Flat fielding is still an open question in SEDM pipeline, and so I make the following assumptions, all need to be tested experimentally.
Tests to perform:
Simple divide image by flat-lamp image. Because of flexure, this approach doesn’t work on all data, but it does work on some (most) data. If the simple divide approach works, then we have to change the daytime calibration step to account for flexure. If test (1) doesn’t work then there are two approaches to the spaxel-level flat that can be tried:
Perform a spaxel-level flat calibration by summing light in a broad wave band.
Spaxel-level flat is a function of wavelength.