autoguiding craig stark

In his article on Starizona’s Hyperstarsystem (ATT April 2007, p. 28), ScottTucker described astrophotography assomething that, at least historically,require a “certain borderline lunacy.”Astrophotography is typically not for theimpatient, casual amateur, nor is it typi-cally for those without fat wallets. Take,for example, the relatively simple,unavoidable fact that the earth rotates.The result of this inescapable fact is thatthe stars move slowly across the sky.“Slow” is a relative term, of course, and ifyou use a large amount of magnification(e.g., by using a high-power eyepiece), youcan see the stars race across the field of view.

The solution to this is, of course,quite simple and well-known. With asingle motor turning the telescope aboutan axis that is parallel to the earth’s axis ofrotation (i.e., that points at the celestialpole), we can counteract that motion andkeep the target from drifting away. Forvisual observation, this is all we need. Itmatters little if, over the course of 10 min-utes, the target rocks back and forth in theeyepiece by a half an arcminute. At200x in a Plossl eyepiece, a half anarcminute corresponds to under 4% ofthe field of view. You’re losing sleep overobserving, not over this slow, small,

wobble of the image.To an astrophotographer, this slow,

small, wobble of the image is a seriousproblem to lose sleep over in and of itself.If we have a camera exposing this entiretime, the wobble will make the starsbecome lines rather than points. Forexample, if we image at a fairly typicalimage scale of 1.5 arcseconds/pixel,instead of the star being perhaps 4 pixelsin diameter, it might be 4 pixels high andover 20 pixels wide.

The solution to this problem is alsoquite well-known. If, in addition to expos-ing our main image we carefully watch astar, we can use this “guide star” to see theerror the mount is making and correct forit as it happens. As the star starts to driftto one side, we send a signal to the mountto either reduce or increase the speed ofthe motor briefly so that the star will bepulled back to the proper position. We cando this either manually, by physicallywatching a star and pressing buttons(manual guiding) or automatically, byhaving a separate camera watch the starand send commands to the mount (auto-guiding).

This sounds simple. Where’s theproblem? Why doesn’t everyone guidetheir mounts and take pictures with nice

round stars? Historically, there have been alot of problems, hurdles, and myths thathave kept guiding out of reach of manyastrophotographers. For example, manyare told that astrophotography cannot bedone on basic, mass-produced mounts.Bigger and more precise mounts have notonly less error overall, but error that is“smoother” and more easily fixed withguiding. These are also priced out of reachof many amateurs. If you must spend$3,000 to get a mount that is consideredentry-level for astrophotography purposes(and still need to guide), many will shyaway. Even if you have a suitable mount,you may be presented with a bewilderingarray of parameters. For example, you maybe asked to enter in the number of arcsec-onds/pixel, the RA and Dec aggressive-ness, their hysteresis, the amount of back-lash in milliseconds, which way north is inthe guide image, whether it is mirrored N-S and/or E-W, how much error to tolerate,etc. While some of these are easily deter-mined, others are not and with so many toset, the chance of user error is high andfinding the source of the error can be chal-lenging.

I remember not-so-fondly trying touse several systems and becoming so frus-trated that I abandoned guiding.

Astronomy TECHNOLOGY TODAY 47

By Craig Stark, Ph.D.

Autoguiding:As Simple as“Push Here Dummy”?

Sometimes I’d hit “guide” and the starwould race off 10x faster than the slowdrift it had shown. Other times, the pro-gram would lock onto a hot pixel andinsist my mount was perfect (funny howthey don’t move) or would work for ashort while and then degrade. My“favorite” night was when it was workingwell in RA but after about 5 minutes ofjoy, drift in declination began to build upand the program and I had a disagreementas to which way “north” was. Its correc-tions only made the error bigger and thestar promptly shot off the chip. I can stillremember the literal pain in my neckmanually guiding was and despite a num-ber of valiant efforts, I never managed toproduce round stars (it didn’t help that Itoo would forget which axes I should keepin the same layout as the buttons andwhich should be mirrored).

