Kenyon Astrophysical Observatory

Successful CCD Imaging

Introduction

CCD Imaging has probably generated more excitement in the amateur astronomical community than any other technological advancement in decades. It has evolved the capture of images and information more efficiently, and quantitatively than other methods. CCD's have found application not only for imaging, but in photometry, and astrometry as well. The benefits of being able to IMMEDIATELY view an image and evaluate it at the telescope have significant advantages. However the benefits of this evolution can only be realized if the observer has a very good understanding of the process and techniques necessary for successful CCD imaging.

I have found that, although not absolutely necessary, a very skilled observer with a solid astrophotography background is important. A great deal of the "telescope driving" required for object acquisition will stretch the skills of the best and most experienced observers with a strong command of the sky. Second, examining 10 arc minute star fields, plotted with Megastar to confirm an 18th mag galaxy, will put to the test your star hopping and celestial navigation ability.

Once you are prepared to collect an image with the ccd, most of the skills needed during and after image collection are very similar to astrophotography and dark room techniques. In particular, the ability to diagnose problems such as focus and guiding errors are virtually the same. The primary difference lies in the digital representation and computer manipulation of the image.

As has been alluded to above, equipment knowledge of not only your telescope, but strong computer skills on the platform of choice are essential. Operating system, environment, file system and I/O interface capability will all come into play during an imaging run at the telescope.

The Telescope mount

Let's look at the basics of successful ccd imaging literally from the ground up starting with the telescope mount. This is the "foundation" for everything you will be doing. Stability is absolutely essential here as the need to eliminate all sources of flexure and vibration allow your telescope to support the ccd head, cables and cryogenics. Short of a permanent mount which is cemented into the ground and isolated from the observing area, there are a number of things which you can do.
With a portable mount, stiffening the tripod or pier is the primary area to concentrate on. Add weight to the tripod at the center tie point. If possible fill hollow piers or tripod legs with wood, sand, epoxy or even cement. This will add weight (maybe a great deal) and you will need to make a tradeoff between increased stability and portability.

[Picture of mounts]

Next, the drive mechanism needs to be backlash free and accurate. Pressure loading the worm and a periodic error correcting (PEC) dive corrector will save many of your images. If your telescope has a long focal length (>60"), this may be the only way to succeed at imaging longer than 5 minutes.

[Picture of worm and gear]

An often overlooked aspect of the drive mechanism is noise immunity and grounding. If you have a permanent site, be sure that the mount, gear plate and metal housing are attached to earth ground. Use a grounded extension cable for powering your electronics, and connect the metal components of the mount to earth ground. If you have a semi permanent site or favorite observing location, considering driving a copper electrical grounding rod into the earth near your telescope. A solid earth ground will help avoid data communication and electrostatic discharge (ESD) problems which can lead to system problems and poor image quality.

The telescope tube itself needs a good evaluation as well. Examine all of the "hard points" or where the tube is attached to the mount and where anything is attached to the tube. Pay particular attention to the focuser and ensure that it will be able to hold the weight of the ccd head without flexing. Next, an electric focuser with very fine positioning will save an enormous amount of time and frustration. As with the mount, connecting the focuser to earth ground is also a good practice.

Last, install the ccd head, and balance the system by starting with a radial balance of the tube. In other words, counter balance the tube on the side opposite the ccd head. Next balance the declination axis (telescope front to back). Either slide the tube forward and back or add counterweight to the telescope. Finally, balance the telescope's polar axis by adjusting or adding counterweights. It is best to have counterweights which are removable so that the telescope can be used visually with minimal effort.

The Optical Axis

The optical axis is the most important part of the imaging system aside from the ccd itself. The optics should be of exceptional quality and very clean. Be sure that the mirrors are not warped by their securing mechanism. A small amount of astigmatism which you have visually worked around before will lead to image problems.

A dark light path which is well baffled will help to improve image contrast. Not enough can be said for culmination, if you are looking for those 18th mag galaxies or a perfect Gaussian distribution of a 5 arc second star image, you must have a perfect optical alignment. The investment in a precision crafted optical alignment tool and laser will be well rewarded. Also, have a second person available who can help with alignment, one person should "adjust" and the other "view". The initial alignment should be done in daylight.

