Author Topic: No astrometric match  (Read 523 times)

DrG

  • Newbie
  • *
  • Posts: 12
No astrometric match
« on: February 28, 2018, 05:42:55 PM »
I have been unable to get an astrometric match for my FITS files.  These are the MicroObservatory files.  An example is the file DeltaCep180129020826 which was recently shared using DropBox.  One problem may be the large field size.  The FOV is 10 degrees.  I know the field center is close because I can identify objects in the images.  I know that my image scale is 64.4 arcsec/pixel.  Unfortunately the FITS header has IM_SCALE = 5.210, but I set the image scale before trying a match.  A potential problem is that this is the image scale for the main camera, but I am using the finder images.  The FITS header knows this since it has INSTRUME = 'Finder  '.

I have tried adjusting some of the Match parameters without success.  I have tried some Cosmetic correction without success.  I am sure I have just not yet tried the appropriate "trick."  Any chance we can get this to work?

For the record, I also have some images from the main camera and they accomplish a match simply without any problems.  Sorry to be causing so much grief...  but I have been waiting for Observatory for decades!  The same is true for all those Mac people out there that do astronomy.  Thank you!

Sander Berents

  • Global Moderator
  • Jr. Member
  • *****
  • Posts: 64
    • Code Obsession
Re: No astrometric match
« Reply #1 on: March 01, 2018, 12:38:47 AM »
Thanks. This is highly appreciated. It would be great if you could spread the word so even more people become aware of Observatory.

Observatory's plate solver currently only supports tangent plane (TAN) projections, and for images with a large field you'll have to reduce the area the plate solver is using. Hence the first thing I did with this image was to set the "frame size" to its minimum value (use only the center 25% of the image) and set the estimated pixel scale to 64.4. No luck. I then solved the image with astrometry.net (it reported the original center to be 2 degrees off), and imported it in Observatory. If you do that and display the overlays, it looks like that 25% setting is not small enough, but even if Observatory would allow you to make that smaller, there wouldn't be enough good stars left in the 650x500 pixel image.

For these finder images, the only "solution" right now is to plate solve these very wide fields with for example astrometry.net and import the result into Observatory. Sorry about that. Improving wide field handling in Observatory is on the roadmap.
Sander Berents
Code Obsession, LLC
http://codeobsession.com

DrG

  • Newbie
  • *
  • Posts: 12
Re: No astrometric match
« Reply #2 on: March 01, 2018, 03:47:06 AM »
OK.  And thanks.  I was afraid the field size might be the problem. 

I am developing some "projects" to introduce people to authentic research using robotic telescopes.  I wanted to start out with the classic variable stars such as Algol, Beta Lyr, Delta Cep, Mira.  But they are too bright and have no useable comparison stars for a normal telescope field.  The MicroObservatory instruments are 6-inch, but these targets are still too bright and the field is far too small.  Field of main camera is 1 degree.  The finder field will work.  10 degree field.  And the photometry tool in Observatory is just what is needed.  (Thank you again.)  I do not really need a plate solve for this work, but thought I would use this feature of Observatory if I could get it to work.  No problem!  Doing the photometry is all I need for this project.  Follow-up projects will involve fainter stars so I can then use the main camera.  And I will then use the Match feature in Observatory as part of the projects/activities.

I am with the Education and Public Outreach group at Sonoma State University and we are partnering with MicroObservatory which is managed by Harvard-Smithsonian Center for Astrophysics.  MicroObservatory can have many hundreds of users in a year, and I suspect many of those users will be on Macs.  I will be spreading the word.  We are also developing projects that can be implemented for traditional telescope-camera systems.  I may be having more questions.

Please let me know if I may assist in any way with the development of Observatory, or with any pre-release evaluations.

Sander Berents

  • Global Moderator
  • Jr. Member
  • *****
  • Posts: 64
    • Code Obsession
Re: No astrometric match
« Reply #3 on: March 01, 2018, 06:08:35 AM »
That's a great project! Are the finder images always unfiltered, or can you also select a V filter? You may also want to include very basic image calibration in the project (in this case probably only dark frames).

What if I add support for the astrometry.net service? That way, if for whatever reason Observatory's plate solver cannot find a solution, you can just click another button to have it send the image to astrometry.net. When it finds a solution, it would then merge the new WCS information. It is something that was already on the roadmap, but I could also try to add it in an upcoming maintenance release so you have it much sooner. I am sure such feature would make other users very happy as well.

Providing feedback and spreading the word is the best way to help the development of Observatory right now. There's no public beta mechanism yet.

Please note that Observatory is available with a volume discount for educational institutions.
Sander Berents
Code Obsession, LLC
http://codeobsession.com

DrG

  • Newbie
  • *
  • Posts: 12
Re: No astrometric match
« Reply #4 on: March 02, 2018, 07:41:39 PM »
The MicroObservatory Finder has no provision for a filter.  There is also no provision for obtaining a dark.  This would need to be done manually at the telescope, I suspect.  I am going to ask if they will do that when they do maintenance.  I may even volunteer to help and show up.  Amado, Arizona is closest to me.

If the astromet feature gets implemented I will add it to my project descriptions.  Also glad to hear about educational discounts.

I originally purchased Observatory for my professional research.  Then I began to discover all the capabilities and began to bring it into my E/PO work.  I predict that Observatory is going to get really big in the educational community.  Amazing that it is now possible to do professional quality image analysis and organization on a Mac!!!

Sander Berents

  • Global Moderator
  • Jr. Member
  • *****
  • Posts: 64
    • Code Obsession
Re: No astrometric match
« Reply #5 on: March 06, 2018, 02:36:36 AM »
A small update: I finished the astrometry.net client code. It is able to work in two modes:
  • Upload the whole image to astrometry.net, let it plate solve it, and obtain the WCS information, or
  • Use Observatory's star detector, upload only coordinates to astrometry.net, and everything else the same. That's a very tiny upload (a few KBs) compared to the whole image.
I still need to integrate it into Observatory's UI, but just wanted to let you know that this is indeed going to be part of Observatory 1.2.2. It is the only new feature for 1.2.2, but a nice one.
Sander Berents
Code Obsession, LLC
http://codeobsession.com