Minor mystery resolved: Perseus cal curve

When I started working on the Perseus data again, I decided to use the Enoch 2006 calibration curve directly. However, it has a very different form than all other epochs. The reason, as revealed below, is that it was not forced through 0,0. Additionally, all of the BGPS data was observed with mean DC ~ 2-3 V, while the Perseus data was observed with mean DC 4-5 V, so the relevant regime is in a very different location. The reference DC bias was much lower, ~2.15 V vs. 4.6 V in the 2005-2007 BGPS and 2.6 V in the 2009 BGPS.