Feature Extraction

Notes

We start by creating the terrain using a simple Perlin Noise with a few octaves.

Next, we connect it to the Height Output to ensure the data will be used, and then extract features using the Get Slope and Get Cavity nodes. We use the data generated by these nodes to select specific regions—such as areas with high slope or regions where the terrain is relatively flat—and paint them using the Color Output.

Finally, using the Fold node, we transform the flat region mask into a selection of all concave and convex areas.

Node system

Result

Crop Pattern

Notes

We’ll start by creating a Voronoi Noise node and setting the Noise Type to Voronoi Index. This gives us flat regions with a single value across each region, which we can isolate using a Select Mask.

This mask defines the area where the pattern should appear. To smooth the boundary, we first blur the Voronoi noise with the Blur node, then reselect our region. Finally, we use Apply Mask to ensure that no values appear outside the selected area.

With our masked zone in place, we can use a Function node to generate the desired pattern and combine it with the Apply Mask node.

To break up uniformity, we then apply some warping to the entire pattern using the Warp node.

Node system

Result

Desert Pyramids

Notes

To place different pyramid-like structures, we start by creating a grid of points and randomly positioning them. We use the Jitter Points Node to randomize their position, rotation, and scale.

At each of these points, we create a pyramid using the Shape Node and spawn it with the Points To Height Node. By ticking the required boxes, we ensure that the sampled shape follows the points data correctly.

Next, we apply some stepping to the pyramids to add texture, making them look more like layered cubes.

We then combine the pyramids with basic noise to give them a less uniform look, as though there’s sand on top of each layer.

Finally, we blend these pyramids with some Simplex Noise to achieve the final terrain.

Node system

Result