Hi Paul,
This is the excellent advise that Enrico gave me about taking exposures over several nights; I hope this will help, Tom Osterholt
The key to long total exposure time *is* to shoot over more than one night.
You don't align and combine with either Photoshop or Paint Shop Pro. You use CCDSoft or MaximDL, or if you prefer, the more exotic and even better Mira AP or Registar.
First things first: practice on taking long total exposures over many nights.
What I'm suggesting is to work on getting a *great* luminance exposure. Don't worry about color; once you learn how to take data over several nights, those several nights can span YEARS. Right now, I'm working on an LRGB set that I started over a month ago. With any luck, I'll be able to complete the set next week. If the permacloud over me does not dissipate in time for my targets to be at their prime this year, there's always next. But I digress.
To be able to take exposures over more than one night, here's what I do:
1. If possible, don't disconnect the camera from the scope. Leave it attached to the scope at *exactly* the same orientation (rotation-wise). Since I set up and tear down every night, that gets dicey sometimes.
2. Take a reference image. This can be a 2x2 or 3x3 image of your target composed exactly the way you want it, with an appropriate guide star already selected.
2a. This is the way I used to do it when I used CCDSoft exclusively. My reference image is an image off the *Autoguider* chip, not the imaging chip. On night 2 and so on, what I do is to take an image with the guide chip and look for the guide star from the reference image. Once the guide star from the reference image is visible in the new image, I look for the coordinates of the guide star from the reference image (for instance, let's say the guide star is at x=120, y=75, but the current image has it at x=200, y=179). CCDSoft has a nifty command where you tell the camera to position the selected guide star to a new location. IOW, you effectively tell CCDSoft to GOTO x=120, y=75. And voila! You will be pretty much set up as you were the night before.
2b. MaximDL style: MaximDL has a "Point Telescope here" command. Even easier than the CCDSoft method, though with the guide star method of CCDSoft, this can be way more accurate (though I don't worry too much about it with Maxim since I actually like the rough placement, which forces dithering. What's dithering? Ahh....). Note that the "Point Telescope Here" command requires some setting up. CCDSoft I think has a similar command, but I've never succeeded at getting that to work.
3. If you *have* to remove your camera every night, or (and this has happened more often than I'd care to admit) the camera rotates of its own accord, you may have to do a bit of hunting and pecking as far as rotating the camera is concerned so you can recompose your image to where it was at night one. I try to avoid this whenever possible, but when I do I again don't worry too much about it, as long as I'm within a few pixels.
OK, so now you've mastered taking images across one or more nights. Problem is, for one reason or another the images from one night are slightly misaligned with respect to images from another night. Guess what? That's actually good!
This is where dithering comes in. Dithering is where you deliberately (via software commands, of course) shift the composition around from subframe to subframe by a few pixels. The idea here is that you want to spread the camera's bad pixels (each and every camera has bad pixels - either a bit too "hot" or a bit too "cold"). Dithering will allow you to statistically remove those bad pixels by putting them in different parts of the image from subframe to subframe.
How do you take advantage of dithering? When you combine your images (and remember to use CCDSoft or Maxim, or one of the more advanced programs), you want to use a statistical reject combine method. Say what? A what? I'm talking Median combine, or better yet, Sigma Reject. Simply put, Median combines are like what some Olympic scoring techniques do: drop the low and high values from the samples, then average the remaining values. For a median combine to work, therefore, you would want to have at the very least three subs. For Sigma Reject, you want at least ten.
Guess what? Median combine is also a means to get rid of the aggravations (meteors, planes, satellite trails) and allows you to use subs with those confounded trails! Since the trails will exist in only one of your samples, the median combine will get rid of those trails for you! Remember, it drops outliers!
OK, onwards and upwards.... meridian flipping. That is the biggest challenge. So you've got your image composed right, you got a nice guide star. You shoot, and you shoot, then it's time to meridian flip. You flip and.... where's that $#@$#@@# guide star?!?!?
At this point, you'll have a few choices to make:
1. Plan ahead. Use TheSky to help you find a composition that will give you a guide star both pre and post flip.
2. Plan ahead. Decide what target you want to shoot pre flip, then maybe pick another target post flip.
3. A dicier method: If you feel frisky, rotate your camera exactly 180 degrees. This is where saving a reference image from your guide chip will help you again. You'll want to rotate, hunt and peck, until your guide chip's image post-flip looks a lot like your guide chip's image pre-flip.
4. Since you're using a refractor, get a guide camera and guide scope. This is how I'm currently managing flips, though it has spoiled me to where I'll almost have to relearn my techniques using the built-in guide chip.
If you manage to use one of the above techniques to image after the flip without rotating the camera 180 degrees, the images will also be rotated 180 degrees. No big deal, just rotate them 180 degrees back before aligning them with a selected master frame.
OK, my fingers are burning up now. Hope this little treatise was understandable and of some help....
Eric
