rank LUT ranges are not consistent in Catalyst? How to clamp?

In Situ Support
1

I am seeing a case with Catalyst (version 1) where LUT ranges vary across ranks. This leads to parallel volume rendering artifacts.

A printf here on a 4 rank runs results in:
rangeColor 4.81899e-05,0.967385
rangeColor 4.81899e-05,0.967385
rangeColor 4.81899e-05,0.962687
rangeColor 4.81899e-05,0.962687

This gives me blocky/processor artifacts as in the attached image. which go away if I hard code the range at the c++ level.

OSPRayVolumeChi_000000p=000.00t=090.00zoom=03
OSPRayVolumeChi_000000p=000.00t=090.00zoom=031024×768 104 KB

I would just clamp the range at the python level but I cannot seem to find a syntax that matches what the clamp to custom lut range button does.

I am on ParaView 5.10.1.

Thanks for any insights you might have.

2

Not sure why the behavior is doing this. The LUT is broken up into colors and opacity in the generated Python script. The colors will look something like:

pressureLUT.RGBPoints = [100.0, 0.231373, 0.298039, 0.752941, 150.0, 0.865003, 0.865003, 0.865003, 200.0, 0.705882, 0.0156863, 0.14902]

Here I have 3 sets of values in my LUT (each set of values is a 4 tuple). The minimum value is 100, the mid value is 150 and the max value is 200. The other parts of the tuple don’t ever seem to need to be modified but kind of seem to have to do with slope and such of the LUT.
Similarly for the Opacity I have:

pressurePWF.Points = [100.0, 0.0, 0.5, 0.0, 200.0, 1.0, 0.5, 0.0]

Here for some reason I only have the minimum, 100, and maximum, 200, values. This is what I get when I clamp my LUT in the GUI and export in a Python script. If you’ve played around with your LUT, you’ll have more values in your Python list and need to do some arithmetic.