hmm i dont see the same flaw in the reflections, and only real difference is the normals would be pointing opposite directions, while in mine, they both face 'forward" towards the ball, odd how things work now

and yes i know both have reflections in the real world, but was no point to use the front one when other things didnt work right :) as for the edges, the frame of the mirror covers them in my model, so was no point to having them either

thanks for the solution though :) reversing the back normal works :D