New Scientist has a fascinating article by Paul Marks today about a new type of imaging technique that avoids the entire analog-to-digital conversion process and shoots an image straight to a bare naked memory chip. I could summarize the piece here, but, honestly, Mr. Marks has done a spectacular job explaining not only how the new technology by researcher Edoardo Charbon works, but he also covers CCD and CMOS image processing in just a sentence or two. Brilliant and worth a read.

The technical analysis is fantastic, but what does the "Gigavision" sensor mean for consumers and videographers? Well, first of all, any commercial application is at least six months away, if ever. There are a couple of technical hurdles to overcome (how to do color, for one, although an armchair scientist like myself thinks that can probably be solved with a Bayer-type filter), even if many of the advantages are quite clear (more accurate than CMOS). There might be two applications, one for mobile phones (because the Gigavision sensor can be smaller and cheaper) and one for professionals (because you can cram a lot more sensor "pixels" onto a chip of the same size). And all of this hinges, essentially, on using current memory chip technology and simply exposing them to light for a fraction of a second. Crazy!

References:
* Cheap naked chips snap a perfect picture, New Scientist, 07 October 2009 by Paul Marks
* Omnidirectional Vision, Camera Networks and Non-classical Cameras 2009
* The Gigavision Camera, technical PDF from Charbon, Feng, et. al.