The generated images should contain metadata relating to:
Array configuration
Missing antennas. The visibilities should be appropriately scaled by the inverse of the error by the correlator, or prior to imaging, in order to reflect the data missing due to e.g. correlator data link dropouts.
Fraction of data lost due to RFI flagging (to be included in the value determined above).
Facet field centers, if facetted imaging is used.
An estimate of the primary beam (perhaps this can be stored in a centralized location, with a timestamp on when it was estimated). If this is model (primarily the cos^2 dependency), then the model can be probably included in every image.
A model of the expected noise for this timestamp, based on sky models? (Very speculative, but allows comparison with what is actually measured).
A discussion is needed on how to accomodate this meta data in QL images (perhaps XMP extensions embedded in png? Move to FITS?). Discussions related to the casacore images will be moved to the imaging pipeline issue tracker.
The generated images should contain metadata relating to:
A discussion is needed on how to accomodate this meta data in QL images (perhaps XMP extensions embedded in png? Move to FITS?). Discussions related to the casacore images will be moved to the imaging pipeline issue tracker.