Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-8889

Add height of CCDs to camera geometry specification

    XMLWordPrintable

    Details

    • Type: Story
    • Status: Done
    • Resolution: Done
    • Fix Version/s: None
    • Component/s: afw
    • Labels:
      None
    • Story Points:
      6
    • Team:
      Data Release Production
    • Urgent?:
      No

      Description

      For some inexplicable reason I failed to include the departure of flatness of the focal plane in the camera geometry. We are going to need this for some approaches to PSF modelling, and certainly for studies of the LSST camera during commissioning. In the short term, we need it to map out the HSC focal plane from focus sweep data (although that is no concern on the LSST project, it is relevant to studies of PSF estimation using HSC).

      I propose that we call the height of the centre of the detector (defined as the point returned by Detector.getCenter()) Detector.piston and that it be measured in mm for consistency with Detector.pixelSize. If Detectors are flat this is sufficient as the Orientation is sufficient to define tilted devices, but we might want to future-proof cameraGeom by adding a representation of warped devices at the same time as this ticket.

      Looking at the code to construct cameras it isn't obvious to me how to add a new field like this to cameraGeom (even with a default value of 0.0) in a non-intrusive way. If it is indeed impossible, the scope of this ticket must include suitable modifications of all obs packages (for some suitable value of "all": lsstSim, hsc, cfht, sdss, decam?)

        Attachments

          Issue Links

            Activity

            Hide
            rhl Robert Lupton added a comment - - edited

            The HSC flatness map looks like

            Show
            rhl Robert Lupton added a comment - - edited The HSC flatness map looks like
            Hide
            rhl Robert Lupton added a comment -

            Simon added his slightly-different proposal after I filed this. They should be merged and one deleted.

            Show
            rhl Robert Lupton added a comment - Simon added his slightly-different proposal after I filed this. They should be merged and one deleted.
            Hide
            swinbank John Swinbank added a comment -

            Simon Krughoff's comment from DM-8943:

            We need to add a piston to the orientation objects. This could be done by storing a Point3D instead of a Point2D in the Orientation object, or by adding a piston attribute. The former is likely cleaner than the latter.

            Show
            swinbank John Swinbank added a comment - Simon Krughoff 's comment from DM-8943 : We need to add a piston to the orientation objects. This could be done by storing a Point3D instead of a Point2D in the Orientation object, or by adding a piston attribute. The former is likely cleaner than the latter.
            Hide
            swinbank John Swinbank added a comment -

            Reviewed for DM-CCB, 2020-05-26. This request has not been satisfied. It also hasn't been identified as a blocker for any ongoing work.

            Show
            swinbank John Swinbank added a comment - Reviewed for DM-CCB, 2020-05-26. This request has not been satisfied. It also hasn't been identified as a blocker for any ongoing work.
            Hide
            rhl Robert Lupton added a comment -

            Its information that Joshua Meyers needs for the focal plane model in the PSF.  If the chips are flat enough, this is all we need (as opposed to a full focal plane map).

            Show
            rhl Robert Lupton added a comment - Its information that Joshua Meyers needs for the focal plane model in the PSF.  If the chips are flat enough, this is all we need (as opposed to a full focal plane map).
            Hide
            jmeyers314 Joshua Meyers added a comment - - edited

            I think this is ready now.  I initially made some experiments that required modifying cbp and obs_cfht (and would have required additional modifications of obs_* until I backtracked), but in the end I think I have this working without so many modifications.  Those experimental branches are renamed tickets/DM-8889  =>  u/jemeyers/DM-8889, and don't need to be reviewed; I'll delete them once this is merged.  Since Jira hasn't figured this out, actual PR links are below:

            https://github.com/lsst/afw/pull/649

            https://github.com/lsst/obs_base/pull/427 

            https://github.com/lsst/obs_lsst/pull/416

            Show
            jmeyers314 Joshua Meyers added a comment - - edited I think this is ready now.  I initially made some experiments that required modifying cbp and obs_cfht (and would have required additional modifications of obs_* until I backtracked), but in the end I think I have this working without so many modifications.  Those experimental branches are renamed tickets/ DM-8889   =>  u/jemeyers/ DM-8889 , and don't need to be reviewed; I'll delete them once this is merged.  Since Jira hasn't figured this out, actual PR links are below: https://github.com/lsst/afw/pull/649 https://github.com/lsst/obs_base/pull/427   https://github.com/lsst/obs_lsst/pull/416
            Hide
            rhl Robert Lupton added a comment -

            Looks good.

            Show
            rhl Robert Lupton added a comment - Looks good.

              People

              Assignee:
              jmeyers314 Joshua Meyers
              Reporter:
              rhl Robert Lupton
              Reviewers:
              Robert Lupton
              Watchers:
              Chris Walter, John Parejko, John Swinbank, Joshua Meyers, Robert Lupton, Russell Owen, Simon Krughoff
              Votes:
              0 Vote for this issue
              Watchers:
              7 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:

                  Jenkins

                  No builds found.