Back when I started CCD Photometry about 20 years ago, my camera was the 20u pixel ST6 that was eventually replaced with the 20u ST9xe that I still use today. This camera is a perfect match for my f/10 Meade 12 inch Classic. Nowadays the available cameras have much smaller pixels (9u and lower) that will not conform to the Nyquist sampling when used with the longer f ratio scopes. I see one very expensive 20u camera that is way out of my price range, but nothing else. Binning is not recommended in photometry, so how can the present day small pixel cameras be used with F/10 scopes and still conform to Nyquist sampling. Is there a way to other than using focal reducers (also not recommended for photometry) to allow the small pixel cameras to be used with long focal length scopes to achieve good photometry? I do know that the newer cameras are more sensitive with far less noise than the older workhorses like the ST9xe, but how do they work out for the long fl scopes? Also since images could be well oversampled, how might this affect the accuracy of photometry?
IIRC, differential photometry requires the star light to be collected in an area between 8 to 30 pixels across and, in order to get that distribution, the telescope/lens would have to be defocused. A CCD camera with a resolution of 512x512 pixels each 20µm across would have a rather small field of view to work with, especially when used in a C14 SCT plus having to be defocused until the target star was at least 8 pixels across. From everything I was told about photometry, it seems like that combo would be the most difficult to work with.
I have used 2 X 2 binning and a focal reducer together to get the right sampling using a 9 micron pixel camera with a C14 to do photometry. I think the improved SNR you get with proper sampling outweighs whatever downside there may be with binning and/or a reducer.
Of course, there has to be some light light loss with the reducer, and a little bit of extra noise from binning, but if you have a 3.9 meter FL telescope and a 9 micron camera I think this is a small price to pay for a larger FOV and optimized sampling.
I'm now using a camera with 25 micron pixels which gives almost the same image scale (1.3 vs 1.4) without using binning or a reducer. The only difference I see in the (raw) images is more vignetting in the images done with a reducer. If you take care to make good flats this will calibrate out. I don't see much difference in the calibrated images or the photometry
I am 99.9% using my CCD in 3x3 binning on a 40cm f/6.8 scope with FL of 2.7m. The original pixel size is 9 microns and hence 3x3 is 27 microns. You can see the results and how hey compare to any other observer in the AAVSO database under HMB.
I do not see the reason not to bin. I am also looking on the new cheap cooled CMOS cameras with Megapixels but small pixels. MOst of the time the software does not allow more than 2x2 binning.
So file size could be an issue and possible off line binning is a need. I would go for it and try it out. You have to look also outside the box.
Thanks for your replies. I am reaching back about 15 years when I recall reading the reason for not binning when doing photometry. As I understood it, the binning with these older cameras does not actually occur until the camera downloads the image. What this means for a 9u pixel camera with 2x2 binning is that although the pixel size is now 18u, the full well capacity does not increase proportionately. So the actual exposure is done on 9u pixels that get binned during the download process. Therefore it is possible for the 9u pixels to saturate during exposure. The potential result when downloaded are 18u pixels each consisting of 4 saturated 9u pixels.
Now, that said, I did some research recently that claims the full well capacity actually is increased when pixels are binned. So I am wondering if the newer cameras with smaller pixels are engineered so the full well capacity actually is increased proportionate to the binning. Since modern cameras with smaller pixels are pretty much the only ones available and are in use by many, I was wondering if this issue with older cameras has been somehow addressed by camera manufacturers so that binned pixels truly become and behave as one larger pixel with increased capacity.
When you say "older cameras" are to referring to the ST-6? Perhaps there was a prohibitive problem with that very early camera.
For the SBIG ST-7 series of cameras (ST-7, 8, 9 ,10, etc.), I think 2X2 binning generally works well. Arne Henden discusses the issue here: