In my last blog I mentioned that focus was my next frontier in imaging. Of course I know how to focus, I mean auto-focus. Until very recently, I have always focused by watching some graph showing numbers of the diameter or radius of a selected star. I'd tweak the knobs until it was as small as possible, and vola - Focus! Well, my years of imaging now spans more than one hand worth of fingers, and I can tell you this method does work, but it also doesn't work. I have some great image sets now from a CCD camera with a beautiful set of red and green exposures, and a set of blue's where the stars are strangely BIGGER than they should be. Where this is sometimes a problem in the red channel because... well, it's red light and that's what red light likes to do, you certainly should not have this issue in blue. I also have a five hour luminance image set of a beautiful northern galaxy, and only the first 1/3 or so of the exposures are in focus. The final few actually have small donuts for some of the stars.
So, it almost goes without saying that focus is an important part of imaging. It must be done, it must be done well, and it must be maintained throughout the night regardless of temperature shifts, or just slippage of the mechanics (camera's can be heavy you know).
Tom has been nagging me, and yes NAGGING me, to get out of the stone age and get myself a decent focuser. Well I finally have... three times in fact, and I'm going to tell you what I've learned. First, is yes, manual focus is simply the stone age. Tom get's to say "I told you so".... again ;-)
In an attempt to get a little more focal length for some smaller objects, I've recently acquired a Celestron Edge HD 9.25. They let me borrow an 11" for a while, and it was just too much for my needs and how I like to image. The 9.25 is more of a "just about right", and gives me the extra focal length I was looking for. Unfortunately, I HATE the way these things focus. The whole mirror lock and moving the mirror thing... I can't believe any imager anywhere puts up with this on any model of Schmidt derived imaging train. So, I locked the mirror down and put an Optec TCF-S focuser on the back.
When you get your first auto focuser, there are some expectations that need to be managed. First is that you may not be able to reach focus for the given travel of your camera/connector combo. I found this was the case with my Celestron/Optec/QSI combination. After a good bit of going all the way in and all the way out, I finally figured out I needed a short extender. Only 1/4 of an inch... and unlock the mirror and move it a bit... there were growling sounds in my back yard that night (why oh why do some people try this the first time in the field!) for about an hour or more before sounds of "Halleluia it works"!
Rule #1: You must characterize your imaging system: Know that you CAN achieve focus, and know ABOUT where focus lies. Not in this reality can you just stick any camera on the back of any focuser, on the back of any optic, and press a button to make it focus for you.
Now it's time to learn to use @focus2. There are a lot of buttons and options and you need to understand what they do. If you just click "go" and accept the defaults it will almost certainly not work. Remember rule #1? For your given optical train, you need to know about how much travel back and forth yields big bloated stars, but big bloated stars, not washed out gray nothing. This will vary... A LOT as you'll see in a bit. For my f/10 2350mm Celestron with the Optec, it was +/- 2000 steps. That was a lot more than I expected. The way you determine this is to take a picture of a nice medium bright star. Put a sub frame around it, and set the focus tools to take pictures continuously. Now move the focuser in and out and watch what the star does. Make note of where the star is a dot, don't stress or fret over where the perfect dot is, this will vary from night to night, and @focus2 will figure it out for you. Get yourself a nice big pretty donut and see how far you had to move out. This is your "Focus Range".
Put that in the @focus2 dialog where it asks for Focus Range. Move your focus to the middle of the range (approximately), turn OFF "Automatically Slew" and "Automatically Adjust Exposure Time". You can leave the samples at 4, and the averaging at 2, and then turn it loose. What happens next is it takes two samples at two locations on either side of the "near focus" point. It then fits a curve to these points that converges very near perfect focus, well within the critical focus zone. In fact, I pitted @focus2 against my well practiced focusing skills, and it beat me every time. Yep, I am humbled, eating crow, and will be doing a lot more autofocusing in the future.
Rule #2: You can't drive to the gas station to get gas unless you have gas in the car. You can't polar align a scope unless you know at least roughly where north is to start, and to achieve perfect focus (even when automated), you do need to know at least where in the ball park perfect focus lives. This is really a corollary to Rule #1.
In the future, I'll blog about those options we turned off. Slewing to a focus star, adjusting the exposure times, etc. There is some training involved here too, but what I just described is where I currently am, and it works very well and is superior to my manual focusing efforts.
My Veloce also has a focuser kit, and it requires a Robofocus electronics package to run it, so I decided I would write the Robofocus X2 plug-in. Along the way... well, let's just say if you see me with a screw driver you should be afraid. While waiting on some parts from Italy<vbg>, I did put together another focusing system based on Robofocus. See the picture included. There is a belt system that turns the camera lens, and it's also robofocus based. My plug-in is ready for testing, so I gave it a shot just last night. Remember Rule #1? Well, this system is MUCH different. The focus position of the camera lens is VERY sensitive. The focus range on this (f/2.8, 200mm btw) optic was only 150. An order of magnitude smaller than my Optec/Celestron combo. Also, unlike the Optec, this lens has considerable slack when you change directions. My @focus2 runs were all off by a good 5 to 10 clicks of the stepper motor, and it was because of this looseness of the system. My Robofocus X2 plug-in settings dialog now has fields where you can put in backlash values. How will I determine exactly what they are? Rule #1... I'm going to have to measure it.
What's the benefit of all this work? The same benefits to learning to polar align, to set up a guider, etc. Better images. And more sleep. Once the system is properly characterized, I can sleep through the night, have it pause imaging and refocus every hour if I want, refocus between filter changes, refocus between targets, etc. Yeah, this is worth it, and I can't wait!
04-02-2013 7:13 AM