Like so many other aspects of this hobby, there are dozens of ways to auto-guide your scope during an imaging run. But the essence of all of them is to sense when the scope isn't perfectly tracking the imaging target and provide a correction. If you rely on your primary imaging train (i.e. your main scope and camera) you will be taking long exposures, which mean that you aren't sensing the need for corrections quickly enough. So you need a second imaging-train of some kind. In my set-up I have a second imaging-train, comprising a guidescope and guidecamera (see Equipment) from which I feed guiding images into a programme called PHDGuiding. This is a superb, free program from Craig Stark and is so easy to use, hence the name: Push Here Dummy.

The screenshot shows PHDGuiding capturing images of potential guide stars around a target. We'll get onto what makes a good guide star in a second, but first lets describe the buttons on the screenshot.

The camera button opens a camera chooser - plenty of choice, including Meade DSI for my setup.

The telescope button opens a telescope chooser. As I use ASCOM POTH to coordinate control of my telescope, I select ASCOM POTH from the list of choices. PHDGuiding will then try to connect to the scope via POTH. NOTE: for some reason PHDGuiding complains if POTH is already connected to the scope. If this happens, press the telescope chooser again. Alternatively, I simply make sure PHDGuiding is the first thing which connects to POTH.

When both camera and telescope are connected, the status bar at bottom right will show "Camera" and "Scope" to indicate all is ok.

The next button (Loop) will start the camera repeatedly taking images of the selected duration (0.05s in this case). When you've identified a suitable guidestar (see below), click the mouse over it and a green box will appear around it.

The bullseye PHD button will start the guiding process (you need to have pressed Stop first if you have been taking images). If you look at the screenshot at the very bottom right part you can see "No cal", indicating that PHDGuiding has not yet calibrated the guiding corrections it needs to make. In this state, when you press the PHD button it will commence the calibration process, first nudging the scope west 20 steps, east 20 steps, north and south. This is the clever part - you don't need to enter image size, focal length or anything complicated - PHDGuiding simply works out how much it needs to move the scope to achieve each pixel shift in any direction.

When calibration is complete (2 or 3 minutes, depending on exposure length), "No cal" will change to "Cal" and a set of green cross-hairs will bisect the guidestar, indicating that guiding is in progress. If you press Stop to halt guiding, then press PHD again, it will skip the calibration and start guiding immediately. This is fine unless you've moved to a different area of sky, where the guiding nudges could be different. In this case, press the "Brain" button to open the advanced parameters button. Check the Force Calibration box, and press "Done". When you next press PHD it will re-do the calibration process.

Just one final thing, if PHDGuiding is in the middle of an exposure, it will not react to any button presses until the exposure is complete. This obviously becomes more noticeable the long the guiding exposure. Which takes us on to.....

What makes a "good" length guiding exposure? 

The essence of good guiding is to sense the need for correction as quickly as possible before your main image is ruined by trailed star images. So you might think very short guide exposures are the answer. Well the problem with that is two-fold:

  • You are limiting the number of 'viable' guide stars - fainter ones simply won't show up with a high enough signal-to-noise ratio (PHDGuiding will complain)
  • More insidiously, however, very short exposures will react to perceived movements in your guide star caused by the 'wobble' induced by poor seeing - we call this chasing the seeing and it causes very erratic and spurious guiding corrections

Obviously, if the guiding exposure is too long, PHDGuiding will react too slowly to any drift and star images get elongated. On balance, I tend to use around 4-5 seconds, which seems to work fine.

One other consideration is that the guidestar shouldn't be too bright. It needs to be the brightest in "its part" of the image to avoid other stars confusing PHDGuiding, but if it's too bright it makes it harder for PHDGuiding to spot the exact centre (centroid) of the guidestar, reducing guide accuracy.

This probably makes it seem more complex than it really is - go for 4-5 seconds and a reasonably bright star - it'll be obvious! In any case, if there's a problem with your guiding, you'll soon see it in your images!


Make a Free Website with Yola.