Calibrating ALIS

``Gör vad du kan -- och uppgiften skall vila lätt i din hand, så lätt att du med förväntan sträcker dig mot det svårare prov som kan följa.

Det är när morgonglansen bytts i middagströtthet, när benmusklerna skälva vid anspänningen, vägen tycks oändlig och plötsligt ingenting vill gå riktigt som du önskar -- det är då du inte får tveka.''

Calibration is the process of answering the following two basic questions:

1. What physical value does the pixel represent?
2. How is each pixel mapped to the observed object?

Answering these two rather simple questions is often no simple task.

In the case of a classical photometer, there is only one detector with a relatively narrow field-of-view. For an ALIS imager there are about $10^{6}$ individual detectors (pixels) covering a moderate field-of-view ( $50^{\circ}$- $60^{\circ}$). Calibrating image data is therefore a considerably more complex issue compared to calibrating a photometer.

An ALIS Imager produces two-dimensional images of size $m \times n$ pixels. Each pixel is represented by a signed 16-bit integer value ($-32768$-$32767$). This value is often referred to as digital output, $\mathit{DN}_{ij}$. The indices are pixel columns, $i$, and pixel rows, $j$, These go from the first to the last pixel in each direction (Figure 4.1).

Figure 4.1: Coordinates and notation for the CCD in the ALIS images. The camera control unit reads out a line from each quadrant in parallel (along the dotted arrows 0-3). The pixels are then sent to the host computer, starting with quadrant 0. As this is finished, the next line is immediately read from the CCD. The host computer then sorts the pixels to form an image as indicated by the X and Y-axes in the figure. The approximate quarters (N, S, E, W) are only applicable to the fixed ALIS stations (i.e. not the bus). The quadrants are sometimes denoted 0-3 and sometimes A-D. The X-direction is also referred to as column or NAXIS1 and the Y-direction is sometimes referred to as row, line or NAXIS2. The shaded areas at the edges of the CCD indicate the bias-pixels, which are shaded from light. These are also called overscan-strip or reference-pixels. The optical field-of-view, $\mathit{FoV_{o}}$, is indicated with a grey circle and is measured diagonally across the CCD. The illuminated CCD (white) defines the field-of-view in x-direction, $\mathit{FoV_{x}}$, and the field-of-view in y-direction, $\mathit{FoV_{y}}$, (refer to Table 4.6).

Note that the number of pixels in each direction should be divided by the on-chip binning factors (Section 3.2.1). Supplemental information is contained in the image header (for example azimuth, zenith-angle, integration time, filter, etc.).