Burning Ship Fractal

The Burning Ship repeats the calculation of the Mandelbrot set with one extra operation. A pixel's horizontal and vertical coordinates give a number to add, called cc, and the point starts at zero. Each jump then makes both of the point's coordinates positive or zero, squares the point, and adds cc. The first of those operations is the fold: a point left of the vertical axis flips to the right, and a point below the horizontal axis flips upward.1

Small Burning Ship with repeated towers above its hull
The Little ship view shows a smaller hull with its own rows of masts.

In the route below, the dotted line shows the fold whenever a coordinate is negative. This picture draws positive vertical coordinates downward to keep the ship upright, so on screen a fold moves a point down. A small change in the pixel can change the route a lot: pick a pixel just left of the ring, and at (−1.8,−0.03)(-1.8, -0.03) the point crosses the dashed circle on jump 7, not jump 12.

Follow one pixel

Pick a pixel to test it. The point starts at 0. Each jump folds and squares it, then adds the pixel's coordinates.

JumpPoint
0(0, 0)
1(−1.76, −0.03)

Jump 1x² − y² + a2xy + b
Pointx = 0y = 0
Fold00
Square0² − 0² = 02 × 0 × 0 = 0
Add c0 − 1.76 = −1.760 − 0.03 = −0.03
New pointnew x = −1.76new y = −0.03

Under the picture, the current jump is worked out by hand and rounded to five decimals, and its fold row removes the minus signs. Next jump feeds the new point back in as the next starting point.

A point that crosses the dashed circle has escaped, so a colored pixel is one that escaped, and its color shows how many jumps that took. Black pixels haven't escaped yet. The computer stops after a fixed number of steps, so some of them may escape later.

The Mandelbrot picture looks the same upside down, and the ship doesn't, because of the fold. Take the pixel (−0.5,−0.5)(-0.5, -0.5) and its mirror image across the horizontal axis, (−0.5,0.5)(-0.5, 0.5). The first jump lands each point on its own pixel. The fold then moves both points to (0.5,0.5)(0.5, 0.5), so squaring gives both the same answer, (0,0.5)(0, 0.5). Adding each pixel's numbers puts them on (−0.5,0)(-0.5, 0) and (−0.5,1)(-0.5, 1), which are no longer mirror images.

A pixel and its mirror image

Pick a pixel. The second ring is its mirror image.

JumpPixelMirror
0(0, 0)(0, 0)
1(−0.5, −0.5)(−0.5, 0.5)
2(−0.5, 0)(−0.5, 1)

From there the first point stays inside the dashed circle and settles near (−0.41,−0.28)(-0.41, -0.28), while the second crosses it on jump 4.

Choose Without the fold and the picture becomes the Mandelbrot set. A point and its mirror image now square to answers that are mirror images too, so the two routes stay mirror images at every jump and both pixels get the same color.

Along the masts are smaller ship-shaped regions. At their centers, the calculation returns to zero after a fixed number of steps, so the same sequence repeats without escaping.1

Between the smaller ships are thin, branching shapes that resemble Julia sets, called embedded quasi-Julia sets.1 Places to explore includes three views from Heiland-Allen’s gallery under Embedded Julia shapes.2

Further exploring

Sources and notes

  1. At the Helm of the Burning Ship.
  2. Burning Ship embedded Julia sets.