Home / Uncle Rod, I Want to Take Pictures
by Rod Mollise 07/29/08 | Email Author
Voice your opinion about this topic in the forums
Sure you do. You think you do anyway. Despite its very real rewards, astrophotography is still the most frustratin’ way for a beginner—or anybody else--to spend time under the stars. Probably always will be. The good news is that with today’s gear it is not as difficult as it used to be to get images that will make you happy. The astrophotography game has gotten easier and cheaper if it still ain’t easy and still ain’t cheap. Yeah, you’ll shell out bucks, and no matter how many greenbacks you sacrifice, you’d better be prepared to take on legions of gremlins who will try their dangdest to make your astrophotos look like hell. Before you can begin the struggle, though, you have to get off square one. What you gonna take them pictures with?
The imaging equipment question I hear most frequently from newbies of late? “Uncle Rod, what should I buy, a DSLR or a real CCD camera?” If you’ve seen my astrophotos, you know I ain’t no great shakes at celestial picture-takin’ despite 40 years of tryin’. But, while my results might make you wonder, I have spent a lot of time tinkering with a lot of different cameras, and can at least share my thoughts on what works for me and what might work for you. What I’m gonna do here is look at the three major groups of cameras and outline their individual strengths and weaknesses. Three kinds? Yep. In addition to astronomical CCD cameras and digital single lens reflexes, I like to put the lower-end astro cams in their own group—in some ways they are markedly different from DSLRs or astronomical CCDs.
At the bottom of the price ladder, cheaper than either mainline astronomical CCD cameras or, in some cases, DSLRs are the bargain astrocams. I’m thinking of the Meade DSIs, the Orion Starshoots, and a few other similar cameras that lurk at the five and dime end of the scale. At this time, Orion and Meade seem to be turning away from their low-priced-spread roots and moving on to more expensive and capable “pro” versions of their cameras (which are actually quite similar to their earlier models in most ways). Some of the cheapies are still being made; however, and all are readily available used.
The first of this new breed of CCD to make its appearance was the original Meade DSI. When it landed several years back, it was an honest-to-god revelation. My initiation into deep sky CCD imaging about a decade ago had been rough. I’d been able to scrape up the funds for a Starlight Xpress MX5, but despite my hopes, what a gull-derned disaster that turned out to be. Oh, it was well built and engineered for the time, and people smarter than me made impressive images with it. But it was just too hard for Unk to get going despite 30 years of playin’ with film. My images were poorly framed, poorly guided, poorly exposed, just poor in general despite me workin’ like a Trojan and sweatin’ like a pig. I sold the MX5 and took up webcam planetary imaging with a SAC7 as my sole concession to the CCD revolution. But I still wanted to do deep sky imaging, and, with film obviously dying, it looked like CCDs would soon be the only game in town.
When Meade debuted the original DSI, which they claimed would allow e’en a goober such as myself to take good deep space pictures “the first night out,” I opened my wallet pretty quick; hope springs eternal, they say. The DSI was, unlike my MX5, a one-shot color camera. That meant it could take a nice color picture with one exposure, just like my friendly Instamatic. To get color with the MX5—and almost all other CCD cameras of the time—three separate red, green, and blue filtered images had to be taken. I wanted color, and this looked like a way for me to get it. The DSI didn’t beat the MX5 in one regard, however; it wasn’t cooled. Most CCD cameras, then and now, are chilled to low temperatures to reduce the thermal noise emission that will make your picture of M13 look like it was taken through a Maine snowstorm. The DSI, according to Meade, wasn’t cooled because it didn’t need to be.
The Deep Space Imager, said Meade, was designed to shut down at least some of its heat-producing internal electronics during exposures. That wasn’t the only innovation, either. Unlike Meade’s earlier “Pictor” cameras and many of the more expensive cameras on the market four or five years ago, the DSI was equipped with a modern, fast USB 2.0 interface. What a relief that would be. I could go inside, pour a shot of Rebel Yell, and watch an episode of The Dukes of Hazard while my serial MX5 downloaded a single exposure to the PC. The software shipped with the DSI was purported to be no slouch either. Not only would it do the usual things other camera control programs could do, Enivsage could expose and stack multiple short exposure subframes to produce the equivalent (well, almost) of a long exposure image—discarding any frames that didn’t make a user-selected “quality” level. That was just the beginnin’. Envisage had features like its “magic eye” focus indicator that some of the most expensive CCD-ware of the time did not boast.
