Sunday, November 01, 2015


Let’s Get Going with PHD2

If you’ve read some of my recent posts on short-sub imaging, you know it’s not always necessary to guide your mount; it’s not always necessary to continuously monitor a star and make small corrections to the scope’s aim to make up for imperfections in the mount’s drive gears. Short sub imaging is especially appropriate for photographing bright objects like the Messiers in the light polluted back yard. Go over 30-seconds, and the gradients from the bright sky will become tough to deal with anyway. Nevertheless, there comes a time, as I said a while back, when you gotta guide.

Like when you are out under a dark and clear sky. There’s no doubt longer sub-frames are better there. Stacking many images can reduce noise and make your pictures easier to process, but stacking will not deliver details not already present to some degree in the individual sub-frames. When it’s dark and you are photographing dimmer targets or want to bring out as much detail as possible in brighter ones, a stack of five or ten minute subs is just better than a stack of fifteen or twenty thirty-second ones.

If you need the details of the auto-guiding game, how to get started that is, go to the link above. Today’s subject is mostly for those of you who, like me, have been using PHD Guiding for a while and now see the handwriting on the wall, that’s it’s time to upgrade to the new version, PHD2, a.k.a. “Open Source PHD.

“Open source?” What does that mean? PHD Guiding, “Push Here Dummy” Guiding was originally written and developed by astro-software wizard Craig Stark (Nebulosity). He got his program going in a big way and it soon became the most used auto-guiding application in amateur astronomy. Craig continued to enhance his freeware program, but incrementally for the most part, and it was clear the program could be taken a lot farther. Its bones were strong, but it could still use a little fleshing out. To that end, Mr. Stark decided the program should go open-source, that its development should continue mostly in other hands:
In 2013, Bret McKee and I started working on a complete rewrite of the open-source PHD code with the aim of setting things up for significant expansion of PHD and of the development team working on PHD. Bret really dove in and did a massive amount of work. He also built up the team that has now gotten PHD2 going. In all of this, the vision of PHD has remained - to be user-friendly, yet provide powerful guiding. I'm delighted to see that this has reached such a mature stage and that it's done so with very little of my direct effort. This is a fully open, team project and it's a real joy to see that it has taken off so well. I will continue to host PHD1 here until all have happily moved over to PHD2.
What does that mean for you and me, though? For all practical purposes, PHD2 is PHD Guiding now. The old PHD is still available on the Stark Labs website, but it won’t, I’d guess, ever be upgraded again. If you use PHD, it’s time to transition to PHD2. This sort of thing is painful when you’ve used a program for years and come to depend on it, but the new one offers some significant new features, and I’d guess many more are to come in the future. It is still freeware by the way if you are worried about that, though the developers are soliciting donations (as they should).

Before you can use PHD2, naturally you need to download it from the PHD2 website (above) and install it . As before, it’s small and the installation is quick. One thing you may be worried about that you don’t need to be? That PHD2 will overwrite PHD. It won’t. If you’re like me (occasionally confused), you’ll probably want to keep the old one on your machine for a while. That saved my bacon one night as you’ll read later. PHD2 works very well, but if you’re as silly as me, in the course of investigating new features you may foul something up and need to at least temporarily return to the old program.

When you start PHD2 for the first time, it’s likely you are going to be disappointed, “Heck, Rod, it doesn’t look any different from the old one.” It doesn’t, not really, except for the fact that it comes up in a rectangular rather than square window. You don't begin to see the differences until you begin playing with the menus and settings, the first of which is probably going to be the camera icon that allows you to select your guide camera.

This was my first clue PHD2 was going to be better. Rather than just a list of supported cameras to choose among, what you get is the “Profile Manager.” Here, you select your guide camera, your telescope mount, and other things via nice drop-down menus. When you are done, you save the current camera/scope/etc. setup as a profile. This is a godsend if you, like me, use both a guide scope and an off axis guider depending on the telescope you are guiding. This is also where you’ll connect your camera, mount, rotator, etc. when you are ready to roll. But you aren’t ready to roll yet as there is more configuring to do first.

You'll access the detailed configuration menus with the familiar Brain icon, same as before, but that’s really all that is the same in PHD2. Setting up the program is simpler and less scary than it was in PHD. What comes up with a click of the brain is a window with tabs. The first of these is “Global,” which, with one exception, you can leave alone for now. That exception is your guide camera focal length. Enter it in millimeters in the Focal Length field. Oh, there’s one other thing here you should notice, “Reset Configuration.” Ticking that resets the program’s configuration to its original values. If I’d paid attention to that I wouldn’t have had to switch back to the original PHD on the first night of the Peach State Star Gaze.

