Wide-Field Imaging with CCD Cameras

splendid Cygnus nebulae
Combining the resolving power of CCD cameras with the wide-field views offered by regular 35-millimeter camera lenses, California astro imager Steve Mandel captured this 4°-wide portrait of the North America Nebula (left of center) and the Pelican Nebula (right of center) in Cygnus with his 180-mm Nikon lens, which he operated at f/4 and coupled to a Santa Barbara Instrument Group (SBIG) ST-10E camera and CFW-8 filter wheel. Mandel digitally combined fourteen 5-minute exposures, each taken through a hydrogen-alpha filter, using Adobe Photoshop and MaxIm DL software.

As an avid astrophotographer using film for many years, I love to shoot wide-field views of the sky. It's less demanding than, say, taking high-resolution close-ups of the Moon or planets, and it gives good results rather consistently.

I started out by doing "piggyback" astrophotography and was very excited when Sky & Telescope published one of my pictures in 1984 — a portrait of Cygnus above the towering pine trees in Yosemite National Park captured with a 50-millimeter lens and guided with an 8-inch (20-centimeter) telescope.

A few years later I purchased a Celestron 8-inch Schmidt camera and continued to perfect my wide-field techniques. The camera's excellent resolution, fast focal ratio (f/1.5), and vast angular coverage (4.5° by 6.5°) make it a wonderful instrument for recording large, faint celestial objects with relatively short exposures. The introduction of fine-grained films, such as Kodak Technical Pan 2415 for black and white and Kodak Pro 400 for color, made such cameras even more attractive to astrophotographers. I had a great time shooting with the Schmidt, culminating with my photos of Comet Hale-Bopp in the spring of 1997. In fact, the results were so good that I was able to make a 20-by-36-inch color print of the comet from a 35-mm negative.

In 1998 I packed my medium-format Pentax 67 camera and portable telescope, and, along with my friend John Gleason, took the gear to the dark skies of Coonabarabran in New South Wales, Australia. We came back with stunning, wide-field color souvenirs of the Milky Way and other deep-southern sky gems. I was now a certified wide-field junkie!

Arrival of the CCDs

During the 1990s CCD cameras entered the world of amateur astro imaging. Intrigued with the CCD's potential, I ordered an ST-6 camera from Santa Barbara Instrument Group (SBIG). I quickly learned that the small size of the camera’s silicon chip (6.5 by 8.6 mm) and the large size of its pixels (23 by 27 microns) weren't ideal for imaging large swaths of the sky. The camera worked fine with long-focal-length instruments, such as the 14-inch Celestron telescope that I had in my backyard observatory, but I needed a camera with either a larger chip or smaller pixels to match the short-focal-length lenses I wanted to shoot with.

Sky Coverage* of SBIG Cameras
Lens focal lengthST-7EST-8EST-10E
50 mm5.3° × 7.9°10.5° × 15.8°11.5° × 17.1°
85 mm3.1° × 4.7°6.2° × 9.3°6.7° × 10.0°
105 mm2.5° × 3.8°5.0° × 7.5°5.5° × 8.1°
135 mm2.0° × 2.9°3.9° × 5.9°4.2° × 6.3°
180 mm1.5° × 2.2°2.9° × 4.4°3.2° × 4.7°
200 mm1.3° × 2.0°2.6° × 4.0°2.9° × 4.3°
300 mm0.9° × 1.3°1.8° × 2.6°1.9° × 2.8°
*The formula given on page 4 of this article for calculating a lens's plate scale does not work well with focal lengths shorter than about 50 mm.

In 1995 SBIG came out with its ST-7 and ST-8 models and, six years later, the ST-10. The physical size of the chips increased slightly, but the pixel size shrank to 9 microns square (for the ST-7 and ST-8) and 6.8 microns square (ST-10). Now fitted with Kodak's new E chips, these cameras’ sensitivity at blue wavelengths improved substantially, greatly increasing their potential for tricolor imaging (defined on the next page). The combination of smaller pixels and short-focal-length lenses worked fine. Although the pixel scales are coarser than people are used to using for telescopic work, I was able to obtain good results.

Color Imaging

taken from the suburbs!

This 2.3°-wide view of the emission nebula IC 1848 in Cassiopeia was obtained by Robert Gendler from his driveway near Hartford, Connecticut, using a 300-mm Nikon lens at f/4 and an ST-10E and CFW-8 from Santa Barbara Instrument Group. It is a combination of individual exposures through a hydrogen-alpha filter (160 minutes) and red, green, and blue (RGB) filters (10, 10, and 20 minutes, respectively). All CCD images in this article are oriented with north up.

There are two ways to produce a color image with a CCD camera. You can either use a "single-shot" color camera, such as Starlight Xpress's MX5-C and MX7-C, or assemble a tricolor image from separate exposures through three different color filters. For the latter you need a filter wheel so you can shoot individual exposures through sets of red, green, and blue (RGB) or cyan, magenta, and yellow (CMY) filters, and then combine them digitally with image-editing software such as Adobe Photoshop, MaxIm DL, AIP for Windows, MIRA AP, or StellaImage. You can get away with using small filters if you place them between the lens and the chip.

I experimented with a narrowband (4-nanometer) hydrogen-alpha filter made by Custom Scientific, which can help resolve subtle detail in emission nebulae. I quickly discovered that this setup recorded even more detail than I had captured previously with my Schmidt camera and hypersensitized Tech Pan film.

I was anxious to try color imaging, but I immediately encountered a mechanical problem. With my SBIG motorized CFW-8 filter wheel in place, I was unable to reach focus when I attached a 35-mm camera lens to my ST-10E. I needed a way to use both my RGB and hydrogen-alpha filters with my existing camera and filter holder.

