For this night, the donut tilts look quite stable, dx is generally stable with typical values around 600 microns, and dy is mostly stable with typical values around zero.
The BCAMs and donuts agree well for dy and xt, but are offset from each other in dx and yt past image 152340.
For this night, the donut tilts are quite stable, while the decenters vary between +-1000 microns. The BCAMs and donuts are offset in tilt for images after 151910, and don't seem to follow any pattern in decenter
I believe that the LUT was not working properly over this night of observing. The donuts and BCAMs look very well correlated over the course of the night, but there are large offsets between the two in dx,dy and xt.
The attachment plot shows the donut measured alignment, from left decenter X (DODX), decenter Y (DODY), defocus (DODZ), tip or thetax (DOXT) and tilt or thetay (DOYT) versus image number for a portion of the images taken on 11/14. The units for decenters and defocus are microns, and for tip/tilt are arcsec. As you can see there is a systematic decenter measured, and a smaller systematic tip/tilt. The images in this plot cover the mini survey taken on 11/14, with small slews and relatively stable conditions.
The donut analysis done here is essentially the same as that done online in SISPI, the main difference is that with more time offline, sextractor is run to select donuts instead of a very quick donut finder, and so more donuts are used. The Zernike fit to the donuts is identical between this offline analysis and in SISPI. Thus the current LUT has the same origin as this analysis, yet we observe a significant measured decenter. Whether this is due to a change in the alignment from the time of the Alt-Az scan where the LUT came from, or is due to a hysteresis effect is to be determined.
The donuts are quite stable in tilt, but show variation in decenter of 2mm in dx and dy. The correlation of BCAMs and donuts is quite interesting here. The dy BCAMs and donuts would agree almost perfectly if the BCAM dy were moved down by 1000microns at image 151350 or so. Lifewise the dx would compare perfectly if the BCAMs were adjusted at image 151340 and 151420.
For this night, the donut tilt is quite stable, and the decenters range over 1mm. Here the BCAMs and donuts compare well in tilt and dy. The dx would agree very well if the BCAMs were adjusted up by 2mm for images before 151050