Home - Research - Software


A program to extract spectra and aperture photometry from multiple sources in multiple XMM-Newton observations.

This program can extract spectra and aperture photometry from XMM-Newton observations, calling Ftools and SAS tasks if needed. The observations may of may not overlap. This program checks if the source and background are in the field of view of any observation, and extract products accordingly. Responses (RMFs and ARFs) are calculated along with spectra.

The spectra and responses can be summed with a companion program (cdfs-sumspectra.pl), available available as part of this distribution.

This program was originally written and used for the XMM-Newton deep survey in the Chandra Deep Field South; see the paper: The XMM-Newton Deep Survey in the CDFS. III. Point source catalogue and number counts in the hard X-rays. Ranalli et al. 2013, A&A 555, A42.

Cdfs-extract is written in the Perl programming language, using the Perl Data Language (PDL) modules. It is free software, release under the terms of the Affero GPL v3.

You may download the latest version or browse (and fork) it directly in github.

There is documentation which you may access with the command: perldoc cdfs-extract.pl

How it works

The basic idea is to write 3 text files:

The programme reads these files and cycles on all sources and event files. A sample set of these tables is included to show the format.

There is a first stage, in which it checks wether the sources and bkgs are in the field of views of the event files. It emits a summary of these checks, both on screen and in the file cdfs-extract.log . It also warns if the source and background regions fall on different chips.

In the second stage, it actually extracts source spectra, background spectra, rmfs and arfs. This stage may be fine-tuned, if needed (just ask me if so).

All extracted spectra and responses are placed in a subdir called Products.

If needed, a separate programme (cdfs-sumspectra.pl) can then be invoked to sum the spectra of one source having multiple obsids (documentaion: perldoc cdfs-sumspectra.pl).

If the sources are not too close to each other and the background has no large variations on scales <20-30" (i.e., you are dealing with a shallow survey such as LSS or COSMOS, rather than deep like CDFS or Lockman Hole) then the sky.txt and bgsky.txt can be automatically generated by the autoregions programme, which computes extraction regions yielding the highest signal/noise ratio.


For any further information, please write to pranalli.github@gmail.com.

Piero Ranalli
Post-doc researcher at IAASARS, National Observatory of Athens, Greece
Associate of INAF -- Osservatorio Astronomico di Bologna, Italy

Last update: 2014/11/10