The herringbone floor problem.....

Skatter is brilliant at random distribution of organic forms, but we use Skater’s grid to create flooring and paving and tiling within our models.

Typically we have to set up a new composition for each tile or floor plank.

With a herringbone floor, it’s only two. One for each floor plank. We can mirror and give a bit of rotation. Brick walls maybe 3 or 4.

We export to Octane where were can apply gradient ramps nodes to render out with random materials.

The object just has a uv map applied in SU.

The new convert to composition allows converting a layout to point distribution which i’d originally thought might simplify this process.

What would be a killer feature for me is to lay out the tiles/planks in a small area and for the plug in to make the grid distribution which I can then apply to different floor walls etc.

Profile Builder 3 has a similar function for its creation of assemblies. You layout the profile or components relative to the red axis and the automatic creation makes a good guess of the distribution of the elements.

It’s a massive time saver and means we use the plug in more. Even if we need to tweak after. 8 times out of 10 it nails it.

Boundary Clipping.

The other is the boundary for grid objects. If there was any way for the boundary to literally clip the object along the boundary, rather than have by origin point or object radius to limit the distribution. Oob tiles kind of does this, but with none of the parametric that skatter offers.

Even if there was an option where any part of the object is within the surface area it’s included would be a help.

At present, we get around this by oversizing our distribution areas and having a border that is a few mm higher to mask it.

Thank again for the great work.

The issue is that Skatter only generates component instances. To clip objects like that, it would need to generate many unique copies of the objects, each with its own altered geometry.
It is something I’ve been thinking about for a while, even prototyping with section planes, etc. But I think it’s just too complicated technically, and it would potentially flood the model with unique components.

Considering how Skatter works internally, this is not possible. It would need to be able to generate insertion points outside of the host surface. But how would it be able to know how to expand the surface?


Thank you for taking the time to reply and for the thoughtful response.

Thanks again for great work developing Skatter with SU.


1 Like