Here are some CCD images of mine. These were done with a 10" LX-200 and ST-7. Click the link to see a larger image and details.
M13 |
Horsehead Nebula |
M105 and others |
M51 |
M94 |
M15 |
M63 |
N6946 |
M81 |
M101 |
M33 (center) |
Moon |
M33 |
N7331 |
Allen's N2403 |
|
I recently got a Takahashi FS-102 refractor, and have started to do some wide-field film astro-photography. I use a Mamiya M645 and the dedicated reducer, so I get most of the 51mm, 4.8 deg. field. I piggy-back the Tak on my Meade, with a Losmandy dove-tail and Parallax rings, and guide through the main optics. Early results, with PPF and a Lumicon DeepSky filter, look encouraging.
M42 |
M8a |
M8b |
M16 |
M17 |
I got my first telescope, a Coulter 8" Dob, in 1992. I soon got a set of DobDrivers, which are stepper motors for the Alt-Az mount, and started writing software to run them. My first system ran under BASICA on a floppy-based 8088 lunchbox machine, and over the years turned into a VB app with C++ motor-control dll's running under Windows. The system allowed tracking and slewing, and in concert with TheSky it was a pretty powerful setup. I later got a Coulter 13.1" Dob, as well, and added voice-control to my software. I used a wireless mic, so as to be un-encumbered.
My first CCD was a CCD-10, which is a TC-211 based camera. It was cheap, and it was a start. The first version of the camera was pretty limited, but after upgrading and new software from Dave Groski it was OK. I fooled around with improving the cooling with a crazy system of hoses, a pump and a reservoir in the freezer, and finally built a second-stage Peltier cooler out of the guts of a Coleman electric cooler from Walmart. I learned a lot, and got a few images, but it was a very difficult system to operate. The small field of view, the backlash in my system, and the lack of really precise control over the telescope made finding and framing targets a real challenge. The lack of guiding capability, along with the problems of field rotation, meant that I was limited to one or two minute exposures. Aligning the frames for stacking is more of a chore with an Alt-Az system, as they need to be rotated, not just shifted.
All of this lead me to get a Meade 10" f/10 LX-200 in May 1997, and an SBIG ST-7 CCD camera soon thereafter. I use TheSky and CCDSoft from Software Bisque, and have solved many of the issues I had faced before. Although it still is a lot of work, from the keyboard I now can easily pick a target using TheSky, slew the telescope and find my target with a 10 second low-resolution exposure. I can then accurately center the frame, or set up a mosaic, and then find a guide star and let it run. I take a few fifteen-minute exposures while reading or watching TV, and move on to the next object.
The capabilities of the equipment now available to the amateur astronomer are impressive, to say the least, and undreamt of a decade ago. It is still a lot of work to get outstanding results, however, and there is a learning curve to conquer.
The system I currently use, a Meade 10" f/10 LX-200 and SBIG ST-7, greatly simplifies many of the procedures required for successful imaging. The software driving the system is an important part of the mix. I use TheSky Version 4 / Level 4 and CCDSoft, both from Software Bisque. Together, they allow me to control the camera and the telescope from one PC in my office. Trips out to the 'scope are minimal. A feature of TheSky that I have found very useful is the ability to display Field of View Indicators. I have set up Indicators for the various focal ratios I use, and saved a Display setting (called CCD) that is zoomed in to 1 degree and upside-down -- that is, North is to the bottom -- to match the image display of the camera. This makes it very easy to compare TheSky's display with what I am imaging. I then use TheSky's Motion Controls to precisely move the 'scope to frame my target. TheSky's ability to set up mosaics is also very helpful.
I use Meade's SuperWedge to mount my 'scope equatorially. Meade has announced a Field De-Rotator for the LX-series telescopes, but it is not yet available. The 'scope is much more stable when mounted in Alt-Az mode, but PEC is not implemented in Alt-Az mode, so I'm not convinced it will be the God-send it might appear to be. In any event, the Meade wedge is just OK -- a better solution might very well be the Milburn Wedge. Various approaches to strengthen the wedge have been discussed -- all I have done is add a second set of bolts to the plate and put washers under the bolts that mount the wedge to the tripod. While I'm on the subject, I also use a ScopeSaver. It's great.