The above was what was in Meade’s big magazine advertising spreads, anyhow. And…well…some of us were of the opinion that the company’s breathless ads might occasionally be a trifle optimistic—at least when it came to what the bunglers among us could accomplish with Meade gear. How did this 300 buck camera/software combo work in the hands of one of these bunglers? I found the DSI to be flat-out amazing. Almost unbelievably, Meade’s claim of great pictures the first night out came true; my first light M13 and M17 looked great--to me, anyhow--and were at least as good as anything I had ever done with film. I was finally a CCDer. Me.
It wasn’t even that hard, though, admittedly, those many years struggling with an OM1, an OAG, and a roll of Fujicolor stood me in good stead. I at least had a vague idea of how to proceed. I set it up DSI and scope in the simplest fashion possible—unguided on a CG5 mount exposing through a C8 speeded-up to about f/4 with one of Meade’s reducers. I mashed the “go” button on the software, walked away, and let the DSI do its thing by itself. I was frankly amazed by the images that appeared on the laptop screen once my sequences finished. This simple set up was able to bring home not just bright stuff like M13 and M17; examining at a shot I took of the Deerlick Group (NGC 7331 and company) revealed the presence of not only the big galaxy and her wee NGC buddies, but little sprites, frighteningly dim PGC galaxies. I was mucho impressed with the camera’s color rendition too. It was, frankly, the first time I had seen anything like “accurate” color delivered by a one-shot camera.
Minuses? The main downcheck was the small size of the chip, 5.59mm x 4.8mm. The CG5’s deadly-accurate go-to was pretty good at putting objects in the minuscule frame at f/4, but there was no denying the resulting pictures were on the small side, and that some of my friends with less well-behaved go-to rigs were havin’ trouble acquiring dimmer targets. Still, who was I to carp when the price of the thing was about 1/3 what I’d paid for my mean ol’ MX5? While some of my Boudreauxs found the Envisage software difficult or even maddening to work with, I never had any trouble with it. Oh, the user interface could have been less clunky and less complicated, but there’s a limit to how much simplification can be done when a program has as many options and abilities as this one did and still does. Oh, how did the no-cooling/passive cooling thing work? Pretty well. The images were noisier than those produced by chilled cameras, but were quite acceptable.
Yeah, I’ve outgrown the DSI, I reckon (it still does yeoman duty as my autoguider), but Meade did not. To their credit, they stuck with their creation, continually upgrading and improving both hardware and software. The current top of the line version, the DSI III with a price tag of $1295.00, moves away from the bargain-novice territory a bit, but it’s still very reasonably priced compared to much of the competition (the previous generation, the $599.00 DSI II, is still available). The big advantage of the III? A 2/3-inch 1.4 megapixel sensor that produces nice big pix. The DSI III is available in both monochrome and color versions, so you’ll need to decide between the greater sensitivity of the black and white camera and the greater ease of producing color of the one-shot model.
Who is the DSI (or the similar Orion StarShoot II) for? For the CCD novice? Most definitely. But not just for the novice; they are for anyone who wants an easy to learn non-intimidating camera. If you can get a target centered on a chip and focused, you are assured of bringing home images that will please—and maybe even produce nods of respect from the imaging gurus down to the club. No, you probably won’t be doing 16 hour LRGB exposures with one, but most of us don’t aspire to that anyway. Heck, given the capabilities of the DSI III (or Orion’s Starshoot Pro), you may find you never need to move up to the SBIG or Apogee league.
