Simple outline of imaging data reduction Science image subtract bias subtract {dark-bias} divide by {flatfield-bias} More detailed outline of imaging data reduction Combine individual bias images → masterbias Subtract masterbias from dark and flatfield images Combine individual dark images → masterdark (for a given exposure time) Create masterflat: Normalize the individual flatfield images, using the mean or median flux level Combine them Science image subtract masterbias subtract masterdark, if relevant divide by masterflat Imaging data reduction including 'superflatting' and fringe correction Combine individual bias images → masterbias Subtract masterbias from dark and flatfield images Combine individual dark images → masterdark (for a given exposure time) Create masterflat: Normalize the individual flatfield images, using the mean or median flux level Combine them Science image subtract masterbias subtract masterdark, if relevant divide by masterflat Create superflat: Normalize the individual (non-background-subtracted!) science images, using the mean or median sky background level, ideally masking all sources Combine them → 'superflat' Heavily smooth the superflat → 'illumination correction' Superflat minus illumination correction → fringe pattern Science image divide by illumination correction (! this can be different in the near-infrared, and/or depending on your data !) scale the fringe pattern to the sky background level... ...and subtract it ! Note: Even without the presence of fringing, it can be useful to create a superflat and apply the illumination correction !