Relative error of the calibrated (top) and the contrast images (bottom) produced after linear calibration with our method (left) and the method of Tlatov et al. (2009, for an image of subset 1. The facular pattern was derived from a Rome/PSPT observation taken on 21/08/2000 (shown in Fig. 7 S1). Also listed are the values of the rms, mean, mean absolute and maximum relative differences by comparing image regions within 0.98 R. The colour bars apply to the images below them and are different for each image.