This is a new image from last night. Compare the ATIK 490 camera to the Canon 6D, same imaging lens at F/4. This is the full frame from the ATIK, which covers only about 1/3 of the Canon frame. The spatial resolution of the ATIK is higher, and the red sensitivity is enormously better. We can begin to see some evidence for reflection nebulae just south and east of the Rosetta.

----

Revision 1: Wow! There are so many different ways we can go. I think this one is cleaner. I try hard not to stray too far from what is actually in the data. But since everything is really black to our eyes, who knows how it should look?

------

The big news about this image is that a helpful hint was given me by one of the PixInsight gurus, who told me how to stop down a Canon lens and keep it stopped down. The Canon 200 mm F/2.8 lens produces relatively poorer star images at wide open settings, and needs some stop (at least one) to remove much of the lens aberrations, and to assure a better chance of getting good focus.

The trick is to hold the little-known FOV button down on a Canon camera body with the lens attached, then, while still holding the FOV button, remove the lens. The lens will remain at the chosen F-stop.

I did that, then reattached the ATIK camera to the stopped down lens to achieve this image at F/4. The results speak for themselves. Just compare this image with the image I recently uploaded of the California Nebula, taken at F/2.8, or the Taurus Dark Cloud image, with the same lens settings. These other images show blobby comatic bright stars, and generally poorer focus that I was able to achieve for this image. This image shows nice tight star images, no coma, and pleasant diffraction spikes, which are also indicative of decent focus. All images were unguided 150s exposures, so star trailing should be of similar proportions in each of them, including for this image taken near the Equator, where tracking errors will exhibit the greatest variance.

I don't like losing the F-ratio speed for exposure efficiency, but the new resuls produce better looking images and can be more easily repeatable in performance with the slower focusing cone at the image plane.

The depth of focus of an optical system can be estimated as 2 * Airy disk size * F-number, or 2 * 2.55 * lambda * F^2, for wavelength of light lambda, and F-ratio F. So at F/2.8 the depth at 550 nm is about 22 microns, while at F/4 the depth is about 45 microns. Believe it or not, while that may not seem like much of a difference, when having to focus manually by turning the focusing collar on the lens assembly, that extra 23 microns of focus depth makes it much easier to achieve a decent focus. Still very touchy, but at least achievable as seen here.

The other trick I found for getting good focus with a Bayer OSC is to focus in 2x2 binning mode so that you have maximum sensitivity, and the starved R channel does not detract from a good focus image, as it will in an un-Bayered 1x1 image.

---

New revision restores the evidence for the reflection nebula to the southeast of the Rosetta.