Data Reduction Frequently Asked Questions
As recommended in the pipeline manual I used reduce.ffmethod = pixel for 860 setup data. Now there is a jump in flux between the flux-calibrated REDU and REDL data. Why?
The parameter ffmethod has to be the same for the response and the science data. The MASTER_RESPONSE curves were created using reduce.ffmethod = extract. The inconsistency creates the jump. You can avoid this problem by using a response created from a flux standard star with reduce.ffmethod = pixel.