Hi!
I have a wishlist item which comes in 2 parts:
1. Average PSF generator: takes a number of stars of the same brightness/magnitude and averages them, generating a PSF
a. The could be a lower limit size and upper limit size settable by the user so that a reasonable number of homogeneous stars of the same size could be averaged. Auto mode could try to pick the smallest/dimmest stars which are not just noise, and the user could set a lower noise threshold maybe.
2. HDR PSF generator:
My best idea of the week: (probably already done by someone)
Everyone knows that stars are so far away, that all of them are actually point sources. So "larger" stars are essentially the product of scattered light from the point source through the atmosphere, telescope lens, DSLR lens, filter etc. The ONLY difference between the bright stars and the dim ones is brightness. The scattering profile is the same. This knowledge gives us an opportunity to define the PSF, at radii at much larger distances than would otherwise be possible with the dim true point sources, (due to noise, and the sheer weakness of the signal). So at some radii every star has an excellent definition of the PSF... not too close to the center which may be saturated, and not too far where it would be too dim or subject to noise. The task then is to combine the ideal profile of the PSF at the ideal radii of each star with the others, dimmer stars defining the center and larger stars defining the areas further out.
Assuming everything is linear, this could be acheived simply by rescaling (greyscalewise) the larger stars so that the inner edge of ideal annular ring (at some radius) is the same brightness as the outer edge (of the same radius) of the ideal annular ring of of the dimmer stars, and so on. Of course the centers of the dimmest point-like stars would be used.
HDR PSF estimation..
My wish!
cheers
Colin
