Camera Physics. Differnet Approches

usually, a ray-cast from the anchor point to where the camera would be ideally. Or a swept volume check (put a sphere around the camera). You stop the camera at the nearest collision (with other object, or a wal, ect).

Quote:

Original post by oliii usually, a ray-cast from the anchor point to where the camera would be ideally. Or a swept volume check (put a sphere around the camera). You stop the camera at the nearest collision (with other object, or a wal, ect).

Can u explain in more detail,
Plz provide link regarding this technique if u have..

Thanks.

There is nothing to it really.

Usually, your camera focuses on point near the player (above the head, shoulder cam, ...), along a direction. You trace backward from that point, up to a maximum distance (say, 3 meters), and query the collision system for the first collision. Then you place the camera at that collision if there is one, else at the max distance.

I dont know physx, so I wouldn't know what calls you have to make, but it should be pretty straight forward.

Oh yeah, you probably need to use a collision callback to filter out objects that are not suppose to obstruct your camera (i.e. the player himself, the car he is driving, ect...).