Sunday, December 18, 2011
You Gotta Guide
Yep, if you want to take astrophotos, long exposure deep sky astrophotos, you have to keep a “guide” star centered for the duration of the shot to make up for telescope misalignments and mechanical gear errors in your mount. Well, that’s not strictly true, muchachos. If you have an uber expensive, permanently mounted setup, or are willing to settle for multiple 30-second – 1-minute subframe exposures you stack into a final image, you can get by without guiding.
But that is not most of us. Most of us are using mid-level mounts, in the EQ-6 to G11 range. And most of us are of the opinion that our results will be better, our finished pictures more noise-free, if we expose each subframe longer. We may still stack multiple exposures into a finished image, but the longer we can go on each subframe, the better.
If you’ve been in the astrophotography game longer than about fifteen years, I don’t have to remind you what a pain guiding used to be. Back in the not-so-good old days of film astrophotography, we guided manually. What that meant was you stared at a dim star in a crosshair reticle eyepiece for long periods, often through a devilish little device called an “off axis guider,” an “OAG”
An off-axis guider was attached to the focuser or visual back ahead of your camera and had a small prism or mirror that intercepted light from the edge of the field and diverted it to the guiding eyepiece. The prism was small enough and close enough to the field-edge that its presence wasn’t noticeable in finished images, even in a 35mm frame. Course, being small and being at the field edge meant it was limited in the number of stars it could see. With an SCT, stars at the hairy edge of the field were misshapen too—often looking more like commas or seagulls than pinpoints. Locating a star bright enough and round enough to use for guiding was a challenge—to put it mildly.
Why didn’t we use guide scopes, separate small refractors (usually), to guide? Some of us tried, but over our long film exposures the chance of flexure was high with our long focal length guide scopes (their magnification had to be high to allow you to see any small “excursions” the star would make). If the guide scope flexed, moved independently of the imaging scope the tiniest bit, stars would be trailed in the final image no matter how carefully you guided. Then as now, no astrophotographer liked non-round stars, so most of us resigned ourselves to OAG hell.
Once there was a guide star sitting in the crosshairs of the guiding eyepiece, the fun had just begun. You stared at that star for the length of the exposure. If it moved off the crosshairs even a little, you pushed the appropriate button on the hand paddle, the hand controller, to bring it back.
The coming of the CCD camera changed everything. The most important thing it did was shorten the length of most folks’ exposures. With sensitive CCD cameras, or even today’s ubiquitous DSLRs, many imagers, including Unk, start hitting the skyfog limit, the point where sky brightness due to light pollution becomes a problem, in two or three minutes. Three minutes is a lot easier to deal with for manual guiding than thirty, but by the time the CCD revolution was a few years old, we found we didn’t have to—guide manually, that is.
The autoguider revolution started with something called the “ST-4,” the first camera produced by the legendary SBIG, Santa Barbara Instrument Group. While the ST-4 could take pictures, its small chip made it mostly of interest as an auto guider for folks doing film imaging. We were skeptical that this (expensive) little widget connected to a box with a few LEDs on it could guide a telescope accurately, but it could—in the right hands. The settings, guide star, and mount all had to be just right, but the ST-4 most assuredly could work.
As the 90s came to an end, the ST-4’s guiding system, which issued relay-closure signals to the mount to guide, just like you were pushing the hand paddle buttons yourself, became the default standard. Mount makers left and right added “ST-4 compatible” autoguide ports to their mounts, and many CCD camera makers furnished ST-4 guide outputs on their cameras.
The ST-4 was there first, but almost any CCD camera could be used to guide a mount, a modern go-to mount, at least. You just needed the right software running on a PC. For a while, the standard was SBIG’s CCDops, the simple program that came with Santa Barbara’s cameras, and which most of us learned on. Then came CCDsoft from Software Bisque, which was definitely a step up, at least as far as features. Finally, there was Maxim DL, which was and is for some folks the king of guiding (and imaging and processing).
