As can be seen, there is a shift in the average ADC from Isobar to 27 GeV (with the notable exception of the last day of Isobar, 129). I have included the file with all of the tiles graphed as above (MIPFits.pdf). The issue is found nearly throughout the range, and is more marked in the closer tiles. As an example, here is tile 4 from the same supersector as above:

It should be noted that tiles 1-3 exhibit this behaviour to a lesser degree, but this may be in part due to how they were initially fit for the Isobar run (see above linked blog for details). Here is tile 2 from the above supersector:

Before delving in to the preliminary analysis done on this phenomenon, here is an interesting aberrancy: West PP1 tile 19. This aberrancy was found in tiles 16-23 for the first several of the West supersectors.

We can see that, just as Isobar day 129 seems to be in line with the 27 GeV run, there are a few days of 27 GeV that seem to be in line with the Isobar run. This issue is highlighted in the previous blog on 27 GeV calibration; interestingly, it is found in several tiles that are all in the same "Set" (either 16-23 for West PP 01-04, or 6-9 for West PP 09-12), all on day range 147-150 of the 27 GeV run.

Initial Analysis
The question is whether this shift is actually data driven or if it's a byproduct of the fitting function. The fit function used is a multi-Landau fit, and then the calibrated ADC value comes from the first Mip peak MPV. In order to research this, I took a look at the code to see what it was actually doing. I know that this should be done before doing any actual data manipulation, but I'm still learning ROOT and my understanding is ramping up with every iteration of what I do. I found that the fits were only being done with 2 Mips, so I raised that to 5. I found that, brushing over maybe 20 or so tiles, the 27 GeV data didn't much change. The Isobar data seemed to have a change of something like ~2 or less ADC in most tiles. This is of course not rigorous; more analysis will follow.

In order to see a quantitative difference, I ran a more thorough analysis on West PP7 TT13. I looked at Isobar day 85 and 27 GeV day 145 as these days seemed pretty representative of the problem at hand. Here are the graphs for this tile over those days, including a graph of the two superimposed which was horribly done by overlaying a transparency because I'm still learning to ROOT (don't worry; the scales are correct):



Quantitatively, we have that the calibrated values for the ADC for this tile/day combination have a difference of 17.6. I redid the data with 5 Mips (as aforementioned) and found that the new delta was 14.5 (27 GeV went up by 0.28, Isobar went down by 2.84). It should be noted that the ADC position for the first peak (not the first Mip peak, mind) is different by 16 between the two as well.

Next Steps
There are a few things to look over, here. We really need to get a handle on why this phenomenon is occurring and what we can do about it. It would seem that this is data driven; this is nice as it means at least the fitting functionality of the affair is processing correctly. As to why the peaks seem to be so off, it could be a timing issue or maybe a issue of the data being biased (or not) for primary vertex. I'm not 100% sure what that means, so I'll also have to find that out (though I have a reasonable idea). I'll update as soon as I have more information to compile.