Here's a short clip of the latest project I've been developing. Essentially its a particle field that reacts to your webcam in order to sketch out these cool drawings. It pulls its opacity/direction directly from the webcam pixels. I put some more screen-shots on my documentation blog.


Swoon from Alex Wolfe on Vimeo.

Finished Wind Interaction Projects

Here are the photos of the finished turbine. This is the final result for my obstruction project from before. "Computer Art" without a computer, something that collects mundane data a represents it visually. The LED's brightness directly relates to the speed of the wind! It makes a beautiful dimming, brightening, dimming, sort of breathing pattern when in use. I originally wanted to make a bigger more extravagant diffuser than the little bird, but it was so cute, and it works very well.

Upcycled Wind Generator

Here are some shots of the wind generators I'm working on to power an installation. Made from an old case fan, PVC, scrap wood, and some steel tubing

Naoqi :: Robot Soccer :: Vision

(Thanks to Kiera for taking this video on her new snazzy Droid : P )

So this is the robot I've been doing research on, we're programming them to play soccer in teams of four! I have a wonderful partner, and we're hoping to destroy everyone else by the end of the semester. Right now all it does is find the ball and kick it, which sounds pretty simple but is actually fairly difficult.

 :: Technical Stuff ::     (feel free to skip if this kind of thing makes your eyes glaze over)
The robot is a "Nao" and it essentially runs a mini version of ubuntu, you can ssh into it wirelessly or with an ethernet cable and copy your compiled code and type commands into the command line. However, its memory is pretty small, so typically all we do is run our written behaviors on the robot itself. The cool part is that you start with basically nothing, you have to control everything from the "eyes" to each joint.

First off, here is what it sees through its "eyes" (read : camera).

The top image is the raw feed, and the bottom is the parsed version the robot actually uses to determine its actions after we run our vision algorithm on it. Right now we're just doing ball tracking, so we determine the purest color of each pixel, (right now the only relevant ones for us are orange and green) and search for orange ones. If there is no orange present, it does a pretty sparse search, only one out of six pixels. Once it locates something that might be a ball, it does a much more refined search in only that area, and will search that area first in the next frame.

This makes our code run much faster and allows the robot to react more quickly to changes in the ball's position since it doesn't have to look at every single pixel each time.

We are also super cool, so instead of figuring out which quadrant in vision the ball is located and then moving accordingly (like we used to do in the picture above), we now do visual servoing (as seen in the video)

The robot ALWAYS keeps the ball in the center of its vision by moving its head if it sees it, and then aligns the body with the head in order to precisely line up with the ball. It then checks if the ball is in its right or left field of vision to kick!

That's all for now, I'll post on goal finding next, and then on my latest and greatest art project : P

Face Synth

faceSynth. Second part of the visual score project, we were assigned someone else's project to make sound/visuals from. I had Gabe's; essentially he had 67 or so images of faces and directions for a percussive piece involving them. I had no intention on doing that, so instead I created this app that takes the RGB values of each pixel in  the face and translates it to visuals/sound.

Eventually I want to rig it up so the audience can take their picture in a booth set up with the appropriate lighting, and their face gets sent to the app in real time, so they can hear their own faces throughout the performance.

Press spacebar to see the face image being parsed, and click to skip to the next one. The sound is still a little wonky/creepy, but I'm okay with it

http://www.andrew.cmu.edu/user/akwolfe/apps/faceSymphOsc/index.html

waves

And here are some from the waves app

This sketch builds off my last one, extending into 3 dimensional space. Its pretty processor hefty so I have two versions

(hmm so i just discovered that this thing runs at blinding speed if your computer isn't worthless like mine. I'll get around to fixing it, promise : /  )

click here to see waves if you have a decrepit laptop clinging to life like me
or
here if you have a beast of a machine that lets you physically travel through ssh tunnels and bend the internet with your mind.

Flow

Alright so here are some screens from the flow visualizer.

This one was just a practice sketch, based off the processing noise() example, found here. The noise() function works off the algorithm by Ken Perlin to generate a " more natural ordered, harmonic succession of numbers". It was used in a lot of the early movie special effects to generate terrain.

Here, I'm use it to make a sort of pseudo "sine" wave like those generated by music. I took their sample, switched out the lines for points, repeated it on the x, y plane and changed the movement from being dependent on the mouse to an animation. Part of a body of work to connect sound to a physical form for concept studio.

I am a god of visualizers!

Here's the hard links to the apps I've been working on for concept studio. Essentially we had to create a visual score that represents some sort of music, and then we traded and had to create sound out of someone elses. Formal documentation later.


Here are my scores
flow
waves


Reactive sound
This one gets kind of loud, I'm working on dynamically creating sound based on the color values of the pixels, but the results get kind of unpredictable. Press space bar to see the face that its currently parsing.
faceSymp

Digitally Fabricated Armor Jewelry

So the intentions of this piece were to explore having a duality of form, and pushing the boundaries of what is socially acceptable. What is the line that separates the girl who wears her Lord of the Rings cloak to every class, and Alanna Zimmer stomping down the Alexander McQueen Fall 2010 runway wearing essentially an intricate breastplate? The goal was to construct a sleeve of armor that can be taken apart the separate pieces worn as jewelry.


Here's the finished chain mail concept jewelry, I'm still modeling the armor part, I made one but it was too delicate so some of the plastic melted off in the acid bath. When you 3D print objects, the printer surrounds them with support material you can dissolve in acid in order to achieve your exact shape. Unfortunately if your lines are too thin, it happens to your model too.

Here's the model and failed part one

Here is the final version, I made the lines in the model a little thicker, they lost a little of the grace and ethereal qualities of the first, but it is also significantly more stable, so there must be some trade offs. I constructed the straps out of a soft suede shammy, laser cut and braided. Each bangle comes in an increasingly larger size, so they can be stacked as armored plates down the arm, or worn individually.

Wolf Ring II

Sorry, I'm just going to revert to the classic template until I have time to write a new one. Ugh finals.

Also I stained one of the rings I made with india ink. You can see the details significantly more clearly. And I printed the first piece of armor! I'll post of a picture of it later after it comes out of the acid bath

 

Here's a picture of the original white version.

Our 3d printer is additive, so it prints out objects by stacking thin layers of plastic into the form you want.  However, if your form isn't stable enough to balance on its own during this process, it also prints out support material to keep it in place that you dissolve later in acid.

I'm also making a couple that I'm painting with my beautiful NARS vampire nail polish. Yum