Ingest task to produce intra-camera variation metrics

[fergusL]

Want a task that will generate meta data for variation with groups of exposures

• Intra-Camera vignetting maps

  • mask all sources within the images to produce sky gradient + vignetting map for each image
  • median combine these and then use it to produce residual maps for each camera
  • residual images can be fitted with a simple plane and the gradient, orientation, min and max are recorded as meta data for the exposure
  • NB: hard to decouple the vignetting pattern and the sky background... differently camera orientations may help to filter out he sky gradient component but with only 10 cameras this may not work that well

• Fit 2D polynomial maps of vector field of PSF parameters

  • Extract sources and psf parameters in each exposure (psf size, eccentricity, orientation etc) and fit a 2D polynomial to sampled vector field
  • Can then repeat a similar process to previous metric, looking for variability between the psf parameter maps
  • This would allow us to see if all psf elongation is orientated in the same direction (indicating dome vignetting) or randomly (indicated a detector tilt issue or something else within the instrument)

• Could use bright time as an opportunity to produce vignetting and psf elongation maps, as the brighter sky background will produce brighter residual gradients than ones produced in dark time