ROSAT Status Report #65:
August 10 1993
Off-Axis Point Spread Function (PSF)
The algorithm describing the off-axis point spread function of the ROSAT PSPC is now available on the legacy anonymous ftp account. The original subroutine for the off-axis point spread function was written by Guenther Hasinger and has had minor modifications made during the testing period. In this Status Report, we briefly describe the software available. An MPE/GSFC technical memo discussing the off-axis psf is in preparation and should be publically available in a few weeks.The modified subroutine is now available (in its VMS fortran form) on legacy under rosat/software/off_axis_psf.subroutine
GSFC has also released some software in the form of an FTOOL to compare the model psf and data (in the form of a radial profile of counts).
FTOOLS are a collection of utility programs to create/modify or examine data files in the FITS format. Currently, FTOOLS is supported under the IRAF environment as an external IRAF package, under Unix on Suns, Ultrix on DECstations using the newer DEC FORTRAN compiler and VAX/VMS as a stand alone Host version. FTOOLS contain tasks which can be used to analyze ASCA data, so this may be a good time for sites to install and try these routines.
The PUBLIC release of FTOOLS version 2.4 is now available in compressed tar file form via anonymous FTP from heasarc.gsfc.nasa.gov. The development version of FTOOLS contains the current release, this is updated weekly. The FTOOLS development version is required for the PSF code.The source is in the /software/ftools/release directory. There you will find the following files:
- develop.dec.v2.4a.tar.Z -- DEC source
- develop.sun.v2.4a.tar.Z -- SUN source
- develop.vms.v2.4a.tar.Z -- VMS source
- Install.v2.4.p Installation Guide
- Users.v2.4.ps.Z -- Users Guide with help pages
- Tinyusers.v2.4.ps.Z -- Users Guide without help pages
- Develop.v2.4.ps.Z -- Developer's Guide
The FTOOLS public release has changed substantially with respect to the release 2.3 of a couple of months ago, so please look into the documentation. NOTE : The installation instructions are valid, but the documentation reflects the current PUBLIC FTOOLS status, NOT the DEVELOPMENT FTOOLS status.
The Installation Guide includes contact addresses in case of problems. (Please use those addresses for FTOOLS problems, not the ROSAT GOF)
The PSPC PSF code resides under the caltools package of ftools, and a brief description follows (more information ia available via online help):
PSPCRPSF (June93)--caltools-- author Rehana Yusaf (GSFC)
- NAME
- pspcrpsf - rebins rpsf data, and calculates theoretical ROSAT PSPC
off-axis radial point spread function
- DESCRIPTION
- PSPCRPSF reads an input FITS datafile containing a radial point spread
function dataset in OGIP standard format, and rebins this input
dataset such that such that each new bin contains a user-supplied
minimum number of counts. The result is written in the form of a
BINTABLE (with extname = OBS RPSF) in the output file (again in
OGIP standard format).
- In addition, PSPCRPSF uses the parameterization of the ROSAT PSPC radial psf to calculate the theoretical/predicted model for direct comparison. The ROSAT PSPC psf is a strong function of energy, thus the theoretical psf is weighted using a user-defined PHA spectrum (which must be in OGIP standard format). An associated EBOUNDS calibration dataset (available within RMF files). PSPCRPSF prompts for background count rate, this is useful if no background subtraction was carried out on the radial profile. The background entered is subsequently folded into the prediction. An energy band is also taken as input, allowing the predicted model to be calculated in the same energy band as the radial profile.
- The theoretical psf is written to the output file as a BINTABLE (with extname = THEO RPSF).
- In addition, PSPCRPSF uses the parameterization of the ROSAT PSPC radial psf to calculate the theoretical/predicted model for direct comparison. The ROSAT PSPC psf is a strong function of energy, thus the theoretical psf is weighted using a user-defined PHA spectrum (which must be in OGIP standard format). An associated EBOUNDS calibration dataset (available within RMF files). PSPCRPSF prompts for background count rate, this is useful if no background subtraction was carried out on the radial profile. The background entered is subsequently folded into the prediction. An energy band is also taken as input, allowing the predicted model to be calculated in the same energy band as the radial profile.
A recipe exists on the legacy ftp account under rosat/doc/pspc/psf.recipe describing how to proceed from the pspc events file to a plot of source radial profile overlaid with predicted model.
In addition to the FTOOL, one needs the development version of PROSCON, which is the (unix only at this point) routine to convert the pros spectral table file to a FITS format XSPEC spectral file. This can also be found on legacy under /software/develop/xandev/spectral/proscon/proscon you will need to copy the whole directory of files across to run it.
(Note: this version of proscon reads a table file and writes out a spectrum in FITS format and a response file (not in FITS format). Prosconf reads a FITS version of the table file and writes out a spectrum and response file (not in FITS format)).
Other PSF software:
PROS 2.2.1 contains the task xray.xspatial.rosprf, which creates an image that models the on-axis psf. Users should be aware that this does not include the break to a power-law in the mirror scattering term, and so will not be a good match to the tail of the psf. In addition, this task does not yet have any off-axis parameterization, and it can only produce a single energy psf image.
Select another report
Return to the ROSAT GOF
Curator:HEASARC Guest Observer Facility
Please use the Feedback link if you have questions on ROSAT.