Here are the EPD calibration results for a smattering of days ranging from day 122 to day 146 for run 24. A good deal of inconsistent behavior between tiles, necessitating a lot of time segmentation. A few days considered did not have good statistics, so I used surrounding times to determine what the MIP should be for these days. Some dead tiles, with 1-2-8 going dead sometime on day 131 or something like that.