REVISION DATE: 14-May-2013 13:03:51

On day 7 plaque data the BSG dark bias for some unknown reason went from ~100 ADU to near zero. This cause a problem in the low pixel end because the data can not be negative (they are unsigned interger 16's). So the negative data is convert to zero. This will cause the light scans on these tracks to be wrong. Below are some graphs showing the problem. The first is an image of the darks. You can see a tilt to the data. The closest edge starts a zero when it should be ~100 ADU. The second figure shows a dark for each of the 4 dark images (BSG and RSG). This show the RSG did not have this problem. THe third graph is a zoomed view of the cross track. You can see the data sloping down to zero. The forth and fifth graphs show the problem for two long track pixels. The first is in track 1 where the problem is worst and the second is from Track 9 where the problem is minumal.

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5