Hi Hanning, I agree, that it should matter there. But indeed I wrote something wrong in my first reply. It should give the correct sign of the determinant. As your test indicates it does not, that is definitely a bug. I will have a look into this as soon as I got a little time. Thank you Flo