Sunday, September 07, 2008
The Guiding Game
How can I adequately describe the pain we used to go through in order to get a fuzzy Kodacolor drugstore print showing a smudge that might look a little like M13? It went like this: you used a smaller telescope, a “guidescope,” piggybacked on your main instrument or you used an off axis guider (don’t ask) to monitor a “guidestar” during exposures. Why did we do that? Then or now, you unfortunately can’t just open the shutter of a camera (be that a CCD cam or a film cam) and do long exposure deep sky imaging. Most telescope drives are not blessed with gears good enough to allow that. Normally, you must guide, you must adjust the aim of the scope by minute amounts throughout the course of the exposure. If you don’t? You get nasty, trailed stars and a picture that goes into the trash can—real or virtual.
Back in the bad ol’ days, what this meant was that you kept a star precisely centered in the crosshairs of the guidescope’s high power reticle eyepiece for the duration of the exposure. If the star drifted the tiniest amount, you pushed a drive corrector of hand paddle button to re-center it. Back in The Day, most drives weren’t so hot—30 arc seconds of periodic error was common and good—so the star did a lot of drifting. But even with a good drive, you didn’t dare take your eye from that eyepiece. Miss one bump and you’d have wasted your time. You sat there and watched that consarned star for 15 minutes…30 minutes…an hour—as long as you could go. No wonder astrophotographers were looked on as the madmen of the amateur astronomy ranks; we had literally been driven nuts keeping that star centered.
Things are much different today, thankfully. You no longer have to watch that dadgummed star—a CCD camera does it for you. Once you get things set up and running, you can do what we all dreamed of doing way back when: open the shutter and walk away to look at the sky with your binoculars, eat a snack, or have a drink of—err…“sarsaparilla.”
Not only do you not have to keep your eye glued to the guiding eyepiece, the guiding quality delivered by one of these “autoguider” cams is far superior to what you can do by eye—better than what I could do, anyway. Even a minimalist setup is capable of guiding with better than 2 pixel accuracy, and many folks do considerably better than that. I find that at the relatively short focal lengths I image at, I can be pretty cavalier about setting up my guide system and--and this is a big deal for me--polar aligning. In most cases my guide cam can keep up with the errors—declination drift—introduced by my approximate polar alignments. I take relatively short exposures, and the computer stacks them later, making up for drift from frame to frame. That means I don’t have to worry about no drift aligning, which is not one of my favorite activities.
Is there no catch? Is there finally a free lunch? Is it really this simple? Just slap a guide camera and scope on your fave imaging telescope and have at it? Yes and no. Once you get into the autoguiding groove, it really is that simple. The devil comes in learning to autoguide, which consists of assembling workable equipment and setting up the software that does the work. Before you can wander around the field annoying your buddies while your CCD cam and autoguider do their thang, you must choose a camera and guidescope, choose an autoguiding program, choose a method of mounting guidescope and guide camera, choose a guide camera, and, hardest of all, get everything working together.
Almost any CCD camera can be pressed into service as a guide cam. There are only two requirements: that it be compatible with whichever guide program you plan to use, and, most importantly, that it is sensitive. A camera that is not very sensitive will soon incite you to hair-pulling as star field after star field lacks a suitably bright star for guiding (that is, bright, but not over-saturated). People have autoguided with unmodified webcams and things like the Meade LPI, but that is hardly an optimum solution, and I recommend it only if you absolutely have no choice. Even then, I suggest you look into modifying a webcam for long exposure. If you have the electronic skills to do the mod, or have a buddy who does, a webcam can work fairly well if it can be made to do 2 – 3-second exposures. Celestron’s inexpensive NexImage webcam is particularly easy to modify.Webcams do tend to be noisier than “real” CCD cameras, but that usually doesn’t matter much for guiding.
You can spend a wee bit more? One possibility is a used CCD camera of just about any kind. A tired old Starlight Xpress MX5 or SBIG ST-237 will work superbly (just make sure you can find a guide software package that supports the old clunker). A particularly nice choice is Meade’s original color DSI, which is what I use. Not only is it cheap—less than 200 bucks on Astromart, usually—it is quite sensitive. I don’t think I’ve ever been unable to find a suitable guidestar in any field, even though I normally restrict myself to guide exposures of 2 seconds or less.
Yes, you can also buy dedicated guide cameras, CCD (or CMOS) cams sold expressly for this use. Two standout examples are the Orion StarShoot Autoguider, and Fishcamp Engineering’s Starfish camera. Both sport one very nice feature that many general-use CCD cameras lack, an ST4 autoguide-output (more on that shortly). Downchecks? The Orion uses a CMOS sensor and is thus a little noisier and a little less sensitive than it could be. The Starfish is also a CMOS device and is also kinda pricey, $695.00 for the uncooled version, more than your stingy ol’ Uncle would like to pay for a guide-only cam. Is cooling necessary, by the way? Not really. Yes, cooled cams do mean lower noise—you are less likely to try to guide on a hot pixel than you are with an uncooled rig like the StarShoot. In truth, though, at the short guide exposures most of us use, cooling is really not necessary.
