SimIm3: Rotation and Mirror Flip.

Now we begin to combine coordinate transformation processes. As I describe below, I apply both and image flip and a rotation. The WCS in the final image is installed correctly: our Ra,Dec values are correctly placed on their targets. However, we do see a difference in the transformed image compared to our SimIm0 example (only rotation).


% SimIm3 Y 45.0 Y

To make these images I used the command:

 
% SimIm3 Y +45.0 Y 

   * Yes, print debug messages 
   * Rotate 80 degrees 
   * Flip about the Y axis 

Here is the header I obtained with hcards_wcs.sh: 
CTYPE1  = 'RA---TAN'           /  Gnomonic Projection
CTYPE2  = 'DEC--TAN'           /  Gnomonic Projection
CRPIX1  =              100.000 / Axis 1 Reference
CRPIX2  =              100.000 / Axis 2 Reference
CRVAL1  =        180.000000000 / Ra  at Frame, J2000 (deg)
CRVAL2  =         32.500000000 / Dec at Frame, J2000 (deg)
CROTA2  =      45.000000000000 / Image Twist +AXIS2 W of N, (deg)
CDELT1  =      0.0002777777845 / Axis 1 Pixel Size (degs)
CDELT2  =      0.0002777777845 / Axis 2 Pixel Size (degs)
CRVALE  =               50.000 / Combined CRVAL error (arcsec)

Our WCS in the final image (right) hasbeen handled correctly: we recover the correct Ra,Dec and our sky compass is pointed correctly. However, we do see a different behaviour compared to a single rotation in SimIm0. In this case, because we have fliped the X values (with clip_imshift_trs.sh) about the Y axis, the resultant rotation of +45 degrees goes in a clockwise (CW) direction, not the counter-clockwise (CCW) direction of SimIm0. This is not big deal since, in practice, we rarely use clip_imshift_trs.sh for performing such transformations. The important thing is that the flip,theta values we used in clip_imshift_trs.s, when fed directly to hcards_wcs.sh, will create WCS cards in the header that allow routines like ds9 and xy2sky to produce crrect astrometry.

Back to calling page