And I also wrote up a little more about the project on the Near Future Laboratory blog, which goes into some more detail about the ideas this project explores. It also includes some pictures I drew to better explain the words I wrote.
Pretty fantastic timing really because I just finished working on the first draft of my thesis paper, which I’ll turn in today. So that means I’ll be able to spend a lot more time this last week just working on experimenting with the actual project more before the thesis committee meeting next week.
Having to describe the project to a public audience (both in the video and the descriptions) has really helped me think about the implications of the project and what I’m trying to do with them and the next steps involved. It also at least got me to come up with a name for the things other than “magic black box.”
I also sort of feel like going through the process of making a Kickstarter campaign itself is sort of part of the greater project since I’m sort of looking at the effects of our network culture, like the little experiments with Amazon’s Mechanical Turk. I also really hope I can reach the funding goal, not just for the funding but for the opportunity to make the rewards for people, as I think collectively those could also be really interesting projects in themselves.
And now for the next 9 days I really need to focus on writing the first draft of my thesis paper…
A slightly more refined (foam core & duct tape) version of the box, which I’ve decided to start calling a Portal. Also a couple prototype videos showing what it could be like to have two people in this virtual layer. I was able to get a very rough, version of it working (see example with Dustin & Rubina). But the chroma key is so crappy and not really working like I would like. I think I just need much better lighting and higher res cameras. So the video below is sort of a half real half fake prototype. I’ve also been thinking about what your hand could do while inside the box, for instance rubbing a surface to navigate the space like a giant lazy susan. Or having push buttons in the space that could trigger other things so you’re actually interfacing with something inside the box?
Super quick little test to see if I could get a “telepresence” system. You can see Rubina & Dustin’s hands in the same space on screen even though they’re in separate locations. Obviously the chromakey is pretty awful cause the second “set” isn’t lit properly at all. But I think it at least gets the idea across as a proof of concept. I also don’t mind the crappy resolution all that much cause it helps add a layer of abstraction to the people.
I didn’t make a second box since I don’t have another monitor right now. I bought a little 7 inch USB monitor (see it face down on the floor) but I’ve been having trouble getting the full screen video working on it. I was thinking the larger monitor was too big before, but I sort of feel like it’s actually not bad. I was hoping the little one would work just so I could have a shoebox sized box, but maybe it’s too small now that I look at it.
I feel like it’s a super simple little system but it has lots and lots of possibilities. I can imagine magnets on the people so you could move them around from below. Or a system that can identify the objects being brought into the world. Or.. lots of other things! Exciting!
First, a couple of things that didn’t really work this morning… I was trying to use our IP cameras to create a panoramic source for the Motion Tracker widget. Unfortunately the refresh rate on those things is just WAY too slow to work properly. So I got frustrated and bought a $30 webcam from Staples, which works wonderfully, although you can’t place them side by side to make a long panoramic..Next test was getting things onto the iPad to use it as a front display. I tried Air Display, Live View, and Ustream. Unfortunately all 3 of those also had a lag time that was not going to work for me. I was also playing around with CamTwist’s various effects. At first I thought they were all just cheesy filters but the ability to dynamically chroma key and layer video on top of each other was really exciting…
And I was finally ready to start making a box! I found the smallest screen in the studio and the smaller end table fit together pretty well. I’m imagining a more refined version would be both smaller and fully enclosed except for a place to put in your hand. But I think this version actually works pretty well for now..
Thanks to Andrew for both lending me his little people and modeling : ) At first I was hoping the camera wouldn’t be so zoomed in, but I think it’s actually kind of nice. I love the scale shift between the human, the interface, and the mini people. I think what I like about going behind the screen to interact with tangible things is that it sort of plays with the layers of virtual/physical. And it also starts to think about interaction beyond “Pictures Under Glass” cause the thing with touch screens is they’re very much still separated by that barrier.
I definitely need a cleaner background and light it better for a cleaner chroma key. But I sort of like the chunky artifacts. Sort of stays true to the loss of fidelity in going from a physical thing to a digital thing.
I’m feeling pretty good about where this is going. I also feel like there’s lots of other ways this could be even cooler. For instance if I could somehow track for position of objects and then have them appear to animate on the screen after you’ve placed them into the space. Or if the box knew what you were adding and could add additional things related to that specific object. I ordered some RFID stuff to try out so hopefully I could use that to identify objects as they pass through the opening to the box. I am also thinking another element I could add in would be “anonymous hands” that also add to your scene while you’re using it. In theory they would be from other networked boxes, in reality they might just be videos overlaid on top.
