...

• No correlation observed
• Swapping of rows 0,1 between panel halves (to check possible mapping issue) does not work either.

Conclusion

• data sample shows common mode effect, which can be handled by standard cm correction, except overpopulated central region.
• in "raw - pedestal" type of images calibration rows do not reproduce common behavior in ASIC's.
• quad 0 looks good - does not show significant difference between banks - hard to judge about calibrows mapping, it might be wrong mapping yet.
• scatter plots do not show correlation either
• Q: configuration of calibrows - gain mode, what exactly do they show?

...