The best results I have seen involve using a laser to position the reflection surfaces in the tube with respect to the focuser first. Then using the optical alignment tool for final positioning of the "concentric ring pattern". I have found myself and from working with very experienced observers, it's hard to beat the human eye / brain combination for this step. Last, after dark and the glass has equalized with the ambient air temperature, a final star test should be done and the alignment fine tuned if necessary.

On the subject of cooling the glass, ccd's with their high quantum efficiencies are VERY subject to image distortion due to tube currents. I have found that a small fan which sets on the floor (ground) or a small table in the observatory that very gently blows ambient air at the mirror (on a reflector) as soon as the telescope site is out of direct sunlight is best. I begin cooling the telescope as early as possible in the evening. This is especially important and cuts into your imaging time if you wait too long on large mirrors.

I have found that keeping the fan running at low speed and pointed toward the primary (on a reflector) all night keeps the mirror very stable from a thermal stand point. It does require repositioning as the telescope is moved.

[Picture of fan on mirror]

The last step in optical alignment should be to align all optical axes with each other. This should include piggyback telescopes, finders and if used, the Tel-Rad (which I highly recommend).

Polar Alignment

This the single most important step in eliminating guiding frustration and obtaining high quality images of long duration. TAKE A LOT OF TIME TO POLAR ALIGN. Use the "drift method" and an illuminated reticule eye piece installed so that the cross hairs align with the celestial coordinate system (RA and Dec) as viewed in the eyepiece.

Begin with a course alignment of the telescope to the celestial pole. Turn on the drive corrector and let it run for a while and verify that all is well and that the RA and Dec slow motion controls function. Select a star in the South and position it at the center of the cross hair reticule. Ignore any East-West movement (this is either drive corrector base frequency or periodic error), and watch for North-South drift. Adjust the mechanical positioning of the telescope with respect to the pole as outlined in the table below. Repeat this for a star selected in the East. Iterate one more time by selecting a star in the South, adjust, select a star in the East, adjust.

**Polar Alignment Process**
Point Telescope	If drift is:	Move scope:
South(meridian)	North	East
South(meridian)	South	West
East(20 deg above horizon)	North	Down
East(20 deg above horizon)	South	Up

I recommend that your alignment be able to hold a star with NO North-South drift for twice as long as your longest exposure. With the SBIG ST-7, which has a maximum imaging time of 3600 minutes, the telescope should pass a drift test in both the East and South for two hours. You will need to make a decision as to how much drift you wish to tolerate. I typically strive for no more than a star width drift during the test. This is in essence 10 arc seconds (worst case) in 120 minutes, or less than 1/10 of an arc second per minute. With the SBIG ccdops software for example, sub-arc second accuracy in the tracking or self-guiding mode (1 minute acquisition time during guiding yields very stable results) is very possible, so this number is a good guideline for the equipment's range of capability.

At this point, the drive corrector base frequency can be evaluated and addressed (if the drive corrector has this feature). Begin by obtaining the worm period from the drive or telescope manual. If this is not possible, mark the worm gear and time one complete revolution with the drive corrector running at the sidereal rate. Verify this time by allowing the worm to run for 3 or 4 periods. The mark on the worm should return to the same position at the end of each period.

Select a star in the South near the celestial equator and place it in the center of the cross hair. Start a timer when the star has been placed under the cross hair and wait for one worm period. Reexamine the star looking for East-West offset from under the cross hair center at EACH worm period. Continue this process for 3 or 4 worm periods. The star should return to the cross hair center at each worm period interval. If not, follow the table below to adjust the drive corrector base frequency.

**Drive Rate Adjustment**
If Offset is East:	Increase drive rate:
If Offset is West:	Decrease drive rate:

At this point if PEC is available on the drive corrector being used, proceed with this correction. I have found with the PEC systems I have used, the best results are obtained by following a training process with 2 or 3 update cycles.

[Picture of Moto-Track V]

As you can tell, this is one of the more lengthy setup steps. If you have a permanent site or observatory, plan on spending up to 3 evenings with this step the first time you align. One evening drifting, one testing and tuning and one just for the drive correction system. I also repeat this process (but can complete it in one evening) following a "heavy maintenance" cycle of the telescope which includes axis, bearing, worm gear cleaning and lubrication. I also validate polar alignment following a mirror removal and replacement for cleaning.

