fits error

Tom Duvall wrote on Mar 28, 2012

I have a fits file that I think should be fine. It is available at http://soi.stanford.edu/~duvall/Iraf/img0.fits. ds9 reads it fine and imheader l+ gives a nice header. But imcopy and imstatistics both give segmentation violation. The beginning of the header is below. Any idea what is going on?
Thanks.
Tom

imhe img0.fits l+ | head
img0.fits[500,500][real]:
No bad pixels, min=271., max=3528.
Line storage mode, physdim [500,500], length of user area 2633 s.u.
Created Wed 09:27:45 28-Mar-2012, Last modified Wed 09:27:45 28-Mar-2012
Pixel file "img0.fits" [ok]
EXTEND = T / FITS dataset may contain extensions
COMMENT FITS (Flexible Image Transport System) format is defined in 'Astronomy
COMMENT and Astrophysics', volume 376, page 359; bibcode: 2001A&A...376..359H
BUNIT = 'm/s'
INSTRUME= 'HMI'
CONTENT = 'Tracked Extracted Patches, made by hg_patch'
TELESCOP= 'SDO'

---

Jason Quinn wrote on Mar 28, 2012

Your file does not cause a segmentation error under 2.16 for linux. You'll want to post the version of IRAF and architecture you are using. The most likely scenario is you are using an old version of IRAF on a 64 bit machine.

Cheers,
Jason

PS The file does however have a fits error:

              fitsverify 4.16 (CFITSIO V3.250)              
              --------------------------------              
 
 
File: img0.fits

1 Header-Data Units in this file.
 
=================== HDU 1: Primary Array ===================
 
*** Error:   Keyword #47, EXPTIME: Type of value "nan" is unknown.
*** Warning: Data checksum is not consistent with  the DATASUM keyword
*** Warning: HDU checksum is not in agreement with CHECKSUM.
 
 71 header keywords
 
 32-bit integer pixels,  2 axes (500 x 500), 
 
++++++++++++++++++++++ Error Summary  ++++++++++++++++++++++
 
 HDU#  Name (version)       Type             Warnings  Errors
 1                          Primary Array    2         1     
 
**** Verification found 2 warning(s) and 1 error(s). ****

---

Mike Fitzpatrick wrote on Mar 28, 2012

Assuming you're on a v2.15 64-bit system, this is a known issue in the FITS kernel caused by the use of BSCALE/BZERO values in the header. This has been fixed in the recent v2.16 release, but unfortunately if you want to patch a v2.15 system you'll need to relink your system with patched 'libex.a' and 'libvops.a' libraries available from

ftp://iraf.noao.edu/iraf/v215/support/<arch>;

Relinking should then just be a matter of

% cd $iraf
% make linux64
% make update

If you still have problems let me know, your image worked fine for me under v2.16.

---

Tom Duvall wrote on Mar 28, 2012

I downloaded the two libraries libex.a and libvops.a, relinked the system.
This fixed my problem.
Thanks very much!
Tom

---

Last post on Mar 28, 2012