×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

New Internal Solver for Mac, Windows, and Linux -- Testing/ Experiments needed

  • Posts: 333
  • Thank you received: 92
Okay, it looks like CCDCiel has tricked me by modifiying the WCS file and removing the line returns. I have rewritten the WCS write routine to binary but I have to check which new problems this causes for the installed base.
3 years 10 months ago #53549

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809

Hi, I had a go at this one. I changed the Downsample setting to 2 and I changed the setting I just added for limiting the number of stars (Han's idea) in the image to 50. The image solved in a little over 3 seconds. I don't know if that is the best you can do, but give it a shot.

3 years 10 months ago #53550
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809

For this matter, if you are using my program on an image, you can tell it not to delete the temp files, you can solve it with one of the other methods, and then solve it with ASTAP and then compare the resulting WCS files for the same image and basically the same solve. Then you can see how the files are saved differently, any differences in the FITS headers, etc.
3 years 10 months ago #53552

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809

Yes, by all means make sure that what you change doesn't break other things. My main concern is that the WCS file should follow the standard format and be usable in the same way the WCS file is from Astrometry.net. If it cannot be opened in other programs that follow the fits standard and are built on CFITSIO, that is really not good. I am not sure about what is causing the files to not work, but I do think the new line characters are a big part of it because the FITS standard calls for the header cards to be recorded 80 characters long with no line breaks.

Now that I released my version 1.0, you should be able to make a change in ASTAP and then immediately test it by running my program, you can try one of the other solvers and try ASTAP and see if they work, and you can compare the WCS files to find the differences including hopefully what makes it not work. At least that is my hope.
3 years 10 months ago #53553

Please Log in or Create an account to join the conversation.

  • Posts: 333
  • Thank you received: 92
The WCS was written by ASTAP as a text file. I can't change this practice since some programs depend on it. So I introduced the command line option -wcs to write a real FITS header to the wcs file. I have uploaded a new version ASTAP 0.9.362b to my webpage only:

www.hnsky.org/astap_amd64.deb

Han
Last edit: 3 years 10 months ago by han.
3 years 10 months ago #53554

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809

I can certainly give it a go. Do you still want me to try it, or have you already made more changes? It looks like a linux Debian installer? I am mostly working on a Mac but I have an ubuntu virtual machine set up where I have been testing for Linux compatibility that I can try that on.

In terms of the list of sources, from Sextrator, I am getting back the star X and Y positions, the mag_auto, the flux, the peak value, the information about the star ellipse, and sometimes the HFR.
So far, I have been able to get astrometry.net to just give me the X and Y positions and the flux. I would like to get more back from it, but so far, no luck.

From ASTAP, I would say the most important things we could get back would be the X and Y position, followed by the magnitude if possible and/or the flux. These are very useful for people doing photometry or astrometry studies. I think this is the bare minimum we might need. As for the star ellipse information, right now I'm just using that to display the star circles in the annotated viewer and for filtering out things that are not stars. But I could see that information being useful for scientific studies and also for analyzing the quality of your guiding or something similar. The HFR is mainly useful I think for focusing routines, so I would think that would only be needed if you want to use ASTAP as a form of sextractor. Although one thing that might be important is to get some idea of the star size for making the flux information useful. Thoughts?

The newly added features that make use of the star lists and solution information to get WCS coordinates of items in the image after solving is something I just recently added and I think it has many possible applications. But I don't know all of them yet. I have a struct in my program called star and it it has a bunch of pieces of information in it. But that could change in the future if more or less information is needed.

Your thoughts?
3 years 10 months ago #53556

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809

That makes a lot of sense. Thanks I will try it
3 years 10 months ago #53557

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809

It worked!!! So I made a new commit to my program to support this new functionality in ASTAP by adding a -wcs to the args. So now ASTAP is just not recording the star positions and information.
3 years 10 months ago #53560

Please Log in or Create an account to join the conversation.

  • Posts: 333
  • Thank you received: 92
Yes I can see it work here with S-solver 1.0.

I did a speed test comparison for an offset of 15 degrees:

Distance 15 degrees:
* S-solver 500 stars ==> 5.0 seconds
* ASTAP 500 stars ==> 5.9 seconds

* S-solver 100 stars ==> 3.2 seconds
* ASTAP, 100 stars ==> 1.7 seconds (set in the ASTAP GUI or use command line option: -s 100


ASTAP has to build the database quads live, so reducing the number of stars becomes very beneficial.

I had again a look to the command line of Astrometry.net and found this:
manpages.ubuntu.com/manpages/focal/en/man1/solve-field.1.html

-d, --depth number or range
Number of field objects to look at, or range of numbers; 1 is the brightest star,
so "-d 10" or "-d 1-10" mean look at the top ten brightest stars only.

Seems very rarely used and I never tried it but should be very effective.

Han
Last edit: 3 years 10 months ago by han.
3 years 10 months ago #53562

Please Log in or Create an account to join the conversation.

  • Posts: 333
  • Thank you received: 92

We have to define a format. Can your software read .axy files?

Han
Last edit: 3 years 10 months ago by han.
3 years 10 months ago #53566

Please Log in or Create an account to join the conversation.

  • Posts: 333
  • Thank you received: 92
If ASTAP export the stars it could also be done the other way around. Importing stars from Sextractor. Could be beneficial for images showing guiding problems.
3 years 10 months ago #53567

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809

Oh sorry, forgot to mention that. Astrometry.net seems to want FITS binary tables for its input XYLS files, so that is what I have been using. Sextractor is perfectly happy exporting exactly what is needed, it has all sorts of options for exporting data table types and columns. So I have just let Astrometry.net's input requirements dictate the format. I have also successfully used AXY files since they are also in FITS Binary format.
3 years 10 months ago #53571

Please Log in or Create an account to join the conversation.

Time to create page: 0.927 seconds