Terrain Reader

"The song still remains which names the land over which it sings." Heidegger

Travelers on digital terrain experience it in different ways. One computes the cross-section of a linear path over the terrain, and use it as an audio waveform, spinning the line of the path about its center (like wearing a propellar beanie with a miles-long elevation-sensing prop). Another slows down when going uphill, and speeds up going downhill. Yet another, like a drunken sailor in a motorboat who passes out at the helm until hitting the opposite shore only to shove off and pass out again, sends out tendrils that search for the shore like retractable curb feelers, and uses the feeler lengths to control musical parameters.

Some examples of (spinning) cross-sections as audio:

Mt. Tamalpais, Marin County, California (98K)

Berkeley, California (120K)

The Grand Canyon, Arizona (126K)

Read an informal history or a conference paper about the project.

Download the source code.

The Terrain Reader was performed at the Exploratorium on Mondays at 11:00 am, from June 20 through Sept. 5, as part of the Exploratorium's "Interactive Sound Studio."

The latest version (for the Macintosh) will appear on a forthcoming CDROM from Compuserve. Here's a sample (500+K) of how it sounds. This version generates MIDI data.

The first "Terrain Broadcast" was performed live on KPFA-FM in Berkeley, CA on Barbara Golden's Crack O' Dawn in August 1990 and October 1991. The first minute or so of the CD mix is available as 8-bit audio.

In a related vein, you can view an archive of scenery collected from Bill's office window, overlooking the San Francisco Bay.

Read what Chief Seattle had to say about the land in 1854. (Actually written in the seventies....)

Here are some images generated from DEM data, from the U. S. Geological Survey. (An ftp archive of DEM data is available on spectrum.xerox.com) The surface was modeled as metallic using Blinn's reflection model. Light source positions vary with each image.