Ray-Traced Shadows on Nanite Skeletal Meshes: Shadows Appear to Ignore Displacement - Bug or Limitation?

[ShiinaE]

Hi everyone,

I’ve been working with ray-traced shadows on Nanite skeletal meshes and noticed something unusual. It seems like the shadows are being calculated based on the mesh’s original geometry before displacement is applied. This results in shadows that don’t match the displaced mesh’s actual shape.

Ray Traced Shadow vs VSM Shadow:

Has anyone else encountered this issue? Is this a known limitation of ray-traced shadows with Nanite, or could it be a bug? I’d appreciate any insights or suggestions on how to address this.

Thanks in advance for your help!

[glitchered]

why not pre-displace or sculpt that? like bake it in the geo. you’d get actually geo without xtra computation for the displacement. anyway…

cvars to test…

“r.raytracing.nanite.mode” will use the highres mesh for shadows instead of the proxy. not sure if displacement is supported for that.

or…

“r.raytracing.normalbias” will mess with the normal contribution. try some numbers. if you go overboard on it, it may reduce overall shadow casting.

and… a welcome to the forum. : )

[ShiinaE]

Thank you for your suggestions! I’ve tried a few cvars, none of them resolved the issue.

While pre-displacing or sculpting the geometry is a viable solution for static details, dynamic effects on nanite skeletal meshes should be pretty cool in sci-fi games i think, infinity detial on tesselated surface. If ray-traced shadows can’t handle displacement properly, it could limit their usability in such scenarios.

Thanks again for your help and the warm welcome to the forum!