Raw frames courtesy of Deep Sky West Remote Observatory in New Mexico, USA. (deepskywest.com) Data obtained with FSQ 106EDXiii / QSI683wsg / Lodestar / Paramount MyT.

4 hrs total integration (16x900s B) for the bottom panel, 35 minutes (7x300s B) for top panel. Both images were acquired around the same time last November. Our observatories are located about 400 miles from each other, both in the Southwest USA. (Tucson, AZ for me, Santa Fe, NM for DSW).

The top panel is a portion of the B channel from my own image of the Flaming Star nebula, while the bottom panel is the same portion of the same sky field taken from the very dark Deep Sky West Remote Observatory site.

The images have been stretched to the same level, and we are looking at a 1:1 magnification. My own camera is an ATIK 490 on a HyperStar C8 with 1.78 arcsec / pixel. The DSW camera has a scale that is just slightly larger at 2 arcsec / pixel, but both images have been geometrically scaled to be the same.

So now, my own exposures were unguided 5 min integrations on a Paramount MyT, and the DSW exposures were guided 15 min on the same kind of mount. Right off the bat, we can see that guided exposures appear superior to my own unguided. (c.f. star shape, not size) But remember we are looking at full zoom images, and these differences would probably not stand out in a whole frame view.

The next thing to see is that my shorter exposures at my site show greater noise in the sky background than the DSW image. In part that is due to the enormous difference in integration times. However, my aperture is twice the size of the FSQ 106, so I gain a factor of 4x right there. That means my effective integration time is just half that of the DSW image. But I sure see a lot more than a square root of 2 in noise ratio between the two panels.

[ ...turns out I'm wrong about that. The top panel shows a noise level only 18% worse than the bottom panel. Not even a square root of 2 ratio. Looks can be deceiving at such a close up zoom.

.. if measure the MAD in a region containing only background sky, no stars, no nebulosity, then my own image has about 2.5 times the noise level in the DSW image. That's more like it. We can account for a factor of 1.4 from the difference in integration time, which leaves another factor of 1.8 due to other causes.

A sky background difference of 2 mag ought to provide a factor of 2.5. So now why only a ratio of 1.8? Perhaps there are optical efficiency factors involved, which makes their 4 hours of integration something less than that - by a square root of 2 factor in noise? which would wipe out my 2x exposure factor?

... could it be? that the only real difference between us is in terms of seeing quality? that they give up in optical efficiency what gain they had over my site in sky darkness? Nah.... eh? No way... they still have a 1.8x noise advantage over me. Note that this has nothing to do with star image quality, just background noise levels in the images.

If they have a 50% efficiency compared to my C8 Hyperstar, then their 4 hours of exposure is the same as my 30 mins, but their 2 mag / arcsec^2 darker sky still gives them a noise advantage of 2.5x. So we are still gaining from sky darkness, not to mention improved seeing condidtions. ... I mean... after all, just look at their fine image! It is a definite improvement over my backyard site.

Darker sky = lower sky background noise in the images = fainter limiting magnitudes & nebular features.

Better seeing & collimation/focus = prettier stars and sharper nebulae. ]

A good part of that difference must have to do with sky background noise in the B channel. My own site is an Orange-Yellow zone, with the past few Moonless nights showing an average deep darkness of 20 mag / arcsec^2. The DSW site has been showing 22 mag / arcsec^2. It is located up in the mountains of New Mexico, some distance away from Santa Fe. No question it is a darker site.

Thirdly, I have to wonder about seeing... My own seeing conditions are probably as bad as 3-4 arcsec, possibly worse. I don't know the DSW seeing conditions, but if the images are any indication, it probably hovers around 2 arcsec or better. I think that is why my stars are blobbier than the DSW stars.

[ umm... my collimation, or lack thereof, could be a bit of the problem here too. Nothing like a good solid refractor from Takehashi to make that a non-issue for their users. C8 users, and members of their tribe, have to reckon with this. Especially touchy on a fast F2 Hyperstar system. ]

Interestingly, when I compare full frame RGB images of the Flaming Star nebula from my own system, with that generated by the DSW system, apart from obvious quality improvements due to all the aforementioned effects, both images show a ton of highly distracting field stars in front of the red nebulosity. A dark site and better observing conditions apparently can't remove that detracting feature. So both images benefit from the star shrinkage trick.

What I found was that taking a Lightness channel extracted from the RGB images, and eroding stars in that B&W image with a morphological filter, then using the result as an L-channel modification of the RGB image -- whew! -- that allows the RGB image to greatly diminish the field stars without changing their colors. Doing the morphological erosion directly on the RGB image would have turned the stars to a softer gray color, and leave an unnatural looking matte.