I have a project that follows the basics of a Pre Rendered background scene. I have a plug-in that creates a depth shader image from some crude map geometry I made, and I can create a High Fidelity environment png that matches the depth shader scene’s collision, (Like Resident Evil and FF7) . I tell my main camera to ignore the geometry and to put my controllable character in front of the Hi-Fi png. Now, the 3rd person controllable character runs around perfectly on this pre-rendered scene, but he casts no shadows. I’ve even tried to make an invisible sphere as a child object that casts a circle shadow under him, but it seems the shadows are on a different culling mask layer? I can’t seem to tell my camera to just bring the characters shadows on the same layer he’s on. Is this a culling mask problem or is this just an issue of using Pre-Rendered backgrounds?
Render your depth-only geometry using a shader that outputs black (or gray) in shadowed areas and pure white everywhere else, then blend it multiplicatively with the background png. This is typically called a “matte shadow” shader.
Here’s one that might (or might not) work for your Unity/render pipeline version combination:
https://github.com/keijiro/ShadowDrawer
I’d suggest looking at ShaderGraph and getting comfortable with how to make shaders, that will help a lot.