Anyhoo, let’s assume you’ve got yourself a guide camera of some kind and are ready to proceed. Next fork in the road is guidescope choice. Step one? If you’re an old-time film astrophotographer and manual guiding casualty like Unk, forget everything you were taught about guidescopes. Which was mainly that the guidescope should be as close in focal length as possible to the main scope. Longer was even better. That is simply not necessary in this day of computers and silicon chips. The average guide program does not care pea-turkey about the focal length of your guidescope, and shorter guidescope focal length has one huge advantage: a wider field and a larger number of guidestars to choose among.
What will work? Almost any small refractor is a good choice. I have successfully used a William Optics 66SD for lots of guiding, but today I am back to Old Reliable, my venerable 9 year-old Synta-Celestron Short Tube 80. This is, in fact, the guidescope I recommend for most ever’body. It features decent aperture, a short focal length, and a good build quality. One caveat: there have been quite a few ST80 clones sold over the last decade (like on Ebay). The original ST80’s focuser is a sturdy all-metal affair that can be locked down solid. Some of the clones have wobbly plastic focusers that will flex during guiding, ensuring your final image sports less than round stars no matter how well guide cam and PC did their job.
Next hurdle? How do you mount the guidescope? What I did for quite a while (after I stopped using that devilish off-axis guider) was piggyback a guidescope on the main scope, mount it on top of my SCT with a pair of adjustable rings. That can work well. Do be sure to get good, sturdy guidescope rings and a hefty mounting bracket. Any differential flexure between scope and guidescope brought on by a less than secure guidescope mounting arrangement, will, again, result in them derned trailed stars. If I wanted to do it right, I would probably look at Losmandy’s or ADM’s piggyback setups (to name just two outfits producing good stuff). After many years of experimenting, however, I’ve some to the conclusion that a side-by-side arrangement as seen in the image above is less prone to flexing and easier to work with. I have produced many, many well-guided images using piggyback setups, though; quite a few with the inexpensive but effective guidescope mountings sold by Scopestuff. I advise against scrimping in this area, however.
Camera? Check! Guidescope? Check! Guidescope mounting? Check! What else? Unless you are using SBIG’s ST4 or STV cameras, which feature built-in computers, you will need a laptop running a guiding software program. It is this program that is the brains of the outfit. It monitors the position of the guidestar on chip and issues guide commands to move the telescope when and as much as required. If you are as cheap as me, you will be pleased to learn you may not have to pay for a program. If you’ve purchased an SBIG camera, the soft that came with it, CCDOPS, will do a good job of guiding. If you’re using any of the top-of-the-line CCD camera control programs like CCDSoft or Maxim DL, these include guiding features that are more than sufficient. What if you are using a webcam? Gravitate to Astrogeeks’ freeware program MetaGuide. What do many, many of us use with great success, however? Another freebie, this one from Stark Labs (maker of the Nebulosity camera control and image processing program), PHD Guiding.
What makes PHD so good? Better than the guiding routines in many expensive CCD camera programs in many folks’ opinions? It’s a no-fuss, no-muss kinda thing. Most guide software requires you to fiddle with quite a few parameters before the software will guide right. PHD, in contrast, usually only requires you to pick a guidestar and push the “go” button. That’s why it’s called “PHD.” In this context, that does not stand for “Doctor of Philosophy” (though the author of the program has one), but for “push-here-dummy.” PHD does allow you to tinker with settings just like the other programs do, but that is usually not necessary. While I have had very good luck with CCDOPS and CCDsoft for autoguiding, these days PHD is about all I use.
What’s next on our autoguiding agenda? Now that you have all this stuff, you gotta hook it all together. Getting the image output of the camera to the PC is way easy; you just connect the camera’s USB output to a USB jack on the laptop (older cameras may have serial or even parallel format connections). Please be aware that some cams, like the Meade DSI, are powered from the USB bus and are also very power hungry. I find I need to connect my DSI to a powered USB hub, not just plug it into the laptop. If I do that, it completely overwhelms the little DC – DC powersupply I use to run the computer from 12vdc in the field. If at all possible, look for a camera that features USB 2 output rather than USB 1—if the camera image format is much above 640 x 480 in size, image download-time may play hob with your guiding over a USB 1 connection.