With portable systems, I allocate at least 2 to 3 evenings per site with the first evening dedicated to the above process but compressed so as to be completed in about 6 hours. This is a stretch, but I have routinely obtained 30 minute, single star width, E-W-N-S drift accuracy with a number of different instruments.

CCD Set -Up

Install the ccd head in the telescope and be sure that all threaded collars are tightened or even better, set screw into place. I have tapped my 2" focuser collar to accept two thumb screws at 120 deg apart to increase the positioning stability when the ccd is installed. NOTE: If you do not know the approximate position of the focusing tube (or mirror position on SCT's), perform the Mechanical Focus step described below.

[Picture of focuser and cables]

Be sure the ccd head is oriented such that the image which appears on the control computer screen has a logical orientation with respect to the celestial sphere. The convention of selecting RA to run horizontally along the X axis and Dec to run vertically in Y is fairly standard. The advantage in this orientation is obvious when you view any star atlas or catalogue database printout from a program like Megastar. You can more easily match the image on the screen to the target object and it's associated star field. The orientation of the ccd head should be noted on the head itself and focuser with some semi permanent mark using tape or a felt tip marker.

At this point, data cables, power cables and cryo/cooling lines (if used) need to be addressed. They should be secured to the telescope near the focuser with enough slack to allow for independent movement of the head in the focuser. This tie down point is also a "tether" for the ccd head if it happens to slip out of the focuser, a simple backup system for preventing ccd damage.

Next is to decide where to exit the telescope with the cabling system. This will depend a lot on the telescope(s) configuration and mount being used, but in general position your control computer so that it is easy to get at the ccd head and focuser. Also consider how you are going to be acquiring objects and how you will access your Tel-Rad, finder etc..

I have found on both German Equatorial and SCT fork mounts, that bringing the cabling system to the hard points where the tube is attached to the declination axis is an excellent choice. This point is a fixed distance from the focuser, and in the case of the G-E, is at a central pivot point which is out of the way for most all sky positions. With SCT's, running the cables down a fork arm results in essentially the same advantage. This cabling path is more of a problem on larger telescopes where the cables may have to run over and down a large tube and thus limit the distance and access to the control computer's station.

One more consideration for the entire observing area, is Electrostatic Discharge (ESD). Be sure that you observe good ESD practices at your control computer and around the telescope. Eliminate material which can generate or build up a static charge such as vinyl chairs and plastic. The electronics, ccd, and computer can be adversely effected by a static discharge. The focuser motor, electric ccd shutters and the servo on your filter wheel can be repositioned by a static discharge. Ground the control computer's work table to earth ground through a 1 megohm (1/2 watt) resistor. If you can obtain an ESD mat (which you should restively connect to earth ground) do it, as this would be ideal. Last, remember to keep your self as static free as possible, especially when moving from the computer to the telescope and vis versa.

[Picture of work station]

Power up the system

Turn on the power to the ccd and control computer. Bring up your control software and start the ccd cryogenic system up. Note here the advantage of the permanent observatory site, as after opening the hatch to the observing chamber you are here!

The cooling of the ccd to a point of thermal equilibrium is now the task at hand. Consult the recommendation by the manufacture, but one hour is not uncommon to achieve best results. This is an area that you should collect data on. Even if your system is open loop (does not regulate to a fixed temperature), watch the ccd and ambient temperature reading. Note the amount of time it takes to stabilize with a minimal use of power for future reference. This step is typically done in parallel with the glass cooling step described earlier.

[Graph of ST-7 cooling rate]

One very important system check which can be done at this point, is to select a bright object and collect very fast, continuous images as if focusing (we'll discuss focusing next). Verify that the system can be slewed N-S and E-W with both the drive corrector hand controller and the control software if this feature exists. Identify and temporarily mark the north direction on the screen AND the hand controller button which drives the scope north. Repeat this for east (or west) as well. I have found this to be an invaluable crib late at night when you need to move 30 arc minutes in one direction or another to acquire that 18th mag galaxy.

