Fractional Brownian Motion is just Perlin noise summed across geometrically-spaced frequencies: fbm(x)=k=0∑N−1gknoise(ℓkx),
where ℓ is the lacunarity (frequency multiplier between octaves) and g is the gain (amplitude multiplier). With ℓ=2 and g=0.5 you get the classic 'pink' fBm whose power spectrum falls as 1/f2, mimicking the statistics of clouds, terrain, and turbulent fluids. The main panel renders the composite 2D field as a heightmap on a slate-to-amber palette; the strip on the right shows the contribution of each individual octave k with label gk, dimmed toward neutral grey when its amplitude is small — so you can read off exactly how much each layer is adding. Drag the cursor vertically to scrub the gain g∈[0.2,0.9]: at low gain only the coarse base layer survives and the field is smooth and blobby; at high gain the fine octaves contribute heavily and the surface turns wrinkled and craggy. Click to cycle the octave count N through {1,2,4,6,8} — going from N=1 to N=8 at fixed gain visibly carves detail into the same large-scale shapes without changing them, which is the whole reason fBm is the workhorse of procedural texture synthesis.