There is a thermal dose volume (with value range 0-300) for which the module generation yields very incorrect isodose surface (for level value 240).

This is a correct-looking result

And here is the result yielded by the Isodose module

The only difference between the two results is that the top one was computed with the default FeatureAngle value (15), and the bottom with the value set by Isodose logic (60), see

Does anyone (@lassoan @gregsharp) remember why we chose this non-default value? There are no comments in the code. I found the commit, however, where this value was set (by Kevin Wang):
dc07b61
The comment was: "Fixed the isodose color in 2d viewer. It probably has something to do with the scalar and normals computation. Once I modified the pipeline to generate the polydata, it works fine now", so it was not clear to him either but there were 2D display issues before this change. That said, it was more than ten years ago...

If there is no good reason I'd remove this non-default feature angle setting. Thank you.