I abandoned guiding in favor of theease of unguided imaging and lived withthe noisier images my shorter exposuresproduced, always wanting a better way. Inearly 2006, I had already written andreleased Nebulosity, a program designedto both capture and process astrophotog-raphy images. Its mantra was to be power-ful, but easy to use. I know first-hand thatmuch of the time when doing this, I’mstanding in a dark field that is either freez-ing cold or infested with mosquitoes. Ontop of that, I’m generally tired and it does-n’t take a cognitive psychologist or cogni-tive neuroscientist to tell you that underthese conditions, you don’t work or think

your best. (If you do feel you need a cog-nitive psychologist or cognitive neurosci-entist to tell you this, consider yourself so-told.) It’s at times like these that one real-ly appreciates a clean, simple, user inter-face that offloads as much as possible fromthe user’s brain to the computer. After onenight of not being able to image as long asI’d wanted, I decided to write PHDGuiding, taking the same approach I’dtaken in Nebulosity. The name comesfrom both its attempt to be intelligent (orat least well-informed and an expert in aparticularly small domain) and to be“Push Here Dummy” simple.

HardwareBefore we cover how to autoguide

with something like PHD, we need tocover the bits and pieces of hardware youneed for guiding. There are three basicthings you need: 1) a second camera chip,2) a means of projecting an image ontothat chip (e.g., a guide scope), and 3)some way to have your computer tell yourmount which way to move. We’ll takethese each in turn.

Guide CamerasThe same pixels you’re using to take a

long-exposure image of a DSO cannot beused for guiding. Sorry, it’s true. Once youdump the charge from the CCD wells toread off an image, that charge is gone andcannot be replaced to let the image buildup more. Some SBIGs have two CCDs inthe same camera head to get around this

48 Astronomy TECHNOLOGY TODAY

ScopeStuffTelescope Accessories & Hardware

FEATURING ITEMS FROM:TeleGizmos Covers - Astrozap Dew Shields

Dew-Not Dew Heaters - Peterson EngineeringAntares - Telrad - Rigel Systems - Sky Spot

Starbound Chairs - Smart AstronomyDavid Chandler - Lightwedge - BaaderScopeStuff Piggyback & Balance KitsRings, Rails, Dovetails, Cables, ATM,Eyepieces, Filters, Diagonals, Adapters

Green Lasers - And MUCH more!

www.scopestuff.com 512-259-9778

EZ BINOC MOUNT KIT5-axis outdoor mounthandles all sized binocs.View standing, sitting orreclining. Kit contains allparts requiring welds,machining, drilling &tapping, plus specializedparts. You supply inexpen-sive pipe & fittings readilyavailable locally.Put yourself in the driver’s seat for $99.99

PETERSON ENGINEERING CORP.TEL: (888) 712-9332 or (401) 245-6679WWW.PETERSONENGINEERING.COM

AUTOGUIDING

mySKY $399U.S. withFREE Shippingto U.S. andCanada.

25 ft. Coiled Cablefor mySKY MeadeTelescope Control$29.95 U.S.

WE INVITE OUR U.S.FRIENDS TO EXPERIENCETHE KHAN SCOPE CENTRE DIFFERENCE!

We Take Trade Ins! • We Sell Pre-OwnedEquipment With Warranties! • We Ship

Anywhere in the Continental U.S.! • We CarryMajor Brands as Well as Unique Products!Check out this Month’s Internet Specials

Antares • Astrodon • Bob’s Knobs • CelestronCoronado • Denkmeier • JMI • LumiconMallincam • Nikon • Sky-Watcher • TelevueUnihedron • Vixen • William Optics • And More!

Our 20th Year!Khan Scope Centre

www.khanscope.com1-800-580-7160 • 416-783-4140

MEADEmySKY$399

problem. With their Star-2000 system,Starlight Xpress used half of the lines (e.g.,the odd lines) to build up an image whilethe other half (e.g., the even lines) wereread off for guiding purposes. But, in nei-ther setup are the same pixels being usedfor guiding and imaging at the same time.

