CCD Imaging Techniques

Introduction

Every now and then I come across an amateur astronomer who wishes to know more about the various techniques for CCD imaging and the best ways to start out in this exciting realm of astronomy. I have found that there are as many unique answers to this as there are unique craters on the Moon, so I have written this article to share my personal experiences with CCD imaging. I will discuss the various problems encountered, their solutions, and the best methods I have discovered, to date, with my particular system. This is not to say that the below-listed methods are the best or the only way to accomplish the tasks. Keep an open mind and experiment! Those looking for more information than that presented in this article should visit my CCD Imaging FAQ.

Equipment

John and C-8 Takahashi FSQ-106n

The primary equipment I use for imaging includes the following:

Though this list may sound like an advertisement for a variety of companies, I know one thing about myself: I am mechanically inept! I have been successful at mirror grinding, but when it comes to building mounts, or doing machining, I am seriously lacking in skill. To accomodate, I use available products, and, when I can, I try them out first before purchasing. I have found that a perfectly good CCD setup can be obtained through the correct matching of various products on the market.

The Takahashi FSQ-106n is a superb wide field apochromatic refractor with an very flat field and excellent color correction. I use it for film, visual and CCD imaging of wide field objects (like M-42). A "must have" item with this scope is a 2" diagonal and the Extender-Q Barlow assembly. For CCD imaging, I recommend a Feldstein adapter to allow a permanent mounting between the CCD and the OTA. This also shortens the backfocus requirement, allowing for the use of the CFW-8 filter wheel. Check out my review at: FSQ-106n Review

The Takahashi CN-212 is an under-rated telescope. For its capabilities, too few people are actually making use of this fantastic instrument. It is a 212mm dual-focal point telescope. By swapping two different secondary mirrors, one can bring the focal point to a Newtonian focus or a cassegrain focus. This gives you a super wide-field f/3.9 scope or a tight f/12 instrument suited for planetary and lunar work. Collimation is challenging once. After that, collimation is maintained very well between mirror swaps. For imaging, the Newtonian focus has a plane flattener or corrector lens in place. Though this scope focuses by moving the primary, the focal adjustments are very fine, and mirrorflop is absolutely minimal. Check out my review at: CN-212 Review

The SBIG ST-7XME is a non-antiblooming (non-ABG) variety. It is more sensitive to light than the antiblooming models and thus reduces integration times by up to 33%. The problem encountered with this is the blooming of bright stars. This requires multiple images to be added together if a bright star is in the field of the chip. Any star brighter than magnitude 8.5 will bloom rapidly (under one minute). Special softwares can be found that will automatically align multiple images then either add them or average them together. I use MaxImDL (Cyanogen) for this and most of my processing needs. Another plus of the ST-7XME is that it has regulated thermoelectric cooling which allows a precise temperature setting on the chip. This makes keeping a library of dark frames an easier task. The USB downloads of images to the PC only take 3 seconds. That makes focussing easy!

Process

Deep Sky Imaging

M51

What follows is the process I use with the ST-7 imager with self-guiding.

  1. Open the dome.
  2. Attach the laptop to the mount and imager.
  3. Turn on the lights, the 12V DC supplies, the PC, the imager and the mounts.
  4. Boot the PC and launch TheSky and MaxImDL/CCD.
  5. Connect the software to the mount and the imager.
  6. Note the time and temperature in the log. Start the ST-7XME cooling to 30C below ambient.
  7. While the ST-7 is cooling, slew the mount to the zenith.
  8. When cooled off, set the CFW-8 to the desired filter and take flat fields. Repeat for all filters.
  9. By now stars are visible. Correct any focus issues. Flats may have to be retaken later if this is an issue.
  10. Find object to image. I usually start with asteroid hunting mosaics. Center the object.
  11. Find a guide star. Start tracking with the ST-7 built in guider.
  12. Image. Repeat until tired.

