Debate Silences
More silence videos... This time its the 2008 presidential and VP debates. This one is the VP debate between Palin and Biden. See my Vimeo page for the other 2008 debates. These were all edited using custom software that analyzes the audio track to decide what to edit out.
sdfTwo prints and a video in Cleveland, OH. The prints include body scans from Pittsburgh, and TV face averages, and the video uses voluminous face data captured in the process of creating the faces print. A link to a review...
APRIL 2007
sdfInteractive video. To start the video playing, the viewer must press a button. Pressing this button causes an electrical shock in the hand, but in return the viewer sees the video of the previous person being shocked. To my surprise, over 300 people interacted with this piece on opening night.
APRIL 2007
Slow Motion
sdfHigh speed camera shooting at 1000fps. The fraction of a second where the subjects react to the stimulation is drawn out into several minutes. We see them produce a wide variety of micro expressions, fear, anger, sadness and confusion.
MARCH 2007
Thesis Exhibition

sdfCarnegie Mellon thesis exhibition. A collection of hybrid faces for our poster. Combinations of Jan, Gunnhildur, Lauren, Dave, Ian and myself. Mostly referred to as creepy.

The Face of TV
sdfCustom software, automatically gather all faces it finds on TV to compute an average. Here I examine eight different cable TV channels, gathering thousands of faces over the course of a day. Here we have Fox News, CNN, MSNBC, BET, MTV, CSPAN, PBS, and CBS.
The Average Art Student
sdfHere a public video installation uses face detection to capture and align images of people in the art building at Carnegie Mellon. After a few days, and hundreds of participants, the average image looks something like this.
MAY 2006
I interrogate my friends. They are wearing headphones that delay their auditory feedback producing confusion and frustration. Bright lights respond to the volume of their voice. But somehow, they manage to fight through the distraction and confusion to say some amazing things.
When Pulse Becomes Pitch
sdf Solo show at AV in Rochester, NY. Collaboration with Andrew Johnson. This piece is interactive. You speak into the speaker horns, and out comes a voice that is not your own. When you speak it speaks. When you are silent, it is silent.
Cycling74 interviews are online
sdf It's an honor to be interviewed for Cycling74, the SF based group that creates the Max/MSP/Jtter software I use. Sue Costabile interviewed artists, designers, musicians and researchers who use the software. Check out the interview...
Oral Fixations at SIGGRAPH 2005
sdf Eight hour animation. Produced at Carnegie Mellon by James Duesing, Jessica Hodgins, David Tinapple, Moshe Mahler, Jao O'Berski, Sang Il Park, Bum Lee. The animation ran all day starting each morning at the entrance to the SIGGRAAPH art exhibition. For this I wrote software to synchronize and composite about half a million HD frames. Also wrote software to algorithmically compose the seven hour soundtrack based on the physics simulation data used to render the animation.
MAY 2005
Scanned Body Portraits
sdfPittsburgh engineers submit to being "scanned" while standing in front of their cubicles. These portraits are made using a high resolution digital SLR mounted on a custom robotic motion rig. A single scan takes 60 seconds, and generates 100 photos as the point of view shifts slowly from head to toe. Custom software then stitches the photos, taking advantage of the parallax errors between photos.
Silence Extraction

Custom software to automatically edit out all but the silences in video. Here are all the quiet parts of the first 2004 presidential debate, over 1200 silences in all. Surprisingly this 90 minute debate consists of more than 15 minutes of silence...breathing, gesture, pause. Also, even without the words, there is clearly a debate of appearances.

Appeared on the Chaise DVD

Media Landscape DVD

A river of media chronicling the images and rhetoric leading up to the 2003 invasion of Iraq. 100 days float by in 15 minutes, starting at the SOTUA and ending with MISSION ACCOMPLISHED. Moments selected by hand from a personal video library of 800 hours of news footage from this time period.

