Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Weighting, Reference, Noise vs Detail
Most discussions in weighting for integration seem to revolve around noise (though it is less clear what the new PSF weights encompass).
Most discussions on picking the reference sub for integration are light on details, and sometimes just say "best".
Many and maybe most integrations could use more data, and are not well up on the curve of diminishing returns, so noise is quite important.
Let's say you have quite a lot of data, are well onto the curve, and are more interested in resolving fine detail as a goal.
Does all the standard advice, and the new PSF ratings, equally address that? Or should one take a different weighting approach? More to my point, I have about 10 days of data on M82, some of which have decent seeing, many that had good transparency and lots of data but relatively poor seeing.
I recognize that noise can hide detail, and do not suggest they are unconnected -- but are they directly, 1:1 connected?
Is there something I should do, weighting wise, or selection of reference image wise, or maybe construction of a synthetic Luminance used for combination, that emphasizes detail perhaps at the expense of some noise?
As a related question: I am combining Ha + Luminance for a synthetic luminance, to try to highlight the jets (?) coming from the center. Is there a different technique, such as doing dconvolution on each separately prior to combine, or do it on them after combine?
Bottom line: if the goal is to bring out structure, and you have a lot of data, is there anything you change in the approach from when you have little data?