Sorry to be a late-comer to this post, but I had a similar issue, and had done some investigation. This is information that I had posted on a different forum, a few years back. This was when I was shooting with Baader narrowband filters (3nm Astrodon's now). I had posted photos Ha, OIII and SII photos. The OII photos had more severe halo's, where the others did not.
Extrapolated from the post:
I had contacted Baader about the reflections, and they indicated that it was likely reflections in the optical train. So that got me doing a little more research.
Take a look at this article, that I found very helpful:
Also take a look at this calculator:
My examples and what I calculated is nearly identical to the example calculation for the 11K sensor. My halo's are 88 pixels in diameter, scope is f5, so my calculated distance is 1.94mm (vs. 1.89mm in the article). The only thing to which this can correspond is a reflection between the sensor and the cover glass, which is 1.4 +/- 0.25mm, according to the Kodak datasheet for the 11002 sensor. Evidently Kodak only used a single layer MgF2 AR coating, as they wanted good transmission into the IR, to match the sensor's response.
Furthermore, the reflected spectrum of silicon (also discussed in the article) shows it to be more reflective in blues. Greens are next most reflective and red & IR being the least reflective. This corresponds perfectly to my examples. My Ha and SII images have almost no halo's & OIII is the worst.
So... In my particular case, these halo's are highly likely artifacts that are characteristic to the sensor & cover glass AR coating. The only reason that they appear more pronounced on the OIII filter is because of the higher reflectivity of green by the silicon sensor.
In your case, the blue halo being worse than the red, sounds like my situation outlined above. You can run through the CCD calculator and see where your reflection lies, but I have a feeling it will be similar to mine, and near the CCD/cover glass interface.
Hope this is useful. It was enlightening to me...
--- In firstname.lastname@example.org, "larryl" <k3fit@...> wrote:
> In the "Larry Leitch" folder are two jpegs. One taken through my blue filter and one through the red filter. The blue one is a double halo. Trying to fix the final image in Photoshop is not working; better to fix the problem early on. Does anybody have any suggestions?
> The scope is a Takahashi Epsilon 180ed and the camera is a SBIG STT-8300. The exposure is 22 5 minute 2x2 subs.
> Larry Leitch