The flux radius is a non-parametric morphology feature we were asked by the Science Working Groups to compute and include in DR1. The implementation was supposed to closely resemble the SExtractor2 (SE2) and SourcExtractor++ (SE++) implementation that many scientists are familiar with. Unfortunately, in the first implementation used in DR1 we made the error of limiting the area for the flux computation to strictly segmentation area only and not extend beyound the segmentation area as it is done in SE2 and SE++. As illustrated in Fig. 15.4, left side, the distribution of flux radius versus total VIS brightness is constrained to lower values when compared to the proper, SE2/SE++-like implementation on the right. Likewise in the DR1 version the spike of the point-like sources about tends towards slightly lower values for faint objects. As indicated in Fig. 15.4 we have switched to the SE2/SE++ implementation post-DR1.

