From Stefano (25th Feb 2016) :
In attachment you see the sequence of PESSTO spectra for OGLE-2013-SN-100. (OGLE_Antonia.png) reduced with the PESSTO pipeline. What attired immediately my attention was the unusual poor S/N ratio in the red part of the spectra in comparison with the blue one. This is mostly due to a poor fringing correction. I then decide to do a test and reducing one spectrum by hand. Since the result was much better, I reduced the complete sequence (see OGLE-2013-SN-100.png in attachment). I reached this result mostly because I used a narrow window around the SN signal. In this way the fringing correction is usually much better. Of course, you cannot use this trick with the pipeline because you need to fully calibrate the complete 2d spectrum. But people should keep in mind that, especially for faint spectra, sometimes an ad hoc reduction is better than the pipeline one.
Standard reductions :
Specifically tailored reductions with windowing around the target