Focusing

This step is probably the step that most all ccd'ers have the greatest trouble with. Part of this is due to the indirect method which must be used. That is to say, in visual focusing and astrophotography, we examine the image with our eye, and adjust the focus either manually or electrically. This is in essence a closed loop system with our brain responding in real time. The ccd focusing step is discrete, not continuous, with a view- decision-adjust-wait interval. Since this is a discrete process at a minimum, electric focus is essential, and digital positioning is extremely useful.

I typically take focusing in three steps, mechanical, course and fine. Even if you are an experienced ccd user, going through these steps can yield information which can be very useful for diagnostic purposes.

Mechanical Focusing

Mechanical focusing is in essence the "calculated focus" step and is used to setup the focus to APPROXIMATELY where it should be. This step should done in the daylight and before final ccd installation in the focuser occurs.

Most manufactures publish the distance from the shoulder of the ccd head to the ccd surface. This is the required back focus distance of the instrument. It is usually around an inch, but can vary if you have added accessories like a filter wheel. If you do not know this number, try to estimate it by looking down into the camera while holding a ruler along the outside of the mating tube. This is the distance that the top of the focuser must be moved back from the telescope focal plane so that the image focuses on the ccd surface.

To set the focuser back the proper distance, point the telescope at the moon, a planet or a very bright star. Run the focuser all the way down to the tube. Then while holding a 2"x2" piece of computer paper perpendicular to the focus tube, find the point where the image of the selected object projects the sharpest on the piece of paper. While holding the piece of paper very still, bring the focus tube up toward the piece of paper leaving a gap equal to the back focus distance obtained above.

[Picture of measuring the back focus]

Note, on SCT's, turn the focus knob until the sharp image is formed on the piece of paper about 1" further back from the tube than the ccd back focus distance. Measure from the end of the adapter on the back of the telescope to the piece of paper. Then SLOWLY adjust the focus so that the sharpest image forms closer and closer to the tube. Continue to measure the point from the back of the adapter to the piece of paper stopping when you reach the required distance.

Now the top of the focuser tube (or SCT adapter) is approximately the correct distance back from the focal plane to begin course focus. Insert the ccd head into the focuser and verify proper orientation as described above.

Course Focus

Course focus depends a lot on the control software you are using. Set up the ccd system to provide the best image possible by using all of the auto features available to you. Auto dark subtract, auto contrast and auto update. In general I like to start with a very fast imaging/download cycle using the lowest resolution of the ccd possible. On or off chip binning with a very fast (0.1 seconds) imaging time to get in essence a "real time" picture. This step is to simulate the closed loop process you would use in visual focusing. Thus the need to have the best image possible automatically displayed.

This is where the alignment of your finder(s) as discussed earlier is very important. Center a very bright star or planet in the finder. Using your slow motion control, center the object on the screen of your control computer. You should be seeing your planet or bright star as a "doughnut" on the screen (if your telescope has an obstruction). If you were very accurate in setting your back focus distance, you may have a very bright blob instead of a doughnut. This means you are very close to proper focus.

[Picture of doughnut focus]

Choose one direction to adjust the focus and continue to adjust the focus in this direction. Note with a piece of tape or jot down in your log the direction you have chosen. Make very small adjustments and observe the doughnut. The doughnut should be getting smaller and smaller with the central dark "doughnut hole" disappearing. If the doughnut is getting larger, then reverse the focus direction and change your direction notation as well.

Examine the screen data summary if you have this feature, and determine if the brightest pixel(s) on the ccd are saturating (i.e. maximum quantization level, 16383 on a 14 bit A to D). Remember with binning, the ccd has an increased sensitivity, so you will find a range of focus over which the pixels will saturate. Adjust the focus back and forth, one small adjustment at a time, examining the image between each change. Determine the center of the range at which the pixel(s) saturate. This may also be the range at which the doughnut appears and disappears as well.

Another help with course focusing is to watch the defraction spikes (if you have a spider mounted secondary) generated by the object you are focusing on. They will become thinner, and extend further out from the center of the object as you reach focus.

Last, there are products on the market to help with focusing. These accessories can be purchased, or fabricated as described below. The device ( a Hartmann mask)consists of a full aperture mask with two 1" to 2" holes cut in it. Space the holes EXACTLY 180 deg apart, and separated by 70% of the mirror diameter. The holes should be positioned at the same radius distance from the center of the mask.

