Weird problem with Image Solver PI. Any suggestions?

[Ouroboros]

There's a weird  little problem I have noticed using Image Solver script with my new MacBook. 

Image Solver chunks on through and produces a solution in the process console.

However, what it is not doing is writing the solution into the FiTS header of the file.  

Weirder still is that SPCC seems quite happy to provide a solution - graphs and everything.  

How is this possible? It shouldn't work should it?

PS. The plate solving result shown in the process console is confirmed by Astrometry Net.  PS is definitely solving the image.  Just not saving it to the image header. 

[Paul M]

Is updating the FITS header a selectable option, maybe? 

[Oddsocks]

11 hours ago, Ouroboros said:

There's a weird  little problem I have noticed using Image Solver script with my new MacBook. 

Image Solver chunks on through and produces a solution in the process console.

However, what it is not doing is writing the solution into the FiTS header of the file.  

Weirder still is that SPCC seems quite happy to provide a solution - graphs and everything.  

How is this possible? It shouldn't work should it?

From PixInsight build 1.8.9-2, released August 14 2023, PI no longer calculates a WCS solution when plate solving but uses its own spline-based XISF compatible solution instead, which is why PI internal processes that require a plate solved solution work as expected but when you export a solved image in FITS format there is no WCS solution in the FITS header since PI no longer uses the WCS standard for any of its processes.

Unfortunately this change is another result of PixInsight’s well publicised aim of divergence away from supporting the FITS standard for anything other than very basic import/export of FITS image data and no longer supporting writing extended meta data to FITS files.

For full details see the release notes for build 1.8.9-2, section “New Astrometric Solutions - Image Solver Script version 6.0”

https://pixinsight.com/forum/index.php?threads/pixinsight-1-8-9-2-released.21228/

For those of us still needing to be able to interact with other astronomy applications outside of the PixInsight environment we need to process the output FITS file from PixInsight through a different vendor application if a WCS solution is required in the FITS header although there is some validity in the argument that since you can’t be certain that the WCS solution written in any FITS image received from a third party is 100% reliable we should always ignore that and re-calculate a WCS solution anyway.

[Ouroboros]

Hello @Oddsocks.  Thanks for that. After a quick skim read of that link can I take it that:

After doing a crop we still need to run Image Solver before using a process like SPCC that needs the WCS*?

The WCS data is stored somewhere in the XISF file.  Can I see that WCS info somehow?

* WCS - World Coordinate System …. I had to Google!  

 

[Ouroboros]

10 hours ago, Paul M said:

Is updating the FITS header a selectable option, maybe? 

Prior to Oddsock’s post I had wondered that. In fact I looked around for one. The answer is no apparently. 

[Oddsocks]

On 28/01/2024 at 08:57, Ouroboros said:

1: After doing a crop we still need to run Image Solver before using a process like SPCC that needs the WCS*?

2: The WCS data is stored somewhere in the XISF file.  Can I see that WCS info somehow?

 

 

Hi @Ouroboros

Re Q1: The answer to that is yes, you do need to run the Image Solver script again if you crop a previously solved image and then run another process after cropping that requires WCS coordinates.

For example, if you run the Image Solver script on an image and then run Dynamic Crop afterwards you will see a pop-up message "Warning: DynamicCrop: Existing astrometric solution will be deleted", telling you that the stored image coordinates will be deleted if you proceed to crop the image, if you accept that warning, crop the image and then try to run SPCC it will fail to start with a message in the Process Console "Error: The image has no valid astrometric solution: <imagename>".

Re Q2: There is no user tool or third-party script that I could find that allows you to explore the meta-data stored within a XISF file, maybe there is something in the developers toolkit but I have not explored that part of PI for a while since platform development is now too rapid for me to keep up with and I tend to use other applications for science imaging in preference to PI.

The only option that I am aware of if you want to read the WCS coordinates in PI is to run the Image Solver script and then take a desktop snap-shot, or physically note the astrometric solution that is displayed in the Process Console when the Image Solver script completes.

HTH

[symmetal]

There is this python tool on GitHub which will read and write XISF metadata which may be useful if you're familiar with python. 🙂

XISF Encoder/Decoder

Alan

[Ouroboros]

Thanks so much, @Oddsocks and @symmetal for both your informative responses.  It never ceases to amaze me how knowledgable and helpful members are on SGL.