Correcting for Chromatic Aberration?

[stu]

To those maybe one or two of you keeping track of the Saga of Stu's Astrophotography, I recently figured out how to connect my camera to my school's 3" and 5" refractor finder scopes so as to capture larger fields of view (my camera on the school's 16"/18" Cassegrain results in a field just shy of 15x10 arcmin). See http://www.bautforum.com/showthread.php?t=37857 for more on this.

So Saturday I finally got to the 'scope when it was a clear night (albeit a 99% full moon). I discovered I couldn't hook up to the 5" because it's a 2" barrel and there was no eyepice projection equipment for that size. I discovered that I could not use my camera at prime on these 'scopes. I finally was able to get hooked up to the 3" refractor, and I went on a "Mini-Messier Marathon" (say that 3x fast) taking 120-sec shots of various M## objects (a post with the spoils of them is forthcoming).

However, I discovered this weekend while I was processing some of them that the chromatic aberration off to the sides is somethin' fierce. In the center ~25% of the view, the stars are nice white/blue/red/yellow balls on the order of 10 arcsec across for my focus/seeing, which is as good as I've ever gotten from Boulder. But closer to the edges, the stars make some lovely rainbows with red radiating inward and purple radiating outward, with the other colors inbetween.

So to get to the point of this post ... does anyone know how to take care of this in an "easy" way, either at the 'scope, or in processing in PhotoShop?

I've come up with two "solutions" and I don't like either of them: The first is to take several images of an extended object, moving the 'scope a little for each shot, and then stitching together the centers of each fram where the aberration isn't detectable. The second way would be to edit each color channel in PhotoShop separately, doing a scale transform on the red, increasing its size to around 100.5% and then doing the same with the blue, decreasing its size to around 99.5% (or whatever it takes to line them up).

Anyone encountered this before and have a better solution?

[Fr. Wayne]

If the abberation is uniform all the way around, it possesses excellent data. The spectrum stretches are good measurements of magnitude and color. But if you are into perfect picture with a simple refracting telescope, you'll be pulling hairs. Trying to locate a convex lens between the objective lens and your camera ain't worth it. Patching 4 of your 25% shots into one may be the cheapest way to go.

[Tog]

I read a fix for this that took the form of an aperture mask. Basically, you get an old lens cap that fits that scope, then cut a circle out of the middle. You will need to play with it to find the right size. If I recall correctly, about 1/2 inch from the edge. That would make your 3 inch a 2 inch, and would extend your exposure times due to higher f/ ratio though. It works by blocking the light from passing through the part of the objective where the light has the most severe bending.
Any chance you could post a pic of the 5 inch set up, maybe we could MacGuyverise* something.

MacGuyverise- Verb. Jury-rig. To take a buch of stuff that has no buisiness being combined, and building something fantastically useful (though usually temporary and hideously ugly) out of it all, Usually with a Swiss Army Knife and duct tape.

[Fr. Wayne]

MacGuyverise- Verb. Jury-rig. To take a buch of stuff that has no buisiness being combined, and building something fantastically useful (though usually temporary and hideously ugly) out of it all, Usually with a Swiss Army Knife and duct tape.

Is that different from "Gerry-rigging something?" MacGuyver was cool.

[stu]

Hmm. I think I'll try the shifting and shooting before I try an aperture mask.

Thanks.

[Tog]

Is that different from "Gerry-rigging something?" MacGuyver was cool.

I was in High school when Macguyver was on. Not many my ge knew what jury-rigging meant. MacGuyverisms they understood. Now it's the other was araound. I think you HAVE to use a Swiss Army Knife for ir to be a MacGuyverism. But that's more if a hunch than a rule