[Picture of a Hartmann mask]

The process is to select a bright object with the mask in place. There will be two images on the ccd control computer screen if the object is not in focus. Adjust the focus as described before until the two images merge to one.

At this point, your focus is probably within a tenth of one focuser knob revolution , OR LESS. The key point here is that in the next step, fine focus, very little adjustment should be necessary.

Fine Focus

The fine focus object selection will depend a lot on the size and sensitivity of the system you are using. The trick here is to select a dim star which will yield an image across a small number of pixels. NOTE: This step is also very subject to the observing conditions and seeing at the time of focusing.

I will point the telescope to a star at the threshold of visual seeing (5th or 6th mag). Be sure it is centered in the finder, and use the slow motion controls to center the star on the screen. Once I have the star on the screen, I will reset the ccd to full resolution and use a reduced ccd area (or cropped) image capture. Also raise the exposure to around 3 seconds or higher) to obtain a more consistent integration result.

Select the smallest cropped area of the image you can that contains the star, but allows for some image shift. This shift will be due to seeing, wind, and focus adjustment. You may need to adjust the cropped image size if you find that the star moves out of the field during this step.

If you are having trouble fine focusing with the cropped image (or you do not have this feature available), it may be necessary to image with full ccd area and resolution. In either case, find a single dim and clearly defined star which does not saturate. The peak pixel values should be between 25% and 75% of your A to D quantization range. It is a good idea to note the magnitude of stars used for this step, as in the future it may only be necessary to repeat from this point on!

The process at this point involves evaluation of the image and associated data, then adjusting the focus and waiting for things to "settle". There is a subjective part to fine focusing, but with practice you will be able to eliminate most of the subjectivity. Remember, the computer is collecting an image, downloading the data then processing and displaying the result. If you change the focus during the image collection step, an erroneous image will be displayed. Also, if you are manually focusing, the telescope will vibrate when you touch and release the focus knob. This is why after an adjustment, examine a few image capture iteration cycles to see the true result.

As before, choose a focus direction and note it. Evaluate both the image and the image data. You should have a star which varies in size and position by a few pixels each capture iteration and (depending on your system) is 5 to 10 pixels across. The peak value of the brightest pixel(s) should be in the 5k to 10k range (assuming a 14 bit A to D). Watch the screen for a few iterations and get a feel for the average image size and data range.

[Picture of stars on the screen]

Adjust the focus in the selected direction a very small amount and wait. Re-examine the image and data for a few iterations, If you selected the correct direction, the star should get slightly smaller, and the average peak value should go up. If not, back the focus up in the opposite direction as close as possible to where you started from (the advantage of digital focus positioning) watch for the data to return to the original values (about). Continue in the direction which shrinks the star image, and increases the average peak value.

The key here is patience! If the seeing is not good, or there are excessive tube currents, the peak values may very wildly along with stellar image size. If this is happening, all you can do is wait for conditions to improve. In any event, if the peak values are varying by more than 10% it will be difficult to obtain a perfect focus.

Now you should have a perfect focus. To verify this take a few 30 second or 1 minute images (to minimize the effect of alignment, periodic error and seeing) of an area of the sky which is rich in stars. Grab these images, and bring them down to the control computer. Now examine these images (with your image processing software if necessary) zoomed in on a few stars to the point where you can resolve individual pixels. It is a good idea to note the Full Width at Half Maximum (FWHM) for future refrence.

[Picture of a perfect star image]

The star images should be round, not oblong (may be alignment or tracking) or triangular (flared due to focus or culmination). Strange asymmetric shapes may be due to the seeing. Ensure that the stellar image is Gaussian in nature. That is, bright in the center and dropping off rapidly as you move radialy out from the center. The reason for taking a few images is to look for consistencies image to image. If you can not find an image with round stars, and all other conditions are favorable, you may want to repeat the fine focus step.

Last, and very useful for the future is to prepare a "ccd focus position" or par focal eyepiece to eliminate the mechanical focus step. After you have completed the evening's imaging run, remove the ccd and select an eyepiece (possibly with an extension tube) which can be slid in and out of focus in the focuser tube. Do not adjust the focuser itself.

