Image Perlin Power Fractal

The pack's plain Image Perlin Noise node gives you a single layer of Ken Perlin's classic gradient noise - smooth, organic, good enough for a lot of texture work on its own. This node is the fancier version: a "power fractal" is the standard term for stacking multiple octaves of that same noise on top of each other at different frequencies and amplitudes, which is how terrain generators, cloud renderers, and procedural texture tools have made richer, more detailed noise for decades (you might also know this technique as fractal Brownian motion). More layers means more fine detail riding on top of the broad shapes, at the cost of more computation.

Where it sits among the pack's noise nodes

WAS actually ships three related nodes worth telling apart:

• Image Perlin Noise - single-octave Perlin, plain and fast, image output.
• Image Perlin Power Fractal (this one) - multi-octave layered Perlin, richer detail, still image output.
• Perlin Power Fractal Latent - the same layering idea, but generating directly in latent space to feed a KSampler as init noise instead of the usual random Gaussian start. Per the README that one "doesn't work with all samplers (unless you add noise)," making it the pickiest of the three.

If you want a texture you can actually look at, composite, or use as a mask, you want an image-space node - either of the first two. This one is the one to reach for when a single Perlin layer looks too flat or too simple for what you're building.

Why you'd use it

Anywhere a plain Perlin layer feels thin: more convincing marble, cloud, terrain, or organic-noise textures; a displacement map with real fine structure for the pack's Image Displacement Warp node; or a busier procedural background than single-octave noise can give you.

The inputs and outputs that matter

The output is a plain image - a noise texture you use directly. Given the "fractal" framing, expect controls around the number of octaves/layers and how quickly detail falls off between them, alongside the usual seed for reproducibility, but I can't confirm exact widget names from a schema so I'm not going to invent them here.

How to install it

Bundled with the full WAS Node Suite:

• ComfyUI Manager - search "WAS Node Suite", install, restart.
• Manual - cd ComfyUI/custom_nodes && git clone https://github.com/WASasquatch/was-node-suite-comfyui, install requirements.txt (portable: python_embeded\python.exe -s -m pip install -r requirements.txt), restart.

Pure procedural math - no models to download.

Common issues & troubleshooting

It's noticeably slower than plain Perlin Noise. That's expected - you're generating multiple noise layers and combining them instead of one, so more octaves means more work per pixel. If it feels sluggish at high detail settings, that's the fractal layering doing its job, not a stall.

Unexpected results run-to-run. Like the plain Perlin node, this is deterministic given a fixed seed. If your texture keeps changing when you didn't intend it to, check whether the seed itself is changing between runs.

Standard pack caveat: WAS Node Suite has been unmaintained since December 2023, and the recurring issue is the whole suite failing to import after a ComfyUI update, usually a stale pinned dependency (BLIP support is the repeat offender). Reinstall requirements.txt against the right Python environment if the suite won't load.