Issue #536: Deep Sky Imaging in Seven Easy (Sorta) Steps

All those cables!

We’ve spent the last several weeks setting you up with a telescope, mount, camera, and guide system. Now it is finally time to get outside with all that gear (assuming you, unlike me, have clear skies) and use it to capture the deep sky wonders of Spring.

Step One:  Set Up

Naturally you’ve got to assemble the telescope and mount. But where should you assemble them? Unless your backyard is really, really horrible light pollution wise, I strongly council you to use the good, old back forty the first couple of times you work with the new gear. You’ll be dealing with a bunch of unfamiliar equipment and some complex tasks, and it’s always easier to do that at home where you can run inside for a look at a manual (or for a quick drink if you get really stressed) under white light rather than squinting at the instructions with a red flashlight at a dark site.

Anyhow, set up tripod and mount and level them. How precise does level need to be for a German equatorial? Technically, you don’t have to level the mount at all. All it needs is to be level enough so that it doesn’t tip over.  In some cases, being close to level can make polar alignment easier, however. Level won’t affect polar alignment, period, but being near level means the mount’s altitude and azimuth adjusters won’t interact. When you move in altitude, it doesn’t also affect azimuth, and vice-versa, making it quicker to dial in alignment.

Next, attach the counterweight(s) to the declination shaft. Always mount the weights first, followed by the telescope. You will be mighty unhappy if you do the reverse, your R.A. lock isn’t secure, and the telescope slams into the tripod. When tearing down at the end of the run, reverse that. Remove the telescope first, then the weights. Once the scope is safely on the mount, install everything you’ll be placing on the tube:  guide-scope and camera, imaging camera, finders, etc.

Balance is very important if an inexpensive GEM is to track well, so spend enough time with that to get it exactly right. First, decide which side of the Meridian you’ll be imaging on, east or west, and balance accordingly. You’ll always want to be slightly east heavy. If you are going to be imaging on the West side of the Meridian, you should be “scope heavy.” On the east side? “Counterweight heavy.”

Polar alignment with Sharpcap...

Balance in R.A. first. Point the scope north, lock the declination lock and, with the counterweight down and halfway up the shaft (if you don’t know approximately where it should be), undo the R.A. lock at least partway and move the mount in R.A. so the counterweight bar is level. Do not let go of the scope. Now, still without completely letting go, allow the scope or weight to rise or sink. When you’ve determined which way the weight needs to go on the shaft, up or down, return the mount to counterweight down position, lock the R.A. lock and move the counterweight as required. Return the mount to counterweight bar level, and see if balance is perfect. Keep on going with this procedure until it is.

Now for the East heavy bit. When you are in perfect balance, move the weight about ½-inch up the shaft if you are imaging to the west, and ½-inch down the shaft if you are imaging to the east. That should be more than enough to ensure the R.A.  gears remain constantly meshed in the interests of good tracking.

For declination balance, return the mount to the counterweight bar level position, lock the R.A. lock and, holding on to the telescope, release the declination lock. When you know which way the scope needs to go in the mount saddle, forward or back, return to the counterweight down position, move the scope (carefully) as required return to the counterweight bar level position, and check. Keep going till the scope is balanced in declination.

What if your mount, like many in this class, is a little stiff on the declination axis and is somewhat difficult to balance? Don’t worry about it too much. Your mount is not tracking in declination, and if you’ve done polar alignment well, PHD shouldn’t have to issue many guide corrections on that axis. Good R.A. balance is far more important; declination balance can be “approximate” without hurting anything.

Step Two: Polar Alignment

If you are using a polar alignment borescope, Polemaster, Sharpcap, or the Kochab’s Clock method, do polar alignment now. None of these methods require the scope to be powered up and tracking, so it’s convenient to do the polar alignment before the mount is all festooned with cables and hand controls (I use Sharpcap these days). Take your time and do as good a polar alignment as you possibly can; you’ll thank me later.

Step Three:  Hooking Up

Connecting to the mount...

Plug in all the cables and the telescope’s hand control. You’ll have at least four and maybe five cords to deal with:  Power cable, serial cable for scope control, imaging camera USB connection, guide camera USB connection, and an ST-4 cable if you’re going to guide through the mount’s auto-guide port. Try to do a neat job with the cords, arranging them so they are not prone to snagging on the mount or tripod—which will ruin your guiding. Don’t forget to attach dew heater strips, dew heater controller, and dew shield if you need them in your environment.