If not using one of these systems, weneed a separate guide camera. Numerouschoices exist for a guide camera since thequality of the image is not paramount.That said, some cameras work better thanothers. In general, the best guide cameraswill be a) monochrome and b) capable oftrue exposures of at least a second or so. Itis true that you can guide using a simplewebcam, but the color filter array over thesensor and the short exposures typicallyavailable on webcams conspire to placereal limitations on how faint a star you canguide on. A one second exposure captures30 times as many photons as a 1/30 sec-ond exposure and a filter designed to passonly red, green, or blue light passes lessthan a third as many photons as no filter.Put these together and a monochromecamera with a one second exposure cap-tures about 100 times as many photons asa webcam running at 1/30 second. Whileyou can stack short exposures on the fly to

get closer to a long exposure (and PHDGuiding does this), your’re facing a differ-ence of five star magnitudes. What thismeans is simply that the right choice ofguide camera will make finding a suitableguide star far easier.

Guide Scopes andOff-Axis Guiders

We need some way to form the imageon our guide camera. One method is touse an off-axis guider (OAG) – a devicethat uses a small prism to pick off a por-tion of the light from your scope that isheaded for an area just outside the maincamera’s sensor. By directing this light toyour guide camera, you can use any starthat’s available here as your guide star.

Off-axis guiders have several advan-tages. First, you don’t need any other tele-scope since your main imaging scope feedsboth cameras. Second, you don’t need toworry about things like mirror-flop or flexin the mounting of your guide scope, soyou’re always sure the main camera andthe guide camera stay pointed in the samedirection. That said, off-axis guiders dohave their own issues. First, they take up abit of focus distance (you need to makesure you have enough inward travel inyour focuser to make up for the thicknessof the OAG). Second, you have to find aguide star at the focal length you’re imag-ing at and in the small part of the sky cov-ered by the pick-off prism.

Difficulty with this aspect is what


AUTOGUIDING

Tandem guide setup. Here, a small guide scope (a modified 8x50 find-erscope) with a guide camera (Fishcamp Starfish) sit next to the mainimaging scope (TMB 80SS). Both are mounted on the same dovetailbar that is attached to the mount.

Piggyback guide setup. Here, a small refractor (William Optics ZS 66SD) with a guide camera (CCD Labs QGuide) ride atop the main imag-ing scope (Vixen R200SS). The guide scope is held in place byadjustable rings that are affixed to the main imaging scope.

Zeke’s Seat’sWHERE BUSINESS IS ALWAYS LOOKING UP

Our astro seats are all TIGwelded to last a lifetime.Great seat - Great price.Four full inches foam paddingfor all night comfort.One handed adjustment.About 10Lbs, 0-33 ins.

#25 Channel Ln. Ft. Myers, Fl. • (239) 693-3691 • [email protected]

The Sky Looks So Sweet, If You

GET A ZEKE SEAT!!We take PayPal, Cash, Check, M/C, Visa & Trade?

Special Purchase

18-24 ins

$119$99

drives most to use a separate guide scope.Guide scopes are typically small refractorsattached to the main telescope (or mount-ed side-by-side with the main scope) andheld in place with adjustable rings. There isno need for the guide scope to be of partic-ularly high quality (or even that it be arefractor), but you do want to make sure itis mounted rigidly and that its focuser canhandle the weight of your guide camerawithout flexing. If modest pressure on yourguide camera lets you move it relative to themain scope, you will be limited in how longyou can expose without your stars trailing.As the mount rotates, gravity acts on thetwo setups and flex will result in it actingon them differently. The guide star maytherefore remain stationary on the guidecamera while the image in the main cameraslowly drifts. Addressing this flex in myown setup let me move from 2-minuteexposures to 20-minute exposures.