You can leave the “Guiding” tab completely alone at first; click the next one, “Camera.” Here, you enter the pixel size of your camera’s chip. Where do you get that? From the manual or from the camera manufacturer’s website. Failing that, you can obtain the required numbers from the spec sheet of the CCD or CMOS chip used in your guide camera (you should certainly find the type of chip your camera has on the camera maker’s site). This is important; PHD2 needs to know your image scale in order to be able to tell you the quality of guiding in arc-seconds. What else is here? You may occasionally find a need to mess with camera gain, but usually only if you are for some reason trying to guide on an overly bright star. Leave everything else the way it is.

Finally, there is the Mount tab. Some of this can also be left alone for now—though you will likely eventually come back here and fiddle with aggressiveness and other values to fine tune your mount’s performance. Most mounts will guide quite acceptably with the default settings, however. What you do need to enter here is the size of the “steps” PHD2 uses during calibration. Unlike the old version, PHD2 will figure this out for you. Click the “Calculate” button, enter your guide scope focal length (or main scope focal length if you are using an OAG), pixel size, and mount guide speed, and the program will give you a value in milliseconds to enter in the calibration step field. This is mostly important if you are using a short focal length guide scope—too small a value here will mean calibration takes forever to complete.

And that is pretty much it. Otherwise, the program mostly works the same way as the original. Connect to camera and mount (using either ASCOM or “on camera”—through your mount’s ST4 port), set your exposure via the drop down, begin looping exposures, choose a good looking guide star, one not too dim and not too bright, click on it, and then click the PHD (archery target) icon and calibration will begin. PHD2 will move the mount E/W and N/S, and when that is done will begin guiding. End of story. Actually, there is one change to the procedure; you no longer have to stop looping exposures before you click on the guide star. In fact, you are now encouraged to select the star while the video is running.

There is also an interesting new feature concerning guide star selection, but it’s not apparent until you examine the Tools menu. If you’re not sure which star in the field of your guide camera is the best candidate for a guide star, you can have PHD2 automatically choose one. Just click, "Auto-select star" in the Tools menu  (you can also tell it to do that with an Alt-S hotkey combo) and PHD2 will do just that. I’ve gotten pretty good at recognizing the look of a good guide star for PHD, but I played with this anyway, and can report the program always selected what I’d also have identified as a good one.

And that is all there is to it campers. If your guiding is good, you don’t have to do another thing. There are plenty of other features, like graphs to show you how your guiding is going, a drift procedure to help you tighten up your polar alignment, and more. But if your guiding is giving you round stars, you can leave all that for later or never.

What if your guiding isn’t as good as you would like, though? Which can happen, especially if you’re guiding a long focal length telescope. PHD2 offers several ways of refining your guiding without causing overmuch heartburn.

The first thing you can do to improve things is to clean up the guide camera's video frames, make the guide stars look better and get rid of some of those pesky hot pixels. PHD2 makes it easy to take and apply dark frames. Just select “dark library” from the Darks menu, set the range of exposures you normally use for guiding, cover guidescope (or mainscope if you’re using an OAG), and hit the go button. The program will acquire your range of dark exposures and apply them henceforth to light exposures of the same length. 

Sometimes darks ain’t enough, however. Most guide cameras are uncooled, and today’s popular high-sensitivity CMOS chips' output can look like a snowstorm at the North Pole even after you apply darks, making it hard to choose a good guide star and making you apt to try to guide on a freaking hot pixel. To further improve guide frames, you can use PHD2’s “bad pixel map” feature, but I find it easier to clean up problems darks won’t quite fix with the surprisingly effective noise reduction setting found under the Brain icon’s Global tab. On a hot summer night, selecting “median” from the drop-down gets the job done. It’s perhaps not as efficacious as a bad pixel map, but easier to do, and I like “easy” as you well know.

Your guide camera’s images look better, but the guiding is still not going quite as well as you’d hoped at 1500mm plus. It’s OK, but like most astrophotographers, you can’t resist examining the DSLR’s stars at a 400% enlargement. Unfortunately for you, they are a little eggy. What do you do? What you used to do with the original PHD was start tinkering with settings:  declination/R.A. aggression, minimum move, etc., etc. It may still be necessary to fool around with these values and particularly with the declination setting that determines whether your mount will be allowed to guide in both directions, north and south, or not.  However, PHD2 may be able to help you determine the other values.

What should these esoteric settings be? For the technically challenged astrophotographers among us, like me,  there’s a way to easily figure this out for a particular set up without just blindly trying values like I used to do with the original PHD. There’s this new feature called Guiding Assistant. Access the Assistant from the Tools menu, let it run for a few minutes, and it will, when you stop it, come up with suggestions for your guide settings. If you think they are reasonable, you can have the assistant apply them automatically. Just be careful before you do that. THINK.

