Slashdot Mirror
An Origami Lens for Your Camera Phone?
Roland Piquepaille writes "Your next camera phone might get a new kind of lens if researchers at the University of California at San Diego convince the cell phones makers. They have designed an 'origami lens' which will slim high resolution cameras. Today, their 5-millimeter thick, 8-fold imager delivers images comparable in quality with photos taken with a compact camera lens with a 38 millimeter focal length. In a few years, these bendable lenses could be used in high resolution miniature cameras for unmanned surveillance aircraft, cell phones and infrared night vision applications."
It's a reflector. Don't ask me how it works, the story and illustrations aren't very clear. But it's not a lens, fresnel or otherwise.
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
Despite what the summary says, the "lense" isn't bendable. It just manages to compress a lot of light-bending capability into a small space by using reflective, rather than refractive optics and combining all the optics in a single crystal. I say "lense" because it's not refractive, so it's not really a lense.
pi = 3.141592653589793helpimtrappedinauniversefactory7
This is actually nothing like a Fresnel lens. Fresnel lenses are based on refraction and tend to give horrible image quality since they have a whole bunch of concentric rings. This lens does things a completely different way. It's really a pretty clever piece of optics. It's functionally equivalent to putting a conical reflector over the imaging device and another reflector at the edge or the lens. They just get a longer focal length by bouncing it up and down more.
ACK NAK RST
Here is one of many greasemonkey script to remove piquepaille stories
http://userscripts.org/scripts/show/5735/ [userscripts.org]
don't cut it off www.mgmbill.org
They are only going to use the Diamond cutter to produce the master for the molded glass lenses. After the master is created cost of molding a plastic reflective imager is pretty much the same as cost of molding a plastic lens. They do need more software but we all know software is free as in beer Right?
**Life is too short to be serious**
This lens has the same problem as any lens-system with a central obstruction; the contrast for medium-scale detail is poor, due to diffraction effects.
Image quality is generally specified using a concept called Modulation Transfer Function (MTF). It is like a frequency response for lenses except the frequency is spatial in cycles per mm rather than Hertz.
Lenses with a central obstruction can have comparable MTF with respect to unobstructed lenses of the same speed, at spatial frequencies near the limit of resolution. However, you try very hard not to use a lens that way because the performance is poor. At the more important, intermediate spatial frequencies, an unobstructed lens has much better performance.
Astronomers have picked up on this idea. They like to use reflective lenses with a central obstruction for viewing stars where resolution limit is the only thing that counts and the perfect colour correction provides an advantage. However, unobstructed refractors are better for planets where you have a distributed image.
It is possible to make reflective telescopes without a central obstruction but the technology is still a little expensive. I expect, one day, they will displace refractors.
Aliasing is another issue using a centrally-obstructed lens with a pixellated image sensor like a CCD or CMOS device. Spatial frequencies above the Nyquist limit (2 pixels per cycle) generate garbage within the pass-band of the detector. A lens of this type concentrates its performance in the worst frequency range for the detector.
There are lots of promising approaches for cheap, compact lenses for cell-phone cameras but I doubt this lens is one of them.