Here’s an initial prototype of the “touch screen” interface. Right now it just uses the motion tracker widget to track motion, though that’s sort of finicky. I’m wondering if it might be better to use something like an IR camera so it tracks more precisely. The monitor in front is also way too big. But I do like that if you were sitting in front of the monitor and put your arm around the back side it would make you basically hug the screen and force you to get pretty intimate with it.
Also, while I was trying to figure out how to get another camera to work as a webcam I came across this tutorial. I wasn’t able to get my Canon Rebel working cause I haven’t been able to find the EOS Utility software yet. But I did get CamTwist going and got pretty excited about the prospects of that alone.
Basically it lets you make a box around an area of your screen which then becomes another “camera” source. So then in the Netlab motion tracker widget you can use it as a camera, meaning it’s no longer bound by the camera on your laptop or one plugged into the usb. So I could even use a bunch of those IP cameras we have from the show.
In the tests above I tried it with a Youtube video, a street view, and a live chat and they all worked just fine, although street view is too low contrast I think. The most exciting part was using the live chat because this means I could actually have people’s webcams control something on my computer, or if I hook up some servos to the motion tracker, something in a physical location. Also interesting that it lets me use both my laptop camera and the CamTwist camera at the same time. I feel like there’s some nugget of an interesting idea in here but I haven’t quite sorted it out yet. It’s a little confusing thinking about all the different camera feeds….
Thinking about some alternate interfaces for editing the views for the roundabout. I had an initial sketch that looked like a diorama but I’m thinking a better one would be one where the screen is in front of your hand, so you can’t actually see the objects you’re holding. maybe being abstracted makes the objects your holding more open to imagination. Super quick test of what it might look like:
I’m also thinking about how the collaborative aspect of it might be even more interesting by having to stick your hand into this thing you can’t see. Perhaps when other people are in the same view you can feel their hand in there moving around too. I’m thinking the glove thing would be important as a way to make it more anonymous but also simplify the representation of the hand so it wouldn’t have to be actual video but perhaps images, stop motion like. Here’s the rough sketch of what I think I’ll start trying to make this afternoon.
Also, some existing similar interfaces.. theres the glove boxes people use for super clean work and then there’s the fact that many surgeons look at screens and not the actual body when doing their work these days…
So one of the things that came up during my meeting was to think about a bunch of different ways I could have something like my Rolling in the Streetview appear in the world.
So this is a a quick initial experiment with a very low-fi “lenticular” animation, although there’s no lens. It’s just two frames, cut up into strips, interlaced, glued down and folded like a fan.
Advertisers use this sort of thing for billboards and posters and direct mailers all the time, but I think the interesting part is using for non-advertising purposes, and having it relate to the space contextually.
I think the end result is kind of fun. I just realized afterwards that I could have printed a photo of the background behind the kitty instead of just having white, which would have made it seem sort of “transparent” and I could have used a clear acrylic stick instead of the wooden one. Perhaps those are things for version 2. Putting it on a stick also makes me think I’d like to make an animated protester sign. From one view it can say “FOR” and the other side “AGAINST” for the flip floppers out there.
Other thoughts that came up while making this was thinking about using the surface of blinds to toggle between two frames. The added bonus to having blinds is that in theory you could see through to the real world and maybe it would appear to be “augmenting” it. Also if it was on some sort of mobile (on wheels, not not phone) device it could be placed anywhere.
I think what I really like about this sort of low-fi version is that it doesn’t require any sort of screen or projection so it works in the daylight and doesn’t really need any technology so people could potentially print these out and take it outside themselves.
A very rough initial prototype of what an interface for these manufactured landscapes might be. At this point it sort of just seems like a fancy knob project. I’m thinking that I could embed some of these little RFID’s into each of the tiles so the car knows which tile its on… The only bummer is that the car, which would be the reader, would have to be considerably larger to fit the RFID reader. Plus a USB cable sticking out the side…
I’m thinking it would actually be easiest to just get a clock motor thing to power the arm so I don’t need to make a servo thing. Plus if it’s just a clock motor it would in theory be easy to coordinate what’s happening between the tangible device and what you see on screen. But on the other hand, the idea of a more fluid idea of time is also kind of interesting….
for the Romans, the length of an hour varied by the season. Twelve hours of daylight in December required shorter hours than twelve hours of daylight in June. The length of the hour varied between two extremes: 45 minutes long at the winter solstice and 75 minutes long at the summer solstice. Our hour divided (no matter what the season) into 60 minutes and 3600 seconds is a creation of the mechanical clock, which of course was unknown to the Romans. The Roman based the time of the day on their observation of the sun and the shadows it created.
Also, a side note maybe there is a separate tangible interface for adding things to a tile, so it’s not all in one thing.
Oh! And thanks to Sal for helping me re-find that reference I was looking for..