That’s the connection from camera to computer, but you will also need a path from camera (or laptop) to the mount for guide commands. There are two ways to go about this. The most familiar to many of the older CCD hands is “ST4.” This uses the autoguide input many mounts feature. The format is relay/switch closures (not serial data). This guiding system is simple and reliable. When the camera wants to move the mount east, it closes the east switch, just as if you’d pushed a button on the HC. Ointment flies? Only that some cameras, like the DSI, do not have autoguide outputs. In that case you’ll need Shoestring Astronomy’s GPUSB or GPINT-PT adapters. They provide an ST4 output from your PC, and plug into the laptop’s USB port or a parallel port, respectively. Why is this type of guiding called “ST4,” anyhow? Santa Barbara Instrument Group developed this guide system for use with their ST4 camera.
You can use ST4 guiding only if your mount has an autoguide input port. What if it doesn’t? The popular Meade LX90 and LXD75, for example, do not. That’s harder. Unless the mount manufacturer or somebody else makes some kind of conversion kit or add-on you can forget ST4 guiding. Shoestring does make an ST4 kit for generic (non-go-to) CG5 GEM-style dual axis drives, and Meade used to make the #909 Accessory Module, which could add an ST4 port to the LX90 and LXD75 (apparently the 909 has been discontinued). That does not mean you have to forget autoguiding if there’s no way to add an ST4 input to your mount, though. You can guide through the mount’s serial port instead. In this setup, the computer issues guide commands over the serial line connected to your hand control (usually). This is just like when you are using the laptop to send the mount on go-tos, only instead of telling the mount “go-to M13,” the computer says “move east at guide speed.” What if your PC does not have a serial port (most new laptops do not)? One of the ubiquitous USB – serial converter cables you find in Walmart or BestBuy will work fine for guiding.
Some folks wonder whether it’s OK to use serial guiding. Isn’t ST4 guiding more accurate? Not necessarily. Some mounts and software will be able to use a serial guiding method called “pulse guiding.” This allows the software to not only move the scope in the appropriate direction, but also to change guide speed by small amounts, making for very accurate guiding, easily the equal of ST4. In fact, even non-pulse serial guiding works very well in most cases. Do be aware that if you need/want to use serial guiding, you’ll probably need another program in addition to your guiding software—ASCOM (free). Generally, most guide-software makers do not include built- in drivers that allow your PC to talk serial to your particular mount; they use ASCOM instead.
One other very important thing? If you are guiding through ASCOM, remember, you do NOT plug the serial cable into the Autoguide port. Lots of folks want to do that—where else would you hook-up to the mount for guiding? A serial cable from the computer ONLY goes to a serial port on the mount. Only if you are NOT using ASCOM, only if you have a Shoestring Adapter or a camera with a built-in guide output, do you plug a cable (an ST4 format cable; Shoestring sells ‘em) into the autoguide port on the mount.
Software Set Up
With all the cables connected, only one (important) preliminary step remains, setting up the software. The instructions that follow are aimed mostly at PHD users, but other software is remarkably similar. Software configuration is relatively simple and mainly involves telling the program how you are going to guide and (if you are using serial guiding) what kind of scope/mount you are using. In PHD, you set up for serial guiding by clicking “ASCOM” in the “Mount” menu at the top of the screen. That done, click on the little telescope icon at the bottom of the screen to bring up ASCOM’s familiar telescope chooser. Select your scope type and fill in its details with the “properties” button.
What about them other choices on the Mount menu? GPUSB and GPINT? They allow you to guide through one of the aforementioned Shoestring guiding adapters if your camera does not have an ST4 output. Note that there are multiple selections for the parallel port (GPINT) adapter: GPINT-378, for example. The numbers refer to the address of your particular parallel port. This can be determined using the Device Manager in Windows. Take a look at the Properties for your parallel port and examine its “Resources.” If that sounds confusing, Shoestring Astronomy has a little (free) utility that will find out what your number is. The last choice on the menu, “On Camera,” is, natch, for you lucky folks who have an ST4 output built into your camera. If you are using any of these guide methods you do not and cannot choose your scope type. That info is not required.
Once all them cotton-pickin’ cables and menu selections is sorted, the next step is getting the guidescope focused. If you have not already done so, click the Camera icon in PHD and choose your camera. Pick an exposure too. How long? To begin with, fairly long, at least several seconds depending on the sensitivity of your camera. It is likely the guidescope will be way out of focus. To begin, click the “Loop” icon at the bottom of the PHD window, and watch the display. What do you focus on? If you’re using a go-to scope, the last alignment star will work just fine. Focus until this bright star is as small as possible on the PHD video display. When it is, observe other stars in the frame, increasing exposure a bit if necessary. Make the dimmer ones as small as you can as well. With PHD and most other guide software, it is not necessary to achieve perfect focus; just make the stars as small as you can by eye. When you’re done focusin’ press the Stop icon. Send your go-to scope to the night’s first target, or push a non-go-to one there, and get your imaging camera arranged.
