next up previous
Next: Astrometric calibrations and co-adition Up: Data reduction Previous: Data reduction

Bias, flat-fielding and fringing correction

The reduction was performed using the task mscred under IRAF[*].

Overscan plus bias/dark removal and normalisation by twilight flats were done CCD by CCD using the standard pipelines of the mscred package. Because of the large field of view and the field corrector, mosaic data are significantly influenced by diffuse light from bright sources in the field, therefore the correction with a super-flat is needed (see Figure 1 for an illustration of these combined effects).

The CCD mosaiced frames in the red part of the spectrum, from the R broad band filter and redward, are affected by a fringing pattern which cannot be simply removed by a normalisation with a super-flat, because the intensity of the fringing patter depends on the average counts and the diffuse light at a particular pointing. A procedure for the optimum fringing removal was developed anew and it is described in Appendix A; the results for an intermediate filter image is shown in Figure 2. The resulting flat-field and fringing corrected images are uniform to better than 1.5 Before astrometric calibration, the background is interpolated and subtracted from each individual frame.

Figure: Section of an image of the CCD No. 6 in the $\lambda 816$-band with fringing (top). The fringing pattern is shown in the middle panel, while the final image, after correction by flatfield and fringing, is shown in the bottom.
\begin{figure}
\vspace{20cm}
\special{psfile=fig2a_alcala.ps vscale=45 hscale=45...
...ig2c_alcala.ps vscale=45 hscale=45 angle=0 voffset=-80 hoffset=-10}
\end{figure}


next up previous
Next: Astrometric calibrations and co-adition Up: Data reduction Previous: Data reduction
Juan Alcala
2002-02-05