For those of us not stationed under desert skies with plenty of time to pursue pretty pictures, and those of us not after serious scientific results, Maxim DL was overkill, however, at 600 bucks. What if we just wanted to guide our mounts? Didn’t need lots of frills, and didn’t want to spend that much money?
A cool little (freeware) program called GuideDog was popular for a while. Not only was it free, it worked. Alas, it had some rough edges and was designed to use modified webcams for guiding rather than standard CCD cameras or guide cameras. Its author stopped developing it by about 2004, too. I kept looking, and eventually heard about PHD, “Push Here Dummy,” Guiding from Stark Labs. Amazingly, that was just what PHD Guiding was. You pushed a button (well, actually a couple of buttons), and it calibrated itself and just, well, GUIDED.
PHD did not control your imaging camera. It did not process images. All it did was autoguide, either through an ST-4 port or a serial (RS-232) port on the mount. It supported a lot of guidecams and CCDs, and was being continuously updated and improved by Mr. Stark Labs, Craig Stark. I used PHD Guiding, a William Optics 66 SD refractor as a guide scope, and the original Meade color DSI CCD cam as a guide cam successfully for several years. But I wasn’t overly happy with the setup aside from PHD.
One thing I didn’t like about it was my guidecam. The color DSI was more than sensitive enough to do the job, but it had a failing: it did not have an ST-4 output. Which meant I had to add an adapter cable, a Shoestring Astronomy adapter cable that converted parallel data from my laptop to the ST-4 switch closures my Atlas mount’s autoguider port understood. Oh, I could have guided from a serial port on the computer to the serial port on the mount hand control, but that still meant another snarfin’ cable plugged into the computer to go with the the DSLR USB cable, the guidecam USB cable, and the DSLR remote shutter control cable. I was kinda tired of the DSI, too; its chip was small. It would have been nice to have a bigger one that yielded a wider field and more guide stars.
Nor was I happy with my guide scope. The WO SD was a fantastic small refractor, but it was also a little on the heavy side, actually heavier than a short tube 80. My rings were kinda clunky, too. They worked, hell yeah, but they were a little light, and I was pretty sure I was throwing away more frames due to flexure than I ought to be. It was also a bit of a hassle to install the rings and 66 SD for an imaging run.
I could have invested in a Losmandy or ADM guide scope rings setup to banish the flexure problem, but you all know how cheap I am, and it would still be a pain to mount them and the 66 on the C8 before every imaging run. Why not go back to an OAG, then? A guide camera will work with one. Uh-uh. Nosir buddy. I had my fill of them tools of Satan back in the 90s.
I was stumped for a while, but then I heard salvation was at hand. Folks over on one of the Cloudy Nights discussion boards were talking about, a Canadian company, KW Telescopes in Ontario, who were selling the combination of a QHY5 autoguide camera (with an ST-4 output), and a guide scope made from a modified 50mm finder. According to the cats and kittens on CN, this system, the QWIQ Guide, was simple and worked well.
I was skeptical. How could you guide pea turkey with that little focal length and aperture? Yeah, I knew guide cameras do not need the obscenely long focal length guide scopes we needed back when we guided by eye, but, still, 50mm? Nevertheless, imagers were reporting the 50mm jobs worked like duck soup with imaging scopes of up to 2000mm focal length. I almost pulled the trigger on the QWIQ, but hesitated. Seemed too good to be true, even though I knew SBIG had produced some workable guide scopes for their cameras with objectives e’en smaller than 50mm.
While I was sitting on the fence about the QWIQ system, I did solve the guide camera conundrum. Orion (Telescope and Binocular Center) had begun selling a nice little guide cam, the StarShoot Autoguider, which was very much like KW’s QHY. For $279.00. Nice big chip, excellent build quality, ST-4 output, and proven to work with my beloved PHD. Orion even included a copy of PHD Guiding in the package.
