Bugs/Features in the ASCA data analysis software
Last updated on 2001 Feb 17
This list is appropriate for HEAsoft v5.0/5.0.1 and supersedes the ASCA bug list for ftools v4.2 Please also check known FTOOLS bugs and A list of known XSPEC bugs.
- ascaarf
-
We have recently found a bug in the convolution routine used in the
release versions (inc. HEAsoft v5.0.1) of ascaarf. Ascaarf expands
the GIS WMAPs to facilitate the PSF convolution using FFT. Usually,
the expanded WMAP is square (e.g. 64x64), but in some cases it is not
(e.g., 64x128) in which case the convolution routine gives a normalization
which is too small by a factor of 2, and the source flux measured with
such an ARF is too large by a factor of 2.
Users can check if this is the case by looking at the screen output or the HISTORY keywords in the FITS header of the ARF files:
HISTORY Input WMAP array has size 44 by 43 bins HISTORY expanded to 128 by 64 bins
If the two numbers on the 'expanded to' line differ, then this bug is affecting your results. - There is also an error in the normalization of the ARF when using the raytrace option with SIS data. The resulting ARF should be multiplied by 0.746.
-
We have recently found a bug in the convolution routine used in the
release versions (inc. HEAsoft v5.0.1) of ascaarf. Ascaarf expands
the GIS WMAPs to facilitate the PSF convolution using FFT. Usually,
the expanded WMAP is square (e.g. 64x64), but in some cases it is not
(e.g., 64x128) in which case the convolution routine gives a normalization
which is too small by a factor of 2, and the source flux measured with
such an ARF is too large by a factor of 2.
If you have any questions concerning ASCA, visit our Feedback form.