[FIXED] Global shift- problem with original coordinates

[daniel]

Still couldn't reproduce the bug.

Here's what I did:
- load a file with big coordinates
- accept the proposed shift (-3.5e5, -5.3e5, 0)
- pick some points with the 'Point List Picking' tool
- export those points as a new cloud
- export this new cloud as ASCII

Shift is correctly applied and the points fall in the right global coordinate system...

Am I misunderstanding something else?

[MatBo]

Sorry for late reply.
i am trying to export only the picked points, not the cloud and there is the problem. the problem is solved if i go to global shift and delete the minus in front X and Y coordinate.

you can see the problem in the pictures below. Watch the original coordinates and the global shift.
this is without changing the minus in global shift

global_shift.jpg (116.18 KiB) Viewed 3799 times

and this is when i change global shift with deleting the minus

without_minus_global_shift.jpg (148.68 KiB) Viewed 3799 times

[daniel]

Okay I think I understand. It is normal to have a negative shift (because the shift is added to the original coordinates). The error is that the original coordinates displayed in the console (and on labels) are wrong! So this is only a display issue. If you export the cloud with the negative shift, it should be ok (you should be able to test it by saving the result as an ASCII file).

I've just fixed the display issue on the trunk.

[MatBo]

i exported the ascii file and also in ascii file the coordinates were wrong.

[daniel]

Do you mean you are using the 'Export (x,y,z)' method of the 'Point List Picking' tool?! Because that one was also buggy ;)

But if you export first the points as a point cloud, then save this point cloud as an ASCII file it should be ok even with an old version.

Otherwise I'm lost ;)

And in any case, I've updated the latest bet version:
http://www.cloudcompare.org/release/Clo ... in_x64.zip

[MatBo]

Hey!

it works fine now.

Thanks :D