Figure 1: Combined, Unbinned Image (no mask)
ANDICAM IR Array Binning
Starting with the 2003B observing semester, IR data taken with ANDICAM on the SMARTS 1.3m telescope will be binned 2x2 in software after readout.
Overview
The IR Array in ANDICAM is a 1024x1024 HAWAII array with 18-micron pixels, which deliver an image scale of 0.137-arcsec/pixel at the 1.3m telescope. Given the typical seeing, binning the array 2x2 to 36-micron effective pixels for a scale of 0.274-arcsec/pixel is a better match. Binning is done in software after readout with reference to a bad pixel mask as described below.
The unbinned original images will be archived to tape, but not delivered to the data server at Yale for pickup by project PIs. Only the 2x2 binned images will be available on the server.
If for some reason you feel you have a very compelling reason to use the original images, please email Prof. Charles Bailyn regarding this matter.
Software Binning Process:
Raw IR images from ANDICAM are software-binned 2x2 pixels using IRAF with the following a procedure:
The first step is to apply a bad pixel mask to the raw images with the IRAF routine FIXPIX. A bad pixel mask was created by dividing a bright J-band skyflat (combination of five bright J skyflats) by a dark image (combination of five 60s darks) using CCDMASK. A typical bad pixel mask (1024x1024) can be seen in this image .
After applying the bad pixel mask, the images are binned 2x2 using BLKAVG. The binned version (512x512) of the bad pixel mask can be seen in this image.
As an example of the procedure, we present before and after images below. Both images have been created by first processing and then combining the individual images. The examples are zoomed in to show more detail.
The first image
was created by combining 13 unbinned J-band images with no bad pixel mask. This is what you would get if you just used the unbinned raw images from ANDICAM.
Figure 1: Combined, Unbinned Image (no mask)
The second image:
shows the same data after (1) applying the bad pixel mask, (2) binning the image 2x2, and (3) combining the 13 binned frames into the final J-band image shown above. Close examination reveals the absence of "snowy" pixels in the binned image, which have been removed by the bad pixel mask application prior to binning.
Figure 2: Masked, Binned 2x2, & Combined
Updated: 2006 May 11 [rwp/osu]