Step Four:  Goto Alignment

Take care of goto alignment now. Do whatever procedure is required to get the mount going to its gotos. How exactly do you line up the alignment stars, though? You can choose one of two methods. You can either remove the camera from the telescope and temporarily replace it with diagonal and eyepiece, or you can use the camera to do the alignment. In the beginning, it may be easier to remove the camera and do the goto alignment with an eyepiece.

If you do use the camera, you’ll, of course, need to turn it on (and the laptop and its software if you are tethering to a PC), and center the alignment stars on the camera’s display or the laptop screen. Since alignment stars are bright, one will allow you to get rough focus, too.

If you choose to use Celestron’s AllStar Polar Alignment, which is built into the hand control firmware, take care of that once goto alignment is done—ASPA requires the goto alignment be accomplished first. If you do a declination drift polar alignment (horrors), now is also the time to do that, since having the telescope tracking during the procedure makes things much easier and is practically required.

Step Five:  Focus

The Double Cluster is an easy and pretty target...

If you goto aligned using the DSLR, you’ve got focus roughed in, and can now do a fine focus procedure. If you used an eyepiece instead of the camera, however, get rough focus with the imaging camera at this time. If the last alignment star was a good, bright one, stay on it and use it to focus.

To get in the focus ballpark, adjust the focuser on the telescope (I am a big fan of remote moto-focus for imaging) until the bright star is as small as you can make it and dimmer field stars begin to appear and sharpen. Exposure? I like one to two seconds; that allows me to see results quickly after tweaking focus. Set camera ISO as high as needed to get a good image of the stars. If you are way out of focus, you may need to max ISO out and increase exposure time till you detect the big round globe of a star (in a refractor). Once it is closer to focus, back off on ISO and exposure for a less overexposed star image.

When the field stars are as small as possible by eye, tweak focus with a fine-focus method of choice, which may be a Bahtinov focus mask, or a focusing routine built into imaging software (like Nebulosity) if you are tethering the camera.

Step Six:  Acquire Target and Compose Shot

Rough focus done, I interface my planetarium program (Stellarium these says) to the GEM. I start Stellarium (or whatever), and connect it to the telescope mount, that is. How? I invariably use the ASCOM telescope driver system, even if the program on the laptop, like Stellarium, has built-in telescope drivers. Why? Because ASCOM includes a little onscreen telescope HC that allows me to move the mount (at different rates) from the computer. That means I don’t have to get up and walk out to the scope and HC, press a direction button to center the object, walk back to the computer, etc.

A Bahtinov Mask makes fine focusing easy...

Alright, time to get on our first subject. What should that be? Even if you are at least somewhat experienced in deep sky imaging, begin with something easy with this new rig. This time of year, perhaps a nice winter open cluster over in the west like M35 or M37 or the Double Cluster. One important consideration given the economical mounts we’re using? Stay away from the Meridian. These GEMs just don’t track well in that area. Don’t image an object that will come within 10-degrees of the Meridian before your sequence is done, and don’t begin imaging an object until it is at least 10-degrees past the Meridian.

Once the scope goto stops, take an exposure long enough to reveal your target to see how the composition of the shot looks. If the subject is not centered, or just not framed the way I want it, I use the ASCOM HC to fix that. I keep the exposures short enough to make framing easy, maybe referencing a bright star in the frame if the object doesn’t quite show up in 1 – 3-second shots.

What if the target object is not in the field of the camera at all when the goto slew is done? That’s not much of a concern these days for most mounts, but if you have a problem, a quick solution is to slew to a nearby bright star, center it with the aid of your finder and “sync” on it. You should then be able to slew back to the target and have it in the frame. Oh, before you do that, be sure it really isn’t in the frame. If the target is a dimmer one, increase exposure and ISO and see if it appears.

When the subject is properly centered fire up PHD2 and get auto-guiding going. The main gotchas there? Make sure the guide scope is well-focused and that the guide star you’ve chosen is neither too dim nor too bright (saturated). When you put the cursor on a star, PHD2 will tell you all about that. Some imagers believe the guide star should be slightly out of focus for best guiding, but I’ve found I get better results from sharp stars.

When the mount is guiding, I go back to the imaging camera and do a test exposure. How long should that test subframe be? That depends on the sky and the subject. If I’m in the backyard, going much beyond a minute causes the background sky to brighten up so much that processing can be difficult later. At my dark site, I’ll expose for 2 – 5-minutes. Exposure also depends on the subject. An open cluster like M37 will be just fine in 30-second – 1-minute subs. The Horsehead Nebula will not be.

