Volume plot doing something weird

[SterlingButters]

Describe the bug, what's wrong, and what you expected.

I was attempting the example posted here:
https://docs.pyvista.org/version/stable/api/plotting/_autosummary/pyvista.Plotter.add_volume.html
and got something that looked like the figure the first time. Now I get something that looks like:
https://pasteboard.co/XENz46LHrFa1.png

What is happening?

Steps to reproduce the bug.

import pyvista as pv
grid = pv.ImageData(dimensions=(9, 9, 9))
grid['scalars'] = -grid.x
pl = pv.Plotter()
_ = pl.add_volume(grid, opacity='linear')
pl.show()

System Information

--------------------------------------------------------------------------------
  Date: Wed Dec 06 12:34:36 2023 Mountain Standard Time

                  OS : Windows
              CPU(s) : 12
             Machine : AMD64
        Architecture : 64bit
                 RAM : 31.7 GiB
         Environment : IPython
          GPU Vendor : NVIDIA Corporation
        GPU Renderer : Quadro T2000/PCIe/SSE2
         GPU Version : 4.5.0 NVIDIA 471.11
    MathText Support : False

  Python 3.9.13 (main, Aug 25 2022, 23:51:50) [MSC v.1916 64 bit (AMD64)]

             pyvista : 0.42.3
                 vtk : 9.2.6
               numpy : 1.24.4
          matplotlib : 3.5.2
              scooby : 0.9.2
               pooch : 1.8.0
              pillow : 9.2.0
             imageio : 2.19.3
               PyQt5 : 5.9.2
             IPython : 7.31.1
            colorcet : 3.0.0
          ipywidgets : 7.6.5
               scipy : 1.9.1
                tqdm : 4.64.1
          jupyterlab : 3.4.4
               trame : 3.3.0
        trame_client : 2.13.0
        trame_server : 2.12.1
           trame_vtk : 2.6.2
       trame_vuetify : 2.3.1
jupyter_server_proxy : 4.1.0
        nest_asyncio : 1.5.5

  Intel(R) oneAPI Math Kernel Library Version 2021.4-Product Build 20210904
  for Intel(R) 64 architecture applications
--------------------------------------------------------------------------------

Screenshots

No response

[SterlingButters]

Of note, there are no issues when I export the plot to html and render in the browser. Appears to be an issue specifically with the plotter.show() method

[banesullivan]

Typically I'd say the artifacts you are seeing are a result of hardware (GPU) but I think a Quadro T2000/PCIe/SSE2 should be plenty good, otherwise a Windows driver issue.

For me, this plot looks like:

[SterlingButters]

Any debugging ideas or driver replacements I can make?

[banesullivan]

Sorry, I should have given some direction in my above comment. This can be really tough to debug on Windows.

One thing you could try is to save that mesh out to a file and try to render it in ParaView:

import pyvista as pv
grid = pv.ImageData(dimensions=(9, 9, 9))
grid['scalars'] = -grid.x
grid.save('myvolume.vtk')

Then load that file into ParaView. Once in paraview, see if you have the same issues (as ParaView and PyVista are both using VTK under the hood). Then in paraview, you can experiment with different volume mappers to see what looks best.

If you can find a mapper or other rendering parameters in ParaView that look good, then we can mimic this in PyVista. If it continues to look bad in ParaView, then we can perhaps seek help from https://discourse.vtk.org/ on how to get this to work