CloudCompare forum

Hello everyone,

First of all I would like to thank for the development of CloudCompare as it is a great software for manipulating/analyzing point clouds. Recently, I think I have discovered something that is not giving me the desired outputs?

Problem 1:
I have this huge point cloud that weighs 3.5gb and is already rotated so to be parallel to the "set front view". The intent is to make a screenshot ("render to file") in order to have a sectioncut/elevation of the building. But when exporting the image (no matter what the "point size" or "zoom" is) it does not display every point of the point cloud.

To better explain here are some images:
A)screenshot to show the density of the point cloud . Note that point size is set to 5. B)render to file of the point cloud. Here, the point cloud has been segmented. The part that is more dense is the small segment. The rest of the point cloud, which is more sparse, is part of the bigger point cloud. In other words, I found that as the point cloud size increases the fewer points are taken into account for making the final image. Inversely, the smaller the point cloud the more points are represented. Is it due to hardware? Or is it something that I am not doing right?
I am running on 96gb RAM, AMD Radeon R9 200 series, 2x Xeon processor.

Thanks in advance,
João

What are the 'render to file' parameters? Have you set a magnification zoom? In this case have you unchecked the 'don't scale features' option?

And by the way, what is the version and OS you are using?

Thank you for the reply.

It happens with any 'render to file' parameters. I can choose 'zoom' from 1 up to 10, change 'point size' from default or 1 up to 10, and still happens. I have also unchecked/checked 'don't scale features' and even turned off the 'Render overlay items'. It makes no difference at all. The only thing that does make a difference is to slice the point cloud into many parts, as if each part weighed approximately 200/300 mbs(no need to save point clouds) instead of the single 3.5gb point cloud.
The machine is running on Windows 7 and CC v.2.6.3.

I have tested on my laptop (8GB RAM, GeForce GT 740M, i7, Windows 10, CC v.2.6.2)
opened a a 1gb point cloud + 0.6gb point cloud together. Took a screenshot of an area and everything was sparse (C). I made a really tiny slice and when a new render was processed surprisingly everything was normal (D) (I was expecting only the really tiny and small segment to be with normal density as shown in screenshot#2 in post #1).
For these I chose 'zoom' 10, unchecked 'don't scale', checked 'render overlay items', 'point size' default.
C) D) João

That's weird...

Can you maybe uncheck the 'load clouds on GPU for faster display' option in the Display settings? That's the only reason I see why the cloud size would make a difference (i.e. it doesn't fit on the GPU memory).

And with the GeForce, changing the default point size doesn't have any effect on the result either?

The 2.7 version is coming out soon. It uses a much newer version of Qt, and the GPU management is much better. Hopefully it will help.

Actually the option 'load cloud on GPU for faster display' was already unchecked. I checked it and tested. Everything seems to work very well now!! /like :)
I have a suggestion, perhaps could the option "load cloud on GPU for faster display" be also available on the 'render to file' window?

About the GeForce, changing the parameters also had no effect. However, I did some more tests today and the problem didn't happen again...kind of weird? I will try run some more tests tonight (load a lot more point clouds and see if it happens again. If so, I will try out the "load cloud on gpu" option on and off).

In any case, thank you very much!!
João

In fact this option is activated by default on all systems... but the ones with ATI cards ;). I still don't understand why it works only with this option enabled, but it may be another ATI mystery... (since the very beginning of CC we have issues with the OpenGL support of ATI drivers).

Anyway thanks for the feedback!