Suspect that this is due to a large time frame between images eg. 30 mins and a large number of images causing the grav histogram array to overrun.
(which is set with 30 minute bins by default, as was designed for a small number of minutes between images.)
Should be easy fix.