Catherine McCurry

Continuing from assignment 2, I dug into 3d modeling and got my feet wet in Maya, Mudbox, and Meshlab. I really liked the quality of my Kinect scans, and figured out how to reconstruct them and shape them so that I could rig them with a skeleton and control them with a Kinect, using OpenNI. The scans didn’t turn out perfectly. I was excited about Mudbox because it simulates molding with clay. Turns out mesh is not clay. In an attempt to separate the legs from each other, and the arms from the torso, I scraped and scraped the mesh, until it became stretched so long and thin that it looked like it was going to tear – so I decided to leave it and move on.

(batwings…) I spent a lot of time with my janky meshes, fighting, fixing, orienting joints, constraining rotation (fat lot of good that did, doesn’t even export from Maya…grumble). I like the look of my mud me. It seems somehow more personal and specific than a perfect copy.

Of course the motion is not perfect. It jerks and lurches, and deforms itself with every movement. I decided to try and wear the mesh anyway, and attempted to put my body in the shape of the mesh as it twisted and contorted. The motion actually would have been better, but I didn’t have a long enough usb extender to put the Kinect directly behind me, so it was catching me at an angle.

Initial Concept:

We imagine a world in which a person’s external form is as they wish it to be, moment to moment. We wear a holographic shield of light in the shape of our own bodies that can be sculpted like clay by manipulating our own flesh. Spikes and spines can be pulled out of backs, imperfections erased, bodies enhanced, exaggerated, and edited.

In our proposed performance, a 3D scan of a performer is projected on top of them, and rigged to puppet their movement. They examine their flesh, and their avatar’s form. They pull, push, and shape their light body until it is as they want it – exaggerated, human, recognizable, subjectively perfect. They exit the room encased in it.

/////

What we did:

Following the Nightmare Kitty tutorial, we were able to control characters in Unity with our physical motions using the Kinect, and to implement some basic gestural controls.

This was pretty simple, actually, and we planned on creating our own mesh, rigging it with a skeleton, and substituting it for the video game style character in the example.

We used Kyle McDonald’s kinectToStl app to save scans of my body to .stl files, then imported them to the 3d modeling program Meshlab to edit them, and export them in a format readable by Unity.

This took a good deal of time and finagling. Although we managed to make and save scans, then trim, and export them as meshes meshes, to finally import them to Unity – we have not yet managed to rig them to a skeleton model. Additionally, the imported meshes do not behave exactly as expected in Unity, but they do look pretty scary (in a Swiss cheesey sort of way).

We realized that we had jumped the gun in terms of using our own models with functioning skeleton models.

During this process, inspired by the sculptures of Antony Gormley, and the aesthetics of the images we were creating, we came up with a couple of alternative concepts.

1) A multitude of beings stares at us from the screen. One of them starts mimicking our gestures. The beings look at him, then at us, and start copying us as well. Gradually, the entire crowd is moving like us, as they stare in our direction with blank, wide open eyes.

2) As we look into the screen, we see cubic versions of our bodies from different perspectives. They follow our movement, but remain in their position and orientation. They do not acknowledge our presence: we are only one more of them.

+

We created a basic body made of primitive blocks, attached it to a skeleton model and added OpenNI functionality to it. We learned a lot about the ways to do (and not to do) this in Unity.

///////

Where we are now:

We began with an ambitious concept and a song in our heart. Though we had some setbacks trying to interface with 3D modeling software and are disappointed that we did not get further, we learned a lot are in a good position to fully execute our project for the final. We aren’t so excited about the look of the block man model, but have a better idea of what we will need in order to animate our scans. Apparently this needs to happen outside of Unity – in Maya, or Blender. I downloaded Maya and Mudbox at the suggestions of Patrick and Greg.

“On Being Memorius” Recall an abstract space from your memory and recreate it using only programmed primitives, color, and lights.

* After lessons, I spent time in his room. In their room, but his brother didn’t talk.
* His mother gave me technique.
* I practiced and got better, but never less awkward.
*
* He taught me to play grunge rock on his guitar. He taught me to wear flannel.
* His room smelled like wood and, like, cool.
* His parent’s room was across the hall and the possibility of meeting his father scared me.
* It took courage to walk up the stairs and into his room.
* I felt annoying but my crush made me brave.

source

((P.S. I <3 vector math.))