In this case, how did we know that yellow was closer to the camera than blue? When are we picking up information about depths of each object?