ImageSolver fails using bottom up FITS

Hi PTeam

I am looking to use the image solver, with eDR3, to accurately plate solve images, including distortion correction. If I use the PI default of top down coordinates, the image solves fine. However, I wish to export this plate solved file for use in Aladin, and AperturePhotometry Tool to identify catalogue stars and calculate instrumental magnitudes. Exporting this top down solve results in seemingly reversed coordinates and the catalogue stars do not overlay with image stars at all.
If I enable bottom up, as I think this will solve the issue, the image solver now fails to solve the image completely. To get the solver to work I must use top down, which seems incompatible with other programs standards. I understand PI can do all this for me but I have a workflow established and would like to keep using external programs, whilst taking advantage of the superior plate solving offered in PI

Thanks

Paul
 
Is anyone able to help? Trying to use Pixinsight plate solved images in external programs but when overlaying catalog data the sources don’t line up with image stars.

thanks
 
This is a well-recognised problem. This thread discusses it. In post #57 on page 3 of this thread I summarise the problem, which results in an incompatibility between PI and some other application when exchanging FITS files (note also posts #59 - #62 in this thread). I also suggest work-arounds, but I'm afraid no "set and forget" solution.
 
Last edited:
Hi Paul,

The problem here is that "other programs" are not able to manage different FITS row orders (bottom-up or top-down), as PixInsight does. PixInsight always includes a nonstandard (but widely accepted as a de facto standard) ROWORDER keyword in written FITS files that other applications should use to know the actual row order of the stored image.

Anyway, other applications are unable to understand and use our astrometric solutions with distortion models based on thin plate splines, so you could not use them anyway, irrespective or row orders. You could only use our linear solutions, which are more accurate than other implementations but cannot model distortions.

We'll try to implement a compatibility mode for FITS WCS keywords in the future, but please understand that we have other priority projects, including more accurate astrometric solutions based on more advanced interpolation devices, which we'll implement hopefully during this year, so this is a low-priority task.
 
Back
Top