Planetary and Lunar Imaging

M51

  1. Open the dome.
  2. Attach the laptop to the mount and imager.
  3. Turn on the lights, the 12V DC supplies, the PC, the imager and the mounts.
  4. I use the ST-237 on the CN-212 with a 2.5x TeleVue Barlow.
  5. Boot the PC and launch TheSky and MaxImDL/CCD.
  6. Connect the software to the mount and the imager.
  7. Note the time and temperature in the log. Start the ST-237 cooling to 25C below ambient.
  8. While the ST-237 is cooling, slew the mount to the zenith.
  9. When cooled off, take flat fields.
  10. By now stars are visible. Correct any focus issues. Flats may have to be retaken later if this is an issue.
  11. Find object to image. Center the object.
  12. Focus, focus, focus.
  13. Image. Repeat until tired.

Image Processing

The integration is just the beginning to a long process that continues well into the next day. Image processing is now recognized as an added necessity to all forms of electronic imaging. There are many varied philosophies about processing, especially when discussing "how much is too much".

Dark Frames

Probably the easiest form of processing that should be done is to subtract a dark frame from the final integration. This should be done first, before any other processing steps. Basically, an integration of the same duration is taken with the cover placed over the CCD imager. This records all the ambient noise inherent to the CCD system, such as hot pixels and gradient heat saturation from internal electronic components. The record of the noise is then subtracted from your final image thus removing it. This is easy to do, and your images will show improvement! An additional suggestion: Take the dark frame integration just before or after your imaging run so that the temperature conditions of the imager are close to the conditions used for the image itself. If you have a CCD which has thermal control, then you cna take many dark frames at a variety of temperatures: Make a library! I have taken 10 dark frames for each temperature (C): 10, 5, 0, -5, -10, -20, down to -50C. That is cold. Used in conjunction with bias frames (a 0-second integration), the darks can then be averaged together then scaled to ANY integration time that you have made. Take 10 bias frames as well... at each temperature.

Flat Fields

This is a more complicated matter. Ideal flat frames are taken with the imager at the exact same focal point and radial position as the final integration. A flat frame is basically a grey image taken of the twilight sky or of a flat grey panel (even illumination is a must). This flat field is then divided out of the final integration. It works to remove unwanted signals from dust specks and vignetting. If you do photometry, flats are essential. Take 30 or more flats which will then be averaged together to improve results. Some software, like MaxImDL, will allow you to use the same library of darks and biases to process the flats. Yes, flats also need to have darks subtracted! Also, if you use filters, you must take a series of flats through each filter you intend to use that evening. Also... you need to take your flats at the same temperature as the imaging session.

Histogram Stretch

One key to a good CCD image is that the integration time should be long enough to get a nearly saturated image. In other words, try to maximize the integration such that the brightest portion of the object nearly saturates the pixels on the chip. This insures a good signal to noise ratio, and gives you plenty of data to process. If you can only halfway saturate the CCD pixels with a short integration, you can add images together or stretch the histogram out. This will bring out faint objects rapidly, but also distorts the data if you are using it for photometric purposes.

Adding Images

The adding of images is a great way to boost overall signal gains using short unguided exposures. If your mount can only track reasonably well for 30 seconds at a stretch, then you can take many 30 second integrations and add them all up to achieve similar results as of a longer exposure. Be aware that noise also increases with image additions, though not as rapidly as the signal information does.

Averaging Images

This is a fine technique to boost the signal to noise ratio of a series of integrations. The image will not get any brighter (as with addition), but it will look finer.

The Results

Drop into the astrophotography page to see a whole selection of images I've taken using the above techniques. I will be happy to discuss any of the proceses mentioned above, and I enjoy learning more about this exciting field of astronomy.

Related Articles

Self-Guiding the MX-5c with the Starlight Xpress STAR2000 System.
CCD Imaging Frequently Asked Questions.

Contents What's Up Feature Papers Photos
Links Observations Spectra Sketches

Last Modified: 3/19/03 8:55p
This page:© Copyright 2005 by John A. Blackwell