I’m building an RTS style game and I really like the latest SimCities camera controls and using this as inspiration. I’m looking at how to wire this up and I see two approaches so far. My preferred approach is to use an invisible character that “walks” on terrain and the camera is attached to a movable boom to this invisible character pivot point very much like the top down sample. My issue is I would need the character to only “walk” on the terrain and be able to go through buildings, trees, etc. I believe I can do that with the collision model.
Is this a viable approach to this problem or should I not use a character at all and just draw ray tracing to the terrain to find the pivot point then do all camera operations based on the ray traced pivot point.