Back in the days of film imaging andmanual guiding, there used to be a rule ofthumb that the guide scope should be at

least half the focal length if not the samefocal length as your imaging scope. Thiswas so that with typical reticle eyepieces,you could see the motion soon enough andreact accurately enough to manually guideout the mount’s error. With CCD-basedautoguiding this rule can be thrown away.Computers are far better at spotting verysmall movements and far faster and moreaccurate in their reactions.

With modern guide software, motionsthat are small fractions of a pixel can beaccurately estimated. Subpixel guiding onlyrequires that a star’s light covers several pix-els (which, even with short focal lengthguide scopes and large CCD pixels, can becreated with a slight defocus). Suppose, forexample, that a star’s light strikes four pix-els in a 2x2 grid and that the star is placedexactly at the center of this grid. Each pixelin this 2x2 grid would therefore be equallybright. Now, suppose the star moves ever soslightly to the right – a tiny fraction of apixel to the right. As its Airy disk’s energy isnow centered up/down but is slightly off

center left/right on this 2x2 grid, the twopixels on the right will get more energy andbe a bit brighter than the two on the left.Move it down a bit and the lower-rightwould get the most energy and the upper-left the least.

It is this basic notion that allows us touse short focal length guide scopes.Personally, I use a 66-mm telescope with a388-mm focal length (William OpticsZenithstar 66 SD doublet). I know manywho use scopes of similar focal lengths andhave even seen excellent results from an8x50 finderscope that had been convertedinto a guide scope (200-mm focal length).The SBIG eFinder accessory for their STVguider is even shorter at only 100 mm! Thedays of very long focal length guidescopesare over.

Getting Guide Signalsto the Mount

At this point, you’ve got an image ofthe stars from your guide scope on yourguide camera. You still need some software


AUTOGUIDING

that will capture these images and figureout what commands to send to the mountand you need some way to get those com-mands to the mount. We’ll take on this lat-ter bit first.

There are two basic ways of nudgingyour mount during guiding. First, if youhave a computerized mount, you can usethe same cable used for controlling themount from your computer (e.g., via someplanetarium package) and forupdating it. Generally, this is a serial(RS-232) cable or a combination of aUSB�serial adapter and a serial cable.When sending commands over this con-nection, your guide software either needs toknow the particular dialect spoken by yourmount (i.e., the specific commands neededto nudge a Meade Autostar vs.a Celestron Nexstar vs. a LosmandyGemini, etc.) or it needs to know how totalk to something else that knows thisdialect. In Windows, the ASCOM plat-form (www.ascom-standards.org) providesthis intermediary and many software pack-

ages rely on ASCOM as a result. Individualdrivers are written that translate “ASCOM-speak” into each individual telescope’sdialect and other programs just need toknow how to “speak ASCOM” to then endup successfully working with a wide rangeof telescopes. The vast majority of ASCOMdrivers are free and most mounts will beable to be controlled with the driversincluded in the download. Don’t expecttech support via a toll-free call, however, asASCOM is a collection of programmershelping the community out in their sparetime.

For many mounts, guiding via the seri-al port can be very effective and accurate.For some mounts, guiding via the serialport (either directly or via ASCOM) is notas accurate as one would like. The reasonfor this can usually be traced to the smallcomputer present in the mount and to thenature of the commands that can be sent.These small CPUs may listen for a newcommand once every quarter second or so,allowing them to spend most of their time


AdvancedTelescope Systemswww.AdvancedTelescope.com

Manufacturing QualityPortable and Permanent

Telescope Piers

Toll Free

(877) 96-SCOPE

AUTOGUIDING

executing commands, keeping up with thedata coming from the position encoders onthe axes, etc. When listening for GOTOcommands and the like, four times per sec-ond is plenty fast. If guide commands takethe format of “Guide East ON” and“Guide East Off”, we run into a problem.The shortest guide pulse we could executein such a system would be a quarter secondas it would take at least this long (andsometimes twice this long) to process thecommands. Modern GOTO setups usuallyhave a more complex command that allowsyou to specify how long the pulse should be(e.g., “Guide East for 79 ms”) to get aroundthis problem. The generic term for this isnow “Pulse Guiding” – an adaptation of amore complex scheme originally designedfor Astro-Physics’ mounts.

