Observing Procedure

ShutterCorrect works by using a series of images of varying exposure time to infer the motion of the shutter and correct for the difference in exposure time the moving shutter will generate across the detector of the camera.

For astronomical applications, this will require "twilight flats," which are exposures of the twilight sky after the sun has set. During this time, the sky as imaged by the camera is roughly a uniform brightness. As the night gets darker, to achieve a given exposure level will require a longer exposure time. For very long exposure times, the perturbation on exposure time induced by the moving shutter is vanishingly small. For more on the actual derivation process of the shutter correction, see Shutter Map.

##Obtaining the Test Images

Things to keep in mind:

• Nonlinearity: it would be ideal to keep twilight flats to roughly a 20% - 40% illumination of the saturation value (10,000 counts or lower for the MMT camera), otherwise we would possibly need to correct for non-linearity effects, which adds another step to the process.
• On the other hand, twilight flats should have a decent amount of counts (> 20% saturation level), since it is difficult (impossible) to derive a shutter map with an underexposed image.
• If you are using a camera with a filter wheel (for example the Sloan set of filters u,g,r,i,z), it does not matter which filter you use as long as you keep it consistent throughout the entire test period. Be wary that you would also like to achieve a wide enough range of counts across the twilight sky and for a range of exposure times ranging from 0.1 seconds to greater than 60 seconds. If you are unsure, a filter like r might be good to try.

###Images

If the dark current of your detector is non-negligable, you will also want to take dark frames at the exposure times of 0.1s, 0.5s, 1s, 5s, 10s, and 60s, or whatever your longest exposure is.

• Bias frames (~10 or more total frames) to remove the overall "bias" level induced by the chip

Time permitting, obtain as many twilight flats as possible at each of these exposure times. 9 or more twilight flats at each exposure time would be ideal, since this will improve the statistics on the shutter map. Anything much more than 9 might be excessive.

• 0.1s twilight flat
• 0.5s twilight flat
• 1s twilight flat
• 5s twilight flat
• 10s twilight flat
• 60s twilight flat or as long as you can manage without saturating the chip.

##Reduction Steps Then, "reduce" these frames in a standard manner using a program such as IRAF (or this could be done with custom routines written in IDL or Python). The reduction means that each twilight frame should be corrected for bias (pixel-to-pixel variation), overscan (variation of count "pedestal"), and dark current. These should be as close to your "science frames" before actual flat fielding, since this program is actually designed to determine the illumination correction, which flat fielding is designed to remove. If you have multiple reduced twilight flats at each exposure time, average them together using a task such as IRAF's flatcombine.

You will feed these reduced frames to the shutterCorrect.py script.

Next: Shutter Map