The DVD can play on its own but is intended to play in staggered multiples... playing multiple looping copies of the video side by side, with the playback staggered so that a group of displays flow one into the other, as one large screen.


APRIL 2003
Video Timeline Exhibition
sdfWar is clearly on it's way. And the media seems afraid or unable to ask hard questions, swept up in the push towards Iraq. I can't escape the urge to get this on tape... all of it. Like a car wreck in slow motion. So I record cable news coverage of the push to war. After MISSION ACCOMPLISHED I have 800 hours of video. I build a video installation, 40 feet long, mapping news footage onto a timeline. A custom screen in the shape of the Dow Jones Industrial average.
JULY 2003
Design Concepts for Email Interaction

Spam is a hot topic these days, but most of the proposed solutions focus on filtering out the bad messages. What if instead we had a new way of interacting with mail such that spam simply became irrelevant... Spam after all only is annoying because it gets in our way. What if it didn't?

Here is a test animation, a time lapse visualization what my 2002 inbox would have looked like in this interface. Mail is grouped into relationship threads. Relationships have buoyancy, they sink or float based on communication activity.

* thanks for help from Matt Lewis

JUNE 2003
News Ticker Analysis
sdf Remember the days before the "news crawler" at the bottom of our TV's? The "ticker" appeared on every new station the morning of 9/11 and, although no longer necessary, nobody is willing to turn it off. In these tests I re-frame the video screen to "fold" at the same rate as ticker. This process distorts the video, but the ticker is immune.
Perspective Folding for HRI
sdf Driving a robot is difficult. Even more so under intense time pressure, in a rescue situation, and all you have is bad video feed to guide you through the rubble. At CSEL we are trying new ways to approach Human Robot Interaction. With this software prototype I test the viability of "perspective folding"... the idea that on a multi-camera system, it is the physical relationships between the cameras that carries important information. Preserve and use these relationships and important perceptual abilities can emerge.
APRIL 2002
Radial Video Extrusions
sdf Medical commercials. I find them visually fascinating. I want to dissect them, take them apart and put them back together. Here I bend and stretch them along their temporal dimension. It's enough to change my perception of the screen. Maybe by making visible the small patterns locked in these sequences, I can generate procedures for finding larger patterns... understanding our relationship to the images, and how they organize us.
MARCH 2001
Volumetric Photography
sdfOur eyes are not cameras. We tend to think of the photographic image as "realistic". We even frame our own vision as a photographic process, but that's not how our vision works. We see as much with our bodies as we do with our eyes. Without movement we are blind. We require a constantly shifting viewpoint in order to make sense of the world. I want to make portraits in this manner so I built a robotic camera platform that moves and records the parameters of motion so that later, images can be constructed using software that treats the change in viewpoint as the image source.
sdf I turned my car into a giant scanner. Footage taken from a steadily moving car of the passing landscape can be processed to produce new images. Much like the sweep of a spy satellite, or a finish line camera, the synthesized images resemble maps in their flattened quality. The path of the car becomes embedded in the image. Time and movement become explicit dimensions of the synthesized images.
Video Cube
sdf Experimenting with projection onto non-flat surfaces. When the projector's "field of view" is known, as well as the geometry that is being projected on, images and video can be transformed to exactly fit the surfaces.
Intelligence Analysis Interface

How do intelligence analysts do their job? How can an interface help them do their job better? In this design task at CSEL, we are prototyping ways of interacting with search tools and analysis tools that can support the cognitive work of the analyst and the analyst team. The broad goal is to help the analyst community understand and reflect on their own analysis process, furthering the ability to coordinate among analysts and to revisit and reuse past analyses.

JULY 1999
Supercomputer Institute

How can we interact with large databases of video? Here I am given an opportunity to build an interface for navigating thousands of video clips. Some of the problems are smoothly shifting the attention of the user between different temporal scales and providing perceptual anchors. The result is an interactive CD-ROM documentary following the experiences of a group of gifted high school students as they come face to face with the fastest computers in the world.