I was lucky enough to receive a StarShoot for Christmas 2009 from the wonderful Miss Dorothy. I got the cam out once shortly after I unwrapped it, and it appeared to work well, but… Late 2009 was when The Herschel Project got underway and I put my DSLR away for a while. With nearly 3,000 dim Herschel Objects awaiting me, I hardly used anything other than my deep sky video camera for dang near two years. With the H Project finally under control as of this past fall, though, astrophotography reared its ugly head again as it always does. Yes, I’d get the Canon Rebel out of mothballs, but I wanted to fix my guide scope problem first.
There was that QWIQ finder-guide scope. If it worked, it would be just the thing. A 50mm finder is a light little thing, and even the somewhat rudimentary Synta finder mount would likely be less prone to flexing than my rings/66mm scope. In a bit of synchronicity, just as I was considering giving KW a call, Orion came out with their version of the QWIQ setup, the “Mini AutoGuider” package. You could get the whole shebang, including the modified finder-guide scope and the StarShoot camera for a nice price, $349.00, considerably less than what KW wanted given the exchange rate at the time. Since I already had the guidecam, it was even sweeter, a mere $89.95 for the Orion Mini 50mm Guide Scope.
Since this set up would be destined for my C8, Celeste, I would have to provide an SCT compatible mounting foot for the finder-cum-guide scope, one which could be had for the not too unreasonable sum of $16.95 (Orion provides a mounting foot with the guide scope, but not with the spacing for an SCT’s accessory holes). That would mean sacrificing Celeste’s original finder, or at least its mounting. But that was OK. Her finder was a well-made Japanese job, but I’d never liked its ring mount. Only its forward ring had adjustment screws; the rear ring had a rubber O ring to hold the finder in place while allowing adjustment. I don’t know if that O ring was just getting old and drying out or what, but the finder would not maintain its alignment for long.
I wasn’t too sorry to see the Celestron finder itself go either. Its polar alignment reticle is not very useful anymore, I rarely use its illuminator, and I like my Orion (Synta) RACI correct image 90 degree finder better. The RACI would fit the mounting foot and would be real sweet on Celeste when I wasn’t using the guider.
In due time the Mini AutoGuider scope arrived, and I was somewhat impressed. It appeared to have been made, or at least professionally modified, for its intended purpose, and included a straight-through 1.25-inch eyepiece holder with three set-screws, a rear dust cap, and the typical “screw-unscrew the finder objective” focus setup. Only minor complaint? Like most 50mm finders, the dewshield/objective end of the tube didn’t extend far enough forward to do much good for either dew or glare.
Getting the original Celestron rings off, the new SCT finder mount on, and the mini-guider installed on Celeste was the work of maybe 15-minutes. It looked super, but how would it act? To find out, I’d need to head to the PSAS dark site in Tanner-Williams. I gave you a report on that run here, but I’d like to go into a little more detail about the guiding end of things.
Set up was easy enough. Mounted the Mini Guider on Celeste, wrapped a 2-inch dew heater strip around its barrel just behind the objective, and inserted the StarShoot camera into the 1.25-inch port. The Mini AutoGuider came with a little parfocal ring to go over the camera’s 1.25-inch nosepiece to allow you to preserve rough focus once you find it. I slid that over the nosepiece. Cabled the StarShoot up, with the guide output on the camera going to the ST-4 autoguide input on the mount, and the USB from the StarShoot to my little netbook, which was loaded up with the latest release of PHD.
First order of bidness was getting the guide camera focused. Brought up PHD and pushed the button that starts the program “looping” images from the guide camera. The Mini AutoGuider’s instructions tell you to focus roughly by loosening the set screws that hold the camera in the guide scope and sliding the camera in and out. I found that unnecessary. Vega was close enough to focus to allow me to get it sharp by using the fine focus method, screwing and unscrewing the objective.
With PHD’s exposure set to 2-seconds, I observed Vega’s image and started unscrewing the objective…close…close…alright! As I neared focus, I was gratified to see dimmer field stars popping into view. I had been a little concerned about the StarShoot’s CMOS imaging chip being sensitive enough to reveal many stars, especially with a dagnabbed 50mm scope. That, it appeared, would not be a problem. Satisfied with focus, I snugged up the finder’s knurled focus lock ring against the objective. I tightened the parfocal ring up against the camera with its (tiny) Allen screws, and focusing was done.