The second way of getting commandsto your mount is required for mounts thatdo not support this command syntax, fornon-computerized mounts, and is evenused by many for their computerizedGOTO mounts (e.g., when using planetar-

ium software to control the GOTO aspectsof the mount). This method uses the “auto-guide” or “ST-4” input port found on mostmounts (developed by SBIG for their ST-IV autoguiding system and adopted now asthe de facto standard). Each direction ofmovement corresponds to one pin on theconnector and a fifth is the “common” pin.In the simplest systems, these pins aredirectly connected to the motors or to yourhandbox’s arrows. Moving the mountamounts to electrically connecting thecommon pin to the desired direction’s pin.

No computer ships with an ST-4output port (and no, you can’t use themodem or network ports). Thus, we needsome hardware to get an ST-4 output fromyour computer. Shoestring Astronomy(www.shoestringastronomy.com) sellsversions that attach to your parallel(GPINT) or USB (GPUSB) ports andplans for the parallel-port version arereadily available on the Web for thoseinclined to build themselves (and thosewho still have a parallel port on their

computer). In addition, a number ofcameras come with ST-4 output ports builtinto the camera. This negates the need forsomething like a Shoestring adaptor as theguiding software can send commands tothis onboard ST-4 port that are then sent tothe mount itself.

Preparing Your MountBefore guiding, you need to make sure

your mount is working well. You do notnecessarily need a very expensive or high-end mount, but you do need to make sureit is working well. What this means is:a) It must not be overloaded. For

mass-market mounts, find the mostweight the manufacturer ships withthe mount and don’t get very close tothis. Some suggest a 50% de-rating ofthe capacity, but I have found thatvery good results can be had with lessconservative de-rating. Unless this is avery high end mount, don’t run it atfull-capacity.

b) It must operate smoothly. If your


AUTOGUIDING

mount has large shards of detritusembedded in grease that resemblesepoxy and if it takes 30 seconds foryour mount to begin to respond whenyou reverse directions due to excessivebacklash, it’s time to clean and tuneup the mount.

c) It must be well-balanced. For guidingpurposes, this usually means a slighteastward bias to the weight balance(so that the motors are workingslightly against gravity).

d) It must be reasonably well polaraligned. Perfect alignment is notneeded but guiding accuracy will besignificantly improved if you have areasonable polar alignment. A roughdrift alignment, a few minutes withan iterative GOTO alignment, oralignment with a good polar scopewill suffice. But, you should not seethe star drifting away with a quicklook through a high-power eyepiece.

Software ChoicesWith the hardware in place, we now

need something that will get the image ofthe star, figure out how much it’s moved,figure out what guide direction(s) shouldbe engaged and for how long, and thensend those guide commands. Many optionsexist here and many are either very afford-able or free. A large number of options existforWindows and there are solutions for theMac and Linux as well. When choosingwhat software to use, make sure it will workwith your hardware (computer,mount/mount interface, and camera) andthat it provides an interface that you arecomfortable with. Since you have manychoices that are free or that have free demos(or that may even exist in other softwareyou have), give a look at several to find onethat you like.

I have tried several, but know oneexceptionally well having written it. Giventhis, the fact that it is available for bothWindows and OS X, that it supports awide range of hardware, and that it is free,PHD Guiding (www.stark-labs.com) will

be used as the sample program in theHow-To shown on pages 54 and 55.

As you can see in the How-To, thereare only a handful of simple steps to getguiding up and going. When I head out toimage, this is all I do and start-to-finish ittakes about five minutes. You don’t needto orient the camera in any particular wayor tell PHD anything more than whatkind of camera and mount you use. If youneed to move the mount (e.g., change to anew target or reframe the current one),just press Stop, start looping exposures,find a new guide star when you’re donemoving the scope, click on it, hit Stop,and then hit the PHD button and it’llstart guiding again.

