Julia Set Fractals

One equation, infinitely many fractals

Launch Julia Set Visualizer Launch Explorer

One Formula, Infinite Fractals

The Julia set uses the same iteration as the Mandelbrot set: z → z² + c. The difference is what you hold fixed. The Mandelbrot set varies c and always starts at z = 0. A Julia set picks one specific c and iterates every point in the complex plane as a starting z. Each value of c produces a completely different fractal.

That's the part worth pausing on. The Mandelbrot set is one image. There is one Mandelbrot set. But there are infinitely many Julia sets - one for every complex number you could choose as c. Some are lacy and connected. Others are scattered dust. The Mandelbrot set is actually a map of which c values produce connected Julia sets and which don't.

Julia Set visualization Julia Set visualization

Good Values of c to Try

Not all Julia sets look interesting. Many are just circles or formless clouds. The best ones live near the boundary of the Mandelbrot set, where connected and disconnected regions meet.

A few classics worth knowing:

  • c = -0.7 + 0.27i - Produces branching, tree-like spirals. One of the most commonly shown Julia sets for good reason.
  • c = -0.8 + 0.156i - Gives a more elongated, organic shape with fine tendrils at the edges.
  • c = -0.123 + 0.745i - Douady's rabbit. Three-lobed structure that self-replicates at every scale.
  • c = -0.75 - Right on the real axis, where the main cardioid meets the period-2 bulb. The Julia set here is a dendrite: connected but infinitely thin, with no interior area at all.
  • c = i - A dendrite with interesting rotational properties. Clean and geometric.

Move c even slightly and the whole structure shifts. That sensitivity is the point - it's what makes animating through c values so effective.

Connected vs. Dust

Every quadratic Julia set is either fully connected (one piece) or totally disconnected (a Cantor set - scattered dust with no solid regions at all). There is no in-between. This is the Fatou-Julia theorem, and it's a surprisingly clean result for something that produces such wild images.

The test is simple: iterate z = 0 with your chosen c. If the orbit stays bounded, the Julia set is connected. If it escapes to infinity, the Julia set is dust. That's exactly what the Mandelbrot set computes - the black region is the set of c values where 0 doesn't escape, meaning those c values have connected Julia sets.

History

Gaston Julia published his major work in 1918, at age 25, while recovering from injuries sustained in World War I. He'd lost his nose in combat and wore a leather strap across his face for the rest of his life. His paper on the iteration of rational functions ran to 199 pages. Pierre Fatou was developing overlapping results independently around the same time - the two were aware of each other's work but approached the problem from different angles.

Neither of them could see what these sets looked like. That had to wait sixty years for computers. When Benoit Mandelbrot generated the first images in the late 1970s and early 1980s, the visual complexity hiding inside these simple iterations was finally visible. Julia's theoretical work had predicted the broad structure, but nobody expected the level of detail.

Exploring It

The visualizer animates through different c values, so you see the Julia set continuously morphing. Each animation mode traces a different path through c-space:

Circular Orbit

Moves c in a circle around the origin. Steady and predictable - good for seeing the full range of shapes as c sweeps through different regions.

Lissajous

Traces figure-8 paths in c-space. The real and imaginary parts oscillate at different frequencies, so the path revisits some areas while skipping others.

Random Walk

Wanders randomly. You get abrupt transitions between very different Julia sets, which can turn up shapes you'd never find on a smooth path.

Boundary Following

Keeps c near the Mandelbrot set boundary, where the most detailed Julia sets live. This avoids the boring regions (solid circles, featureless dust) and stays where the structure is richest.

Interactive

You control c directly with keyboard or touch. WASD moves through parameter space, Q/E zooms. Best way to find specific shapes you like and study them.

Further Reading