On Mon, 19 Nov 2001, David P. Huenemoerder wrote:
> In chandra-users, you wrote:
> >Hmm, this gets me started. I'll try your suggestions for Sherpa. But in the
> >meantime I have a couple of further questions regarding what you said about
> >pile-up:
> >
> >[...]
> >> Pile-up model in Sherpa is only available for 1D analysis. The
> >> modification of PSF by pile-up needs to be folded into your psf image.
> >
> >I'm a bit lost here: how do you propose I modify my PSF file for pile-up?
> >Also it seems to me that doing things this way would force one into assuming
> >that all sources have exactly the same fraction of piled-up pohtons. This
> >isn't the case in my data: there are four sources affected by pile-up, with
> >fluxes (and presumably piled-up fractions) that vary by about a factor 4.
> >
>
>
> I think the only way you could do this is to do spectral modeling of
> each source independently with the pileup model to determine the
> unpiled number of counts. This doesn't give you the piled psf,
> though. For that, you might have to perform some forward-folding
> iteration to determine the unpiled count density vs radial distance to
> then estimate a pileup fraction vs radius for the model, then perturb
> the model psf accordingly.
>
> As Aneta said, we don't have any spatial pileup models.
>
> Hope this helps.
That's one possibility. But I have both a 0.4s frame-time dataset and a 3.2s
frame-time dataset, with the former nearly unaffected by pile-up but only 1/4
the length. So I have another possibility: to just ratio the two datasets
and then correct the longer, 3.2s frame-time dataset by the fraction of
photons lost to pile-up. Of course this won't do anything spectrally, but it
will be a try at getting the deep data to have a PSF which agrees with what
mkpsf gives me.
Any thoughts on doing this?
Eric
-- Eric S. Perlman E-mail: perlman@jca.umbc.edu Joint Ctr. for Astrophysics, Physics Dept. Phone: +1 410 455 1982 University of Maryland, Baltimore County Fax: +1 410 455 1072 1000 Hilltop Circle WWW: www.jca.umbc.edu/~perlman Baltimore, MD 21250
This archive was generated by hypermail 2b29 : Thu Dec 13 2012 - 01:00:06 EST