On Wed, 1 Jun 2011, H. Tugca Sener-Satir wrote:

> I am having a strange situation with figaro commands and their spectrum 
> orientation. I attached a figure, I hope this will help to clarify my 
> explanations.
> 
> I started with an image tagged as original, vertical to x-axis. I 
> rotated it 90 degree by using kappa-rotate command. To subtract the sky 
> background, I use polysky and polysky thinks the spectrum is oriented as 
> shown in the image (on the 1st region). But when I want to extract the 
> object spectrum by using extract command, it thinks the spectrum is 
> located on the 3rd region (y<0) as shown on the image. 
> 
> Position difference wouldn't be a problem if I hadn't use gaia to decide 
> where the object spectrum starts and end. I mean, when I examine the 
> spectrum by using gaia, gaia thinks the spectrum is oriented just like 
> the way polysky considers. To use that limits, I can make very simple 
> subtraction and reach useful values but I wonder what is the mystery 
> behind this coordinates & orientation...
> 
> Any idea for this strange situation would be very appreciated.

Hi H. Tugca,

the problem you have looks like an inconsistency between these two FIGARO 
programs when dealing with NDF origins (and maybe the WCS information, 
FIGARO is now old and not compatible with KAPPA/NDF WCS). When you 
KAPPA:ROTATE your image the lowest left-hand corner is given a origin with 
respect to the origin of the unrotated image (1,1) so you now have 
(1,-NX), looks like extract understands this, but polysky doesn't.

The simple fix is to KAPPA:ROTATE your image and then do:

    % erase rotated.AXIS ok
    % erase rotated.WCS ok

To remove the old coordinates, and then:

    % setorigin rotated '[1,1]'

to reset the origin as well. That should give you an NDF with the simplest 
possible coordinates.

Cheers,

Peter.