space Reflections in Unity
Picking up on some of what I touched on in the previous article, reflections can be another expensive process for Unity calculate (as far as 3D environments and lighting are concerned). Reflection probes, like light probes, offer a way of measuring and storing some expensive lighting information.
The Unity Documentation describes a reflection probe as ‘rather light a camera that captures a spherical view of its surroundings in all directions’. The captured image is then stored as a cubemap, which can then be used by any objects (within a specified radius) with reflective materials. Reflections on the object can then change convincingly according to this information about their environment.
I’ve also noticed we can choose whether a probe is for a baked, custom, or realtime setup. By choosing baked, the reflection probe won’t capture GameObjects at runtime that have their ‘Reflection Probe Static’ flag disabled. Baking should give us the most performant results at runtime, but if we do want to capture dynamic GameObjects in a baked Reflection Probe, we’ll need to select Custom and enable Dynamic Objects. In this event, objects not marked as Static in the inspector will be baked in to the reflection.
If we’re looking at having reflection effects for mobile or lower-end devices, reflection probes with a baked setup will most likely be the way to go. On the other hand, if we’re developing for higher-end devices or desktops, we may want to consider the Screen Space Reflection effect (part of the Post-Processing package). This is a considerably expensive technique, but can yield some great results and handle some of the subtler reflections which can really make a scene feel more authentic.
I’ve included a series of screenshots below from a scene in the project I’m working on as part of the GameDevHQ course, which should hopefully give us a useful comparison of a few different combinations:
