×

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

Bi-monthly release with minor bug fixes and improvements

StellarSolver failing

  • Posts: 91
  • Thank you received: 18

StellarSolver failing was created by Dean

I did sudo apt update && sudo apt dist-upgrade as of tonight. I am running SM OS on rpi 4B+ 8Gb.
I have tried multiple option permutations to get the PAA so solve and it eventually times out
and fails. I am using ASI533MCpro as my main camera, ASI120MMmini as guider attached
to ZWO OAG, so the CCD to plate solve with is the ASI533. Prior to this upgrade which had the
old PAA using on-board astrometry everything worked fine. This is extremely frustrating as
tonight is the first clear night here since Dec 31 2020! I have Ekos debugger running in case
of a crash and have verbose logging going to /tmp/astrometryLog.txt ... Can someone please
help me salvage this evening ? I cant even do a basic two star alignment model: nothing will solve.
3 years 1 month ago #66901
Attachments:

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

  • Posts: 91
  • Thank you received: 18

Replied by Dean on topic StellarSolver failing

a bit more info. I focussed with bhatinov mask using the ZWO EAF hand control then refined the focus
with the Ekos focus module. My Ekos profile is correct and has not been changed since before last update.
I then changed my settings for PAA / StellarSolver:
source extraction method: Internal SEP, solving method: Internal Solver, options profile: 1-Default
I have under Scale & Position Use Scale and Use Position unchecked. The Load all indexes in Memory is
now unchecked in the profile. PAA works but is very slow and now I cant get the solver to plate solve on
a star in Cassiopia to work on building my alignment model.
3 years 1 month ago #66903

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

  • Posts: 1208
  • Thank you received: 559

Replied by Hy Murveit on topic StellarSolver failing

Dean, Sorry you're having issues tonight. Don't know if this is actionable advice, but I too have issues from time to time with solving, though recently StellarSolver and I have been on good terms. My approach is to have ASTAP and the internal solver both available, and if one fails I use the stellarsolver control to go from 'Internal Solver' to 'Local ASTAP' and back to see if one works.

Yes, I also try the scale and position checkboxes. I also try and see if it's detection that's failing or solving. With regard to the parameters, I am suspicious of some of the stellarsolver parameters, and in particular don't allow any of the "keep" parameters to be 0. I use Initial Keep = 1000 and 'Stars Keep #' = 500. I also double check the image for focus, and sometimes increase the gain and/or exposure to make sure I have a reasonable image.

Good luck!
Hy
3 years 1 month ago #66904

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

  • Posts: 388
  • Thank you received: 17

Replied by Ronald Scotti on topic StellarSolver failing

Do you get an Astrometry log, what does it show? I always have that checked, but most of the time it never writes an Astrometry log. I have had crashes associated with the solver, not just non-solves. I originally downloaded a large number of index files (I am also running on an Rpi4 with 8 gb of RAM) because I was solving with the main camera (ASI533) and main scope SCT 9.25 at f/10; the FOV was very small. I was getting errors that indicated a possible memory leak as it tried to load to many index files. I have since switched to solving with my e-finder (ASI120 on an Orion 9x50 finder scope) which has a much more reasonable FOV (1 degree) and I reduced the number of index files to only those required for that FOV. I wish I could say this solved my crashing problem, but it did not; I have had multiple crashes during PA, but fewer so once I get away from the pole. I am wondering if solving around the pole is problematic with certain FOV?
I have not been able to find the right configuration of settings, equipment and index files under which plate solving works well all the time. But then I am using plate solving much more as it has become an enabler for me to use the equipment.

So I too am very interested in what people have to say about settings and results.
3 years 1 month ago #66905

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

  • Posts: 1208
  • Thank you received: 559

Replied by Hy Murveit on topic StellarSolver failing

Yes, I'm imaging tonight, and I checked and I have that file. My log settings are

and my StellarSolver settings are:

So, I'm using ASTAP. The contents of the file tonight were:
> cat /tmp/astrometryLog.txt 
/opt/astap/astap -o /tmp/externalSextractorSolver_22.ini -speed auto -f /tmp/solvere06fadc7df14428b9d5a2eca3c677a20.fits -wcs -z 0 -fov 1.545 -ra 5.62319 -spd 89.8366 -r 5 -log
480 stars, 372 quads selected in the image. 408 database stars, 316 database quads required for the square search field of 1.5°. Search window at 100% based on the number of quads. Step size at 100% of image height
Using star database g17

That's not something I tend to look at.
Hy
3 years 1 month ago #66906
Attachments:

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

  • Posts: 91
  • Thank you received: 18

Replied by Dean on topic StellarSolver failing

Hy, can you post some documentation on the drop down Options profiles or provide a link
to that information ? I was able to get plate solving going but much slower than
before my apt updates tonight. Hoping for insight on fine tuning.

thanks,
Dean
3 years 1 month ago #66907

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

  • Posts: 1208
  • Thank you received: 559

Replied by Hy Murveit on topic StellarSolver failing

Dean, I have some intuition on this, but really I mostly look to Rob for directions on this,
beyond what I said above.

