for a current project I splitted the scene into multiple parts. Starting with a “init” scene which only contains cameras and lights. On runtime I append the different geometries without any lights or cameras. Problem I’ve got is that the shadows doesnt seem to work. Current workaround I’ve found is creating a plane in that “init” scene, which covers the whole area where I need shadows to be.
Just wanted to ask if there is any other “proper” way to get around this? Maybe something like a “updateShadowMap” function?
Unfortunately, this light type is not so easy to support for dynamic loading. We calculate size of the shadow volume in the original scene and export it in the asset, so it does not get extended in the engine itself. The method you’ve chosen (with additional plane) works pretty good though.
Anyway, I’m going to add this issue in our TODO list.