CloudCompare forum

Hi! We have been using cloud compare with .LAS files that have extended classification codes that range between 0-122 (thus ignoring some of the flag bits) for quite a few years now. We need to have at least about 60 class codes for what we do so the standard bits0-4 which is feature codes 0-31 isn't sufficient for what we do. Recently with the newest cloud compare update, we were confused because of some changes that appear to have removed two features from Cloud Compare and maybe there are some hidden settings or tips and tricks you could suggest?

(1) File class codes appear to be limited to 0-31 with the remaining data split into flag bits. Is there a way to ignore flag bits and view the actual class codes as a full byte like we could in previous versions?

(2) We used to be able to double click any point to create a new camera orbit location and now we have to click the tool to select a new camera orbit position which is much clumsier. Is there a way to re-enable the double click option for new camera orbit position?

Thanks! :)

So for the easiest: the double-click feature still works (at least on my side, with the latest 2.13.beta versop, ;). Which version are you using?

And for the classification field: with the LAS 1.4 format, you can normally use 'extended classification' field (8 bits, with codes ideally following the guidelines on page 30 of the specs: https://www.asprs.org/wp-content/upload ... _4_r14.pdf)

So if you want to save your classes you should definitely target point format 6 or higher (and version 1.4). And to manage old files where you ignored the flags, then you may want to reload them with an old version of CC, and save them as BIN files, to open them in the more recent version?

Hi Daniel: Thanks for your response. For some reason, even though I have notifications enabled for my email, I didn't see any notification emails so I apologize for the delay in response on my part. Anyways, I checked my version of cloud compare and I am running the 2.13 beta version as well and it doesn't allow me to double click to update the camera orbit position but with the 2.13 alpha version it does. I have attached a screen recording video for you here: https://drive.google.com/drive/folders/ ... drive_link

Regarding the other issue with the .las versions, we have terabytes of the older .las files that are saved as .las-1.2 file format that have ignored the flag bits and instead just utilized the entire classification byte for feature codes. With the alpha version of cloud compare it still allows us to view the entire byte but the new beta version of cloud compare does not... So I loaded the 1.2 files into the alpha version of cloud compare and saved them as version 1.4 .las files and then loaded them into the beta version of cloud compare and it worked great! I realize that the older .las files that we have are somewhat breaking the rules but it would be really nice if the new cloud compare versions would provide a setting during import or in the main interface somewhere that would allow for some way to still utilize the old data that we have without having to convert each one. Especially since the older versions of cloud compare can handle that. I think I saw an import setting called decompose somewhere that did that in a previous version? Thanks again for the suggestions! :)

Yes I remember this setting. To properly manage all LAS file versions, on all platforms, we had to change the way LAS files are managed. We now use a different 3rd party library. Thomas Montaigu had to rewrite everything from scratch. I don't think we would have the time to change anything before the release of the 2.13 version sadly. But we may be able to do something later.

Are you able to proprely pick points with the point-picking tool?

Can you try with a more recent beta version? (10th of February ideally ;).

Hi Daniel: Thanks you for considering the feature and I completely understand that development takes time. I did download the latest version after you made the suggestions and then I tried the point picking and that tool seems to work just fine. I also tried the manual "pick rotation center" tool and that seems to work well too but the auto pick rotation does not respond to double click and will only auto pick the screen center if there are points there. Thanks, Joe

Oh, can you disable the 'auo-pick rotation center' feature?

See https://www.cloudcompare.org/doc/wiki/i ... ew_toolbar (the 'Automatically pick a point as rotation center...' option)

Hi, regarding the double click re-center feature, I think I've experienced this issue or something similar in several recent versions. Sometimes the double click will register, but the center of rotation is not where I double clicked, it's up and to the left. I found a work-around that worked for me, and it has to do with the display resolution of your computer. CloudCompare rescales depending on the resolution set for your monitor in the settings (I'm on Windows), and the click feature works properly when I change the scaling from e.g. 250% to 150% (such that all icons get smaller). It's only a bit annoying in that I have to use a magnifying glass to see what I'm clicking on ;)

Interesting. It's true that this part (the magnification) is more or less properly managed by Qt... It has always been an issue. Sadly, I'm not sure what can be done on our end...

Hi Daniel: After quite a bit of testing. I found some interesting comparisons between the 3 latest versions of cloud compare that you might be interested to know about. Here is a link to a small sample file that I used for testing: https://drive.google.com/drive/folders/ ... drive_link

CloudCompare_v2.13.0: This version appears to have a different zoom level for icons and text than the previous versions on my 4k TVs that I use for monitors as all text and icons are much larger. The double click for orbit location function doesn't appear to work right but that may be a function of zoom level like lweidner mentioned. Also, this version is very legalistic when it comes to ASPRS standards for classification codes since importing LAS v1.2 files with classification codes larger than 0-31 are incorrect. I wish the "decompose" option that appeared on previous versions was still available.

CloudCompare_v2.12.0: This version works well with large classification code numbers even with LAS v1.2 files since it has a "decompose" function on import that allows for all 8 bits to be used for classification rather than just bit 0-5 and the remainder for flags. However, I discovers that if I import an LAS v1.2 file and save it as an LAS v1.4 file then the coordinates are changed for some reason which can be proven by importing both the original and the newly exported files together... Very strange...

CloudCompare_v2.11.1: This is the latest version that works perfectly for everything that we do with none of the above listed problems. :)