Arduino, c++, Processing
comment 1

Four Letter Words [Arduino, Processing, c++]


About a year ago, Rob Seward created the Four Letter Words piece. The original video now counts about 111k views on vimeo and has been blogged by numerous sites out of which I think Pieter and Rhizome were the first. Earlier today, I got an email from Rob about the latest video he made, that was projected on a screen hung between two trees, with several other sound and video installations in the woods nearby. It was made using After Effects with sound in Ableton Live and using the FLW installation as source material (see bottom of this post). I thought the project needed a re-visit, looking at ins and out of how it actually works, what makes is tick as they say. After few emails back and forth with Rob, here are the details:

The installation consists of four units, each capable of displaying all 26 letters of the alphabet with an arrangement of fluorescent lights. The piece displays an algorithmically generated word sequence derived from a word association database developed by the University of South Florida between 1973 and 1998. The algorithms take into account word meaning, rhyme, letter sequencing, and association.

There’s a mac mini running Processing that sends alignment data to 4 arduino boards (one for each letter) that are chained together. The positions of the lights are stored in an XML file. There is an app that allows Rob to tweak the positioning in case anything gets out of alignment (see first image below). There’s another app just takes what you write on the keyboard and sends it straight to the machine – that’s what was used in the A-Z section of the video. The third app reads lists and sends words to the sculpture. Rob describes it as a bit more complicated than he thought it would be because there are certain transitions the sculpture cannot do without intermediary positioning of the lights. For example, if S goes to D, the top and bottom lights will collide, causing the machine to jam. The processing app makes sure that none of these problem transitions occur without inserting an intermediary arrangement of the lights that allows them to move safely.

For installations, the words lists are derived from some C++ apps Rob wrote. You can find more information about them here (second image above).  The words you see in the video are put together by association. Thus DEER goes to HUNT goes to KILL. KISS goes to LIPS. The words that it’s choosing tend to have more negative associations.

The other two images above show text with english-like letter ordering (see third image above). Rob made it by modifying a Markov-chain ruby script. The software, written in Processing also places 4-letter words adjacently (fourth image).

The installation in total includes 4 arduinos, 20 servos, 8 Step motors, 24 3.9 inch CCL (cold cathode) lights and their inverters. Each arduino has 2 steppers, 5 servos, and 6 lights to control. There are 2 custom shields on each arduino – one for the lights and one for the motors. I wrote a library to operate the servos and stepper simultaneously which you can download here (github).

While the piece was conceived with idea of displaying algorithmically generated lists, it was designed with flexibility and expandability in mind. The individual units can be connected ad-infinitum, and are theoretically capable of displaying any length of text. While Four Letter Words deals with a specific range of content, the technology can be easily expanded for future textual experiments.

Thanks Rob!

Rob Seward is an artist and programmer. His work has been exhibited at the Blanton Museum, Austin; CVZ Contemporary, New York; Center For Opinions in Music and Art, Berlin; and Nova Scotia College of Art and Design, Halifax. He has lectured at the Centre Pompidou, Paris; Columbia University; and Location One, both in New York. He holds a master’s from the Interactive Telecommunications Program (ITP) at New York University’s Tisch School of the Arts. Before getting his master’s, he worked in collaboration with composer Fred Lerdahl creating software based on the Generative Theory of Tonal Music. Rob lives and works in New York City.

Previously: Kunst Bauen [iPad, iPhone, oF, Mac] – “interactive artwork

  • Paul_Rand

    the action of the type is excellent!