The anisotropy of the solid-liquid interfacial energy, gamma(sl), plays a key role in the accurate prediction of growth morphologies in metallic alloys. This interfacial energy anisotropy can vary duo to alloy composition, especially when that of the pure solvent is weak. Recently Gonzales and Rappay, [1) showed the influence of an increasing zinc content on the growth direction of aluminum dendrites, which varied progressively from < 100 > to < 110 > as the zinc composition changed from 10 to 90 wt%. At the onset and end of this dendrito orientation transition (DOT), textured seaweeds were even observed. While this DOT could be simulated by phase field modeling with a change in the anisotropy of gamma(sl), seaweed could not be reproduced [2]. In order to explain this disagreement, it is necessary to have direct access to the anisotropy parameters. A combined numerical/experimental methodology to determine the needed experimental data is presented. It is based on inverse methods applied to 3D equilibrium shapes obtained by X-ray tomography. The gained anisotropy values are evaluated in a phase field code featuring a description of the interfacial energy anisotropy by a development up to the third order of the spherical harmonics for a cubic system. The hitter enables to model growth directions out of {110} planes, which was not possible with previous models.