Hello, thanks for reading! I’m working on a project based on Rendezvous with Rama, set inside a 50 kilometer long cylinder world. Since I’d rather avoid to-scale geometry for what would essentially be a background, I’d like to have a 1/1000 scale cylinder with its own camera that moves with the character at 1/1000 the distance.
the inside of the cylinder would be rendered first, along with environmental effects, then the foreground objects.
is this the right approach, or is using a geometries at huge scales acceptable?
If this would work, how can I get a camera to render first and then another camera’s view overlayed on top?
I’ve worked out a pretty neat solution to this. I followed the directions from a Youtube video: https://www.youtube.com/watch?v=JSRsQpRfDlk
in order to create a custom skybox with an editable material. I then created a scene capture cube, assigned it to a render target cube textureand used that as the skybox material’s source texture.
The scene capture actor is now located inside a miniature cylinder with a 2 sided material applied to it. Now, the scenecapturecube will move 1 unit for every 1000 unit the character moves and the cylinder will rotate 1 degree every time the character moves 44.4 units.
One question I have about this method is whether I can have two scenes running at once, one with the scenecapturecube, cylinder and a light source, and the other scene being the main game scene, using the rendertarget cube texture generated from one scene as the texture for a skybox in another? Is this possible?