[vtkusers] Blue artefact on complex scene with VTK 8.1 + QVTKWidget + OSX and Intel chipset

Simon Esneault simon.esneault at gmail.com
Wed Jan 31 10:33:19 EST 2018


After further investigation, it appears that the problem only occurs
when SetUseDepthPeeling is enabled on the 3 vtkRenderer and does not occurs
when we disable this option...
It used to work fine with VTK 7.1, so this might be related to the new
implementation of the depth peeling that was written for VTK 8, no?
Shall I fill a bug for this ?

Thanks

2018-01-30 23:48 GMT+01:00 Simon Esneault <simon.esneault at gmail.com>:

> I forgot to say that the same code does not produce any blue artefact on
> an Nvidia or ATI card, and we were not able to reproduce this visual glitch
> on Windows either.
>
> If that can help ...
> Thanks
>
> Simon
>
> Le 30 janv. 2018 10:17, "Simon Esneault" <simon.esneault at gmail.com> a
> écrit :
>
> Hello community
>
> After the migration to VTK 8.1 (from VTK 7.1), we are seing strange blue
> artefact in one of our scene as it can be seen in the following snapshots.
> And I'm having real hard time debugging that ...
>
> https://pasteboard.co/H5irkgl.png
> https://pasteboard.co/H5irBQ4.png
>
> The context is as follow :
> - Qt 5.3.2 on OSX  and Intel chipset (Iris or HD > 4000)
> - VTK 8.1
> - 3 layered vtkRenderer in one vtkRenderWindow in a QVTKWidget
> - GPUVolumeRendering on the first layer, some vtkActor/Polydata on the
> second layer, and some vtkActor2D/vtkTextActor on the third layer.
>
> This bug only appears after the view has been "reparented" (the user can
> choose different "layout") and is triggered when we add a new vtkPolyData
> to be rendered. If the user move the scene, all goes back to normal. It
> look like we're seing some garbage blue buffer, but I have no clue where to
> look at...
>
> Anyone have experienced something like this ?
>
> Thanks a lot
>
> PS: I know QVTKWidget is deprecated and we should move to
> QVTKOpenGLWidget, but our app is quite big and moving to Qt >= Qt 5.4 has
> some other implication so that is not solution for now ...
>
> --
> ------------------------------------------------------------------
> Simon Esneault
> Rennes, France
> ------------------------------------------------------------------
>
>
>


-- 
------------------------------------------------------------------
Simon Esneault
Rennes, France
------------------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://vtk.org/pipermail/vtkusers/attachments/20180131/3a1db8cc/attachment.html>


More information about the vtkusers mailing list