You do want to take that next step up? Or you just want to start out with something a little more sophisticated? Up one tall tier from the DSIs and their sisters are the cooled astronomical cameras. What’s that you say? Some entry level cameras like the Orion StarShoot are cooled, too? That’s true, but the next bunch, represented by SBIG, Starlight Xpress, Apogee, FLI, and quite a few others, are different. Most of ‘em don’t just have solid state Peltier coolers like the little guys, but regulated coolers. Not only that, their Peltiers can often be supplemented with water cooling when necessary. Why the regulation? If the camera’s temperature is not kept steady, noise reduction from subtracting dark frames will not be nearly as effective. Water coolin’? That is not a new idea for amateur cameras. Way back at the beginnin’ of the CCD era in the early 1990s, folks used windshield washer pumps and buckets of cold water to carry heat away from their homebrewed Cookbook CCD cameras (but not me--I admit I never did finish buildin’ mine). In modern CCDs, this liquid cooling allows the temperature to be kept low under high ambient temperatures (like down here in Possum Swamp).
The biggest plus for the big boy cams ain’t their coolers, though; it is their chips. They are, yep, big. Some are considerably larger than even the nice-size sensors in the DSI III and the StarShoot Pro. More importantly, most of these high-falutin’ CCDs are also more sensitive than the video camera chips and similar that have often been used in the bargain cameras. They are frequently of better quality as well, with fewer hot/dead pixels to make for trouble during image processing with Photoshop.
Deciding to go for caviar instead of Vienna sausage for a first camera is often just the start of a confusing journey. Not only is there a passel of companies selling cameras in the 1000 buck and up range (and that is the range is for cooled astro-cams), each outfit makes a bunch of models. Which one? If you insist, I’ll identify my ideal step-up/better CCD, but mainly what you should do is eliminate from consideration any of these manufacturers’ bottom-of-the-pile models. Sure, SBIG’s ST-402, for example, is a well-made, reliable camera, but for your $1500 smackers you’ll be getting a relatively small 6.9 x 4.3mm monochrome chip. If you really need to stay on the cheap side of the street, I advise seriously considering the DSI III or the StarShoot Pro. They have significantly larger sensors that produce bigger images and are easier to work with. Nice, big high resolution images are not the only reason to choose bigger chips, however.
Lots of boys and girls get started in CCDing, and lots of them never make it very far. Some just decide they prefer simple visual observing to the cart-out-a-ton-o’-gear game, but many are defeated by their cameras. The main reason for that, usually, is they scrimped or chose unwisely and bought something with a too-small CCD chip. Small sensors present two very disheartening and difficult hurdles to the CCD newbie. The biggest stumbling block is that, as mentioned earlier, it’s hard to get objects in the field of a small chip. Even if you have a fairly accurate go-to system that will place DSOs in a 25mm eyepiece every dad-gummed time, you may—no, you WILL--find it frustrating to center targets with a small chip cam, even at short focal lengths. Hair-tearingly frustrating. Also, once you’ve got said target acquired, you will quickly discover the “magnification factor” imparted by a small sensor makes even minor guiding errors very noticeable. Your images will have egg-shaped (or worse) stars unless they are accurately guided.
So, if you want to do this right from the beginning—or at least don’t want to do a lot more work to get good pictures than you hoped—my advice is be prepared to bite the bullet and spend around 3 grand for a camera that will make learning imaging a little easier and will grow with you as your interests become more sophisticated. There are many cameras that will fulfill these things, and you can go all the way up to big, ol’ SBIG ST11000s (make sure a real big chip matches the resolution of your scope before you spend needlessly) and even bigger and more sophisticated rigs from SBIG and others. What do I think will please an awful lot of folks, though? My pick is the SBIG ST2000. While this one is available in a one-shot color version, my advice is get the monochrome model. Not only is it more sensitive (the built-in filters of one-shot color chips mean they will always lose out to black and white CCDs when it comes to sensitivity), it is more suited to scientific tasks should you decide you want to do more than take pretty pictures.
The ST2000 has just about everything I’d recommend for imagers old or new. Start with a good-sized sensor, the 11.8 x 8.9mm Kodak KAI-2020, which boasts nearly 2 megapixels and delivers an expansive 1600 x 1200 pixel image. It’s got a very effective, regulated cooler; it’s available with the easy to learn and capable CCDSoft camera control/processing software; is compatible with just about every other camera control program sold; and, maybe nicest of all, it can be ordered with an onboard guide chip.
