I am looking for a camera and would like to know the critical specifications to look for. I know well depth is one measure, but should I be looking more at (well depth)/(read noise)? I another words, if I get lower noise can I tolerate a small well depth?
For instance: The QSI-660 cameras have a well depth of 20,000 and a read noise of 5e- for a S/N of 4,000. It is a fast camera with very low noise, so you can probably accumulate a lower noise photo over the same time sequence as from a large well CCD. Does this logic apply to photometry, or just to pretty picture taking? I realize the pixel size of this Sony CCD is small compared to my typical night sky seeing, however I can bin2 and get 9um pixels which will be a better match and I still get a good image size.
The KAF-8300 (QSI-683) has simular well depth but is much noisier. On the other side of the coin the KAF-1603ME (QSI-616), has a large well depth, good S/N, but suffers from smaller image size.
I have looked at SBIG, but QSI seems to be using more up to date CCDs, is of high quality, and at a better price point.
Any advice is greatly appreciated.
Frank
Hi Frank,
Like you, I think QSI makes some excellent cameras.
Your logic is sorta right. There are some caveats. Lower well depth means that you have to stack multiple flatfield images to reach the magic "one million electrons" necessary to give you S/N=1000 per pixel, which is the value to shoot for to keep the flat's noise from significantly contributing to your science frame. This also highlights one of the other reasons for deeper well depth: you want to have as much dynamic range as possible so that you can measure bright and faint objects on the same frame without saturation. Looking at S/N per pixel is fine, but remember that the noise value you quote is only the readnoise, and for most wide-band filters, you will usually be dominated by sky noise and not read noise. Finally, pixel size is important. If you are buying a camera, then you have the option of selecting the pixel size to properly match your seeing conditions. Using small pixels means binning, as you suggest, and I strongly prefer to not bin if possible, as binning adds another layer of complication (what happens if one of the 4 pixels that are binned together is saturated, and the others are not, for example). The usual rule of thumb is to get 2 pixels per typical seeing fwhm. You can go either side of that, of course, and still do good photometry.
There are lots of other considerations, such as readout time, QE response (the Sony chips tend to be poorer in the red than the KAF chips), camera size and weight, camera cooling, how guiding will be done, what kind of shutter is involved, etc. In general, ANY commercial camera will give you good photometry, so you can't go wrong. Good luck!
Arne
I am a little surprised at the current offering of QSI cameras. I'm trying to work myself up to buying a 6303 camera so I've been watching the QSI site on and off for some months now waiting for the 700 series cameras (which was supposed to include a 6303 version.) It looks like the 700 series has disappeared and many of their cameras are now Sony chips. While the cameras seem to be excellent I don't think the Sony chips are ideal for photometry.
Rick Wagner
WCR
Hi Richard,
It is a little strange that QSI does not offer a 6303 camera. I can only guess that there is a parts issue, or a size issue (I think it is physically bigger than any other sensor they support). Sonoita uses the STL-6303 camera, which works well (and you might find a used one). The only drawback with the 6303 is that it does not have as good a red response as some of the other KAF chips.
The Sony chips are pretty good, with low noise and now reasonable QE. I've seen a lot of good photometry from them, especially from Europe/UK. The limiting factor to me is the smaller physical pixel size.
QHY makes a nice looking camera; I'm working on getting one of their cameras for testing.
Arne
Thank you for your replies. From what I gather the physical size of the SOny CCDs equates to offering the same FOV as the KAF chips of the same size, but with higher resolution. This makes sense in the commercial camera market where small size is a virtue,and lenses support the resolution gain. However the same FOV in a C11 with a high resolution CCD will mean nothing as my seeing conditions will not support it. Thus a 9 um pixel CCD (with a 0.7 arc-sec resolution) is better suited, even though it comes with a lower resolution of 1.6 MP. Also the 100,000 e- well depth will give me the best dynamic range.
The QSI-616 seems like the best fit for the C11 at this time.
I would guess that if I went with a shorter focal length telescope, a F/2.2 like the Rowe-Ackermann Schmidt Astrograph, the discussion would fall more toward the Sony CCDs.
Thank you again for your insights.
Frank