cannot reproduce Failure to complete ImageSolver script

M

Matt3am

New member
Hello,

This happens every time I use the ImageSolver.js script:
Image solving succeeds using triangle similarity - converges after 4 iterations get to "Saving keywords ..."

Hangs at "Saving keywords" with blue Windows activity indicator. Locked out of PixInsight, with CPU running at 17.5% (highest running process on the system).

Does not recover. After 10-20 minutes tried "Pause/Abort" in the console window. Get "Running - Abort Requested". Sits in that state for several minutes (3-5?) and then finally exits with successful plate solve. So its not a show stopper, but serious impact on user experience, and just a lot of time.

I see this with poor exiting behavior with other processes as well. For example - On conclusion of SPCC the process continues to run after the solution has converged with the application read-locked and spinning Windows circle of "whatcha-doin?" and Java burning 17%+ of CPU while I go make a coffee or whatever ... Sometimes it recovers before my patience expires.

My guess is that it is writing to disk all of the calibrated files? And that's why it churns so long? I have a lot of subs. Am I right about this? If so, how to I make it just plate solve, or color correct the image I care about? Or is there something else going on?

Note that I am a brand new user - So could very well be pilot error.

Thanks.

System:
Newly installed version of PixInsight
Windows 10Pro
Intel I7 8 cores, 64G physical RAM, internal SSD with plenty of room (2TB)
Windows Swapfile setup 8G with max 64G
 
Last edited:
Hello @Matt3am
ImageSolver is typically used in "Active Window" mode where it will try to solve only the active image. Is that what you are doing or are you using it in the "list of files" mode (or are you talking about WBPP)?
 
nico1038 said:
Hello @Matt3am
ImageSolver is typically used in "Active Window" mode where it will try to solve only the active image. Is that what you are doing or are you using it in the "list of files" mode (or are you talking about WBPP)?
Click to expand...
Just running it with active window - my cropped master-light. Running it from the script menu, default settings, then hitting the "OK" button.

Just did it again as a sanity check. Console window says all is good:
"* Successful astrometry optimization
Saving keywords ... "

Then busy ... for a looooong time ...
 
Matt3am said:
Just running it with active window - my cropped master-light. Running it from the script menu, default settings, then hitting the "OK" button.

Just did it again as a sanity check. Console window says all is good:
"* Successful astrometry optimization
Saving keywords ... "

Then busy ... for a looooong time ...
Click to expand...

Something is not right. If you are trying to solve the active image then Pix is not trying to write to you calibrated files.
Are you sure you are on Pix 1.8.9-2 build 1605 and Image Solver version 6.2.3?
 
nico1038 said:
Something is not right. If you are trying to solve the active image then Pix is not trying to write to you calibrated files.
Are you sure you are on Pix 1.8.9-2 build 1605 and Image Solver version 6.2.3?
Click to expand...
Yes, v1.8.9-2, build 1605, ImageSolver v6.2.3

I think I may have solved this myself. Or should I say "solved my own error"!

I was neglecting to fill in the object identifier search: "M 101 Pinwheel" in my case. I thought it would solve off the Metadata in the FITS header. Well it turns out, if you leave that field blank it will still solve, and give you the success message. But for some reason it hangs when trying to write the keywords.

With the correct entry for the object and the object search resolved it worked perfectly. So error on my part. But I find it curious that it would complete successfully, but hang like that if you leave it blank.
 
Matt3am said:
Yes, v1.8.9-2, build 1605, ImageSolver v6.2.3

I think I may have solved this myself. Or should I say "solved my own error"!

I was neglecting to fill in the object identifier search: "M 101 Pinwheel" in my case. I thought it would solve off the Metadata in the FITS header. Well it turns out, if you leave that field blank it will still solve, and give you the success message. But for some reason it hangs when trying to write the keywords.

With the correct entry for the object and the object search resolved it worked perfectly. So error on my part. But I find it curious that it would complete successfully, but hang like that if you leave it blank.
Click to expand...

If the (correct) coordinates are in the header it is not necessary at all to search the object. It's very curious...
 
To add more detail if someone wants to reproduce. I had cropped the master-light, and successfully done a DBE. When I cropped I lost the FITS data, so I copied it over from the original master.

And the solution is definitely "good" even with it hanging .. Since I can complete SPCC afterward. But like I said, hang only happens if I fail to put the object name in Plate Solver Image Parameters Search. Note that those new "catalogue" searched image parameters are replacing the ones from my FITS header
 
You must log in or register to reply here.
Back
Top