Re: Updating RA/DEC fits headers with true location?

[Robert Creager <Robert_Creager@LogicalChaos.org>]

When grilled further on (Tue, 24 Aug 2004 09:54:45 -0500),
Michael Koppelman <lolife@bitstream.net> confessed:

> There is always WCSTools, too:
> 
> http://tdc-www.harvard.edu/software/wcstools/
> 
> Michael
> 

Hmmm...  WCSTools does not appear to be in the best of shape (code wise).  I
downloaded and started playing with it.  It segfaults, doesn't call it's own
internal functions correctly, and from my brief looks, is kind of a mess.  Of
course, I'm biased since I write code 24x7 ;-)  I submitted a couple of patches
to the author, but have not heard back yet...

But, it almost works well!  The solution, once I modify the FITS headers a bit
(I swap CRVAL1 and CRVAL2, swap CTYPE1 and CTYPE2, and change the sign of CD1_1
and CD2_2) is really good.  Anyone know how to run imwcs that might be able to
direct me (to get the correct data)?

Michael, how far off were the 'bad' solutions you came across?

From my test image, the star x,y is 1255.36, 2021.21 (y right on
the edge), and the calculated location (from match) was 19:15:12.3 39:42:55.6. 
The raw image headers were degrees off.  After using imwcs with the star data
from match (x,y RA,DEC), the x,y position using xy2sky measures at 19:15:11.6
39:43:39.4.  Using sky2xy (with APIS projections), the original location is
called out at 1249.6, 2020.4.

That seems pretty good to me.  I may automate some measurements to get more
information, across multiple images.

Cheers,
Rob

-- 
 17:25:53 up 13 days,  3:48,  3 users,  load average: 2.33, 2.18, 2.14
Linux 2.6.5-02 #8 SMP Mon Jul 12 21:34:44 MDT 2004

PGP signature