IEEE Spectrum - North American - March 2014 - 38

are just 17.5 or 35 millimeters on the diagonal. To make an arc lamp
brighter, the bulb itself must become physically larger, including
the arc that produces the light. But increasing the size of the arc
makes it harder to focus the light onto the chip. In practical terms,
we've already hit the wall in terms of getting more light from arc
lamps through digital cinema projectors.
Lasers do not share this limitation. All of their power can be
easily focused onto a small area, and essentially all of that power
is used. That's not the case with white light: After the three primary colored beams are separated from the arc light, the rest of
the visible light spectrum, as well as a lot of infrared and ultraviolet radiation, is wasted. It is dumped within the projector,
which must therefore dissipate a lot of heat.
Lasers have other advantages, too. They can be very efficient
electrically, last 20 000 to 50 000 hours, have near-constant output, and are highly controllable. Also, because lasers are compact
and do not get very hot, they can be packaged into a system small
enough to replace the xenon lamp assembly in an existing projector.
These considerations have long intrigued projector makers. The
laser technology itself goes back more than a decade, when U.S. and
German companies developed laser light sources to go into flight
simulators for pilot training. IEEE Life Fellow Peter Moulton and I
conceived the company Laser Light Engines to commercialize the
laser system Moulton developed for the U.S. Air Force Research
Laboratory. This system used infrared laser diodes to pump a laser
crystal, which produced another infrared laser beam. That beam
went into a series of nonlinear optical crystals that converted the
infrared into the red, green, and blue beams. Nowadays, the company uses an aggregation of semiconductor laser diodes to produce
the red and blue laser beams. The red beam is generated with gallium arsenide-based diodes, with quantum wells of aluminum
gallium indium phosphide. The blue are gallium nitride diodes,
with indium gallium nitride quantum wells. The green comes
from a high-powered, frequency-doubled, diode-pumped laser.
Early on, though, it was far from clear that lasers were the way
forward. Their biggest problem was a shimmery image artifact
called speckle. Instead of a patch of solid color with completely
uniform brightness, early laser projectors produced images with
sparkling surfaces that seemed to dance and move, especially
if you moved your head. Speckle occurs because the surface
roughness of most movie screens is on the order of a wavelength
of visible light. So rays of laser light reflecting from the screen
constructively and destructively interfere with one another.
Thus laser projectors seemed dead on arrival until Laser Light
Engines, which is based in Salem, N.H., solved the speckle problem in 2010. It developed several solutions before settling on one:
broadening the spectral bandwidth of the red, green, and blue
beams enough to avoid speckle. For that it uses a proprietary
nonlinear optical process, which effectively reduces the coherence of the laser beam, widening the bandwidth of the colored
beams from about 0.1 nanometers to 10 to 30 nm.
After speckle was tackled, the challenges became more conventional: delivering as many as 600 watts of total laser power
while achieving the desired figures for projector lifetime, energy
efficiency, and cost. These goals are 50 000 hours or about
10 years, 10 white lumens per wall-plug watt, and an acquisition
38

MaR 2014

nORTh aMERICan

SPECTRUM.IEEE.ORG

cost lower than that of a lamp-based projector, plus the many
bulbs needed over its lifetime (these lamps cost about $1000
apiece). Laser Light Engines is close to achieving all of these figures; the last goal will be the most difficult, but I am confident
that it will be achieved in three to five years.
All the major projector makers have now joined Laser Light
Engines in building laser-illumination systems. The brightest
of these projectors can put out 70 000 lumens-several times
as many as an arc-lamp projector. That brightness is more than
enough to offset the losses caused by 3-D. This past November,
Laser Light Engines and NEC demonstrated projectors with this
new light source at the Technicolor facility in Burbank, Calif.
Higher brightness benefits not just 3-D but 2-D movies, too.
The reason is that more brightness means a greater range of
luminance, or brightness, from sunlight bright to deep black.
But to take advantage of that wider range, software specialists
will have to increase the number of levels of digital encoding between bright and dark pixels, to create a smooth ramp in brightness over that extended range. This increase in "bit depth" per
pixel in turn will require huge increases in digital bandwidth,
but it'll be worth it. Today, movies don't even come close to displaying the natural contrast that the human eye can see.

in the longer terM, digital cinema and laser projectors will
far transcend the boundaries of traditional film. Today's arclamp-based projectors can produce only about 40 percent of
the colors that most people are capable of perceiving, whereas
laser-based systems can reproduce up to around 60 percent.
Laser illumination can also project much more saturated colors because its red, green, and blue beams can have much narrower spectral bandwidth than filtered lamp light.
This vastly greater, brighter, and more saturated palette will
translate into movies that are more vivid than anything possible today. But such an advance won't come easily: More colors
will require coordinated changes to global standards, and that
won't happen without a lot of arguing over how "wide" to go.
More colors will require more bits, which will in turn require
more bandwidth to and within the projector.
Higher contrast and color rendering aren't the only factors that
will increase the size of movie files. Movie directors are starting
to use frame rates higher than 24 per second, the standard since
around 1927. Peter Jackson's The Hobbit: An Unexpected Journey
(2012) was filmed at 48 frames per second, as was its recent sequel,
The Hobbit: The Desolation of Smaug. A movie's frame rate has a
huge influence on how the viewer perceives motion, and it also
increases perceived contrast and resolution. Higher frame rates
allow the appearance of fast-moving objects to remain supersharp and can eliminate the jerky effects that can arise when
the camera or the subject is moving. But as with higher dynamic
range, there is a price to pay. Showing more frames per second
increases not only the quantity of data in the movie file but also
the data-transmission rates necessary to project that movie.
Yet another possibility for future movies is greater spatial
resolution, or pixels per scene. This resolution is limited by
spatial-light-modulator technology, either DMD or LCoS. Most

http://SPECTRUM.IEEE.ORG

Table of Contents for the Digital Edition of IEEE Spectrum - North American - March 2014