3. Image Processing & Calibration
PART 1: TAKING AND PRODUCING A GOOD IMAGE
Several sources discuss the general methods for taking and producing a good image. Following is a general description of the steps involved in taking an image with an SBIG camera.
- Decide on the target. Something rising or near the meridian will be in the sky longer giving you time to take more images. An object that is setting will not be visible for too long causing you to rush to take the needed images.
- If you have access to multiple telescopes, decide on the telescope. It is important to try to match the camera to the telescope.
- Turn on the CCD camera and adjust the temperature to 20 or 25 degrees below the ambient air temperature (see the recommended settings for your camera). During the summer it helps to have a fan blowing past the cooling fins of the camera. In the summer, if it takes more than 95% of the power to cool the camera, then the temperature is set too low and must be increased.
- While the temperature of the camera is adjusting, start and set up the telescope. Make a directory for the images and change the telescope parameters as needed.
- Hopefully, it is not yet fully dark. Point the telescope towards the evenly lit dusk (or dawn) sky and take several (>5) images. These are the flat-field images. Take several series each at different exposures that will be used throughout the night when taking deep-sky objects.
- By now it should be dark. Find and center the target object in the telescope. Observe for several minutes to make sure the alignment and tracking are correct. [Optional -- sketch and/or photograph the object before or after taking the images for later comparison.]
- If you do not have a flip mirror system, do a preliminary focus by focusing the telescope with the smallest eyepiece available or by using a parfocal eyepiece .Take the eyepiece out and put in the camera (with or without a projection assembly, color, or polarizing filters).
- If you do have a flip mirror system, your camera and eyepiece can be independently focused and then the view flipped between the two.
- Use your software to focus and position the object. This is the hard part. You might have to take several shots at different exposures to find the minimum where the object will be visible. Keep the auto contrast on. Then using that exposure, take several shots while focusing the telescope. When the object is almost focused, use planet mode to fine tune the focus. Once the focus is achieved, go back to full frame and center the object on the chip by carefully moving the telescope. For many objects, you can also use focus mode to find an exposure that does not saturate the chip and is still short enough not to show any drift in the telescope during the exposure.
- You are now ready to take images (light frames). Use exposures close to but not past the saturation point of the object. Bright planets and the moon can saturate the chip in less than a second while deep-sky objects may take several minutes to overexpose. Single images and images in series should be taken.
- For each set of images at a certain exposure, take a corresponding set of dark images at the same exposure. It is important to do this at roughly the same time as the other images so that the temperature of the camera is the same for both light and dark frames. Also a set of bias frames can be made while the telescope is covered but the exposure time is zero- well almost zero at 0.01 seconds. The dark and bias frame may also be made earlier before it is completely dark by covering the CCD camera during the exposures. However, the camera's temperature must be the same when taking any images.
- Once you are done taking images, you can start processing them. First, average each series of flat-field, dark, and bias images. For example, if you have seven series- three flat at X, Y, Z exposures, three dark at X, Y, Z exposures, and one bias. Averaging each series will result in seven images. From hereon, flat, bias and dark shall refer to the averaged image series.
- Make corrected images of the flats by taking a flat and subtracting the bias frame. Then subtract the dark frame of the same exposure time as the flat frame.
Flatcor(X) = flat(X) - bias - dark(X)
- If the camera did not shift between exposures of an object, you can also average those light images.
- Open either a single or averaged light frame. Subtract the bias and the appropriate dark frame.
Lightcor(X) = light(X) - bias - dark(X)
- Now 'flat-field' lightcor(X) by flatcor(X). The resulting image, if focusing was done properly earlier, should be sharp and clear- a good image. The background and range might have to be adjusted, but that is a minor operation.
final image = lightcor(X)/flatcor(X)
Further processing is optional. Most of the image processing functions in many programs are for aesthetic purposes and are therefore not recommended for images that are to be used for photometric or astrometric measurements. However, images that are not going to be used for measurements might benefit from more processing. 'Playing' with the image and functions is rather haphazard. Knowing what each function does and applying only those that are necessary should produce a pleasing image.
PART 2: APPLICATIONS OF CCDs
CCDs were first designed as an electronic analogue to the magnetic bubble device, a type of memory cell. However, the potential of the CCD as an imaging sensor was far greater than as a memory device. With the early and enthusiastic backing of JPL, CCDs have been quickly assimilated into the field of astronomy. Although CCDs are common in astronomy, they are also used in other disciplines such as physics and chemistry. CCDs are also becoming a part of everyday objects like camcorders.
Because CCDs are a nearly perfect sensor, better images are obtained than with standard film. The digital images can also be analyzed with a greater amount of precision and accuracy then photographs. For example, photometric and astrometric measurements can be made from the same image.
To learn more about taking and processing CCD images...
Berry, Richard. Introduction to Astronomical Image Processing. Richmond, VA: Willmann-Bell, Inc., 1991.
Berry, Richard. "Image Processing in Astronomy." Sky & Telescope April 1994: 30-36.
Berry, Richard. "Working in the Digital Darkroom." Astronomy August 1994: 62-67.
Burnham Robert. "Virtual Sky." Astronomy March 1994: 70-77.
George, Douglas. "Stretching." CCD Astronomy Fall 1994: 8-11.
George, Douglas, Ajai Sehgal, and L. Robert Morris. "Resolution: To The Max." CCD Astronomy Spring 1994: 8-11.
Gombert, Glenn and John Chumack. "Catching Comets with a CCD." Astronomy February 1995: 72-75.
Gilliland Ronald L. "Details of Noise Sources and Reduction Processes." Astronomical CCD Observing and Reduction Techniques. Astronomical Society of the Pacific Conference Series Vol. 23. Howell, Steve B., ed. San Francisco: BookCrafters, Inc., 1992. 68-89.
Hayes, Brian. "Scanning the Heavens." American Scientist November-December 1994: 512-516.
Kaitchuck, Ronald H., Arne A. Henden, and Ryland Truax. "Photometry in the Digital Age." CCD Astronomy Fall 1994: 20-23.
Meyer, Eric and Herbert Raab. "CCD Astrometry." CCD Astronomy Winter 1995: 20-22.
Newberry, Michael V. "The Signal to Noise Connection." CCD Astronomy Summer 1994: 34-39; Fall 1994: 12-15.
Newberry, Michael V. "Recovering the Signal." CCD Astronomy Spring 1995: 18-21.
Newberry, Michael V. "Dark Frames." CCD Astronomy Summer 1995: 12-13.
Reader Reports. "Capturing Colors with a CCD." Astronomy December 1993: 89.
Reader Reports. "Full Moon Tricolor CCD Imaging." Astronomy June 1994: 81.
Schild, Ralph E. "Coloring the Electronic Sky." Sky & Telescope February 1988: 144-147.
Skiff, Brian A. "CCD Observing Projects for Amateurs." Webb Society Quarterly Journal No 92. Authors Reprint. April 1993: 10-16.