During the 20120909 OL420/LED data set collected by Mike I found that for Track 8 there was a ~4000 (ADU/pix/sec) drop which was about 10%. After looking at the matricies you can see that there was a readout problem with Track 8 on image 7. I rewrote my program to removed any track with zeros in them. This showed that Track 11 on image 6 in the same file had the same problem. This greatly improved the day to day stability for track 8. Track 11 is was kind of on the edge of the track so it did not effect the data enought to see.

Figure 1

Figure 2