Attendees:


Notes:

  • Merlin Fisher-Levine 
    • No major calibration work.
    • Rubin-TV isn't the issue.
  • Christopher Waters 
    • Calibration issue.
    • Parity with eo_pipe.
      • Correlation of overscan => crosstalk signal?
    • Need to make tickets.
    • CVT slides.
    • CTI with new ISR Task?
      • Discussion on Tuesday.
  • Agnès Ferté 
    • Machine learning conference last week.
    • obs_fiberspectrograph
      • Merge just the butler element?
      • Still need to merge the rest of the package.
    • New ISR Task unit tests.
      • Config differences create the majority of image differences.
      • How to handle configuration settings for calibration production?
  • Andrés Alejandro Plazas Malagón 
    • Porting Astier code with larger lag values.
      • Paper is based on HSC.
      • Subtracts long-range covariance from the core.
      • Slack discussion.
      • Testing with Alex Broughton.
      • Large covariance solving doesn't converge in reasonable time.
    • Additional electrostatic fit ticket that is also in testing.
  • Eli Rykoff 
    • Overscan refactoring for new ISR.  DM-41316.
      • Splits serial and parallel overscan into two tasks.
      • Fixes unit tests.
      • Current parallel overscan doesn't work with transposed detectors.
      • Parallel overscan masking is being phased out as well.
      • New code inherits/shares code with old code.
    • Masking:
      • Hot column bleeds into the edge by the parallel overscan region.
      • Defect detection ignored edges, so these were not being flagged as defects.
      • Cold defects from flats might catch edge roll off, but hot defects from dark frames should never be excluded from any region.
      • This ties into the calibration issue that will hopefully get resolved this week.
      • Current defects will include pixels with e-field issues, and when that is fixed, we will redo the defects.
    • Fix defects.
      • Separate ticket to change the configuration options?