×

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

Bi-monthly release with minor bug fixes and improvements

[SOLVED] StellarSolver failure in alignment module

  • Posts: 300
  • Thank you received: 57
Last night I found myself completely unable to use the Align module because StellarSolver failed in "Capture and Solve"

Every.
Single.
Time.

Very odd because I can save an image at the same position with the same settings and "Load and Slew" nails the solution very quickly.

I've attached a screenshot to illustrate the problem. Image preview shows (faintly) a nicely centered M33. Scale and position are correct. Same image in "Load and Slew" solves almost instantly.

Can somebody help me figure out what I'm doing wrong?

I did a quick demo of the problem and saved logs. Basically this short session demos the alignment failure with "Capture and Solve" followed by a quick capture with successful "Load and Slew"

Ekos log is attached.

The astrometryLog.txt is too big for the forum. Here's a link to astrometryLog.txt.gz in my Dropbox

File Attachment:

File Name: log_22-51-36.txt
File Size:37 KB
Last edit: 3 years 3 months ago by Scott Denning.
3 years 3 months ago #64845
Attachments:

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

  • Posts: 300
  • Thank you received: 57
Very strange:

Can't "Capture and Solve" in alignment module

but "Load and Slew" works fine.

Using StellarSolver in KStars 3.5 on both R Pi 4 and also amd64, both running ubuntu 20.04.1 LTS with mate desktop

Capture, Save, then "Load & Slew" is OK for now as a workaround but pretty much takes the scheduler off the menu!

StellarSolver is working great, except when automated as part of the Alignment module. Logs provided in original post above.


Last edit: 3 years 3 months ago by Scott Denning.
3 years 3 months ago #64875
Attachments:

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

  • Posts: 1119
  • Thank you received: 182
I think your FOV is wrong. With the ASI1600MM and a focal length of 488mm the FOV should be ~ twice that size. That much of divergence is outside the tolerance of the Stellar Solver.

Jo
The following user(s) said Thank You: Scott Denning
3 years 3 months ago #64890

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

  • Posts: 85
  • Thank you received: 19
Is this not working in more than one setup? It sounds like you've got a RPi4 and a computer with otherwise identical setups?

My recollection of how Load & Slew works is that it solves the loaded image, slews the mount to the putative centre of the loaded image, then takes a further image, solves it, reslews and so on.

So in this, is it taking and successfully solving those subsequent images?

Reading through your log, it doesn't look like it is.


One other thing I noticed was a difference in downsampling between the loaded image and the captured ones. The loaded image is being evaluated at "Scale range: 3.85544 to 5.78316 arcsec/pixel".

The first autocapture by contrast: "[...] arcsec per pixel range: 1.94069 to 3.8505"

The second (I think) autocapture: "[...] arcsec per pixel range: 1.45551 to 2.88787"


So from my understanding, as this range decreases, you need an increasingly large library of astrometry FITS files.

And that brings me to one last thing in your logs: "Total Size of Index files: 2.44583 GB"

I've got like 18 GB worth. So perhaps you don't have enough of the higher precision FITS files installed?
3 years 3 months ago #64891

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

  • Posts: 85
  • Thank you received: 19
I just saw El Corazon's response after posting mine... I agree, your FOV appears to be out. And that probably ties in to my discovery of the discrepancy in image scale in your log.
The following user(s) said Thank You: Scott Denning
3 years 3 months ago #64893

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

  • Posts: 300
  • Thank you received: 57
Thank you both!

Feeling sheepish ... this FOV was left over from a different setup.

Marking SOLVED
Last edit: 3 years 3 months ago by Scott Denning.
3 years 3 months ago #64906

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

Time to create page: 0.530 seconds