When you are ready to begin guiding, hit the Loop button again. Exposure should be no longer than two seconds if at all possible. If your mount is not precisely polar aligned, and/or has significant Periodic Error, keeping the guide exposures, and thus the guide updates to the mount, short will help keep you on track. Now, examine the display for a reasonably bright star. If one is not found, you may have to adjust the aim of the guidescope (or the whole rig if need be). Only increase exposure if you can’t find a decent guidestar without moving the imaging camera off target. Remember, this star does not need to look like Sirius; it just needs to be obvious to your eye that it is a star. One caveat: make sure this star is not close to the edge of the frame.
When you have a good guidestar candidate, press Stop and then click on the star. When you are ready to begin guiding, press the archery-target icon on PHD. The program, like all other guide software, will then do a calibration; it will move the mount in all four directions (assuming you have declination guiding enabled) to see how the star moves on the guide camera chip. That is why you must not choose a guidestar at the edge of the frame—if calibration moves slide it out of the frame, calibration will, naturally, fail. Don’t be impatient; calibration can take a goodly while with PHD. If all has gone well, calibration will finish without any kind of error, and PHD will begin guiding without any further input from you. There will be a small box around the star, and the program will keep the guidestar in that box (you’ll see changin’ numbers at the bottom of the screen indicating guide corrections). If it looks as if ever’thing is as should be, take an exposure with the imaging camera of the duration you intend to use for subframes. Examine the resulting image carefully. Stars (those at the frame center, at least) nice and round? Success!
What then? Shoot your target. When you are done, press Stop on PHD before moving to the next target, otherwise PHD will make unhappy sounds because its guidestar has been lost. The same thing will happen if the wind blows the scope off target or your buddy, Goober, comes up and taps on the tube: “Wut’s this here camera, Unk Rod?” Chances are, though, that all will go well. Autoguiding is a fairly complex process, but PHD is about as bullet-proof a piece of software as I have ever seen. What if all your stars look like eggs, though? Here are a few tips to get you over the rough spots:
- Balance, balance, balance! This is probably the number one cause of poor autoguiding. Always balance the scope so that it is slightly heavy to the east, so the drive gears pull the scope along. That ensures the main drive gears stay closely engaged and lessens the magnitude of hiccups.
- Start small. When you are first learning to guide, begin with a 1000mm focal length imaging scope or less. You will cry far fewer salty tears than you will if you start out with a C14 at f/11!
- If your stars don’t look great, calibrate. A good calibration is the path to good performance, and one calibration will likely not last all night long. Some programs, like CCDOPS, insist you either enter the declination of each new guidestar or perform a new calibration (I generally have best results by doing a new calibration). PHD does not ask for this, but it does want a new calibration if you move from one side of the Local Meridian to the other.
- If nothing else works, you may have to adjust “advanced settings.” In PHD, clicking the Brain icon brings up a list of cryptic stuff like “aggressiveness,” “RA hysteris,” “Dec algorithm,” and “Calibration step.” In my experience, the only one you may need to fool with in PHD is “Calibration step.” If your candidate guidestar is far from the celestial equator, you may need to increase this value.
- I’m frequently asked whether it is a good idea to do a PEC, “Periodic Error Correction,” recording on the mount (if it has this feature) before attempting to autoguide? The answer is probably not. In some cases a PEC recording and an autoguider actually seem to work against each other. Most guide setups can take care of a mount’s periodic error without any additional help anyway. I will say I have seen a few telescopes—mainly large aperture fork mount SCTs—that simply could not be made to autoguide without doing a PEC recording first to bring down egregiously large RA excursions.
Does this sound like a lot of work? It is, I reckon, but once you get the hang of it it’s easy and adds just a few more steps to your imaging routine. It is gull-derned worth it, too, as you’ll agree when you see all them nicely guided subframes begin to roll out of your imaging camera. Is it always as simple as Push Here Dummy? Usually it is, whether you are using PHD or some other program. Naturally, in this here blog I could only draw the vaguest of outlines of the difficulties you can potentially face; if you do run into problems, as always, thank god for the Internet. There’s a Yahoogroup devoted solely to autoguiding issues, Autoguiding. All the guide software I am aware of has good online support. Go here for the group devoted to Stark Labs software (PHD).
I just wish I could get out and give my autoguiding setup a workout. We got through Gustav with only copious rain (9-inches on Selma street…or was that the River Selma?), but we are now cautiously watching bad, ol’ Hurricane Ike. I keep telling myself this too shall pass, that clear skies and calm are coming. Am I a cockeyed optimist? Stay tuned.
Awesome Tutorial Uncle Rod. I am about to start guiding for the first time. Your fantastic tutorial is very much appreciated. Thanks, Philip A Cruden Billion Planets Quest.Post a Comment