[Shiraz (Ray)]

Hi Scott. Hope some of this is helpful.

As Colin points out, if you image for an hour you will collect the same number of photons regardless of gain settings or sub length - you don't get more photons by using higher gain and with the 1600 you actually get less noise at higher gain.

If you use short subs/high gain, you will collect relatively few photons in each sub. When you stack the short subs, you will get back all of the signal, but will have read noise from the larger number of subs (each one contributes read noise). To keep the total read noise manageable, you will need a low-read-noise chip if you want to use short subs - the 1600 is suitable.

If you use high gain with the 1600, you will have limited well depth. However, you will have a lot of full wells to add up in stacking (one for each sub), so you get back the dynamic range over most of the gain region, particularly since the read noise drops with higher gain. With broadband, you could use any gain from 0 to about 150 with appropriate sub lengths and get essentially the same results in the same total time. The big advantage is that you can use very short subs at high gain - makes guiding etc easier. suggested gain/sub table attached, along with graph of total well depth.

With narrowband, the basic rule is always use the longest practical sub length to get the lowest read noise. However, balance that against the need to keep from saturating bright stars, so choose the longest sub that will give you acceptable stars. FWIW, I have found that 5 minutes under moon is OK with an f4 system at gain 100 - under dark sky, I am currently testing 10 minutes at gain 200.

edit: also, the "unity" gain setting is actually very high gain cf other chips - the 1600 behaves much like a normal CCD when it is used at gain 0, so using it at gain 139 is really pushing it into a region that is not reachable by conventional CCDs.