Crashing using the newly updated Photometric Mosaic script

John, Rob

I disabled the beta 1 script and then installed beta 2. (Just to confirm, the script was headed displaying "Beta 2").

I used my Blue registered panel 1 and panel 2 files for the test.

With panel 1 as the reference the two panels merged and the script ran as expected.

I then switched to having panel 2 as the reference and PI crashed and closed immediately.

I then reopened PI, disabled the script beta 1 and ran the feature script beta 2, opened the Blue panels 1 and 2.

This time I selected panel 2 as the reference - the script ran as expected. I then cancelled the script.

I then re-opened the script and selected panel 1 as the reference. The process console as it progressed began reporting in red font that it was reading outside the image. The script did produce the merged mosaic and PI did not crash.

I then repeated the merge test again by re-opening the script with panel 1 as reference - merged successfully. Then repeated once again with panel 2 as reference and this too was successful. The process console each time reported in red font that it was reading outside the image.

I then closed PI and repeated the test using the Blue subs, with panel 1 as reference beta 2 merged the two files successfully and the process console did not report any 'reading outside of the image' in red font. Then repeating once again with panel 2 as the reference it merged the two panels and the process console this time did report in red font 'reading outside the image'.

I then repeated running the beta 2 script for the Red panels 1 and 2. The beta 2 failed to merge, crashed and closed PI with panel 1 as reference. It happened too quickly for me to read what was being reported in the Process console.

Hope that helps John.

Barry
 
Last edited:
John, Rob

I disabled the beta 1 script and then installed beta 2. (Just to confirm, the script was headed displaying "Beta 2").

I used my Blue registered panel 1 and panel 2 files for the test.

With panel 1 as the reference the two panels merged and the script ran as expected.

I then switched to having panel 2 as the reference and PI crashed and closed immediately.

I then reopened PI, disabled the script beta 1 and ran the feature script beta 2, opened the Blue panels 1 and 2.

This time I selected panel 2 as the reference - the script ran as expected. I then cancelled the script.

I then re-opened the script and selected panel 1 as the reference. The process console as it progressed began reporting in red font that it was reading outside the image. The script did produce the merged mosaic and PI did not crash.

I then repeated the merge test again by re-opening the script with panel 1 as reference - merged successfully. Then repeated once again with panel 2 as reference and this too was successful. The process console each time reported in red font that it was reading outside the image.

I then closed PI and repeated the test using the Blue subs, with panel 1 as reference beta 2 merged the two files successfully and the process console did not report any 'reading outside of the image' in red font. Then repeating once agin with panel 2 as the reference it merged the two panels and the process console; this time it did report in red font 'reading outside the image'.

I then repeated running the beta 2 script for the Red panels 1 and 2. The beta 2 failed to merge, crashed and closed PI with panel 1 as reference. It happened too quickly for me to read what was being reported in the Process console.

Hope that helps John.

Barry
That is really interesting. The red console text is a message I created in Beta2 which is triggered if the StarDetector method is attempting to read outside the image. Beta2 is also applying a mask to StarDetector to prevent it detecting stars too close to the edge. That the mask only gets applied correctly some of the time is very interesting, and may hint at where the problem is.

Regards, John
 
I have fixed a problem in the StarDetector star mask. Hopefully this will prevent the reported crashes.

I have attached PhotometricMosaic V3.2.1 Beta3

Either uninstall Beta2, or copy the source code over the top of it. If you are running the correct version, the console will display:
=== Photometric Mosaic 3.2.1 Beta3 ===

Thanks for testing it for me
John
 

Attachments

  • PhotometricMosaic.zip
    136 KB · Views: 53
Thanks John. I'll be able to test over the UK Easter bank holiday weekend & will report bank. Thanks for your perseverance.

Regards,

Barry
 
John, Rob

I have tried various combinations of each registered panel 1 and 2 for the R, G and B; sometimes with panel 1 as reference and sometimes panel 2 as reference.

I am pleased to say that the script has worked flawlessly (even seemed to complete slightly faster too).

Good work John.

Many thanks,

Barry
 
John, Rob

I have tried various combinations of each registered panel 1 and 2 for the R, G and B; sometimes with panel 1 as reference and sometimes panel 2 as reference.

I am pleased to say that the script has worked flawlessly (even seemed to complete slightly faster too).

Good work John.

Many thanks,

Barry
Great news!
After I have added a few extra performance improvements, I will upload the new version to the PixInsight platform.
Thanks, John
 
Back
Top