So your showoff neighbor brings home a new 36 megapixel Nikon SLR, and your previously top-of-the-line 18 megapixel gadget starts to seem... inadequate. The insolence! The injustice! What can you buy to put that jerk in his place?
How about raising the stakes an order of magnitude with a 960 megapixel supercamera?
Researchers at Duke University and the University of Arizona report in this week's Nature they've built a functional prototype gigapixel-scale camera. The Orwellian-sounding AWARE-2 uses a 98 sensor-array mounted behind a single aperture to capture an enormous image in one go.
Why is this such an advance? It has to do with the data problem. As the megapixel race scales upward, the technical obstacles shift from a problem of how to pack more sensors onto a chip to a problem of how to manage staggering amounts of data. Past a point, wires simply can't handle the information deluge. Gigapixel cameras can only work when the data is captured in smaller, more manageable units.
Previous gigapixel systems worked around the problem by stitching together serial captures from robot-driven cameras that pan across a scene, taking minutes or even hours to composite a final picture. These break the problem up over time. Great for landscapes, less so for applications requiring a unitary moment.
The advantage of the multi-sensor camera like AWARE-2 is in parallelizing data capture. It breaks the problem up over sensor space. With data streaming through 98 channels at once, the problem of data capture is rendered tractable and a gigapixel image becomes possible in a single exposure.
What are the chances Duke will send me a demo?
Not to use, mind you. I don't think the technology is all that applicable to my usual macro fare. I just want to walk around in public with one of these strapped to my neck.
Brady, D. J. et al 2012. Multiscale Gigapixel Photography. Nature 486: 386–389 doi:10.1038/nature11150