the fake/force analysis step needs to be finished. Paul has made some progress, but more work is needed. This is still a NO-OP in the production code. The equivalent efficiency analysis needs to be implemented for stack and diff processing as well.

The synthetic grizy DVO database used for astrometric and photometric reference has two known errors:

  • the astrometry is saved as floats, with the resulting precision truncation
  • the photometry has surprising trends with magnitude (and unsurprising trends with color)

the astrometry can be fixed by replacing with values from 2MASS where possible (using, eg, catdir.2mass as the input source). the photometry may represent astrophysics, in which case it may vary as a function of position on the sky. At best an average fix could be applied. More study is needed (Eddie Schlafly & Doug Finkbeiner). A correction could be made to the data or on the fly when the calibration is performed.

We need to build y-band fringe masters for GPC1 / Run 3 data. This will probably involve some debugging of the fringe-creation system, which has not been tested in some time.

Note: See TracRoadmap for help on using the roadmap.