Projection Technology in Digital Cinema

by Michael Karagosian
©2004 Karagosian MacCalla Partners, all rights reserved worldwide
Published in the June 2004 issue of INS Asia Magazine

Technology has clearly advanced to the point to where electronically projected images can replace the venerable 35mm film projector. However, 'can' is not necessarily an indication that it 'will' replace film anytime soon. As has been pointed out in this series of articles, the number of obstacles that stand in the way of a world-wide rollout of digital cinema are not few. Certainly, the high cost of the projector is among these, providing a pillar of encouragement for those with potentially low cost projection technology.

The projector of choice today for digital cinema trial systems is based on the Texas Instruments m25 DLP Cinema™ light engine. This "2K" projector can produce an image having a 2048 x 1024 pixel array, with a color gamut and contrast ratio sufficient to duplicate images as seen with film projection. Not all may agree that the image quality is up to the best of film projection. But recent demonstrations in Hollywood of the ASC/DCI StEM (Standard Evaluation Material) projected side-by-side in both digital and film have satisfied most 'golden-eyes' that digital projection today can do a remarkable job of emulating the film experience.

TI's DLP™ technology is unique among methods of digital projection. The technique is based on an array of digitally-controlled mirrors, one mirror per pixel. Light is modulated by flipping the mirrors such that the light value of each pixel is proportional to the duty cycle of its mirror. The technology used to manufacture the chip is called MEMS, which stands for Micro-Electro-Mechanical Systems. The micro mirrors and their mechanical support are fabricated on the DLP chip. By flipping the mirrors in response to the input signal, the device digitally modulates light reflecting off the chip. An important aspect of this technology is that the data bits of the digital image are never converted to an electrical analog signal. Instead, the digital image data directly determines the number of flips each mirror makes.

While most quality projectors today accept image data in digital form, DLP is the only technology on the market whose projected image is the direct result of digitally modulated light. With DLP, it is our eyes that integrate the pulses of light into an analog signal. The response time of the mirrors is quite fast, and the rendered image has no persistence. An interesting result is that DLP projected images can be very revealing, exposing image artifacts that a technology such as CRT (cathode ray tube) will conceal. Another point on the plus side for DLP Cinema is its proven reliability, without which there wouldn't be a digital cinema discussion today. However, there is a negative side to DLP. MEMS technology is expensive, and difficult to extend to higher-resolution arrays. Also, the mirrors are manufactured with a noticeable gap between them, which affects the visual "fill" factor of the projected image.

Other technologies exist, however, which could also find application in digital cinema. LCOS (liquid crystal on silicon) is an imaging technology embraced by several manufacturers, both for consumer applications as well as professional projectors. The technology employs a layer of liquid crystal grown on the highly reflective aluminum layer of a silicon circuit. The reflective layer is pixilated, through which the opacity of the liquid crystal area immediately above is controlled by means of an electric field. Light is modulated by reflecting it off the LCOS device, which must pass through the liquid crystal layer before and after it reflects.

In terms of switching speeds, liquid crystals are not a particularly fast technology. Their response time is slow when compared to the speed of a MEMS mirror, and so cannot be digitally toggled on and off in the manner of DLP devices. Instead, the liquid crystal is operated as an analog device. The electric field that controls the opacity of the crystal is created by an analog voltage. This requires that LCOS devices convert the digital image data into an electrical analog signal for the display to operate.

LCOS technology has certain cost advantages. Liquid crystal silicon technology is cheaper to manufacture than MEMS, and projectors built with this technology require less optical glass than DLP, further reducing the cost of the product. A notable feature of LCOS is its fill factor. As the liquid crystal layer has no boundaries between pixels, fill factors greater than 90% can be achieved, eliminating the "screen door" effect observed with some display technologies. But LCOS also has its problems. The requirement for analog voltages invites all of the problems associated with analog circuitry, including drift with temperature, affecting color and intensity stability. A difficult issue is the temperature sensitivity of the liquid crystal itself, which imposes a warm-up time for the projector. When a beam from a 6000 Watt xenon lamp lights the crystal, temperature gradations will occur, producing unwanted color variations across the projected image. This makes LCOS difficult to tame for the high brightness, large screen applications typical of cinema.

Another technology that is often discussed for cinema applications is laser projection. Laser projectors come in two flavors: those that use lasers to draw the image directly on the screen, and the second where the lasers are used as a coherent and efficient light source for conventional imaging technologies. There are pros and cons for each approach, but all laser projection methods share the advantage of efficiency. As a light source, lasers are inherently monochromatic, meaning they produce light which predominantly has a single wavelength. To produce three primaries, three lasers are employed. Conventional xenon lamps, in contrast, produce a full spectrum of light, extending quite far into infrared. The infrared frequencies must be filtered out without blocking visible reds, and then further filtering is performed to produce the primaries. Consequently, an enormous amount of energy is thrown away when using conventional light sources, and this energy is dissipated within the projector. The use of lasers will produce a cost savings in power consumption, dissipate less heat in the projector, and optimize the energy transferred to the imaging device. Lasers are not exactly cheap, but when their cost is amortized over a 10-year lifetime, they can be cost effective when compared to the cost of replacing high wattage xenon lamps over the same period.

All laser projection devices, though, have a common problem called 'speckling'. When the coherent light of a laser reflects from the optically rough surface of the screen, visible peaks and dips in light intensity are perceived by the eye. These variations in intensity occur as spots randomly positioned both spatially and temporally. The result is a random pattern of light and dark spots superimposed on the projected image, which is not an acceptable situation. One of the challenges of laser projection is to eliminate or sufficiently minimize speckling.

Laser projectors can also pose a safety issue. If using the laser to draw the picture on the screen line-by-line, a very powerful laser must be used. If the beam is continuously moving, it may not impose a danger to the audience. But if the projector should misbehave and the laser beam become stationery, the potential for serious injury exists. For this reason, lasers as light sources and not as display devices may become the smarter application in theatres. Some companies propose to solve the problems of LCOS in high brightness applications, for instance, by using laser light sources.

Digital cinema is only in its infancy, and projector technologies will lead the way as this new application develops and matures. This article presents a few of the technologies that have potential in this marketplace. Let's see what the future brings.