For serious astrophotography the camera is attached to a telescope on an equatorial mount, and accurate motorized tracking compensates for the rotation of the earth while taking long exposures. The necessary equipment is bulky and expensive; however, many deep sky objects can be photographed surprisingly well without tracking or mounting the camera to a telescope.
Here are some results from using 35mm and 50mm lenses on a Nikon D5100 body mounted on a photo tripod, without any tracking. Exposures up to 4 seconds produce no noticeable star trails with the 35mm. With the 50mm a safer choice is 2 sec, depending how far from the pole the object is located. Close to the pole you can get away with longer exposures before star trails become noticeable. Cropping to about 800x800 from 16 Megapixels (4928x3264) still provides acceptable image quality, and stacking 25 images improves the signal to noise ration by sqrt(25) = 5. The free DeepSkyStacker application is used for aligning and stacking. GIMP is used for final touchup, such as removing some of the light pollution.
The Nikon D5100 has the nice interval timer function that lets you set the number of frames (up to 999); after that you can continue your visual observing while the camera keeps shooting away. ISO above 1600 introduces serious noise and leads to noticeable deterioration in individual images. However, image stacking gets rid of some of that noise. Even ISO 6400 is often acceptable after stacking 30 or 50 images. Using the NEF Nikon raw format instead of basic JPG improves the results slightly but results in much larger files and longer processing time in DeepSkyStacker.
Acceptable results were obtained with the Nikkor 50mm f/1.8 and 1600 ISO when to object is in the center of the frame. Stopping down to f/2.8 improves image quality noticeably, especially off-center with much less coma. Going to f/4 yields further small improvements; in this case ISO is set to 3200 and exposure to 4 s; close to the celestial equator 2 sec. The 50mm has a distance scale with a fixed stop at infinity; however, the infinity mark is not completely accurate, it focuses a little 'beyond' infinity. You have to back off a tiny bit, which probably varies from lens to lens. On mine it is about 1/2 mm.
The Nikkor 35mm f/1.8 provides wider views. Stopping down to 2.8 and shooting 50 frames at ISO 1600 and 4 seconds exposure time provided the best results here.
Focusing is critical and difficult. Unfortunately, most current lenses do not provide a distance scale, making focusing very tricky in low light. Autofocus will work on very bright stars, but the results are unreliable. Reasonably accurate focus can be achieved by using live view and zooming in on a few stars just bright enough to show on the screen. These stars will vanish from the display when only slightly out of focus, thereby providing a good reference. An articulating display is convenient.
Still, the whole process is time-consuming and frustrating. Therefore, lenses with distance scale like the Nikkor 50mm f/1.8 are useful, once you know how much you have to back off from infinity. This Nikkor is of high quality and also cheap since it is a simple design; note that it is AF instead of AF-S, meaning that it does not have an autofocus motor, so it does not autofocus on the mid-level cameras like the D5100.
Western Australia is blessed with dark skies. Even a few miles from Perth CBD the light pollution is moderate; large and bright objects such as the Eta Carinae nebula and open cluster NGC 3532 are easy and rewarding targets.
This part of the sky is rich in deep sky targets, such as the Southern Pleiades (lower right), and NGC 3114 (upper right).
The Large Magellanic Cloud is an easy naked-eye object from dark sites, but like most galaxies it has low surface brightness, and so close to the city center even the camera and the stacking process cannot resolve much detail.
A trip to a darker site about 50km from Perth was rewarded with improved image quality, and some spectacular visual observing. While photographs provide accurate detail they cannot truly reproduce the visual experience of looking through the telescope and seeing distant galaxies, nebulae, and star clusters directly with your own eyes. Thanks to Bert and Hilary of ASWA for an amazing observing night!
Another constellation image back in Austria, really bad conditions with huge drifting clouds. Nikon D5100, 50mm 1:1.8 D, 30 frames, 2 sec, f/2.8, ISO 3200, quality basic JPEG, no dark/flat/bias frames; DeepSkyStacker on the full size images, output scaled to 50% in GIMP. Wide open at 1.8 the results are good, but stopped down to 2.8 the quality seems a little better. In spite of bad conditions the three globular clusters M5, M10, and M12 are clearly visible. Cartes du Ciel (Sky Charts) on Linux was used to identify the stars and DSO in the image.
Galaxies M51 and M101 are clearly visible in this image of the big dipper. No spiral arms of course, but M51 shows two distinct cores, and the fact that the faint fuzzy M101 shows up at all is not too bad for a 50mm camera lens mounted on a cheap tripod with no tracking at all. Same as above, 30 frames with 2 sec each, f/2.8, ISO 3200. The D5100 does a good job with noise reduction even at ISO 3200. The old infamous Nikon star eater effect does not apply to the newer Nikon cameras which use a different algorithm.
This part of the constellation Cygnus shows the aptly named North America Nebula NGC 7000. Like most large but faint nebulae it is a difficult visual target that suffers heavily from light pollution. Photographically the outlines become visible. Deep Sky Stacker has many options yielding somewhat different results. M 39 is a large open cluster and a fine binocular target. Cygnus lies in a beautiful part of the Milkyway that is always a pleasure to scan with a binocular or a small telescope. Nikkor 50mm 1:1.8 D, 50 frames, f/2.8, 4 sec, ISO 1600.
Open clusters in Ophiuchus: I4665 and NGC 6633 are nice binocular targets. I4756 is actually in Serpens Caput. Globular cluster M14 is better viewed at higher magnification. It appears almost starlike in this image. Nikkor 50mm 1:1.8 D, 50 frames, f/4, 4 sec, ISO 3200. Stopping down to f/4 should mean better focus, but the result does not differ noticeably.
My first take on the teapot: the large Milkyway patch M24 dominates the middle, to the right there is the rich open cluster M23, below that the Trifid Nebula M20 with the small open cluster M21 close to the upper left; further below the eye is drawn to the spectacular Lagoon Nebula M8 with the dark lane clearly visible. To the left of M24 there is the open cluster M25, and below one of the finest globular clusters, M22; to the right and closer to Lam Sgr (the tip of the pot) there is another smaller globular, M28. Above M24 there is the small open cluster M18, above that the magnificent Swan Nebula M17, and further above the equally impressive Eagle Nebula M16.
The pair of M8 and M20 are nicely framed in 8x40 or 10x50 binoculars. Their eerie glow is enchanting, and the fact that they are so low in the sky helps a lot with binocular observing.
The images were done with ISO 6400 and f/4; 4 seconds exposure and 100 frames. The result shows that trading ISO for f-stop is futile: while the stacking removes much of the noise the quality is effectively going down because of the deterioration in the individual frames. The rule of thumb seems to hold: for good quality at high aperture, stop the lens down one step from max.
M7 (center) and M6 (right and up) from a nice dark site. Quality is much better, especially so close to the horizon. One can see why M6 is called the Butterfly cluster.
Another take on the M24 region. The Moon was still up, so quality is not that good. Maybe next time.
The Swan is a very rich Milkyway region, M27 can be found near the center.
The cropped image shows the dumbbell shape more clearly. However, it looks slightly out of focus; maybe the hard stop at infinity of the 50mm is not completely accurate? Update: it is not. You have to go back just a tiny bit from infinity, maybe 1/2 mm or so.
Are the individual frames also slightly out of focus, or is this simply the limit this lens will go for astrophotography? After all, the situation is very demanding; extremely high contrast, bright point sources with black background. A tough job for every instrument, and certainly not a standard task for a DSLR camera lens.
