We suggest the user to execute the data reduction recipes using parameter
defaults and all the reference and master calibrations indicated in
this manual.
The following is a list of currently-known problems with UVES recipes, and
workarounds, if available:
Due to the difference in light paths between the flat-field lamp
and the observed target the blaze function and interference fringe
patterns are not fully corrected. At sufficiently high S/N this leaves
residual patterns in the final spectrum at the scale of one order
and at the scale of the inteference patterns.
When flat-fielding in pixel-pixel space (non-default option),
the statistical error bars from the flat-field are not
propagated properly in the optimal extraction. If the object S/N is very high
(comparable to the flat-field S/N) this causes a systematic bias of
the final spectrum error bars. For this reason the default value
reduce.ffmethod=extract
is recommended.
Line tilt correction (correcting for the fact that the
dispersion relation depends on slit position) is not supported for
2d spectrum extractions, only for average/linear/optimal extractions.
The recipe uves_obs_scired
reduces only the first provided
science frame. To reduce several science frames, the recipe should
be run several times.
![[ ESO ]](http://www.eso.org/images/eso-logo.gif){kind=logo}