×

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: 219
  • Thank you received: 41
Jo, thanks for your help. I was aware on how to go back on git, this was the reason I've told Rob that perhaps, this kind of changes will be best applied as a squashed commit. Because as its currently merged, if you go back in time, you will not recover exactly the same repo. I hope, that as you have found in your compilation, the leftovers doesn't interfere too much with the rest.

But I'm starting to put my feet on thin ice, because I that I'm not working actively on the project, should not be talking about its dynamics... sorry, it's a professional deformation :)

Ok, let's go to help stabilising the product. Rob if you need any other test, or do you think that I can try something, please, don't hesitate to ask me

Rafa
3 years 5 months ago #61598

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

  • Posts: 106
  • Thank you received: 4

Turned on Alignment logging.

So M57 is a tough one:

2020-10-14T17:36:55 Solver Failed
2020-10-14T17:36:55 Solver was aborted, timed out, or failed, so no solution was found
2020-10-14T17:36:55 Starting Internal StellarSolver Astrometry.net based Engine. . .
2020-10-14T17:36:55 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2020-10-14T17:36:55 Scale range: 0.638705 to 0.780639 arcsec/pixel

2020-10-14T17:36:55 Set odds ratio to solve to 1e+09 (log = 20.7233)

2020-10-14T17:36:54 Configuring StellarSolver
2020-10-14T17:36:54 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2020-10-14T17:36:54 Stars Found after Filtering: 15
2020-10-14T17:36:54 Removing stars smaller than 4 pixels
2020-10-14T17:36:54 Stars Found before Filtering: 79903
2020-10-14T17:36:19 Starting Internal StellarSolver Sextractor. . .
2020-10-14T17:36:19 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

These are my settings:



Where do I have to go from here.

This is a compressed jpeg file of M57:



No field flattener on this small scope (C90 Mak on a HEQ 5 Pro).
Powered by

GNU / Linux
Git
KDE neon
KStars | EKOS | INDI

and some cheap hardware
3 years 5 months ago #61601
Attachments:

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

  • Posts: 2876
  • Thank you received: 809
Cerro Torre,

Your image solved in about 1 second using Parallel Small Scale, but yeah FastSolving took forever and I just cancelled it. But I compared the profiles, there is more than one difference, it's not just the parallelization. I noticed a couple of other things that I did differently. I just tried changing one and it solved pretty quickly.

Try this, change the minwidth to 0.1 in the FastSolving profile, save your changes, and try solving with FastSolving. Let me know if that works better.
3 years 5 months ago #61604

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

  • Posts: 106
  • Thank you received: 4
Hi Rob,

no luck:
This is the output using the FastSolving profile with "Min Degree Width" set to 0.1

2020-10-14T18:33:49 Solver Failed
2020-10-14T18:33:49 Solver was aborted, timed out, or failed, so no solution was found
2020-10-14T18:33:48 Starting Internal StellarSolver Astrometry.net based Engine. . .
2020-10-14T18:33:48 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2020-10-14T18:33:48 Scale range: 0.637219 to 0.778824 arcsec/pixel

2020-10-14T18:33:48 Set odds ratio to solve to 1e+09 (log = 20.7233)

2020-10-14T18:33:48 Configuring StellarSolver
2020-10-14T18:33:48 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2020-10-14T18:33:48 Stars Found after Filtering: 43
2020-10-14T18:33:48 Removing stars smaller than 4 pixels
2020-10-14T18:33:48 Stars Found before Filtering: 49692

I even tried the ParallelSmallScale profile without success.
Powered by

GNU / Linux
Git
KDE neon
KStars | EKOS | INDI

and some cheap hardware
3 years 5 months ago #61606

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

  • Posts: 2876
  • Thank you received: 809
You probably have different index files than I do. I have all of them from 4204 to 4219, and all of the 4100s
3 years 5 months ago #61607

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

  • Posts: 106
  • Thank you received: 4
These are my index files:
Powered by

GNU / Linux
Git
KDE neon
KStars | EKOS | INDI

and some cheap hardware
3 years 5 months ago #61608
Attachments:

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

  • Posts: 106
  • Thank you received: 4
That's interesting:

on the jpeg version of M57 your change in the FastSolving profile works indeed:

2020-10-14T19:38:20 Settling...
2020-10-14T19:38:05 Slewing to target coordinates: RA (18h 54m 20s) DEC ( 33° 03' 38").
2020-10-14T19:38:05 Solution coordinates: RA (18h 54m 20s) DEC ( 33° 03' 38") Telescope Coordinates: RA (13h 13m 00s) DEC ( 90° 00' 00")
2020-10-14T19:38:05 Solver RA (283.39564) DEC (33.02976) Orientation (89.57861) Pixel Scale (0.63525)
2020-10-14T19:38:05 Field parity: neg

2020-10-14T19:38:05 Field rotation angle: up is 89.5786 degrees E of N
2020-10-14T19:38:05 Pixel Scale: 0.635247"
2020-10-14T19:38:05 Field size: 54.8598 x 36.6061 arcminutes
2020-10-14T19:38:05 Field center: (RA H:M:S, Dec D:M:S) = (18:53:34.954, +33:01:47.135).
2020-10-14T19:38:05 Field center: (RA,Dec) = (283.396, 33.0298) deg.
2020-10-14T19:38:05 Solved with index: 4108
2020-10-14T19:38:05 Number of Matches: 19
2020-10-14T19:38:05 Solve Log Odds: 137.781
2020-10-14T19:38:05 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2020-10-14T19:37:18 Starting Internal StellarSolver Astrometry.net based Engine. . .
2020-10-14T19:37:18 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2020-10-14T19:37:18 Set odds ratio to solve to 1e+09 (log = 20.7233)

2020-10-14T19:37:18 Configuring StellarSolver
2020-10-14T19:37:18 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2020-10-14T19:37:18 Stars Found after Filtering: 593
2020-10-14T19:37:18 Stars Found before Filtering: 593
2020-10-14T19:37:17 Starting Internal StellarSolver Sextractor. . .