Point the telescope at a bright star or planet and while viewing the object, position the eyepiece in the focuser tube at the point of sharpest visual focus. When a sharp image in the eyepiece can be seen, gently tighten the thumb screws on the focuser and mark the eyepiece barrel (or extension tube) with a semi permanent mark.

[Picture of par focal eyepiece]

In the future, reinstall the "ccd focus position" eyepiece in the focuser to the mark and adjust for a sharp image in the eyepiece. Remove the eyepiece and install the ccd head. If you have been very careful in marking, positioning and focusing, the ccd should be focused. This can also be helpful when attempting to acquire difficult objects as will be discussed later.

Another common problem is that the focuser (or mirror on SCT's) may move when the telescope is slewed to the opposite side of the sky. To verify that this is or is not a problem, take a few images 30 deg above the horizon at each major cardinal point (N, S, E, and W). Examine each set of images as described above. Once again changes in the stellar image may point to a needed tune-up of the telescope focuser or mirror mounting system. If you are unable to correct these problems immediately, you will have to focus after each major telescope repositioning.

Acquiring the Object

There are a number of alternatives to getting the telescope with the ccd to point to the desired object, and they depend somewhat on the actual set-up being used.

Piggyback Telescope

This is where the ccd is installed in a telescope which is piggybacked (mounted) to a second, usually larger instrument. If the two optical axies are coaxialy aligned, simply find the object (or star field) in the larger telescope, and begin imaging with the piggyback system. The trick here is having the two telescopes pointing to the EXACT same spot. Often, the mounting hardware used to mount the piggyback telescope is difficult to adjust precisely. I have used an illuminated eyepiece with adjustable X-Y positioning of the reticule to solve this.

[Picture of piggyback telescope]

Align the two telescopes as close as possible optically before installing the ccd in the piggyback instrument. Install the ccd and begin imaging as in the focus step. Point the telescope with the ccd at a bright star or planet and use the slow motion controls to position the object in the center of the control computer's screen.

Install the illuminated reticule eyepiece in the larger telescope. Adjust the X-Y position of the reticule cross hairs to coincide with the targeted object. Now position any desired object under the reticule cross hairs, and it's on the ccd in the piggyback telescope. If you plan to remove the illuminated reticule eyepiece, mark it's orientation in the focuser for proper reinstallation.

There are a few disadvantage of this set-up, weight, two optical systems and so forth. Although, this is good way to start developing your ccd skills, especially when the smaller piggybacked telescope is used for imaging. The larger scope is used for object acquisition and tracking, and virtually all of the ccd head removal is prevented so focus and orientation are static.

Imaging Accessories

There are a few accessories on the market that are similar to astrophotography accessories which have flip up mirror and the ability to mount an eyepiece at 90 deg to the ccd imaging axis. Focusing is the same as described above, but after the ccd is focused an eyepiece is inserted in the flip mirror path and focused independently.

[Picture of flip mirror]

This accessory allows for a static ccd installation (no removal for object acquisition), however they are costly and require approximately 2 inches of back focus. If an off axis guider is added, the ccd head is now mounted a significant distance from the tube. This has implications for balance, torque on the focuser and (in the case of SCT's) limited positioning.

Single Telescope

If you are not planning on imaging with either of the previous options the use of the "ccd focus position" eyepiece described earlier in the focus step will be very important. Also, the marking of the ccd head and focuser for orientation will come into play as well.

With this method, you may need to removed the ccd head occasionally to verify the targeted object or star field. Before you do, remember, you will need to obtain flat field images if you have previously collected any light images. This will be described later!

Imaging Tools

Regardless of what your setup is I cannot hold in higher regard the usefulness of good database mapping tool which uses the Hubble Guide Star Catalogue and a robust complement of deep sky catalogues. I have used (or own) most all of the one's on the market today and have not found any better than Megastar (ELB Software) or TheSKY(Software Bisque). The database access is fast and they have many features for the ccd imager. This includes being able to flip and invert the field of interest as well as position a ccd image frame computed for the ccd and telescope being used. I have a second computer online with the CD ROM version of Megastar and TheSKY at all times during imaging.

