Gain and Offset

Hi Bill,

You might not have gotten a reply because this is posted in the General forum.  A better fit is the Instrumentation and Equipment forum.

You typically optimize for maximum dynamic range when doing photometry, as you want to be able to measure bright and faint stars simultaneously on an image.  Most vendors give a dynamic range vs. gain setting graph to help you in this process.  For the IMX571 sensor that is used in the ASI 2600 camera, there are two peaks in that curve:  gain of 0, which gives the greatest full well depth, and gain of 100, which gives a significantly smaller full well depth, but much better readnoise.  The reduced readnoise is a dramatic, sharp drop at gain=100.  At gain=99, the readnoise is 2.49e-; at gain=100, the readnoise is 1.11e-.

For photometry with wide-band filters like Sloan or Johnson/Cousins, you are usually sky-noise dominated with exposures of a few minutes.  So readnoise is less of an issue.  For photometry, using a gain setting of 0 also lets you take longer exposures on bright stars without saturating, which in turn lessens the effect of scintillation.  So many observers use the gain=0 setting.

For a recent observatory that I'm helping to set up, the focal length is so long that we bin 2x2 for normal imaging.  This increases the full well depth, since you are combining 4 pixels, but also increases the read noise by sqrt(4) = 2x.  In this case, we are using gain=100 to reduce the read noise, and the full well becomes about the same as a single native pixel, so the end result is a camera that works like the native ASI 2600 (IMX-571), but with 4x fewer, larger, pixels.  The fewer pixels also helps in reducing processing time and storage.

So the answer is that either of the two settings work.  My preference, if binning, is to use gain=100.  My preference, if not binning, is to use gain=0.

I have the full calibration set for the ASI 2600 and will post a review towards the end of the month in the Instrumentation and Equipment forum.

Arne