Editing the Shadow Volume

[Image: The multiple-shadow casting cube by Niloy Mitra and Mark Pauly].

Spotted via New Scientist is an amazing new computer model that allows designers to create objects based on the multiple and highly specific shadows that those objects will cast when lit from different angles.

Seen above is one, relatively mundane example of the technology, by Niloy Mitra and Mark Pauly: three paintings by Andy Warhol are being cast from the same object. “Their computer model can calculate the object shape needed to cast up to three distinct shadows simultaneously,” New Scientist explains. The designers call it “editing the shadow volume.”

Niloy’s and Pauly’s accompanying video is amazing:


But what if we could do this with a glass tower in midtown Manhattan? Or if there was an elevator moving upward through an all-glass shaft, and as the lights in the lobby around it switch on and off, different—often wildly unexpected—shadows are cast within the building?

What are the architectural possibilities of multiple-shadow casting design?

You hook this modeling software up to huge CNC-milling machines, and then you attach the whole assembly to a warehouse-sized block of plywood. You come back one week later to find a sprawling labyrinth of immersive three-dimensional shapes carved directly and seamlessly into the wood, like the mathematical spires of some alien cathedral—it’s an extraordinarily beautiful landscape of precision-cut wood—but it’s only when the lights go off above you and a wall of klieg lamps on the northern wall switch on that you see the jaw-dropping shadows this wooden landscape can cast. But then those lights turn off, and the eastern wall lights up—and more, incredible, seemingly contradictory shadows appear. Then the west wall.

Each time, an impossibly unique scene of shadows is displayed, often too complex to be believed. It is Wayang Kulit for an age of semi-intelligent milling machines and theatrical light.

Or perhaps someday the perfect, cinematic object will be designed: it rotates in all directions amidst a battery of programmed lights, and the shadows that it casts are narrative, moving scenes in a two-hour film, displayed on the walls around it.

Instead of DVDs, we will store our movies in the cuts and grooves of milled wooden objects. Mahogany harddrives. Spirit-objects brought to animate life by angled light.

17 thoughts on “Editing the Shadow Volume”

  1. The movie idea is amazing and weirdly opposed to the procedure for making bullet time effects. The idea of combining arcs of cameras, arcs of lights and these mahogany objects – imagine the dizzying bizarre filmic possibilities.

    Or a rotating physical manifestation of video feedback with moving parts, hinged cubes and blocks sliding up wires or tiny balls that inflate to make pixels.

    Really enjoy these speculative posts, thanks.

  2. Perhaps slightly off topic but with the recent Carl Sagan Day in mind I was thinking what one of these shadow objects would look like carved into a hypercube? How much information might be contained in it? and what the shadows cast into three dimensional space would look like?

  3. This reminds me of a script I tried to write a while ago in architecture school. The idea was to calculate what size and depth of opening was required in a 3D volume in order to cast a patch of light on a given surface within the volume, and then allow this function to repeat through multiple iterations.

    A machine-gun chain of times and dates ensures that a sculpture is bathed in light at the same morning hour every day of the year. Or a pool is covered by a perforated roof, through which the sun casts complex spiralling patterns at it moves throughout the year.

    Or perhaps, even, the sun learns to track your movements, following you from bed to your favourite coffee spot, from the shower to the door,

  4. Looks like the cover of Godel, Escher, Bach by Douglas Hofstadter- He created a tricharacter of GEB ON A BANDSAW to illustrate the interlocking themes of his book.

  5. The model represented in the first image of this article is very simple to create. In Rhino 3D for example, you'd simple extrude each image to form a 3D volume, and use the "boolean difference" tool to subtract this volume from a cube. Repeat for the other two dimensions, and presto. Great idea though, definitely.

  6. Editing the intersection is definitely the interesting part. One artist has made this inner space the focus of his career: Larry Kagan.

    You can do this with any computational geometry tools that allow intersection and extrusion, but keep in mind you might need to model the light as a point source (conic shadow) rather than using orthographic rays (cylinder/standard extrusion).

Leave a Reply to Pete Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.