Alternative triangular lattice gives different results

Hi everybody,

I tried simulating a photonic crystal slab with holes in a triangular 
lattice using lattice vectors other than the ones in the tutorial on the 
MPB site.
Where the tutorial uses [(/ (sqrt 3) 2) ; 0.5] and [(/ (sqrt 3) 2) ; 
-0.5] as real lattice vectors, I tried using [1 ; 0] and [0.5 ; (/ (sqrt 
3) 2)].
My reciprocal lattice vectors are then [(/ (sqrt 3) 2) ; -0.5] and [0 ; 
1], the M-point should be at [0 ; 0.5] and the K-point at [(/ 1 3) ; (/ 
2 3)] (both in the reciprocal basis).
So it's basically the same structure as in the tutorial, yet I get 
different results (with a difference in frequencies up to 0.002). My 
question of course: where does this difference come from? And if both 
structures really should produce the same results, which one is right? :-)
Thanks,

Elewout