But why is the fits file not solved correctly? The only difference I can think of, is the fact, that the fits file contains non stretched data; only the view is stretched?

The ParallelSmallScale profile solves the jpeg file of M57 too.
Powered by

GNU / Linux
Git
KDE neon
KStars | EKOS | INDI

and some cheap hardware
3 years 5 months ago #61610

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

  • Posts: 1119
  • Thank you received: 182
That makes sense! I also have problems solving with FITS files. Some work right away, others not at all and lead to a crash. The strange thing is that a FITS file of the same object, same framing, will crash, another solve right away. No difference between the files, except that they were taken a few minutes apart from each other.
3 years 5 months ago #61612

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

  • Posts: 2876
  • Thank you received: 809
Wait, Cerro Torre, I think I misunderstood your experiment that didn't work, you just said:

"on the jpeg version of M57 your change in the FastSolving profile works indeed"

That means that was NOT the file you were trying to solve that didn't work? Was it actually a FITS file?

All that I saw was the Jpeg so that is what I solved. is there a copy of the FITS file for testing? maybe I missed that...
3 years 5 months ago #61618

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

  • Posts: 2876
  • Thank you received: 809
Remember that one thing about FITS files is that they could contain Scale and position info. Jasem and I just added a method to kstars yesterday that added back the ability to perform a load and slew with the information in the FITS Header as a starting point. Are you using the up to date version of KStars that includes this capability or are you using a build from the day before that had to blind solve them? And are you trying to use the Position and Scale info in the image for solving? That makes a big difference. For example, if we do use position and scale info in the FITS image and it is wrong, then it won't solve at all.
3 years 5 months ago #61619

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

  • Posts: 2876
  • Thank you received: 809
El Corazon, can you share a file that works and one that doesn't?
3 years 5 months ago #61620

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

  • Posts: 152
  • Thank you received: 20
I'm ready to test here. I prefer to build from source. However, I've been unable to do this since the integration with Stellar Solver. All dependencies per the README are met, but something must be missing.

This is the error I get up to the latest commit to master:
/home/schwim/src/kstars/kstars/ekos/align/align.cpp: In member function ‘void Ekos::Align::startSolving()’:
/home/schwim/src/kstars/kstars/ekos/align/align.cpp:3286:26: error: ‘class StellarSolver’ has no member named ‘setSSLogLevel’; did you mean ‘setLogLevel’?
         m_StellarSolver->setSSLogLevel(SSolver::LOG_NORMAL);
                          ^~~~~~~~~~~~~
                          setLogLevel
/home/schwim/src/kstars/kstars/ekos/align/align.cpp:3286:49: error: ‘LOG_NORMAL’ is not a member of ‘SSolver’
         m_StellarSolver->setSSLogLevel(SSolver::LOG_NORMAL);
                                                 ^~~~~~~~~~
/home/schwim/src/kstars/kstars/ekos/align/align.cpp:3286:49: note: suggested alternative: ‘LOG_NONE’
         m_StellarSolver->setSSLogLevel(SSolver::LOG_NORMAL);
                                                 ^~~~~~~~~~
                                                 LOG_NONE
/home/schwim/src/kstars/kstars/ekos/align/align.cpp:3296:26: error: ‘class StellarSolver’ has no member named ‘setSSLogLevel’; did you mean ‘setLogLevel’?
         m_StellarSolver->setSSLogLevel(SSolver::LOG_OFF);
                          ^~~~~~~~~~~~~
                          setLogLevel
/home/schwim/src/kstars/kstars/ekos/align/align.cpp:3296:49: error: ‘LOG_OFF’ is not a member of ‘SSolver’
         m_StellarSolver->setSSLogLevel(SSolver::LOG_OFF);
                                                 ^~~~~~~
/home/schwim/src/kstars/kstars/ekos/align/align.cpp:3296:49: note: suggested alternative: ‘LOG_ALL’
         m_StellarSolver->setSSLogLevel(SSolver::LOG_OFF);
                                                 ^~~~~~~
                                                 LOG_ALL
kstars/CMakeFiles/KStarsLib.dir/build.make:2247: recipe for target 'kstars/CMakeFiles/KStarsLib.dir/ekos/align/align.cpp.o' failed
make[2]: *** [kstars/CMakeFiles/KStarsLib.dir/ekos/align/align.cpp.o] Error 1
make[2]: *** Waiting for unfinished jobs....
CMakeFiles/Makefile2:432: recipe for target 'kstars/CMakeFiles/KStarsLib.dir/all' failed
make[1]: *** [kstars/CMakeFiles/KStarsLib.dir/all] Error 2
Makefile:140: recipe for target 'all' failed
make: *** [all] Error 2

Any idea what I could be missing?
3 years 5 months ago #61621

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

Time to create page: 0.465 seconds