reset oddity(?)

I was looking at the Fowler 7 tests that Mike did.  The files are (per email):

7 N's, 5 S's, 6 N's, 1 D                file c:/data/CD014/cd14_2013mar20/cold.lab.00016.fits
7 N's, 17 S's, 6 N's, 1 D               file c:/data/CD014/cd14_2013mar20/cold.lab.00017.fits


Mostly I was looking at the data in 00016, but what I noticed appears to be true in 00017 as well.

For most of the differences, data values are fine.  There is, however, a strange behavior during frames with resets.  The plot below plots:

white: 00016 12thN-13thN vs 13thN - D
red: 00016 4thN-5thN vs 6thN-7thN
blue: 00016 10thN-11thN vs 9thN vs 10thN
green: 00017 10thN-11thN vs 9thN vs 10thN
yellow: 00017 12thN - 13thN vs 13thN - D

You can see that the ones involving resets (white, yellow) have two states while the red, blue, and green, which do not have resets, are fine.  

A separate plot of just even row differences (ie, 00016 12thN-13thN vs 13thN-D just even rows) and just odd rows shows that the odd rows are the ones behaving differently.  Remember that IDL starts with Row 0, so evens are the first row.