If I haven’t mentioned it already, the instructions that come with Orion’s Mini AutoGuider set up are sufficient. Barely. They do have a serious omission. Nowhere do they tell you you will need to adjust a very important setting in PHD.
The way PHD works is this: you begin looping exposures, and, when you can see a suitable guide star on its display, you press the Stop-sign icon and hit the Guide icon. Before the program can begin guiding on the first star of the evening, though, it must calibrate. It must move the telescope north, south, east, and west to see how the mount reacts. There are several settings under the “brain” icon that concern calibration. The one we are interested in when using the mini-guider is Calibration Step.
This setting determines how long the program will pulse, will move the mount for each calibration step. Normally, with a typical guide scope setup, the default, 750 milliseconds, is fine. With a very short focal length scope like the Mini Guider, it needs to be changed. Leave it as short as it is, and PHD will need to make many steps during the calibration before it sees sufficient movement to complete the process.
Luckily I had read about this need to change Calibration Step on one of the Cloudy Nights boards. PHD might calibrate correctly with the default settings, but it would likely take a long time to do so. By changing the setting to 2000ms, increasing the step size, the calibration completes in a reasonable length of time. The Orion instructions really don’t have much to say about any PHD settings; but this is likely the only one that will need to be changed. If you do need help with PHD, the best place to get it is the Stark Labs Yahoogroup.
Alrighty then. Clicked on a star close to M13 (the cluster was visible as a fuzzball in the StarShoot frame), PHD calibrated, and immediately began guiding (you know it’s guiding when it puts a green box around the guide star). All I had to do was get the imaging scope going with Nebulosity which I will talk about in greater detail in a blog entry devoted entirely to it. Real Soon. I stared at the netbook screen nervously until the first couple of two-minute subs had come in, “Cool! Nice round stars at f/6.3!” (That’s an enlargement of the star field below.) M13 was way too low in the sky to deliver much, but at least I got my traditional yearly shot of the Great Glob.
Does that mean every single subframe I took on this evening was perfect? Hell, no. Are they ever? Mine aren’t anyway. I did throw out a few on the three targets I essayed, M13, M27, and M57—but only a few, a very few. It’s a waste to throw away any subframes, of course, so I did some troubleshooting by the light of the next morning.
First thing I did was check EQMOD’s (the program I use to run my Atlas) ST-4 guiding setup. Turned out I’d left the ST-4 guide rates for Right Ascension and declination at EQMOD’s rather low default, “.25,” which, if not crazy, is substantially slower than what I normally use, .50 or even .90. That may have been a contributing factor; we’ll see how it goes at .50 next time out.
Anything else? During the run, I had noticed the cables from the StarShoot, the USB and ST-4 cables, had got awful stiff in the (for us) cold low 40s F temperatures, and had hung up on the scope/mount a time or two. I’ve secured the cables’ to the finder stalk with a Velcro strip, and, again, we’ll see how it goes next time.
Magazine Plug: For once, this is not in Unk’s own self interest. I have never once appeared in Astrophoto Insight. The talents of the imagers who write articles for this superb e-zine are so far in advance of mine as to make my paltry efforts downright laughable. But that is OK. I have learned one hell of a lot from AP Insight. It is a treasure trove of how-tos and reviews. Go get it, campers.
Next time: It’s Ho-ho-ho and mistletoe and presents to pretty girls! As is our custom here, the next blog entry will appear on Christmas Eve rather than on Sunday. See y’all then!
Links to this post:
Interesting and useful blog, as always, Unk. I use a Philips webcam on the 50mm finder that came with my 200 mm f/5 Newt and it works a treat, even on a CG-5 mount, which is hopeless without autoguiding. For software I like Metaguide because it has more tricks up its sleeve than PHD. Thanks again, Mike.Post a Comment
Links to this post: