This problem happened with standards (during the year2 reduction) and ASASSN-14lp. When the observed (raw) spectrum goes into PESSTOFASTSPEC or PESSTOEFOSC2DSPEC, the output 2D image result corrupted (see below). However, the raw is not corrupted and you can process and extract the spectrum with iraf. It looks that there is no correlation with pixels position of the corrupted pixels (it doesn't happen at a precise position) or brightness (one of the standard in which we got the problem was fine the day after with same exposure time).
This problem happened with standards (during the year2 reduction) and ASASSN-14lp. When the observed (raw) spectrum goes into PESSTOFASTSPEC or PESSTOEFOSC2DSPEC, the output 2D image result corrupted (see below). However, the raw is not corrupted and you can process and extract the spectrum with iraf. It looks that there is no correlation with pixels position of the corrupted pixels (it doesn't happen at a precise position) or brightness (one of the standard in which we got the problem was fine the day after with same exposure time).