
2014.07.03

NOTE: 2 questions on these measurements: chisq and ubercal dist?

I think I have worked out all of the issues in the mag / flux
calibrations for stacks and forced warps.  It took me a few passes to
get the right combinations of the 3.56 or the 8.9 offset to the mags.
But now, the stack psf mags, warp psf mags, stack flux, warp flux, and
mean psf mags all agree at the 0.1 mmag level.

I still have some concerns regarding the flags and astrometry.  I am
not set on my laptop to test the primary cell stuff (well, I could
probably reproduce it, but easier to wait until I'm home).  I need to
review the 'waterfall' and check that the choices make sense (and
write it down for review).

2014.07.02

I am trying to run a (rather artificial) test of relphot
specifically, I have taken a single smf and an associated warp cmf and
modified the headers so I can load them in DVO and run relphot.  Some
issues I've hit:

* needed to specify PHOTCODE for the stack and fforce fake cmfs
* needed to modify IMAGEID and MJD-OBS so the chips associated
  correctly with different gpc1 images
* modified relphot/MosaicOps.c to explicitly skip stack and warp by
  photcode when matching chips into mosaics (kind of a hack as it
  won't work with other cameras)

2014.07.01

thinking about the issues with relphot for the new PV2 / PV3 schema.
I now have 3 separate sets of mean photometry for each of the stages
cam/staticsky/warp.  I need to make sure I am handling everthing
needed.

Calibration sequence:

* relphot_images / relphot_parallel_images : calculate the zero points

Loop around with the standard sequence, but only fitting for the mean
camera mags and camera zero points.   At the end of the iteration, fit
the stacks and warps to the new mean object mags.  

Some things I've started to worry about:

1) QSOs & other variables.  Am I detecting and avoiding these well
enough?

2) what is the appropriate ubercal weight (10, 100, 1000?)