As you likely know, even if you’re new to imaging, you can’t just set up the scope, slap a camera on it and expose for as long as you want. Even if you are well polar aligned, small irregularities in most mount’s gears, the “Periodic Error” you’ve heard tell about, will make your stars trail in the final images. How do you get around that? You either combine many short exposures as I did with my DSI (works) or take longer exposures (much better for keeping noise down). To take long exposures, you will need to guide. How do you do that? You use another CCD camera, usually a small inexpensive one, to monitor the position of a “guide star” and make small aiming corrections via the scope drive when that star drifts off center. It’s possible to do this by mounting the “guide camera” on a small “guide scope” (usually a refractor) piggybacked on the main telescope, but that requires more gear, and care must be taken that everything is bolted down securely or you’ll get trailed stars no matter how well your mount is autoguided.
SBIG’s proprietary autoguiding system makes taking longer exposures a snap. Guiding is done, as with a separate guide scope, by another CCD camera. In this case, though, that’s an additional CCD chip mounted alongside the main (imaging) chip in the same camera housing. In the ST2000, it’s done by a little CCD identical to what was used in the company’s pre-ST402 entry level cam, the ST237. The ST237 chip monitors the same field as the imaging chip through the same telescope, making guiding easy and effective. The combination of the onboard guide chip and SBIG’s good software means I can easily do 15 minute exposures (that go very deep) with my humble and inexpensive Celestron CG5 mount. How much will this goodness hurt? A “dual-chip” ST2000 will set you back about $3500.
So, you have to be willin’ to shell-out three grand plus to get More Better Gooder than a DSI? Not necessarily. There is a third path to astro-imaging enlightenment, Grasshopper, in the form of the digital single lens reflex, the DSLR. When these cameras first appeared (well, the first popularly priced ones) about a decade ago, astrophotographers were eager to try ‘em. With their removable lenses and standard lens mounts, they could be used with the adapters and accessories we already owned. They might be more convenient to use than an astronomical CCD, too, since they could be used without the services of an expensive laptop computer and its complex software. Unfortunately, early tests were not encouraging: lots of noise due to the lack of cooling, and not much sensitivity due to the CMOS chips most used in lieu of CCDs. The early cameras couldn’t even expose for more than about 30 seconds, meaning many noisy frames had to be stacked into final images.
That would have been the end of the DSLR story for astrophotography if the camera makers had stood still. But they didn’t. All the biggies—Nikon, Olympus, Pentax, Minolta, and, especially, Canon—continued to improve their DSLRs, reducing noise, lengthening exposure time (before long most cameras sported a “B” exposure setting just like their film ancestors), and improving sensitivity and spectral response of the chips. One maker, Canon even went so far as to not only acknowledge that its cameras were being used for astrophotography, but actually produce one (for a brief time) designed for celestial picture taking, the Canon 20da.
In addition to the other improvements, DSLR makers continued to increase the size of their sensors, and today all cameras sport CMOS chips that are at least “APS” sized, just a wee bit smaller than a 35mm frame; typically delivering resolutions in the 4000 x 2600 pixel range, larger than the chips in all but the most obscenely expensive astronomical cameras. Soon, DSLR-crazy astrophotographers were turning out color pictures that rivaled the best tri-color shots done with astronomical cameras costing five times as much or more—the average mid-range DSLR can be had for less than $1000 with a decent zoom lens.
Should you steer toward the DSLRs instead of the SBIGs then? Maybe. If nothing else, it’s much easier to justify the purchase of a DSLR to a non-astronomer husband or wife than it is an astronomical camera. Not only will it be cheaper, it can be used to take pictures of the kids the next time y’all pack ‘em up and head to Dollywood for a vacation. There is also the fact DSLRs are much easier to learn than astro cameras. While some advanced imagers use DSLRs in much the same fashion as astronomical cameras, running their Canons and Nikons from laptop PCs and “calibrating” images with dark, flat, and bias frames just like they do with SBIGs and Apogees, it is possible to get extremely nice DSLR pictures with without worrying about any of this complicated stuff. As mentioned earlier, a laptop is not needed to take long exposures; the only additional item required is the camera manufacturer’s remote “cable” release (to hold the shutter open in B mode). I once asked one very advanced STL11000-totin’ astrophotographer I met down Chiefland way what he thought of DSLRs. Though I’ve been impressed by these cameras and have come to love my Canon 400D, his reply still surprised me, “Rod, if all I were after were pretty pictures of showpiece objects, that’s what I’d use. There is no reason to buy anything else anymore.”
