Focusing a CCD Camera

John A McCubbin

Principles

1. Star images are used to focus
2. Focusing should be done in its early stages at any convenient resolution
3. Focusing should be done in its final stages at full resolution using a subframe of the chip
4. Focusing exposure time should be at least four seconds to offset atomspheric turbulence
5. The star is most focused when it reads the highest pixel saturation
6. Focused stars look round
7. You must pass perfect focus on both sides to know what perfect focus is
8. You can't get a good image or guide well without near perfect focus
9. You need to make yourself a "gross focusing" eyepiece

Method

So how do you focus that camera with no eyelevel finder? Well, since a CCD is simply an accumulation of very tiny photocells, you use the reading of the individual photocells to tell you when the most light is concentrated at a single point, thus telling you that you have achieved perfect focus. Since most software programs include a "focusing" routine, hopefully these steps will be fairly obvious. I use CCDOPS v.4 from SBIG or MaxIm DL/CCD, from Cyanogen Productions. You should open your own particular software and look through its menus as I describe the process. It should translate fairly easily to your particular setup.

Step one is to focus your telescope on a star or object using a medium focal length eyepiece. I use an 18mm Celestron Ultima eyepiece, not because it's a great eyepiece, rather that it just gives a middle of the road magnification for my refractor. (I also use it to star align my digital setting circles just prior to my imaging session.) I usually find, center, and focus on the object that I intend to image, because it's easier to find and align the object through an eyepiece as opposed the the CCD camera. I then firmly tighten the clutches on my mount. Then I exchange the eyepiece for the CCD camera. If you are using the camera without a color wheel, the you will have to move point of focus further out than the eyepiece, even further if you have a color wheel. I use a refractor; so this is an obvious direction of movement. I move the camera further out from the back of the telescope. On an SCT it's a little trickier you will have to determine your direction by trial and error, since the mirror moves, not the camera. You should write down which movements of the focusing knob move the mirror in a specific direction.

I start with the camera binned 3x3 so download times are minimized. Start your find and focus routine that is part of your particular software. I use the "pause mode" so I can examine the image and clearly read the maximum pixel value in the field. The maximum value in the field represents the highest pixel reading on the area of the chip selected. If you start with a full frame binned 3x3, then it's the brightest point on the chip. At first stars will look like huge round out of focus blobs, round, but undefined. Through an SCT, they look like doughnuts. Move the focal point out, take another image. The blob should look smaller. If they don't, you went the wrong way. Repeat this until the stars visually look pretty good. At this point you are only close, but not at focus most likely.

Change the camera to full resolution (unbinned). Hopefully your camera has a planet mode or a mode whereby you can select a smaller sub-portion of the chip for download. This allows you to focus on area of the chip at full resolution and avoid long download times. It cannot be emphasized enough, that focus cannot be achieved without using a full resolution mode. It is best to select a medium brightness star, especially one with a few dimmer stars around it. Once this smaller field is selected, then return to the find and focus routine. At this point your exposure time should be at least four seconds to offset atmospheric turbulence; five is even better. Watch the max pixel reading and also watch the shape and brighness of the dim stars in the image. Telescopes with optical aberrations in them will distort star shapes on either side of focus, but at perfect focus they will look fairly round, or at least as round as they are going to get.

To achieve perfect focus you should watch and remember the max pixel readings. Continue downloading images while changing focus until you are sure that you have passed the focal point that gives the highest reading. You will notice that as you near focus, max pixel readings will increase, as you pass it they will start to drop. Try to watch the values close enough that you know what the highest reading is when you pass it up. Once readings start to decrease, then reverse directions and make only very small focusing movements. This is done more by feel than seeing your focuser move. Make only minute focusing movements. This takes practice. You will probably pass it again the other direction before reaching a repeatable highest average reading. At this point, you should take three or four images at each point of focus and mentally average them to insure that you absolutely know whether your readings are declining or rising. Atmospheric distortion will cause values to fluctuate. Watch for stars, especially the dim ones, to take on their sharpest and roundest shape. When you reach the point of focus with the highest average max pixel readings and the roundest stars, you are there.

You may then begin your imaging session in perfect focus. As long as you don't touch the focuser or remove the camera you should remain in focus. Be aware that temperature fluctuations will cause the focal point of many telescopes to change. So if the temperature is changing a lot, you will probably have to refocus.

At the end of your imaging session, remove the camera and insert an eyepiece. Without moving the focuser knob, slide the eyepiece in and out till it looks crisply focused. Score a mark on the barrel of the eyepiece so that in future imaging sessions you can use this eyepiece to start much closer to the focal point. You will probably find that the barrel of your eyepiece is to short and you will have to screw in an extension to the barrel. If you keep a color wheel on your camera, you can't find an extension tube long enough, so I use a 1 ¼" diagonal and find this is fairly close, but slightly outside the focus of my ST-8 and CWF-8. Extension barrels are available through SBIG or most telescope/CCD camera retail companies.

When you next start your imaging session, insert the eyepiece only to the scored mark, then start focusing in the unbinned, full resolution mode, thus skipping the gross focusing stages described above.

Good luck and hope this works as well for you as it has for me.

Back to Resources Page

I consider your heavens, the work of your fingers, the moon and the stars, which you have set in place ... Psalms 8:3