Accurate polar alignment is an important issue, both for tracking accuracy and pointing accuracy. I have found that the best trade-off is when I only have a fairly accurate alignment, so I don't drift align. If I slew from one part of the sky to another, I might need to go out to the 'scope to sync it, but my 'scope has a large amount of Dec backlash and the system tends to track better if slightly mis-aligned, as the Dec corrections all tend to be in the same direction. I use a modified iterative method, which is based on the manual, my own observations, and discussion on MAPUG, especially Phil Perkins. I have embedded a large tile in my lawn, and put two legs on the walkway and the third on the tile. The ScopeSaver has a cross-check level which is much more accurate than the bubble in the base of the 'scope, so I know that my level is pretty good. I start the 'scope with the tube at 0HA and 0Dec, which is where it expects to be. I then slew to a bright star and sync. Next I slew to Star 19, Polaris, and use the azimuth adjustment on the wedge to get Polaris lined up in the finder. I ignore altitude at this point. Next I slew back to a bright star and use the keypad to center it in the finder, and sync. I used to only take out half of the error in both Alt and Az each time, because errors in one turn into errors in the other and taking out half of each error is the fastest route to the correct solution if one doesn't know which is which, but I realized that the level is much more accurate than the compass, and that most of the error is really in Az. Therefore, I now ignore the Alt error for the first few iterations and take out all of the Az error, and go right back to Star 19 and adjust the azimuth again. After two or three iterations I switch to an illuminated reticle eyepiece and do the half of each error trick for another two or three iterations. I don't bother with Meade's Polar Alignment mode at all any more. There is no need to wait between iterations, so it takes about fifteen minutes to get a good alignment.
I want a secure connection between my 'scope and CCD camera, so I use a Meade Basic Photo Adapter. This has t-threads on one end and a slip ring on the other. Balance is achieved through the use of a Losmandy 2-D counterweight system.
I have long been a firm believer that it is important to match the focal length of a telescope to the pixel size of a CCD camera. I use both a Celestron rear-cell 0.63x reducer-corrector and a Meade 0.5x t-mount reducer. To my dismay I recently discovered that I had miscalculated the reduction ratio achieved when I first got my ST-7. It turned out that the t-mount reducer was only achieving 0.8x when mounted directly to the front of the ST-7. What messed me up is that when combined with the rear-cell reducer I got about 0.5x, which is what I expected from the t-mount alone. When I used TheSky's Image Link feature and got 1.4 arc-sec pixels, I thought I must have used just the t-mount alone, and went merrily on my way. Nagging concerns led me to recently remeasure my field of view, and I discovered the problem. Now, I use the 2" barrel supplied with the ST-7 as a spacer, as it has t-threads on both ends. I now have a choice of focal lengths and effective pixel sizes to choose from, with f/5 (1.4 arc-sec) and f/3 (2.4 arc-sec) the ones I use most. The object here is to balance the sensitivity of the system against the resolution. It is generally accepted that the best trade-off point is when the pixel size is half of the effective seeing.
It can be very frustrating trying to get a CCD system into good focus. I use a Hartmann mask. It helps reduce image shift to approach final focus by turning the focus knob counter-clockwise, so I start from that side of focus. I use an alignment star and 3x binning to achieve coarse focus -- when the donut-hole shadow of the secondary closes. Next I take a few second full-resolution image, pick a dimmer star and drag a box around it. Even with the mask, it is easier the judge accurate focus with a dimmer star. I then place a Hartmann mask over the front of the 'scope. My mask is cardboard, 11" in diameter, and has two 3" diameter holes placed on a diameter at 3.5" radius. The mask functions as a split-image range finder. I adjust the focus knob to merge the two star images obtained when out of focus into one round image. IMPORTANT: don't forget to remove the mask :-) With TheSky I can float the Focus control panel over CCDSoft, and thereby run my JMI focus motor while watching the display. This is a big improvement over the endless trips back and forth I used to make. It takes some practice to be able to tell how far is far enough without going past proper focus. Once I turn the knob clockwise some slop gets in the system, but if I approach final focus from the right direction, while pointing near the zenith, I don't have much trouble with mirror shift causing either focus shift or image shift. Once good focus is achieved, the Hartmann mask allows me to parfocalize an eyepiece very accurately, and then find that focus again later. Judging focus with an eyepiece without a Hartmann mask is an iffy proposition, as it is difficult to keep one's eye relaxed to focus at infinity. With the mask, it's a snap.
The next step is motion calibration. The ST-7, with its integrated guide chip, is a pleasure to work with. Once the imaging chip is in focus, I know that the guide chip is as well. I don't have the freedom to move the guider around to find a guide star that I would have with an off-axis guider, but I know exactly where the guide chip is. The images the guide chip sees are as good as those seen by the main chip, which is not the case with an off-axis guider. With TheSky, it is easy to slew the 'scope to put the bright alignment star that is still near the center of the image chip's field onto the guide chip. I have found that it is important to use a bright star for calibration, so that the system doesn't get confused. I put the star towards the lower right-hand corner of the guide chip to ensure that it doesn't move out of the field during calibration. I usually do motion calibration near the zenith, while I'm setting up, and haven't found the need to re-calibrate unless I go to the extreme of the Dec I can reach. I tend to calibrate with the Aggressiveness set to 10 and image with it set to 8 or 9 -- I feel I get better results if the system undercorrects a bit each time, rather than overcorrecting and "hunting".