I think Rob gave some guidance in one of the StellarSolver threads a while back.
I couldn't find it, but there's this giant thread:
indilib.org/forum/general/6845-new-inter...eriments-needed.html
I don't believe there's been a thorough doc written about it yet.

Honestly, the first thing I try when I get in trouble with StellarSolver is switching it to ASTAP. www.hnsky.org/astap.htm
Install that, and then point the StellarSolver options Solving Method --> 'Local ASTAP'
3 years 1 month ago #66909

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

  • Posts: 388
  • Thank you received: 17

Replied by Ronald Scotti on topic StellarSolver failing

Hy,
I am running on an Rpi4 with 8 Gb of RAM; Kstars 3.5.0, Astroberry server 2.0.3.
I loaded ASTAP and the H17 index files at /opt/astap. It solves a stored image, but did not solve any new images tonight. It was cloudy and the images were poor (even after dark subtraction and 10 second exposure), while the internal solver did solve, occasionally. There may be something wrong with my options.
First, I never got an Astrometrylog file. My Logging page looks like yours except I put the log file in my /home/Astroberry/feb4/ directory.




Second, why does your log file indicate external Sexstractor, when you selected "internal SEP"?
I tried both Internal and external SEP, but ASTAP never could solve. I guess I am not sure how to set these up to work together in Kstars/EKOS.
I had to disable parallel solves, as ASTAP does not support that. I saved one of the images and opened up ASTAP (external to Kstars) with just that image, but it did not find any stars; so naturally it would not solve. I do not have Sexstractor loaded on my RPi. So ASTAP is trying to identify stars on its own.

I am mostly curious why I am not getting a log file, the rest could be explained by option settings.

thanks,
Ron
3 years 1 month ago #66944
Attachments:

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

  • Posts: 1208
  • Thank you received: 559

Replied by Hy Murveit on topic StellarSolver failing

1) Looks like you should be writing a log, assuming that the directory /home/astroberry/feb4 exists. However, in your note you capitalized /home/Astroberry/feb4, but in the screenshot it is not capitalized. Case definitely matters in Linux. If that's not it, well, we're clearly running different versions of the program. I'm running the latest software, and it looks like you're running Astroberry, which is at 3.5.0 I've heard. Perhaps there was a bug Rob fixed having to do with that? Don't know. You might ping Rob (@rlancaste)?

2) Don't know why the log is named "/tmp/externalSextractorSolver_22.ini" but it seems like it's just the name of an output file that ASTAP is being asked to write. I believe ASTAP is doing all the extraction etc.

Hy
3 years 1 month ago #66946

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

  • Posts: 91
  • Thank you received: 18

Replied by Dean on topic StellarSolver failing

Hy, thanks for responding, I will dig deeper in the thread and see what I can learn.
I find it odd that within one month of development the solve speed has gone down
for my software environment. At first the new stellarsolve was literally unbelievably
fast .. as in OMG! Ron, I had my settings for logging as Default instead of File and unchecked
the specified file path... I found the astrometyLog.txt in /tmp
3 years 1 month ago #66947

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

  • Posts: 388
  • Thank you received: 17

Replied by Ronald Scotti on topic StellarSolver failing

All, thanks for your responses. I was hoping that your capitalization idea was the answer, but no. Initially, the Astrometry file was set to write to /tmp/Astrometrylog, but that gets overwritten; so I changed it to my home directory (/home/astroberry/Astrometrylog or a current session directory) with the capitalization correct ( I just mistyped it in my response). So it is a(nother) mystery. It has written that log, once in awhile, but I cannot isolate or repeat the conditions when it worked.
I am testing today using simulators for scope and camera and set the log file to /tmp/Astrometrylog, it did a 'phoney' solve and did write the log! Maybe there is a character length issue with the log location setting. There are so many settings in Kstars/Ekos and that makes it difficult for this user to make sure he has them all correct. I understand that this log can be quite large and writing it to /tmp prevents it from eventually clogging up the system. So the question is how to deal with that and still have the log available. The only time I really need the log is when Kstars crashes during a solve (which is almost always the only time it crashes on me). <y suggestion would be to write this log file to the same directory as the other log files, for that day, and sequenced as they are so that you don't have to keep changing the name after a crash. Then there could be a built in a directory remove when a dated log directory is more than a few days (5-10) old. I do this with my Rpi Allsky camera and it keeps the memory there manageable. It would be one less thing that the user has to set up correctly, just check a box.

Thanks for all your efforts on this project. I will leave this log location in tmp and see if that continues to allow it to be written during 'real' sessions.
3 years 1 month ago #66967

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

  • Posts: 91
  • Thank you received: 18

Replied by Dean on topic StellarSolver failing

Ron, slight typo in my prev response: file should be /tmp/astrometryLog.txt .
If you are concerned about the file size, when you are done with KStars, in a command shell you can
truncate the file to zero length with
truncate -s0 /tmp/astrometryLog.txt
3 years 1 month ago #66969

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

Time to create page: 1.310 seconds