On the 1st of November I turned in a residency application to Harvestworks – an organization in Soho whose “mission is to encourage the creation and expand the dissemination of digital media artwork.”The residency is tied to a project. Here’s the concept of the project in video form:
Cool, no?
The idea behind the installation is that your image is volumetrically rendered live, right in front of your eyes. If you’re at the exhibit with your friends, they could walk into the rendering chamber and interact with your vapor phantom, and you could possibly even walk through one another.
Because the signals are entirely digital – this project could be set up to work on the other side of the globe. I could step into a capture chamber in New York and create a vapor phantom in Tokyo.
Well, right now this is all in the concept phase, and it’s difficult to come up with any ideas that have much traction w/out a real setup. Regardless, some questions that I think only real testing can answer: Would color work? Could you record and capture in the same chamber, or would the volumetric lighting create a nasty feedback loop? How many projector/camera systems do you need? What kind of detail in objects work and what kind don’t? Will it look better up close or far away?
I can’t answer these questions now, but I can render some more virtual 3d mockups. Closeups, with color, more systems, etc…
Unfortunately, these animations render very slowly. For parts of the video above, my machine rendered at about a half an hour for a second worth of footage. So I guess this just means I’m going to be learning about this idea very… very…… slowly.
Albert, your blog is the coolest thing I have ever seen. And way smart. You were on the podium at graduation, weren’t you?
we tried that a few different ways, mainly using an array of video projectors to create the light field.
The problem is that there’s nothing that you can put into air (i.e. fog/haze/water vapour) which doesn’t scatter most of the light forwards.
This means that the lights pointing towards you look a lot brighter than the lights pointing away from you.
Then we tried having all the projectors pointing towards the viewer but there was always too much mess. The overlap region was never significatnltly brighter than the trails which led to it.
We then tried something with moving head lamps (Martin Mac 250′s) to create moving stars in space (basically looks similar to your 3D sketch). Calibration issues were a pain but even after they were resolved it was a bit too expensive and not very effective (for what we were after anyway).
I’d be happy to lend a hand on this one though. i think the effect would be best in a large space with moving mirror lamps (e.g Martin MX 10′s) to create volumetric wire drawings in space.
Our original concept was to try and be able to highlight a single point in space, and if that worked then we could use direct control of DLP chips to achieve 10,000′s of fps moving that point (or multiple points) around to draw images.
aha, i’ve seen the second half of your video now (the skeleton bit) with sound
this is exactly what we tried to do just over a year ago to the day.
sadly, it doesn’t work. you just get a mess of light (we experimented with digital projectors and a jem hazer)
To get it to work you need huge beams which converge onto tiny points which can then be moved around in space (i.e. pull the focus to move the image forwards and backwards).
Using a large parabolic mirror/fresnel lens it is possible to make the image effectively converge from a large projector to a small image, in which case working with a single projector to begin with (with digital zoom controls that can be manipulated over rs232 or usb) is probably the best way to start.
What you think?
Elliot