|
At the current state of the art, I can say with some confidence CMOS (regardless of manufacturer) do not surpass CCD sensors in scientific and extremely faint object imaging.
Pixel to pixel variations inherent in the CMOS readout architecture is the problem and only seems to be tamed by statistical noise reduction techniques (e.g. dithering sub exposures).
Binning CMOS pixels is also not the same as binning CCD pixels, where the former is effectively a post-readout combination as opposed to an effectively real/larger pixel size at the capture stage with CCD’s.
But the reality for many is, they will never perform millimagnitude photometry or spectroscopy or image 23rd magnitude objects.
CCD’s will not disappeared either, but it is likely they will become very expensive as the number of fabrication lines making them dries up.
Technologies such as emCCD (read single photo detection) while eye-wateringly expensive at present, may trickle down into the amateur sphere in the next decade.
This leaves us with CMOS. My main gripe with made in ROC cameras is the lack of build quality. QHY are likely the best of the bunch
(they also use industry grade, rather than consumer grade sensors in their larger cameras) but even they lack the precision (i.e have sloppy fittings) and elegance (70 fiddly parts needed to install a filter set!!)
of the engineering that you get long established manufacturers such as SBIG/FLI . But they can deliver great results and won’t break the bank.
Compared to the diminutive Texas instruments Ti211 192x164 pixel sensor that I took my first CCD images with in the early 1990’s,
you are now spoilt for choice, but maybe focus a tad more on specification rather than price.
If a camera won’t do what you want, you’ve wasted you money no matter how cheap it was.
|