Producing a CCD Image: Corrections, Calibration, Subexposures

[ian2012]

I am aware that for a CCD image to be produced, various corrections have to be applied due to the limitations of CCDs in astronomy.
One such correction is the fact that CCDs are not perfectly linear since each detector pixel has a maximum size - it can only collect so many electrons. So sources brighter than a certain number of counts/pixel will have unreliable fluxes (overfilled pixels will bloom - overflowing).

What is of interest to me is: there are various additional processing stages for a CCD image: calibration, subexposures, etc. Could someone explain these stages, other instrumental effects that need to be corrected before a final CCD can be produced? Why the need for numerous subexposures?

Thanks.

 

[S.Vasojevic]

That is a general question aimed at vast field. You should be more specific about what kind of image or measuring you are trying to produce.
Anyway, what all light detectors have in common, is that they can't handle arbitrary wide range of luminosities, and produce linear output. Curve that plots output in relation to input is consisting of three parts. Bottom part is called 'heel', and represents part when sensor or film is beginning to respond to light input, and it is not linear. Next part is straight line which is at slope of 45 deg (if gamma is 1), which is linear in response, if you neglect noise. Then comes the 'shoulder', again non linear part, which represents part where sensor is approaching saturation.
In order to get properly exposed image you should put whole scene on that linear part. That is the problem when you are dealing with wide range of luminosities. There are a few techniques that can be utilised to achieve this, without significant loss of details, or information. One of them is gamma manipulation. You take your image with lower gamma setting (<1), and afterwards you extrapolate it with gamma > 1. This is good technique, but it has limitations. Others may include multiexposures which are later combined into one image.
If you are interested in more of this stuff, field that goes in depth is called densitometry, which refers to optical density of film, and similar mediums, but is equally usable in case of electronic detectors.

 

[ian2012]

Thanks for the response.

To be specific, I am interested in an optical CCD image of the extragalactic sky. I am analyzing an image taken at the Kitt Peak 4m telescope using the CCD mosaic camera, using a Sloan r band filter (central wavelength 620nm). The image was taken as part of the optical observations supporting the SWIRE infrared survey using Spitzer.

The image I am dealing with, however, has already been fully reduced and ready for scientific analysis. I was just wondering what instrumental effects need to be corrected before this final image was produced (& subexposures?).