Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-16802

ISR order-of-operations not uniquely defined

    XMLWordPrintable

    Details

    • Type: Improvement
    • Status: Won't Fix
    • Resolution: Done
    • Fix Version/s: None
    • Component/s: ip_isr
    • Labels:
      None

      Description

      The order of ISR operations is not identical between cameras, and beyond that, it is not clear if any of the orders currently used is completely optimal.  An optimal order should be determined, and the operations rearranged to match.  This will require new calibrations to be constructed for any camera that has a change in the order.

        Attachments

          Issue Links

            Activity

            Hide
            czw Christopher Waters added a comment -

            After the ISR/obs_package unification, the ordering of ISR steps is:

            1. Integer to float conversion
            2. Bad amplifier and SATURATED/SUSPECT pixel masking
            3. Overscan correction
            4. Optional crosstalk correction
            5. CCD Assembly
            6. Bias correction
            7. Variance calculation
            8. Linearization
            9. Crosstalk
            10. Saturation trail widening (for HSC)
            11. Brighter-Fatter correction
            12. Dark
            13. Optional fringe correction
            14. Straylight
            15. Flat/gain normalization
            16. Defect masking and interpolation
            17. Fringe
            18. Bad pixel/camera specific masking
            19. Attach vignette model, transmission curves, and distortion model

            Based on @rhl's writeup on CCD (steps not listed in that talk are surrounded by brackets: [])

            1. [Integer to float conversion]
            2. [Bad amplifier and SATURATED/SUSPECT pixel masking]
            3. Overscan correction
            4. [CCD Assembly]
            5. Bias correction
            6. Crosstalk correction
            7. Linearization
            8. Charge transfer correction (not currently implemented)
            9. Dark
            10. [Saturation trail widening]
            11. Brighter-Fatter correction
            12. Variance calculation
            13. [Straylight]
            14. Flat/gain normalization
            15. [Defect masking and interpolation]
            16. Fringe
            17. [Bad pixel/camera specific masking]

            This updated order removes the optional orders.  Nearly all calibration products will need to be remade to be mathematically correct, although I don't think the differences will be that large (the linearity order shift likely introduces the largest change).

            Show
            czw Christopher Waters added a comment - After the ISR/obs_package unification, the ordering of ISR steps is: Integer to float conversion Bad amplifier and SATURATED/SUSPECT pixel masking Overscan correction Optional crosstalk correction CCD Assembly Bias correction Variance calculation Linearization Crosstalk Saturation trail widening (for HSC) Brighter-Fatter correction Dark Optional fringe correction Straylight Flat/gain normalization Defect masking and interpolation Fringe Bad pixel/camera specific masking Attach vignette model, transmission curves, and distortion model Based on @rhl's writeup on CCD (steps not listed in that talk are surrounded by brackets: []) [Integer to float conversion] [Bad amplifier and SATURATED/SUSPECT pixel masking] Overscan correction [CCD Assembly] Bias correction Crosstalk correction Linearization Charge transfer correction (not currently implemented) Dark [Saturation trail widening] Brighter-Fatter correction Variance calculation [Straylight] Flat/gain normalization [Defect masking and interpolation] Fringe [Bad pixel/camera specific masking] This updated order removes the optional orders.  Nearly all calibration products will need to be remade to be mathematically correct, although I don't think the differences will be that large (the linearity order shift likely introduces the largest change).
            Hide
            czw Christopher Waters added a comment -

            The goal of this ticket was to attempt to ensure DECam and HSC use the same ISR operation order.  As it appears from early data that LSSTcam may want to use a different unique order, I don't think this goal is still likely to be reached.

            Show
            czw Christopher Waters added a comment - The goal of this ticket was to attempt to ensure DECam and HSC use the same ISR operation order.  As it appears from early data that LSSTcam may want to use a different unique order, I don't think this goal is still likely to be reached.

              People

              Assignee:
              czw Christopher Waters
              Reporter:
              czw Christopher Waters
              Watchers:
              Andrés Alejandro Plazas Malagón, Christopher Waters, Meredith Rawls
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:

                  Jenkins Builds

                  No builds found.