Each of the two CCDs is read-down through two amplifiers, and IRAF treats each exposure as if it were four exposures, taking full advantage of the multiextension IRAF format developed for Mosaic. Because there are four different amplifiers in use, each section will have a slightly different gain and slightly different bias level. However, these differences disappear completely in the reduction process (discribed in the Reduction Section ), and one should relax and not worry about it. The charge transfer efficiency is sufficiently similar that we have not seen differences in the point-spread-function between the left side and the right side of a CCD. (Nor have we seen them between the left CCD or and the right CCD; the CCDs are very co-planar.) However, if you look at a dome flat exposure you should not be surprised to see four different levels. This should not normally appear in the object exposures that are automatically displayed through the DCA, as the display has been flat-fielded. The difference in gain between the four amplifiers is small (plus or minus 7%), but enough to be disconcerting to the first-time user.
Because the Mini-Mosaic uses the multiextension format, you must specify which extension you mean in order to use the usual IRAF commands, i.e.,
would plot the middle row (line 2048) of the first 2048 column numbers of the ``left" CCD. Similarly
would plot the right half of the left chip. Even a command like imhead requires you to specify which CCD you are talking about,  or . ( imhead obj007.fits will show you a global header associated with the frame.)
However, you will probably find it considerably more useful to take advantage of special commands that have been written to treat the data as if it were a single image. For instance, although you could display the left side of the right CCD by using display obj007.fits, you would probably find the following much more useful: mscdisplay obj007
This will display all four extensions arranged in the proper geometry; i.e., so that you will see two 2048 X 4096 images with a modest gap between them.
Furthermore, you can now interact with this displayed image using certain tools that will act as if you are indeed looking at a single large image. For instance, the multiextension version of the familiar imexamine is mscexamine. If you run mscexamine you can move the cursor around, pick out a star, and do a radial plot using ``r" just as you would have with a single CCD and imexamine.
The mscexamine task provides some of the most powerful diagnostic quick-look tools within IRAF. If an image that you wish to examine is already displayed in the ximtool window, simply type mscexamine. Wait patiently without moving the mouse and a blinking, round cursor will appear on the display. Place the cursor over a star, and strike the r key, and you will be presented with a radial plot of the star, along with the values of a fit to the stellar profile. The last number displayed is the FWHM in pixels, quite useful for determining the best focus. Other very useful commands include l for making a line plot at the position of the cursor, and c for making a column plot at the position of the cursor. Other useful cursor strokes are shown below.
By default, the size of the image buffer for the display is 2048 X 2048, so that when the DCA automatically displays a new image, or you manually issue a mscdisplay image, you are not seeing individual pixels when you expand on the image tool. You see a larger area, and what you do see is the average of several adjacent pixels. We find this is a good compromise. However, if you really would like to see individual pixels, (or close to it), do a set stdimage=imt4096 in the window from which you are issuing the display commands. You can also use the DCA monitor to change the default display buffer, but changing it will mean it takes longer to display. (To do so, select ``edit" and then ``display params". A GUI screen will appear that allows a change in stdimage for the autodisplay.) Note that there will be some slight ``aliasing" at the default imt2048; your bias frames may appear to have a cross-hatched pattern. Careful examination of your data will reveal this is an aliasing artifact from the display.
If you are familiar with how to reduce CCD data with IRAF, the WIYN Mini-Mosaic will present no challenge. Special versions of flatcombine, zerocombine, ccdlist, and ccdred have all been made in order to deal with the multiextension formats transparently. (Tololo observers familiar with quadproc will be happy to find the higher efficiency of ccdproc.)
If you are unfamiliar with ccdred, you should consult ``A User's Guide to CCD Reductions with IRAF", available on the Web from:
Here are the basic steps you will need to reduce your data.