Quote:
Originally Posted by PRejto
Hi Ray!
I really appreciate your reply to what is now a slightly old post. "Old" only in the sense that I have been reading quite a lot and have other cameras to consider, such as the cameras with the KAI04022 CCD, and perhaps a ICX694 chip. I'm sure that last camera would raise eyebrows; I only mention it because over at CN there have been some big discussions re the "myth" about resolution vs FL if one images with a great mount and minimal required guiding meaning terrific polar alignment. I saw some images with the ICX694 taken a f10 (2000mm fl) that looked pretty decent to me. I think the argument runs along the lines that the CCD is much more sensitive thus allowing quicker sub capture which might minimise star bloating. Well, that's the theory anyway. I would appreciate some additional points of view re this logic, or lack thereof... I think the ICX694 would be a very good match for my TEC140, but then so is my KAF8300, so I wouldn't bother unless it was superior at longer FL.
Anyway, I would appreciate an opinion of the KAI04022. I read that it isn't very sensitive to red and Ha, but I've seen some beautiful images with this chip especially those by Dietmar Hager: http://www.iceinspace.com.au/41-465-0-0-1-0.html and http://www.stargazer-observatory.com/bilder/
Re the 8300, no, I have not tried it yet. I need adapters and a guiding solution and I just don't want to invest, find out it doesn't work so well, and start over. Thought I do some serious asking first. Re binning, If I read your comments correctly, binning at 2x2, or software binning to 2x2 would give similar results assuming the camera takes care of 2x2 binning correctly (or better results if the camera doesn't handle binning properly). But, wouldn't capturing at 2x2 be a lot quicker and allow shorter exposures? I wasn't aware that software binning could increase SN ratio. Can you please explain how that is done? I assume it isn't the same thing as just reducing image size 50% in photoshop...or is it?
EDIT: I found some huge posts on the subject of software binning at CN....I think I know enough now that no further comments are necessary. Like everything in this lovely hobby the answers are varied, complex, not exactly clear, and full of conditions. So what else is new?
Sorry for so many questions! Many thanks,
Peter |
Hi Peter
It sure is a can of worms. I have been trying to approach the subject of camera choice from the perspective of matching the pixel size to the expected FWHM of stars produced by a seeing limited scope (which yours will be). You can still produce very nice images if the pixel matching is not right, but you will either get quick results with some loss of detail (if undersampled) or spend much more time imaging (if oversampled) with no extra detail. To get close to optimum sampling in Australian seeing conditions and your scope, you would require a sensor with pixels up around 10microns or so. Your only option is the Kodak line of sensors and the specs of the 4022 are typical of their products. The 694 is not well matched to the image scale produced by your scope and you would probably need binning to get the best out of it.
However (always a qualification!!!), the guys on CN who have used the 694 are reporting that it outguns the Kodak chips even when undersampling. These claims are not surprising, since the 694 has a very much higher QE and much lower read noise than the 4022. It was interesting to read the comment from one of the camera makers that in essence said
why worry about low QE, you just have to image longer.
And I guess that sums it up. You could choose either chip and be happy with it. The Kodak chip will give you a much bigger field of view and will produce images with detail that is limited only by the seeing. The 694 will also be seeing limited, but will give you an oversampled system with smallish field of view. However, the 694 is so efficient that you could need very much shorter imaging times in good conditions.
The much greater well depth of the Kodak chips is often noted as a benefit over the 694, but the dynamic ranges of the two chips are about the same, so I do not think that this parameter is an issue - as you noted, the CN discussion suggests that the shorter 694 subs can compensate for the lower well depth - that makes sense.
in summary, on your system
- Ultimate image quality is fairly likely to be about the same for both chips
- the 4022 has a much bigger field of view than the 694
- with binning, the 694 would produce 1.5mp images. The 4022 without needing binning would produce 4mp images. if you get exceptional seeing conditions, the 694 could produce 6mp images.
- the 694 should be able to get images in much less time than the 4022.
Please bear in mind that this is based only on untested theoretical considerations - I haven't used either chip and there might well be some important aspect of either that I have not included in the analysis. However, the musings of others (who have used them) on other forums show the same trends. I will add sky brightness to my model - would expect extraneous light to bring the 694 back closer to the 4022 and will post results in the next couple of days if that would help. sky brightness should not affect anything much if you are doing narrow band imaging though. Need also to find out more about RBI on the Kodak chips (not sure if it affects the 4022
edit: it doesn't) and charge leakage, which could possibly affect the small 694 pixels more than the Kodak ones (haven't found any data on this yet).
It's not an easy choice - suggest that you could probably be happy with either chip - you just have to decide what is more important to you:
- good images from a significantly larger field of view plus well established technology
- good images from significantly shorter imaging times with relatively untried technology.
Good luck.
regards ray