I’ve handed over character controls over to Nin for a bit. While I wait for him to reach a state when I can merge his animation-related work back in, I’m working with a simplified non-animated placeholder so I can clearly see how shadows will cast on the character.
Here’s a tip for quick and usable third-person controls: Divide Mouselook controls into separate X and Y, with X on the character and Y on the following camera. For Unity, this means make the camera a child of the character container, placing one MouseLook on the character (restricted to MouseX) and another MouseLook on the camera (as MouseY). Make the whole thing a prefab for easy re-use. This is a simple, crude camera scheme, but it works well for getting up and running around.
I’ll have to revisit my adaptive camera controls later and re-apply everything I learned from UDK last year. That will come after I’m done with light-detection and once we have the proper character back in place, nicely animated.