scalar fields getting damaged

[PablerasBCN]

just to double check,

I did interpolate the original Intensity into the broken one, it looks seamless.

PCV also got damaged,





PCV was computed prior to separating the ground so they should blend seamlessly

so, as of now, it looks like at random moments, upon merging the clouds, the values of some SC are shifted, PCV I've seen a lot Intensity I had not noticed yet until this moment.

[daniel]

Yes, it seems there's still a bug somewhere... If you ever found a way to trigger this bug in a reproducible way, don't hesitate to send me an email!

[daniel]

Ok, the issue should be fixed now! See the updated 2.14.alpha version.

[PablerasBCN]

tx Daniel!! such a tricky one.

I'll update to the new beta and be aware to see if it ever triggers again

[PablerasBCN]

sadly, we're back

I merged the two parts while I had Intensity enabled so I was able to see it fail in front of my face.

Checking the other scalar fields the error also occured in other SF.

I uploaded to drive, a couple of sample images of damaged SF, clearly visible seam on what should be semless SF.

Also I included an screenshot of a couple of screens that likely will be a clue of what breaks. As when trying to save as .las some stuff does not add up like it seems like PCV SF's have been duplicated. Also some warning symbols on som SFs.

In addition 2 .bin one with the damaged data other with the undamaged data, that is not too far back from the damaged data in terms of work. I just did segment some points for cleanup and upon merging the bug kicked in.

https://drive.google.com/drive/folders/ ... sp=sharing

hope this data helps to find the issue.

CC version Jan 11 2025

[PablerasBCN]

There is a chance that the exact conditions prior to the merge could be replicated.

if "numbers of returns " SF is enabled, the one part has bright green, Segment out some noisy points on the left, and you should have the exact original 2 parts.

Then, transfer original SF data to the parts by importing the undamaged bind and interpolating the damaged SF's and it should be the exact scenario prior to the merge.

I tried but takes forever. Let me know if you whant this so I leave it overnight.

[daniel]

Ok, this time the issue was when SFs were loaded from a BIN file (with some conditions).

This should now be fixed with the latest 2.14.alpha release online.

Thanks for the feedback!

[PablerasBCN]

Thank you so much Daniel for the fast fix.

I'll update and report if ever comes back. Cheers!