The first time I used PHD2 with my EQ6 mount in the backyard, it was a resounding success. Perfectly round stars in the 3 – 5 minute exposures I customarily use. Easy as falling off a log. Uh-huh. Fast forward to the recent PSSG. Same mount, but the stars, while fine, were not quite perfect. What was different? I was using my Edge 800 SCT at f/7 rather than my standard C8 at f/6.3. I thought the increase in focal length, while small, was still maybe enough to cause the degradation. I ran guiding assistant for a few minutes. Applied the changes and gave it another go. Result? Substantially worse guiding than what I’d had before I "fixed it."

If I’d known about the option to restore the program’s default values, I’d have used that, but I didn’t, so I switched to the old PHD, which was, luckily, still on my hard drive. The result was guiding about the same as I’d had initially with PHD2. Good enough, but nothing to write home about.
Next evening, more rested—my troubles had come on the first evening of the star party when I was tired from the trip and set up—I set about to figure out what had gone wrong. First thing I did was have a look at a star field in PHD2 (I’d gotten it back to the defaults by this time) to see how the guide camera’s images looked.

Hmm…well darn. Instead of its normal reasonably sharp stars, my old Orion StarShoot was delivering faint fuzzballs. The guide scope was badly out of focus. How? While this 50mm telescope focuses by screwing the objective in and out and snugging up a knurled ring against it, and normally holds focus well, apparently the trip had been enough to throw it out. I had to change focus quite a bit to get the stars looking as good as they should.

Focus attained, guiding became sterling again. Actually, if I’d listened to the Guiding Assistant, I’d have figured this out on the previous evening. When I ran it, it kept talking about the guide star being too dim. When I tried it on different stars in the field it kept complaining about star brightness. The reason for that was that they were out of focus and I wasn’t seeing that with my tired, bleary eyes.

So, the bottom line on PHD2? It just works. Don’t get the idea that it will necessarily make auto-guiding easy in the beginning, though, Joe and Jane Novice. You’ll inevitably have gremlins to exterminate. Things like flexure, cable drag, backlash, and on and on. That said, PHD2 is the new king of guiding software and makes the process as easy as it can be. Even if you’re new to the game, just get it. I said this article was mainly aimed at people experienced with PHD, but that doesn’t mean you shouldn’t start out with PHD2 instead of PHD. There is no downside to the new one; it is just better.

What got you is that the EdgeHD is *sharper*, so you can see imperfections that would not show up with lesser optics.

Covington's Law of Guiding: Sharp star images are not round. Round star images are not sharp.


I'm stealing that one, Michael. LOL

Oddly I have found that PHD v2 has not been a simple transition across. With identical settings and hardware my guiding is woeful. So I had to revert back to PHD v1 which gives me good guiding. I have tried ASCOM & ST4 on PHD v2 and its still the same, weird. I do intend to try and trouble shoot but my clear skies are next to none at the moment...
In general I get good results with my 14" EdgeHD guiding OAG. Usually -10-15 minutes subs with round stars. Focusing is crucial. Although it will guide on a SLIGHTLY unfocused star, but to a point. As you've said those fuzzies are not welcomed. Another problem is a tiny FOV for a OAG, hard to find a guiding star, especially compounded with a problem of under-focused star. Also I guided with QHY-5L-II-Color. For some reason it was very slow- to non-responsive to download via USB. Had to switch to Atik Titan, mono (cooled!). No more USB troubles, so far... I guess all these points are not related directly to PHD2. One is, though. It is hard to come back to previous calibration data, except for the very last one. And I always have to redo it after the meridian flip (probably need to tweak my MesuMount200/SiTech settings further to alleviate). Nevertheless it would be nice to have a library of calibration sets, not just last one.
Thanks for this writeup!
Rob, thanks for this wonderful guide. I have a Synta finder guider with QHY5L-II. PHD2 software shows the calculated step size as 7300ms. Is this correct? I am a bit confused. A friend told me it doesn't look right, so I used a step size of 1800 with which calibration completed in about 5min. I am still surprised about the 7300ms. My guide scope FL is 162mm and pixel size is 3.45microns.
Your mate is right. Around 1200 or so is good. Might check the numbers you put into PHD...
Uncle Rod,
Wonderful write-up, and mui helpful-o. To be honest, I'm not thrilled to death with SkyX's guiding. Calibration is challenging, to say the least, and I'm using an 11" EdgeHD on a Paramount ME that is finely tuned, so, realistically, it shouldn't be. We'll give PHD2 a go, and thanks, again, for your always-insightful advice and down-home writing -- so refreshing.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

stats counter Website Hit Counters