Laser-cutting experiments

These are sheets of laser-cut perspex. Each one is cut with a cross-section of the same point cloud file I have been using for these experiments. I am wondering about options to make them more conducive of light: optical adhesive or mineral oil both might work.

posted March 15, 2011 | Comments (0)

New sculpture render

I have also been using Maya to experiment with new finishes for cast metal sculptures. I would love to make this piece in polished bronze. I’ve smoothed the file, to give the data a liquid quality that I think would work well with this surface.

posted February 22, 2011 | Comments (0)

Presentation options for point cloud project

I’ve been using Maya to render potential installation configurations for this new body of work. One option would be to project still images in anaglyphic 3D on screens, in a darkened space. Careful illumination of mounted prints would also work.

I like the fact that the viewer has to cooperate in some way to bring an anaglyphic image into sight. Even the act of focusing and moving around to see the image’s depth is a participatory one.

posted February 17, 2011 | Comments (0)

Disaster City: project update

While looking online for LIDAR data to visualize, I came across this site. I was immediately fascinated by Disaster City, a training ground for emergency responders in Texas which contains replicas of crashed airplanes, derailed trains and collapsed buildings. I was reminded of Tom McCarthy’s novel Remainder (one of the best analogies of art-making I have read) and the subject matter seemed perfect for the eerie quality of LIDAR. The fact that the unsettling imagery would actually be of simulated disasters was intriguing too.

Disaster City have hosted a number of robotic training exercises which have yielded large datasets for researchers to download and work with. Dave and I spent a weekend looking at the 4 GB binary logfile, and once we obtained the original scanner calibration file he was able to write a utility to interpret it, to a certain degree (see 1 second of partially decoded data, above.)

But there are always limits to visualizing others’ data, and not just commercial or copyright restrictions. In this case the data was captured by a Velodyne LIDAR scanner, which is a very different instrument from the Leica scanner I used at RPI to make the LIDAR images in the Works on Paper section of my site. While it’s a great device (the Google self-driving cars use a Velodyne, as did Aaron Koblin’s Radiohead video), the resolution wasn’t appropriate for the scale and stillness of the works I want to create. A Velodyne is designed to work in motion, whereas the scans we made with the Leica took 2 1/2 hours and contained millions more points. While I think the idea of remote visualization is fascinating, this is one job I’m going to have to do myself, for full control over the results.

So, it’s time for a trip to Texas. I’m planning to go in April or May and do a range of photographic capture; if I can hire a Leica or similar scanner I will, though with the results I’ve been getting from photogrammetry recently, that would serve as a good approximation if not.

I also think it would be fun to volunteer as a disaster victim for a day… It reminds me of this project by artists Demitrios Kargotis and Dash McDonald. (via WMMNA).

posted February 16, 2011 | Comments (0)

Body/Traces: now in 3D!

Forgive the many updates, I have been working on all kinds of things this month and am just getting around to documenting them now.

Seeing Body/Traces on the screen at DNA got me thinking. We never fully utilized the fact that the scans we rendered were three-dimensional. The frame rate we worked at meant that moving camera effects were unconvincing, and the rendered result was not immediately recognizable as coming from a 3d scanner.

I have seen the work of several artists working with various forms of stereoscopic 3d, and while there is no doubt that some of these forms create optically awkward viewing experiences (Roger Ebert is not a fan), the other-ness of 3d viewing space actually seems uniquely suited to the ghostly figures we captured with our scanner for this work.

The 1:1 scale and human presence was so important to the presentation of our installation; I think the depth of 3D view might really add to the experience of presence, especially in a darkened room with only a few viewers.

As a first step, I’m re-rendering Body/Traces in Maya, using the stereo camera rig that was added to Maya 2009. Next I want to do the same with the point cloud data I’ve been working with more recently. I’m going to continue looking for other plugins that provide point cloud support within Maya.

| Comments (0)

The Childrens Portico: part 2

Here is the exported .ply file in Meshlab, ready for cleanup and quantization. Now that it’s in this form I can slice it up into layers using the same quantization code I used for the last beaded sculpture, to turn it into another piece rebuilt using colored beads on nylon monofilament. We are still working on preserving the vertex color you can see in Meshlab; I’m hoping the next sculpture will use colored beads, not just black.

I am going to make this piece about twice the size of the last one, so around 20″ x 15″, and with several thousand beads. It should take about the same amount of time though, as I’m much quicker this time and I won’t be hosting an open studio while I work!

I would love for this to be a room-sized installation, and am writing exhibition proposals for that now. At that scale I’m going to need some interns…

I am also investigating laser-cut acrylic and optical adhesive as an alternative support for the beads (I can use the same Inkscape .svg slice files I used as a guide for the beads last time). In fact the test sheets are being cut this week, so I’ll report back soon…

posted February 15, 2011 | Comments (0)

New sculpture source material part 1: The Childrens Portico

Here is another Brooklyn architectural oddity, reconstructed in 3d from photographs: the Childrens Portico, an architectural remnant which has been preserved and now sits on Pratt’s Brooklyn campus, where I teach.

View the point cloud data in Photosynth here (install required). Don’t forget to turn on point cloud view (on the right in the bottom menu bar).

| Comments (0)

Photosynth experiments

I have spent the past month taking massive quantities of photographs in Brooklyn, for 3d reconstruction in Photosynth. Surprisingly, it’s not a bad tool for reconstructing 3d geometry. Noah Snavely’s Bundler and Arc3D are both more sophisticated of course, but while I’m learning a whole new way of working with 3d data, Photosynth has been useful for these initial tests. Microsoft do not allow you to download the original 3d file, but they seem to tacitly encourage others to do so, and I have been using Christoph Hausner’s SynthExport tool for that purpose.

Here is a reconstruction from photographs of the reflecting pool that is part of the Prison Ship Martyrs Monument: click below to see it at Photosynth.net. (Install required.) I am rendering the point cloud file, which you can view by changing your view options in the menubar on the bottom right of the Synth viewer.

| Comments (0)

Body/Traces at DNA’s Frameworks Festival

It’s been screened in Los Angeles, Tokyo, Troy, Boulder and Burgundy, but Body/Traces has not yet been properly screened in New York City. Those of you who want to see it on the big screen now have the opportunity, at Dance New Amsterdam’s Frameworks Dance Film Festival.

The screening takes place on Sunday, February 13th at 3pm. DNA is at 280 Broadway, 2nd floor (enter on Chambers St).  Tickets are $5 in advance. Click here to purchase.

posted January 31, 2011 | Comments (0)

Lightbox prototype

This is another piece that resulted from my MAD residency. It’s the same 3d file in the piece below, a 3d point cloud file that has been quantized — ie sliced into horizontal layers — and then output as svg files. I printed those files from Inkscape onto layers of acetate which were then framed in this lightbox.

Displayed in this way, the image is flattened, but you get a hint of its 3d origins in the varying opacity in the layer stack, and the accumulation of dark points around the edges.

The original 3d data was kindly shared by Steve Seitz, Noah Snavely and others at the University of Washington, and was part of the Building Rome in a Day project.

posted December 20, 2010 | Comments (0)