Can it really be that simple?Yes. For many, it is this simple and

help is available from other users andmyself via the Stark Labs Yahoo Group(http://tech.groups.yahoo.com/group/stark-labs-astronomy-software/) should youhit snags. If you want to make it a bit

more complex and you want to tunethings up a bit more, you can make anumber of adjustments by entering theAdvanced setup (brain icon). Over adozen settings exist in this dialog and oddsare you’ll not need to use any of themto get things up and going. Here aresome you may want to explore, however,as you try to fine tune your guidingaccuracy:a) Calibration step: During calibration,

PHD sends short pulses to the mount.If you’re using a very short focal lengthsetup, you may want to increase this sothat more movement takes place oneach step. (If PHD doesn’t detect muchmotion after 60 tries, it gives up).Conversely, if you’re on a very longfocal length setup, this step may makethe star move so much it is lost, so tryreducing it. The default works well forthe majority, however.

b) Dec guide mode: By default, PHDguides in declination. When you are


AUTOGUIDING

Feather Touch Focuser…“Get the Original, Get the Best, Get the Touch!”

Starlight Instruments, Inc.847-985-9595 • www.starlightinstruments.com

Available for:• Newtonians• Refractors• Schmidt-Cassegrains

Micros available for:• Celestron• Meade• Takahashi• Tele VueAll units can now be motorized withour new Digital Feather Touch System

Quality Focusers and Accessories for the Amateur Astronomer

Continued on page 56

PHD Guiding: How-ToPHD Guiding is designed to be as close to “Push Here Dummy” as possible. When you start it up, you are presented with a

single screen. Getting guiding up and going is a matter of a few simple steps.


1) Tell PHD what kind of mount interface you will be using. From the Mount menu,select the appropriate item. PHD will remember this choice, so odds are you only needto do this once. The default is ASCOM on Windows and if you’re on a Mac, you’ll noticethe only available choices are for the Shoestring GPUSB and for your camera’s onboardST-4.

3) Connect to the mount. Press the icon that looks like a telescope (second from the left).If you’re using ASCOM, a dialog will appear and ask you to choose which mount you’reusing. At times, the correct choice isn’t entirely obvious. For example, if you’re using aMeade LXD-75, you’d select “Meade LX200 and Autostar” since the LXD-75 uses theAutostar system (which uses the LX200 protocol). If you’re using an ST-4 adaptor, nodialog will appear and it will connect directly to the adapter you’d indicated in Step 1.The “No scope” in the right-hand portion of the Status Bar will become “Scope” and onthe left, it will tell you that the mount has been connected.

2) Connect to the camera. Press the icon that looks like a camera (far left) and a dialogwill appear asking you what kind of camera you have. Select the appropriate camera(many more are available in Windows than in the OS X dialog shown here) and hitOK. If all goes well, the Status Bar on the bottom should tell you the camera was con-nected. In addition, on the right side of the Status Bar, the “No cam” indicator willchange to “Camera.”

4) Press the “Loop” button (third from the left, looking like a green looping arrow). Thiswill start capturing images from your camera and displaying them on screen. The defaultexposure duration is 200 ms (0.2 s) and if you’re near a bright star and close to focus, thisshould let you see something. PHD will automatically stretch the image for display pur-poses (the slider next to the buttons controls the gamma - a mix of brightness and con-trast) so that you can see your stars. If you can’t see any (and you’re pretty sure someshould be there), odds are you’re out of focus. I find it useful to increase the exposureduration (pull-down next to the buttons) until I can see the large, faint, out of focus starand to adjust focus from there. I may drop the exposure duration back down to makefine focusing easier, but don’t fret too much about focus. If you still can’t see something,put an eyepiece in place of your guide camera and make sure you’re on something andthen rack the focuser around with the camera on until you can see stars.


