“…We have now tested this procedure on several data sets (Table 11), collected on both organic and inorganic materials, and it seems to work. The straightforward approach with = 0.3, B iso = 0-1, HM = 20-5, = AUTO , 500 cycles and 100 runs produced interpretable maps that could be identified with the R SF value for a Ba phenylboronate (novel structure, sample kindly provided by Danielle Laurencin, University of Montpellier, France), for the molecular sieve ZrPOF-EA (Liu et al, 2011), for two tricyanomethane derivatives (Š išak, McCusker, Buckl et al, 2010), and for the sugars dl-ribose (sample kindly provided by J. D. Dunitz, ETH Zurich) and d-arabinose (Takagi & Jeffrey, 1977). Three organic structures, all crystallizing in the space group P2 1 like d-ribose, required starting phases from an approximate model.…”