Slashdot Mirror
A Look Beneath the 'Surface'
Hajsky writes "Ars Technica takes a closer look at Microsoft's new "Surface" tabletop device. Turns out that Surface isn't actually a touchscreen at all, but uses five cameras located in the table's base to record movement and touch. 'The five cameras are near-infrared devices, but that's not because they are trying to read heat signatures from fingertips (or other body parts) on the table. Instead, it's because the entire surface of Surface is bathed in light; by illuminating the top of the table, the cameras can easily see when things are placed on it. Shining colored light across the surface of the table would spoil the effect that Microsoft wants, so near-infrared light is used for invisible illumination.' The whole setup runs on a Core 2 Duo and off-the-shelf hardware, and can handle 52 simultaneous touches." jfanning wrote with a link to an overview of similar technology used in the Helsinki 'CityWall'. The article she provides discusses the unique public display, and has an in-depth video on the way these kinds of setups work.
Please note, I use WIndows, I use Linux, and I use Mac OS X, and am an apologist for none of them, but to put it bluntly, I have read much about this supposed innovation and it leaves me wanting tremendously. It requires so many ifs and ands that I would find it impractical in all but the most specialized settings.
For instance, from what I understand (though I may very well be mistaken), the table can read various things on it. That's great, but your body (or any part of it) better not be obscuring the infrared bath, otherwise it will look like your arm also is supposed to be indexed on it.
Also, the whole multitouch thing seems a bit overrated. Now, before anyone jumps on my back, I understand its implications to graphic manipulation, photo manipulation, engineering, and architecture, but it seems like it would be a rather trivial thing to implement on a touchscreen device, if there was a need. First off it has to be big, to allow the artist to use it comfortably, but then the artist has to keep it clear of any clutter, less it mistake the clutter for fingers.
Not only that, but the damned thing isn't all that new. I remember a documentary on Discovery about "anti-terrorism" efforts in the United States. The Department of Defense had something very similar in that they could call up any place in any city via live satellite imagery and manipulate it in real-time via their hands. They claimed it was the most advanced such display in the world, and frankly it was damned cool. It didn't rely on a bathing infrared light and cameras, it was touch sensitive, and you could use one hand to adjust pitch while simultaneously adjusting yaw with both hands.
Frankly, this device is sounding more and more like the iLoo (joke) from Microsoft.
So if I wanted to use this thing as a coffee table, which in many homes sits in front of the couch, I would then have to run a powercord across the floor to this 'table'? Nice.
How much power does this thing consume? Wonder what happens to it if you spill a drink on the surface.
It seems like it could be frustrating, but more likely it'll just be imprecise. Parallax will only throw the thing off by so much, so the smart thing to do would be to design all the controls to be big enough that an approximation would be sufficient. After all, you're supposed to control this thing with your fingers, and not with an ultra-precise stylus.
I use Illustrator all day at work. I'd hate to have to use my fingers, instead of the precision of a mouse pointer. Even zoomed in, with more complex artwork, the miniscule mouse point is too large and cludgy to be as precise as I'd like. I can't imagine using a mouse pointer that's 1/2 inch wide.