... >reduced by about 0.7, (or 2.5*log((1-cos(ph)/2) for this formula. log((1-cos(ph))/2) ... >Something does not LOOK quite right there. Bjoern