Braxtogoo

Can confirm I am having this issue as well.

> https://github.com/godotengine/godot/blob/607b230ffe120b2757c56bd3d52a7a0d4e502cfe/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered.glsl#L1414 > > The issue lies probably in either the inverse matrix or normal, the formula for SH apppears to be correct. The probe gizmos apply the same formula...

That's wild, and now I'm trying to think of any useful applications for this behavior. 😅 Can't think of any yet but regardless these are interesting findings. I don't even...

Interesting update, while updating the description to be more accurate and getting footage for the new gif, I found something else. Moving the position of the LightmapGI node doesn't affect...

There is a tiny part of me that wants both of these behaviors preserved because if you had a spherical room full of dynamic objects you could spin the LightmapGI...