Making An Adapter

hooking up
The author's wide-field imaging adapter (1) allows you to connect any Nikon lens (2) made for 35-mm cameras to any SBIG ST-7E, -8E, or -10E camera (3) fitted with an SBIG CFW-8 filter wheel (4).
Steve Mandel
My solution was to design a low-profile adapter for my CFW-8 that would hold a Nikon lens. I attached a bayonet mount for the lens to a modified CFW-8 front plate. This accessory — which I now manufacture for other astroimagers and sell as the Mandel Wide-Field Imaging Adapter — brought the lens close enough to the CCD chip to be able to bring stars to a focus on it.

Admittedly, this arrangement did not bring the stars into focus when the lens was set at the "infinity" mark on its barrel. But I quickly learned to ignore the lens marking and find focus as I would with a regular telescope and CCD camera.

Excited, I aimed my new wide-angle camera to the sky, and I was immediately astounded by the images it delivered, especially those obtained through the hydrogen-alpha filter. I began imaging deep-sky objects that I'd never even seen before. For the first time I captured the full extent of Barnard's Loop in Orion (S&T: May 2001, page 140) and the very faint emission nebula Sh2–264 between Betelgeuse and Bellatrix. With a normal (50-mm) lens I got a whopping 11.5°-by-17.1° field! You can see more of my images online at www.galaxyimages.com.

I gave my prototype Nikon/CFW-8 adapter to Gleason, who took it to Australia for a field test. He returned with some remarkable images (see his Web site at www.celestialimage.com). Based on the positive results we achieved, I started making commercial CFW-8 adapters exclusively for Nikon lenses (S&T: December 2001, page 36). Recently, Brady Johnson, an amateur astro imager based in Toronto, has made similar CFW-8 adapters for Pentax, Minolta, and Olympus 35-mm camera lenses. Of course, you shouldn't hesitate to contact the manufacturer of any CCD camera you may own to find out what solutions they may have devised since the time of this writing.

To calculate how much sky your camera will cover with a given lens, divide 57.3 by the lens's focal length in millimeters to get the lens's plate scale in degrees per millimeter, and then multiply the scale by the dimensions of the CCD's pixel array in millimeters. For example, a 300-mm lens will have a plate scale of approximately 57.3/300 = 0.191 degree per millimeter. When coupled to, say, an ST-7E camera (with a 4.6-by-6.9-mm pixel array), the camera will record about 0.9° by 1.3° of sky (see the table on page 2 of this article for an array of examples).

Shooting with a Lens

glorious Eta Carinae
The beautiful Eta Carinae Nebula was imaged by John Gleason from Coonabarabran in New South Wales, Australia, with a 180-mm Nikon lens at f/4 and an ST-10E CCD camera. This view is a composite of four 3-minute exposures taken through a hydrogen-alpha filter. Click on this image to see the entire 3.2°-wide field.

Using the camera-lens adapter and a CCD camera is easy. You simply attach a lens to the adapter; mount the setup piggyback on your telescope; center the target object in the camera's field; and focus. I check my focus by choosing a medium-bright star and adjusting the lens until the pixels that make up the star's image register the highest possible brightness value. (If you use a telephoto lens with a built-in ¼-20 tripod socket and mount it directly to the telescope, you can facilitate initial centering on the target by putting a regular 35-mm single-lens reflex camera body on the telephoto and looking through the camera’s viewfinder.)

The SBIG cameras' self-guiding feature works well with short-focus lenses. In fact, because of the large field of view, choosing a guide star is relatively easy.

I prefer lenses with fixed focal length to zoom lenses, whose optics tends to have slower f/ratios. It's also a good idea to stop down the lens at least one f/stop; this often creates sharper star images. (For example, if your lens's fastest focal ratio is f/2.8, then shoot at f/4, and so forth.) Also, you can't beat a high-quality lens for minimizing coma, chromatic aberration, and other optical flaws. And since the CCD's sensitivity extends beyond the visual range, especially at infrared (IR) wavelengths, you need to use some type of IR-blocking filter. Otherwise, a star image that appears crisp to the eye may actually end up bloated on the CCD image.

robo focuser
For precise focusing, the author uses a modified version of Technical Innovations' Robo-Focus computer-controlled motorized focuser.
Steve Mandel

Precise focusing is essential since the lens's depth of field is very small at fast f/ratios. In other words, when you’re shooting at f/2.8 or f/4, focus becomes very critical — you're dealing with tolerances measured in thousandths of an inch. I soon got tired of focusing manually with my big fingers; it took me forever to find the lens's "sweet spot." To solve this problem I modified my Robo-Focus, a remotely operated focuser motor drive from Technical Innovations (www.homedome.com), and attached it to my camera setup using a homemade bracket and a long rubber belt that ran from the gear on the motor shaft to the lens’s focusing barrel. I adjusted the bracket so the motor provides only enough pressure to focus the lens without flexing its barrel. This system allows me to focus the lens from my computer to within 1/1000 inch, using software supplied with Robo-Focus. (Details on how to modify the Robo-Focus are given on my Web site.)

You can try the "luminance-layering technique" used by Robert Gendler to produce excellent tricolor images of galaxies and nebulae (see Sky & Telescope's July 2001 issue, page 133, or go to Gendler's Web site at http://robgendler.astrodigitals.com). Or you can shoot through a hydrogen-alpha filter to obtain black-and-white images, as Gleason and I do.

Wide-field imaging is fun and relatively easy, and it opens up — dare I say it — a wide new field for those using CCD cameras.