Second, a set of digital setting circles are very valuable, not only searching for dim objects, but I have found them to be a great time saver for acquiring comets and asteroids. An added bonus here is that a number of commercially available software packages, including Megastar and TheSKY, can interface directly to these products.

[Picture of shaft encoder attachment]

Invest some time in preparing for the imaging run before sitting down to the control computer. If you do not have the luxury of an on line database, plot two sets of maps for each intended object. First plot one about twice the field (and limited to the magnitude) of your finder or visual telescope with a wide field eyepiece in it. Don't forget to orient the map properly if you use a star diagonal in the finder. Plot a second map that is about twice the field of the ccd/telescope system that you will be imaging through. Plot the ccd image scale on this map, and set the stellar magnitude filter to the maximum magnitude limit (dimmest stars) of the database.

The advantage here is clear. You can quickly locate the object's field with the Tel-Rad, finder or digital setting circles and validate the position of the targeted object with the first map. Next, go to the control computer and grab a 30 second or 1 minute image to confirm the object or star field with the second map. If your finder is well aligned, you will be close enough to identify the star field if not the object itself. Compose the image to appear as you would like it to appear on a final print with the slow motion controls.

[Picture of maps]

The key point here is to understand how the sky will appear on the screen of your control computer. Spend some time verifying orientation by imaging a few of the brighter, well known Messier objects and comparing the orientation to your maps (either printed or on line). Conduct this exercise in different parts of the sky (N-S-E-W) to familiarize yourself with the positioning orientation changes which will occur depending on the type of mount you have.

Successful Imaging

Now you are ready to begin successful imaging. Allow time at the telescope for all of the support activities necessary to obtain the highest quality images.

I have a library of dark frames previously collected at different exposure times and tempratures. This has proven to be the most efficient use of telescope time as it allows me to process an image while another is exposing.

Even if you have a temperature regulated ccd and adaptive dark subtraction software, take a few dark frames occasionally. I usually collect a trio of dark frames (to average together) after the ccd and mirrors have cooled, in the middle of the evening and after my last image. These dark frames provide you with a very good diagnostic tool to assure you that the ccd system is functioning properly. Compair these dark frames with those captured on a previous imaging session or with library frames to assure your self that everything is OK.

With respect to flat field frames, twilight in the eastern sky as the sun sets or the western sky in the morning can be used. However, if you need to remove the ccd head a few times during each imaging run, you will have to generate an "artificial twilight". I prefer this method of collecting light frames inside the observatory (at the observing site) at the end of the evening or as needed preceding a ccd head removal.

I have a piece of cardboard covered in 3 layers of white "flipchart" paper which I mount on a wall in the observatory. I illuminate it from above and behind the telescope such that there are no shadows cast on the white card. The light source is a small (10 watt) white refrigerator bulb mounted in a reflector with a few layers of white paper over the front of the reflector for diffusion.

This allows the collection of flat field frames (3 or 4 to be averaged) with a 30 second to 1 minute exposure. This longer exposure time eliminates the gradation which occurs if the flat field exposure is very short (this may happen at exposures times of less than 1 second). Don't forget to collect dark frames of the same exposure time as the flat field exposure. Collect 3 or 4 of these as well to be averaged and subtracted from the averaged flat field later.

Now your ready to begin image processing the light frames. This subject, however is an entire art and science in itself, and has been delt with often in CCD books and magazines.

Conclusion

Being active and refining your skills is probably the most important advise I can share. You can attend events with other ccd'ers like the big star parties staged around the country annually. Form SIG's (Special Interest Groups) within your astronomy clubs to share what works and doesn't work with each other. If you are not connected, get access to the Internet and sign up for the ccd mail or news groups. There are a number of (and growing) home pages which cater to the ccd community.

Following every element of these best practice discussions will not guarantee successful ccd imaging, but it will minimize the things which can go wrong. A lot of your own knowledge will come from "practice makes perfect". If you want to succeed at ccd imaging, you have to get out there and collect photons, and do a lot of it. So good luck, clear skies and good imaging.

Dave Kenyon

Kenyon Astrophysical Observatory

dkenyon@starstream.net

Thanks again for visiting the Successful Imaging Page!
Return to the KAO Home Page here