As I usually say about any astro-gear, however, it ain’t all gravy. There are some pretty substantial minuses, too. Not only are DSLRs less sensitive to light than astronomical cameras, even the under $1000 brigade, they are much less sensitive to red light. All DSLR manufacturers place an infrared blocking filter over their cameras’ sensors. CMOS chips, like almost all electronic imaging chips, are overly sensitive at the red end of the spectrum. Without a filter to remove the deep-red, terrestrial images are hard to color-balance, and pore li’l Bobbie Sue will look like she fried in the Sun on your trip down to the Redneck Riviera at Gulf Shores, Alabama. So what? This filtering makes it hard to get good exposures of dim red nebulas. DSLRs can be modified to remove these filters, or purchased with them already removed, but this is a job for experts like the good folk at Hutech, and this service begins to drive the price of the “cheap” DSLR into SBIG regions. The good news, however, is, despite what you may have heard, it is quite possible to get credible images of even faint clouds like IC434 with an unmodified camera—it just takes longer exposures and more processing work.
What else do I turn my nose up at? The heralded “don’t need a laptop” thing turned out not to be strictly true. Sure, you can run the camera with just a remote release, but a laptop makes life much easier. Remember how hard it was to focus an SLR on even a bright star? DSLR viewfinders are even dimmer and smaller. Yes, the latest models do have “live view”—a constantly updating video display on their LCD screens--that makes focusing easier. These screens are relatively small, however, and it will always be easier to focus using the large display of a laptop running a DSLR-centric camera control program like Nebulosity. Software like “Neb” makes DSLR imaging easier in other ways too. Since most imagers stack numerous relatively short exposures (5 minutes or less, typically) in the interest of keeping thermal noise down, it’s nice to be able to tell the laptop to have the camera take a series of exposures and be freed to walk around the observing field and annoy your buddies while PC and DSLR do their thing without supervision.
Finally, if, unlike Your Silly Old Uncle, you have dreams of doing real science, a DSLR is not the camera for you. Their one-shot color CMOS sensors are neither sensitive enough nor linear enough to make most scientific pursuits practical. Nobody says you have to have only one type of camera, though, and even if you add a top-of-the-line astronomical CCD to your stable later, you will still find uses for the DSLR on the ground and in the sky.
If I were just starting out in imaging and were able to spend 800 hundred bucks or so, I would glom onto a DSLR in a real quick hurry, no doubt about it. It’s just so easy to get your feet wet with one of these cameras. Even if you are not sure you are ready for the autoguided - prime focus bigtime, you will still be able to take some knock-your-socks-off pretty color pictures of the heavens with a DSLR right away. Buy a cheap piggyback bracket for your telescope and a remote release for your camera, do a halfway decent polar alignment, slap that Canon on the scope, and start squeezing off 30-second – 1-minute frames through your “kit” telephoto lens. A little processing of these images and you will literally be jumpin’ for joy.
How do you process images? And how about that autoguiding stuff? Do you need PEC, too? Do really have to spend all that money for Adobe Photoshop? These things are subjects for a whole ‘nother blog entry on another day. Or, more properly, a whole book like Mike Covington’s or Ron Wodaski’s or Jerry Lodriguss’. Whether your rig is a simple DSI or a complex SBIG, astrophotography is unavoidably a complicated and often maddening pursuit, as I’m sure every astrophotographer you’ve talked to has warned you. You can’t read and learn too much. There’s a long and steep learning curve, and you can’t expect to master this craft in a week or a month or a year. Don’t let that scare you off, though. It’s also a wonderful and infinitely rewarding part of our avocation.
Once you understand it will be a long, long time, if ever, before you see your images in Sky and Telescope’s “Gallery” section, and, like me, just look on your pictures as souvenirs of your deep sky travels--like that snapshot you took of Aunt Lulu up at Rock City--you will be one happy camper, egg-shaped stars or no. Yes, you can go online and look at a million beautiful and perfect amateur and professional pictures of The Whirlpool Galaxy. But I guar-ron-tee none will look half as good as your first underexposed M51. No, it will not be perfect, but it will be yours.