Skip to main content

mada's CCD Image Calibration practices for Exercise #4

3 posts / 0 new
Last post
mada
mada's picture
mada's CCD Image Calibration practices for Exercise #4

Greetings, All,

Below is a summary of my imaging and reduction practices. I welcome any helpful criticism.

Adam Mills

 

Schedule and circumstances have left me with what may be an unusual observing procedure, but one which I try to improve as my limited experience grows.

I observe from a location which, while not remote, is still something of a drive from home. Generally, I am unable to get out and set up until after dark. This precludes taking flats at dusk, and often places a premium on time once the gear is ready to observe. Dark and bias frames are sometimes taken on cloudy nights with the camera set up on my balcony, though I take them on the night of the observations normally. The scope often needs some time to continue cooling after the mount is aligned, so I may use this time to take bias and dark frames once the camera is cooled.
Typically the camera is cooled to about 20 deg C below the lowest temperature expected that night, which uses around 60 % of the cooling capacity. Though this may not deliver optimal results, it ensures stability and allows me to focus on the targets.

Once the scope has had a reasonable time to cool I begin taking data images of my targets for the night. How long this phase lasts will vary depending on how much darkness I have left, battery capacity (particularly during long and cold winter nights) and of course if the weather holds. Multiple images are taken of each target through each filter. At least three, and up to ten, with more for shorter exposure times.

If one factor or another compels me to stop well before dawn, I then finish taking darks, and take flats for whichever filters I have used. Since the system is disassembled each night, flats must be taken for each observing session.

When the dusk or dawn sky is unavailable I take "headlight flats." I reflect one headlight of my car off a piece of styrofoam and into the scope, which has a white t-shirt placed across the aperture to diffuse the light. Half-filling the CCD wells takes up to two minutes for B, 30-60 seconds for V, and about 30 seconds for Rc and Ic, depending on how well aligned the elements are.
This system is clearly less desirable than sky flats, but these do seem to be better than nothing. At the very least, they eliminate the worst of the dust donuts and vignetting in my system.

When conditions permit, I finish taking darks and wait to take flats shortly before dawn. I find it challenging to do this in all four filters, taking 9-10 flats each, as the sky brightens faster than the eye alone might notice. Starting with 30 second exposures, I may be down to 3-4 seconds by the time all filters are completed. These sky flats are also taken with a white t-shirt stretched across the aperture as a diffuser.

If required, after taking flats I will finish taking required darks before the sun rises. Darks taken at this time appear to be of much the same quality as those taken before twilight begins.
Darks are normally taken for each exposure length that I anticipate using, or have already used during the night. Thus bias frames are not usually required for image reduction. However, I often take them anyway, if only to check on the health and consistent operation of the camera.

Image calibration and photometry is performed using either AIP4WIN or MPO Canopus/Photored. I find AIP easier to use for smaller volumes of images, or experiments, and the MPO software easier to automate for larger volumes. As many images seem to contain cosmic ray hits, darks are median combined into a master, which is then applied to individual data and flat images of the appropriate time. Flats are then median combined into a master and applied to data images of the same filter. Ensemble differential photometry is then performed on each data image. If a time series was performed on a target, photometry from each image will be reported, along with the error reported by the software. Otherwise, photometry results from consecutive images of the same target in the same filter are averaged and this mean value reported as the magnitude, and standard deviation calculated and reported as the error.

Having taken this course, I will likely make a few changes to my procedures.
First, I will experiment more with using scaled darks and bias frames rather than taking darks for each exposure length. I hope this will simplify the imaging process and will investigate what impact it may have on quality.
Second, I plan to do something about my flats. Though I have never formally compared frames resulting from the two approaches of "headlight" and sky flats, I intend to do this and determine how much difference they make in the quality of my photometry, or the lack thereof. Also, I plan on assembling a light box to use for taking flats. If I can produce something suitable, it will allow flats to be produced with consistent exposure times and quality, regardless of when I need to take them.

Aaron Price
Aaron Price's picture
morning flats

[quote=mada]

If one factor or another compels me to stop well before dawn, I then finish taking darks, and take flats for whichever filters I have used. Since the system is disassembled each night, flats must be taken for each observing session.

[/quote]

Have you tested taking flats at the start of the session and comparing it with your end-of-session flights, as you did with the darks?

mada
mada's picture
Flats

[quote=Aaron Price]

[quote=mada]

If one factor or another compels me to stop well before dawn, I then finish taking darks, and take flats for whichever filters I have used. Since the system is disassembled each night, flats must be taken for each observing session.

[/quote]

Have you tested taking flats at the start of the session and comparing it with your end-of-session flights, as you did with the darks?

[/quote]

No, I haven't tried that so far. It sounds like a good idea for comparison, though. Perhaps next time I'm able to get out!

Log in to post comments
AAVSO 49 Bay State Rd. Cambridge, MA 02138 aavso@aavso.org 617-354-0484