When I blast a primitive group with an external merged bounding object, and disable render visibility for the bounding object by enabling the render flag for a Null inside the SOP, the group selection is no longer working in the render and nothing is blasted.

In the viewport all seems fine but when I render with Redshift or Mantra the groups are ignored.

The same goes when I only disable the Display visibility of the bounding object, the group selection is ignored in the viewport, but works fine in the render.

I attached an image with a Viewport and Render view, on the left the render flag of the bounding cube is disabled, on the right the display flag is disabled.

Doesn't seem logical to me. Is this a known issue, or am I missing something ?

Cheers,

R.

The Display and Render flags don't define the visibility itself

They are marking the geometry that will be output during Viewport Display or during Render respectively in case you don't have dedicated Output SOP with Output flag

So what you are doing is essentially telling that object that during Render this object has no geo, therefore it has no bounds and hence the group created from it's bounds will not contain any elements to delete
Same way if you flip the flags and now the Display has no geo

You are probably doing Object Merge in your sphere geo reference Cube Object and using it's geo as bounds
However if you reference just the object itself it will return geo following the flag (Display, Render, or Output)
Causing behaviour you se

The safest way may be to Object Merge a specific SOP from within the Cube Object whose geometry you want to use therefore getting stable results regardless of the flags

And also if it's just render/display visibility you are after I'd use Display flag in the Object level
And Force Objects in the render, therefore you are not messing with SOP flags

Thanks, when connecting the merge node of the bounding object to a null SOP instead of the object root node, the render or display toggle has no effect on the calculation. So that solves the problem.

But still, the logic goes over my head. Why would one have render and display settings have effect on the calculations ? It’s like having an object set to wireframe in the display, render as a wireframe as well.

And why is dis- and enabling display and render settings so convoluted ?

But that’s the magic of Houdini I suppose

R.

toonafish
Why would one have render and display settings have effect on the calculations ?

Because they are not render and display settings of the object
They are flags that marks what is calculated during Display and what during Render

So they are exactly meant to have effect on what geometry is output during cooking for Display/Render

It's not the most intuitive mechanism since it's not easy to preview in the viewport what would happen during Render
I prefer not to separate those flags or even enforce some outputs with Output SOP especially in subnets for consistent results