5) Set the exposure duration to somewhere between one and three seconds. This is a nicerange for exposure durations while actually guiding as it is long enough to let atmospher-ic turbulence (seeing) blur the star to a nice average position and yet short enough to fixthe mount’s errors and not let it get too far astray. Don’t be tempted to use very shortexposures here as you’re more likely to end up “chasing the seeing” and trying to fix errorsdue to turbulence rather than errors due to your mount. Seeing changes faster than youcan move the mount, so that’s a race you’ll never win.

6) Click on a star. This will be your guide star. If the star is too bright, PHD will tellyou so in the Status Bar. Too bright is as bad as too dim. A box will appear on the starthat should turn green. If it is orange, PHD can’t locate the star. If it is green, PHDhas locked onto the star.

8) Press the PHD button (fourth button, targetwith an arrow in the bulls-eye and labeled“PHD”). Orange crosshairs will now appear onthe original lock position and PHD will enter itscalibration phase. During calibration, PHD triesto move the star first in RA and then in Dec as itwatches where the star moves. It needs to get it tomove a bit in each direction to get a good estimateof how it moves when guide commands are sent.So, you should see the box remain on the star andmove during calibration (the crosshairs will stayfixed). Once calibration is done, the crosshairswill turn green (they may move a bit from the orig-inal location), the status bar will say “Cal” and itwill begin guiding automatically. Start takingimages in your main camera.

7) Press the Stop button (fifth button). If you forgot to select a star, you can do so now.But once you have pressed Stop, the image will no longer loop and update in real time.

Step 8 Start Step 8 End

not aligned directly on the pole, thestars will drift slowly in declination.PHD will attempt to guide this outand to do so intelligently. Its “resistswitching” algorithm tries to determine

which way the drift is and to stay onone side of the worm gear, switching tothe other only when a lot of evidencehas built up saying it’s guess of the driftdirection is wrong. If you know yourdrift is only to the north or south,however, you can skip all this guessingand tell PHD to only guide in onedirection in declination.

c) Noise reduction: Some guide camerashave a lot of noise (e.g., hot pixels).Passing a filter over this can eliminatethe hot pixels without dark frames andwithout significant loss of accuracy inestimating the guide star’s location. Try3x3 median if there are a lot of hotpixels in your image.

d) Aggressiveness and Hysteresis: Theseare values found in a number of guidepackages and control what proportionof the measured error is to be used(aggressiveness) when determining thelength of the pulse to send and howmuch of a “history” (hysteresis) is to beused. Odds are that the current error is,in truth, the same as the last error asour mounts’ periodic errors are roughlysine waves that change very slowly. If it’sgoing too fast now, it’s likely going toofast the next second as well. Together,these two can be adjusted to balance outhow rapidly the mount attempts torespond to errors.

e) Force calibration: If you’ve moved to a

very different portion of the sky, thecalibration PHD will no longer be valid.Check this and the next time you tell itto guide, it will re-calibrate.

ConclusionsMy goal in writing PHD Guiding and

releasing it as freeware was to get more peo-ple to be able to extend their exposures, godeeper, and make better images. To getmore people autoguiding, the user experi-ence had to be simple, yet the underlyingguiding had to be accurate.

I take real joy in seeing first-light shotsfrom new users posted to groups ore-mailed to me. Some of my favorites comefrom users who say, “I wanted to startwith something easy, so I tried 1 minuteexposures. Since that worked, I set it for 10minute exposures and collected imagesall night.” That a user is naive enough tothe historical difficulty of autoguiding toeven consider jumping to 10 minuteexposures after a 1 minute “test shot” (Iremember struggling to get 1 minuteshots!), and that he can produce round starson the first night or so out, puts a smile onmy face.

As it can be easy enough for new usersto quickly get going and powerful enoughfor experienced users to image as long astheir light pollution permits, PHD showsautoguiding can be both simple andaccurate.


AUTOGUIDING

Continued from page 53

autoguiding craig stark

Documents