
Date: Thu, 09 Nov 2000 09:15:40 +0100 (MET)
From: Peter Kretschmar <Peter.Kretschmar@obs.unige.ch>
Subject: Interactive background handling
To: Stefan Larsson <stefan@astro.su.se>, Sami Maisala <smaisala@astro.helsinki.fi>, Niels Joergen Westergaard <njw@dsri.dk>, Juhani.Huovelin@astro.helsinki.fi, Eero Esko <eesko@astro.helsinki.fi>, Carol Anne Oxborrow <oxborrow@dsri.dk>, Allan Hornstrup <allan@dsri.dk>
MIME-version: 1.0
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from QUOTED-PRINTABLE to 8bit by uhuru.dsri.dk id JAA02985


  Hi everybody -

  it seems we are on a good way for discussions about interactive background
handling. Thanks to all for their contributions. To the many points, that
have been stated, I would like to add the following comments and ideas:

  The basic idea, to which I fully agree, seems to be that we have tools to:
- interactively select model information from the background "database";
- visualize these models in various ways;
- subtract these models from existing shadowgrams, spectra, lightcurves;
- visualize the results of the subtraction in several ways.
In principle I see no problem in having a step like this in Offline 
Scientific Analysis (OSA), as an alternative to the j_bkg_model step. This 
means that instead of running the full analysis script in one go, one could 
let it run up to j_cat_extract, then do a background determination step 
interactively, and afterwards continue with j_image_bin, j_imaging, ... 
(Or alternatively, after running an analysis with auto-bkg, step back to
the BKG level but keeping note of derived source information).
The only problem with this, as Sami has pointed out, is that it is not 
obvious how one could do this kind of work for multiple Science Windows in 
a not too painful fashion.

  Several additional ideas have been mentioned, that - in my opinion at 
least - need further thought and clarification:

* Removal of source contributions

  In principle this should be done, but it is not at all obvious _how_
  one would actually do it. Does the interactive background tool take
  functions from Stefan's or Niels Jrgen's software to model these
  source contributions? Should we have a whole set of tools to manipulate
  shadowgrams?

  I should mention that it is possible to load the code of an executable 
  into isdcroot and execute the functions contained from the isdcroot prompt. 

* Study of effects on final images, spectra, lightcurves

  Maybe you are all aware of this, but I'd like to emphasize that the only
  reasonable way to see the effect of a different bakcground model on your
  scientifc results is to run the subsequent analysis steps with the new
  (improved?) background model as input.

* Background model from sky images

  This is an attractive idea (I see where I have no sources) but how
  is it supposed to work? Would this mean some sort of interactive
  image deconvolution, where I control the iterations and set a 
  background level in the image? Or would it mean a convolution of this
  area on the detector?

  Obviously we have a lot to discuss, see you in a few minutes!
  Peter


  
  


