Re: Why is the Brightness clamp going away?
Re: Compatibility was lightweight, why sunsetting it?
Bundling these 2 questions as the answer is similar.
(@ittrgrey, @az09az99096, @GuilleLopezD, @Metallochrome, @monkeymario1000, @ittrgrey, @Enzotsky, @Beyond_5D , @YTFTAshPlays)
Hey everyone,
I’d like to add more clarity into why we are sunsetting Compatibility, and removing Brightness clamp. In a nutshell, we have to balance the pain points from the sunset of Compatibility (i.e. losing brightness clamp and light anisotropy) against helping the vast majority of our developers who are having important issues with the Lighting system’s scalability.
The main problem we’re trying to solve is that automatic scaling, which reduces Lighting quality when an experience runs at a low framerate, often does not respect developers’ intent – it doesn’t make the best performance decision for many existing experiences and it breaks their artistic intent. It also over-aggressively cuts other features such as draw distance.
To respect artistic intent and improve scalability at once, we need to:
- Separate artistic decisions from the underlying Lighting Technologies as much as possible
- Make sure that any artistic controls work as intended across different Lighting Technologies to preserve artistic intent
Compatibility is not a Lighting Technology (the Technology is Voxel), but it is an artistic choice. The main artistic difference that could be supported across all lighting technologies (Voxel, ShadowMap and Future) is its tone mapping.
As some of you pointed out, the brightness clamp and anisotropic lighting are also part of Compatibility’s aesthetic. The brightness clamp was not extracted into a property because:
- It would have incoherent visual results on other Lighting Technologies.
- It would also impose significant complexity to our lighting shaders which would impact the performance of every experience, even those not using it.
- As a scriptable property, it would fully invalidate the lightgrid every single time the property would change, causing light flickering and performance instabilities.
On the other hand, we tested the migration on a high number of places using Compatibility and we found it is possible to achieve a similar result by adjusting local lights, even if not a perfect 1:1 match. Since similar visual results to Compatibility (although not perfect) can continue being achieved and considering the downsides of the clamp when used with other Lighting Technologies, we decided to sunset it.