Corpus Callosum

Having activated the built-in open-gl support in Processing, I modified the Soundscape sketch to generate three dimensional soundscape images. The movement of the audio waveforms through space is simply dictated by the user’s mouse movement as the software runs…

Since the waveforms are now being rendered in 3D, I had to turn on anti-aliasing for them to be visible. As a result, the graphics generated are not nearly as sharp as the 2D version, which has that great pixel-art look to it. Still fun though.

Also, I should take this opportunity to add it has been pointed out to me by a friend that what I’m doing/trying to do here is not entirely new:

Robert Hodgin did this landscape-generating sketch.
Mr.Doob created a dynamic version of the Joy Division cover art in Flash.

Nice.

Today I created a new processing sketch that I’m really excited about. The images above were generated using audio samples from the Chemical Brothers song “Saturated.” Playing around with the ESS sound library to visualize live audio, I was inspired to create a dynamically-generated implementation of this iconic Joy Division cover art for their first album, “Unknown Pleasures”:

On a side-note…

Though I, as probably just about anyone else that has seen this image, have always assumed that it is a visualization of audio, after having done a bit of research, I discovered that it is actually a depiction of “successive pulses from the first pulsar discovered” scanned out of the Cambridge Encyclopedia of Astronomy. Here is the page form which the original image was taken:

I really love the layered effect of this graphic, and the way it evokes images of a mountain-landscape. Running the application full-screen while listening to music makes for an especially absorbing experience…

I will post the actual functional version of this sketch online as soon as I can figure out how to do that

Here are a few more examples of interesting images created with my TimeScan software. I’ve modified the code to allow the user to define the length of the image exposure. The examples below are pretty quick scans (I believe most of them are one-minute long). You can achieve some pretty interesting and fun effects, reminiscent of what you might see when putting your hand on a flat-bed scanner and moving it around while scanning an image. The difference here though is that since the background is still, it remains looking perfectly normal, unlike the people in the scene which get wildly distorted in space, making for a disturbing, surreal effect.

My next goal is to achieve a much higher image-quality. So far I’ve just been using the built-in webcam on my MacBook which doesn’t capture great images. The other issue with using a webcam is that it has built-in automatic exposure correction, making it impossible to capture dramatic changes of light within a scene over time. And I have no access to any of even the most basic features of a simple digital camera. I’m looking into potential ways of hooking-up a professional SLR to a computer to use as an image-capture device.

The other day I heard a segment on NPR about some researchers at Stanford University that have created a “Frankencamera” that allows developers to directly access its CCD from where one could then manipulate the functions of the camera to achieve any number of effects. Using technology to explore the potential of digital photography, similar to what I’m trying to achieve with this project. Wish I could get my hands on that Frankencamera!

Here are some samples of my early experiments with Processing. This is a program I wrote, tentatively titled TimeScan. The idea is simple. The software writes an image from a video-input device (in this case a web-cam), pixel-by-pixel, over a set period of time. What results is a photograph that documents a passage of time. It can be viewed as an info-graphic, in that scanning the image top to bottom with your eye, the user can deduce what occurred in the scene over a period of time:

Sleep Pattern

This was produced by leaving the camera on me while I slept over a period of approximately 4 hours. What interests me here is what sort of new information can be deduced form this type of image. For instance, you can tell which positions were more comfortable than others based on how long they are held. You can tell that I began in a long period of stable rest, then devolved, after about 1 hour, into a period of great unrest.

Adding a time-line to the Y axis of this image would help expose this information more:

Still Life

This was a still-life set up by a bathroom window over an approximately 4-hour period during sunset. You can see bands of different brightnesses produced as clouds moved across the sky, blocking and revealing the sun’s light. Also, notice the three saturated bands towards the bottom of the image, most-likely produced when the light was turned on in the room by visitors.