×

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

Bi-monthly release with minor bug fixes and improvements

StellarSolver works great, until "plate solve failed"

  • Posts: 222
  • Thank you received: 20
Boy was I impressed with the StellarSolver! Getting the polar alignment sequence shot and solved in 30 seconds? Sign me right the heck up.

Until...it didn't. My initial solves were shot with tree branches and stuff, and the solver worked great. Pointing, capture and slew to target, peerless.

Meridian flip, and then it's as if someone broke the piece marked "Solver" in the box. No better feedback than "Solver failed".
  1. Clear starfield in sharp focus
  2. Parking and starting over didn't help
  3. Parking and power-cycling the Pi and starting over didn't help
  4. Switching some of the options for StellarSolver didn't help

Fortunately my mount's hand controller knows about NGC 281 and got it within the visual field, so I only lost about an hour dinking around. I could try to do logs but I'm not touching anything until the current imaging session finishes.

Any ideas? Anyone else encountering this?
3 years 4 months ago #63549

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

  • Posts: 2876
  • Thank you received: 809
Did you turn on Alignment logging in the Ekos logs?

If that doesn’t provide enough information, the astrometry logging can be turned on and it can print enormous amounts of information about the solve
3 years 4 months ago #63550

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

  • Posts: 222
  • Thank you received: 20
Does that work for the new StellarSolver in 1.5.5 too?
3 years 4 months ago #63551

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

  • Posts: 2876
  • Thank you received: 809
Yes that is precisely what I mean. StellarSolver can’t give you any feedback unless you turn on alignment logging.
3 years 4 months ago #63552

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

  • Posts: 222
  • Thank you received: 20
OK. Not having followed the development, I thought that astrometry and StellarSolver were two different things. Know better now.

One thing I hadn't thought to try previously was a solve off my guide camera. Since moving to an OAG I'd had lots of trouble getting sufficient stars to guide on till I figured out the correct angular position where I could extend the prism enough, so it didn't occur to me. Tried one last night, came back about as fast as my finger left the mouse button (1280x960, no wonder!). After that I uploaded a FITS to nova.astrometry.net so I could get astrometry's idea of my scope/main-camera's FOV exactly right. Hand-entered that, boom, plate solving was back and worked just fine all evening.
3 years 3 months ago #63674

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

  • Posts: 2876
  • Thank you received: 809
Yeah, you could say they are different things, but they aren't as well. Really you could say that I did two things with StellarSolver. 1. I made an internal solver library with an internal build of astrometry.net that requires no external temporary or configuration files. 2. I made a library that can solve using many different methods like using the internal solver library, using ASTAP, using external astrometry.net, and using an online astrometry solver.

So you can say that StellarSolver is astrometry.net, but it is also more than that. I also made a bunch of tweaks and enhancements that can make blind solving blindingly fast compared to before.
3 years 3 months ago #63687

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

  • Posts: 222
  • Thank you received: 20
2.6 seconds, yeah I'd say that counts as "blindingly fast" from my experience!

Thanks again.
3 years 3 months ago #63688

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

Time to create page: 0.268 seconds