Troika Cloud for Airport Terminal
This digital cloud made from flipping pixels is beautiful.
Anyone has a project like this we can build?
More at the troika website
This digital cloud made from flipping pixels is beautiful.
Anyone has a project like this we can build?
More at the troika website
Comments
dave on 24 January 2008
It is beautiful - reminds me of the Aleph reorganizing vision installation.
Dave
Jonathan on 29 January 2008
It’s neat. Wouldn’t want to do the maintenance on it though!
Making displays where you have to build each pixel is hard work (I just don’t understand how Daniel Rozin managed his mirror displays with over 900 motors), but a lowish resolution would be possible (maybe using the MIDI board for the control).
I quite like the idea of a display with a robot arm that flips the pixels. [Very low refresh rate, though!]
I’ve also toyed with the idea of doing an instructable [LED] display, which is sort of the same kind of area.
Perhaps we ought to have a brain storming session sometime on crazy ways to make a pixellated display.
pmarshal on 31 January 2008
It’s really fascinating to look at. I’d like to see it for real. It somewhat reminds me of a program I wrote a long time ago for visualizing magnetic fields. Magnetic fields with multiple poles make beautiful curved shapes, very organic. I wonder if it could be possible to create a volume with large moving magnets inside with the surface actually responding directly to the magnetic fields cutting through them? The effect might be quite stunning.
admin on 03 February 2008
We had some fun with Gaussing monitors, where the magnetic field is shown through the color and shape distortion of a cathode ray tube (ordinary PC monitor). I don’t know how you’d scale this up though. I suppose there may be some very cheap way to detect magnetic field orientation, and then you could tile the surface, but not sure what this technique could be. I’m reminded of the parasitic use of LEDs as touch sensors, by exploiting their capacitance. Then you could use Tricolor LEDS to show orientation.
Patrick on 04 February 2008
I have a particular fascination with tiny, very powerful, and relatively cheap, Neodymium magnets. Tiny magnets mounted directly in gimbals with two degrees of freedom (like navigation gyroscopes) could be used to directly sense the ambient field. You could have spherical sensors with each side painted contrasting colours (black/white, blue/red, etc.) or flat mirrored surfaces like the Troika cloud. With mirrored sensors against an interesting background, you could get some interesting images; a mosaic of tiny reflections either concave (magnifying) or convex. Alternately, instead of gimbals, perhaps you could have a flexible balloon like surface with magnets directly glued to it. Here the freedom of motion would be limited, but you would have a very reliable mechanism. Albeit a rubberized surface may not have a very long lifetime as a standing exhibit. On the inside, the driving magnets could be permanent magnets on a mechanical mount, or fixed electromagnets with varying currents and direction.
Jonathan on 05 February 2008
The kind of component you’re imagining here, Admin, is a Hall Effect sensor. It senses the magnitude of a magnetic field. It also has an orientation, so if you put three together, mutually orthoganal, presumably you’d get the readings you’re after for the vector. (I’ve never used one, so don’t know much about their actual use.) They’re about 30p-60p depending on quantity (have a look at Rapid’s website). Since the output is analogue, you could miss out the digital stuff altogether and make it a purely analogue display. Problem with any of these ideas is cost: do a small 32 x 32 display and it’s 1024 pixels which, at a couple of pounds for each pixel, immediately takes you up to the area of a few thousand pounds: fine if you’re funded, hard on the pocket otherwise (unless you’re good at blagging stuff, of course).
Patrick on 08 February 2008
I like the idea of the Hall effect sensors. To get around the cost, you could have a string of LEDs between each sensor. Since the field is smooth, you could use simple resistive networks to divide the voltage between each sensor pair, i.e. there would be an intensity gradient between a high and low voltage sensor. You would still see the patterns and you could get away with only a fraction the number of sensors. To further save costs, instead of entirely covering the surface with LEDs, you could have discrete lines between each sensor, so you would have a coloured net like effect, which might be very visually appealing.