So, I'm all set up and ready to go. Here is where TheSky is really a great help. The large database of objects, along with the accurate depiction of the field of my system and the size of the objects, lets me pick a likely target. I then use TheSky to slew the telescope to the target. A ten-second exposure, in 3x binned mode, is enough to see most Messier objects. I float TheSky's motion control panel over CCDSoft, so I can slew the 'scope in one arc-minute increments to frame my target.
Once I'm happy with the composition, I go to self-guided mode to find a guide star. I feel I get the best results if I can use two second guiding exposures, but I have gotten good results going to fifteen seconds. If I don't find a good guide star at first, I use TheSky to find one. Since I know the position of the guide chip in relationship the the image chip, and have an accurate depiction of the image chip in TheSky, it is usually easy to find a good guide star. Often a good guide star is only a few arc-minutes off the chip, but with TheSky I can tell which way to go. If more than one star of similar brightness shows up, it is important to use enough exposure for the system to reliably tell the stars apart.
Several factors effect the S/N and how "deep" one's results are -- the efficiency of the ABG ST-7, sky brightness from light pollution or the moon, tracking accuracy and thermal noise among them. One of the great aspects of a CCD system is that all of these factors can be combated with more exposure. I have found that I need _a lot_ of exposure to get the results I want, so I usually take a series of fifteen minute exposures. If I am working on a particularly bright object I use five minute exposures. I have tried CCDSoft's "Track where you were" feature, but have found that I get better results if I let the 'scope drift a bit between exposures. I therefore just "Track where you are" several times. I want each exposure to be long enough to have a good S/N, but not so long as to saturate the detector, or as to be a drag to throw out if it's not up to par.
Now that I have a series of raw frames, the fun starts. The first issue is raw frame calibration. Each raw frame needs to be adjusted for thermal noise and sensitivity variations. The process is straightforward: take a series of dark frames of the same length and at the same temperature as your raw frames, and average them together for a master dark frame. One dark frame is good for all raw frames taken at the same temperature and length, so a library can be built. The next step is flat fielding: a picture of an evenly illuminated surface is taken and averaged into the dark-corrected raw frame. I use the method outlined in the SBIG manual -- a white posterboard is set at 45deg. in front of the 'scope, and a light is held above and parallel to the tube, pointing towards the rear. A small white card is held in front of the light at 45 deg, parallel to the posterboard. I take a series of raw flat fields and matching darks, dark adjust each raw flat and average them for a master flat field. This is a picture which shows the sensitivity variations of the optical train, and needs to be recreated for all combinations of reducers, filters, smudges, etc. My Meade t-mount reducer causes some slight vignetting, and there is a flaw in an internal coating of my Celestron reducer, and some variation across the chip, but all of these artifacts flat-field out.
Once a series of calibrated frames are created, it is time to prepare the final image. There are many software packages and method savailable, but here is the general process I am currently using: I do my basic image acquisition and calibration in CCDSoft, and use it to resize each calibrated raw frame to three times its original size. I then use the Centroid tool to mark a star in each raw frame and add them together. I feel I get better results by adding the frames instead of averaging them. Then I resize back to the original size, and do a mild unsharp mask. I blend the two images together, from 70/30 to 50/50 so as to sharpen it up a bit without getting too "processed" a look. I then switch to DeepSky to adjust the high, low and gamma. I also use DeepSky to create mosaics. I have MIRA A/P, but don't use it as much as I should. There are some things it can do that are difficult or impossible to do with other packages, but I try to avoid most of those issues and call a raw frame "bad" if it has those problems. I'm really just interested in taking pretty pictures, at this point. The manual alone is worth the price of admission, though.
People:
http://www.primenet.com/~handy
Howard C. Anderson
http://www.efn.org/~mbartels
Mel Bartel's Home Page
http://people.delphi.com/sb635
Steve Bell
http://www.fred.net/rick_b/others.htm
Rick Bright's Home Page
http://kencar.simplenet.com
Ken Carter
http://www.astropix.com
Jerry Lodriguss
http://home1.gte.net/kmilburn/index.html
Ken Milburn
http://www.astrocruise.com/
Phil Perkins
http://www.isomedia.com/homes/cvedeler/space.htm
Chris Vedeler
http://www.frazmtn.com/~bwallis
Brad Wallis
http://www.sfwebb.demon.co.uk
Steve Webb
http://members.aol.com/wmti/ccd.html
Rob West's Page.
Places:
http://www.astromart.com/
ASTROMART - The Global Astronomy Marketplace
http://www.system.missouri.edu/atm/
ATM Archives
http://woland.mit.edu/ccd/
CCD Archives
http://www.sbig.com/
SBIG Online - Homepage
http://www.bisque.com/
Software Bisque
http://www.skypub.com/
SKY Online - Sky Publishing Corporation
You are visitor #
