WTTM Flatfields, Zeros, and BadPixels

WTTM Flatfields

 
  1. To this point in time, we have no indication of any issue that would suggest that, in BVRI,  dome flatfield exposures would not allow a user of WTTM to reduce his/her data.  There are no observed artifacts, other than hot or dead  pixels and where the field vignetting gradient is high.  There are no indications of scattered light degrading the reduced data.
  2. A detailed look at the effects of the tip/tilt mirror's position in tip and tilt on the flatfields has been carried out and is presented below.  As long as the user avoids objects, the regions,  that fall onto the steeply sloped vignetted zone, the data should be handled fine by the flatfield.  Also noted is that the un-vignetted portion of the CCD is uniform and is not dependent on the tip/tilt position; effects are variant at the 0.01% level across rows/columns on the chip.
  3. As of have yet to do a detailed, quantitative analysis of the flatfielding performance for the WTTM filters, broadband and Sloan filter sets.


Discussion:
A question to be answered was, do the flatfields depend on the position of the tip/tilt mirror?  To answer the question, we obtained flatfields at various, and arbitrary tip/tilt mirror positions (the mirror can be commanded to different tilt positions) and then compared the images by dividing each flatfield by a flatfield at a nominal 0,0 position (no tip/tilt).  Ignoring variations in the flatfield lamp intensity variations, ~1.4%,  the flatfields in this test were identical except at the regions where vignetting begins.  Note:  The maximum tip/tilt range of the mirror is +/- 5000mv.  Typical variations on sky are +/- 1000mv min/max, but in wind buffeting conditions, larger values of tip/tilt displacement will be seen.

The graph below is intended to show the variation at the very edges of the non-vignetted fields.  Based on this analysis, it is strongly recommended that one maintains ~50pixel boundary from the edge of the vignetting region to ensure that no tip/tilt related effect is introduced via flatfielding.  (Higher Resolution Plots:  Vignetted Region: Column Comparison    Vignetted Region: Row comparison )
Vignetted Region: Column ComparisonVignetted Region: Row Comparison
From the graphs, one can see that along columns, top of image as nominally displayed, the vignetted region changes illumination properties significantly as the tip/tilt position changes.  It is for this reason that it is recommended that a 50pixel buffer be kept from the vignetted regions of the chip when analyzing one's data.

As for the un-vignetted region of the chip, a similar analysis was done.  Again, each flatfield image was divided by a standard flatfield image with no tip/tilt, 0mv/0mv.  The resultant images were then analyzed in the middle columns and rows, and a linear function fit to the sub-section.  In each case, there was only very small detectable slopes of 0.01% or less along the full, un-vignetted region.  Graphs of the residuals to the fits are presented below and show no higher order variation.
Un-Vignetted Zone: Column ComparisonUn-Vignetted Region: Row Comparison

Zero

Please note, we have not been successful in eliminating a light leak at the dewar/filter wheel interface.  When obtaining calibrations, make sure that you have the dome lights off.
Bright light on the dewar should be avoided.

Note that we observe no light leaks, scattered light, in twighlight or bright moon.  Very bright twighlight, scattered light on the zero images is observed.

Bad Pixel Mask


At this time, not done.