One important thing to remember is that while you’ll be stacking many shorter sub-frames into a finished exposure, you still have to have each individual exposure long enough to pick up all the detail you need. Stacking subframes will make the final result smoother and less noisy, but will not show any detail not present in the individual subframes. Longer subframes are always better.

Get your guiding going on...

What should camera ISO (nee “ASA”) be set to? Normally, use as low a value as you can to capture the detail you want while keeping noise down. The higher the ISO, the noisier the image will be (and the brighter a light polluted background). I rarely go above ISO 1600, and try not to exceed 800 in the backyard if possible. Naturally, ISO and exposure time interact. In general, I’ve found it better to go with a lower ISO and a longer exposure when possible.

Also examine the test exposure for signs of star trailing. Assuming PHD is not going wacky on you, you’ve got a good polar alignment, and the seeing is OK, that should not be a problem at the 400 – 600mm focal lengths we’re using. If the stars don’t look right, go back to PHD and make sure it’s still guiding well (bring up the graph as shown in the image here). If it isn’t, you’ll have to troubleshoot.

If the stars are eggs or worse, first make sure the values you’ve entered for the guiding parameters are close to those we outlined here. One variable that can change from night to night is the guide camera’s exposure. While mounts in this class tend to do best with 1 – 1.5-second guide exposures, if seeing is not good a somewhat longer one can improve guiding. There’ll be less tendency for PHD to try to guide on movement caused by seeing.

Step Seven:  Expose

Time to do what we came for, take an exposure sequence. Set the laptop program or the intervalometer to take a number of subframes at the exposure value you determined was best. How many? As many as possible. Even on an easy object like M37, more subs always make for a better looking finished picture. I generally aim for 20 - 30.

Before beginning the sequence, though, let’s put the dark frame question to bed. Since a DSLR’s sensor chip is not cooled, dark frames are mandatory to eliminate the false stars of thermal noise. There are two ways to subtract dark frames from subs, manually or automatically.

If you go manual, finish the imaging run and then, just before packing up for the night, cover the telescope objective and shoot subframes equal in number to the maximum number of lights you’ve taken in a sequence. For example, if you did one 20 and one 30-subframe sequence, take 30-darks. If you used different exposures on different sequences, you’ll have to do more than one sequence of darks—dark frame exposure values need to be the same as their corresponding lights. The darks will be subtracted from the light frames during image processing. This way of working is certainly acceptable, but, in addition to being labor intensive, it isn’t as effective as it could be in my opinion.

Me, I go automatic on dark frames for a couple of reasons. Not only do I not have to worry about messing with darks at the end of the evening or during processing, I think automatic darks are more effective.

How do you do auto dark frames? DSLRs allow you to select a mode called “long exposure noise reduction” (or a similarly named menu item). Engage that, and the camera will take a dark frame after each exposure and automatically subtract it. Yes, that means an imaging run will take twice as long as it otherwise would—30-minutes of subs will take an hour to complete—but I think the results are just better.

Expose!

Why would the results be improved by taking a dark after each light? Because the temperature of the DSLR’s sensor chip will vary throughout the night. Ambient temperature will drop throughout the evening, and, as an exposure sequence goes on, the internal temperature of the camera due to its electronics will tend to rise. To be most effective, a dark should be taken as soon after its matching light subframe as possible, so the temperature it was exposed at is close to that of the light frame.

OK, set that computer or intervalometer for the number of exposures at the required exposure value, push the “go” button and…and…wander around and do something else while the exposure sequence completes. I usually just go inside and watch TV. If I’m at the dark site, I’ll cadge looks through my buddies’ telescopes. I’ll come back periodically and see how things are going—especially how PHD is guiding—but I rarely encounter problems unless clouds have moved in and my guide star has been lost, temporarily or permanently.

Once the sequence is finished, it’s time to go on to the next target. How many targets should you do? That’s for you to decide, but I tend to believe fewer targets, maybe just one or two per evening, and more subframes (and maybe longer exposures) is the way to go.

Done, I’ll pack everything up and head for home if I’m at the dark site, or, if I’m in my secure backyard, I’ll just cover the refractor and GEM with my Telegizmos cover and only take the computer inside—which is a much more pleasant way to end an evening under the stars than having to disassemble everything and carry it back into the house when I’m tired.

And next? Next is processing the images, but that is a story for some other Sunday.