FTS data processing lessons from postvibration CQM testing

1
FTS data processing lessons from post-vibration
CQM testing
2
Funny interferograms

• Some interferograms from CQM were messed up
  (left) or warped (right)

3
Messed up interferograms

• Checks can be implemented in the Java pipeline
  once the first interferograms have been measured
  in-flight and the kind of noise that occurs can
  be qualified better.
• For now Build a place-holder into the pipeline.

4
Warped interferograms

• Dark pixels should be useful to indicate drifts
  of the baseline.
• The signal from the dark pixels will be divided
  out from measurements. How can/must this be
  calibrated to correctly measure the flux?
• Should we attempt to remove any other drifts in
  interferograms - through fitting polynomials of
  1st/2nd/ order?

5
Identifying glitches in interferograms
6
Deglitching interferograms

• Replace several points around the skewness
  outlier with the average of the other
  interferograms
• Impact on averaged spectra?
• Does this algorithm flag too many glitches?
  Criterion for glitch identification!?