Currently the pipeline uses the average airmass of the first frame as airmass for the products of a sequence of frames (e.g. NODDING or OFFSET data).

top Flux calibration of UVB data

The combination of two different flat fields (D2 and QTH) causes a discontinuity in the response curve which results in a small peak in the flux calibrated data at about 360nm (see plot below).

top Efficiency determination

In some cases the determination of the efficiency and response curve fails although the data are good. In addition there are sometimes spurious efficiency values for some orders (see plots below, between 360nm and 400nm [VIS, left] and between 1050nm and 1200nm [NIR, right]). This problem is currently under investigation (2011-05-24).

top Object at slit edge in STARE mode

Automatic processing does not handle well objects at the slit edges. If the object is clearly visible the user may try manual processing with the option --localize-method=AUTO and a carefully chosen value for --localize-slit-height (half height of the extraction slit). A good starting point is 0.25-0.33, which means that 0.5-0.67 of the slit length are sued for object extraction, while the remaining part may be used for sky background determination.

top Order merging

There are sometimes problems with the merging of orders, which may cause jumps in the flux at order overlap regions. PIs are therefore strongly encouraged to check their processed merged data carefully before analysing them. This problem rarely appears anymore (2011-03-02).

top Background correction

The interorder background may be overestimated for the rightmost order of VIS data, as there is very little space between the orders (2009-11-18).

top Sky background correction in STARE data

The sky background correction in STARE produces arrow-like features in the sky region (irrespective of MEDIAN or BSPLINE method) for high S/N data like telluric standard stars (2009-11-12). Below you find NIR examples in echelle format and in the rectified merged format. Maskin the slit edges with the parameter sky-slit-edges-mask (default 1.5) can help. This problem has been solved with pipeline version 1.1.6.

top flux calibration and response function

Several limitiations of the pipeline concerning the data reduction of flux standard stars and the derivation of response functions in SLIT mode have been solved and implemented in pipeline version 2.0. This version is in operation since 2013-06. Details of the improvements are given in Moehler, S. 2014, A&A 568, 9.