CCD vs. CMOS

Dave Litwiller

The CMOS marketplace will consolidate, CCDs will continue to advance and cameras will become more integrated — all of which should benefit the user.

One of the most significant imaging technology events will occur in 2002 because of business circumstances: consolidation of the CMOS imaging marketplace. The past five years has been a period of great enthusiasm for the potential of CMOS imagers, which were promoted in varying degrees as performing as well as CCDs while being more highly integrated and less expensive. They were heralded to displace CCDs in traditional applications and to enable entirely new applications, especially in consumer and near-consumer imaging products.

In some cases, CMOS imagers have risen to prominence, particularly for those applications that don’t need the high image quality of CCDs. In applications such as PC videoconferencing, games, automobiles, desktop scanners, bar-code scanners and security cameras, they are becoming the norm.

However, the realities of a demanding marketplace are taking their toll on the world of CMOS imagers. A few years ago, a good but untested argument for these imagers triggered rampant experimentation by both producers and customers, creating distorted market behavior. Cheap capital from investors subsidized products and allowed consumers to buy at prices well below true costs. This created artificially high demand, and many buyers were drawn in by curiosity despite limited or uncertain benefits.

This state of affairs is giving way to the challenges of operating a sustainable, competitive CMOS imager business. In many applications, when image quality matters more than integration, CCDs remain a more suitable technology. They are superior to CMOS imagers in performance-driven applications ranging from the consumer realm — such as video cameras and digital still cameras — to high-end uses in industrial, scientific and medical imaging.

As a result of the limited success of CMOS technology in performance-driven applications, a contraction of players in the CMOS imager industry is already under way. This trend will gain momentum during 2002, as participating companies exhaust the eager capital of a few years ago. The best CMOS business units — those with substantial differentiating technology that lowers costs and improves application-specific performance — will continue and thrive. Intermediate firms will consolidate — subject to the egos of management — to try to achieve critical mass in technology and market share to become efficient. The weaker and most overhyped firms will suffer selective failure, and buyers will experience more limited product choices. Only profitable, sustainable product lines will remain, but the post-shakeout CMOS imager industry will emerge more solid and will offer customers better support.

The days of start-up CMOS imager businesses coming to market with many new devices in a short period are giving way to a more realistic pace of product development and launch, based on cash flow rather than on capital. The era of breakthroughs and unlimited potential is gone. The demands of running a business rather than an incubator for an exciting technology have set in. Rational, streamlined product lines and producers will remain, delivering an incremental and more sustainable pace of CMOS imager progression in 2002.

A web inspection system uses 15 Eclipse line-scan cameras to detect defects as small as 15 μm.

At the same time, CCDs continue to advance. And steady performance strides in speed, dynamic range, image uniformity, imager size and extended wavelength response will continue to characterize CCD development and introductions.

Whether they use CCD or CMOS image sensors, cameras will continue to shrink in size because of reduced component power dissipation, higher component integration and smaller component footprints. Camera integration will increase, resulting in richer user interfaces, enhanced programmability and migration of system functions to the camera. These improvements will make cameras easier to use and accessible for a broader range of applications. In industrial uses, for example, camera improvements will continue to deliver increased automation of manufacturing and quality control operations.

Standardization of camera interface through the Camera Link protocol will increase competition because users will be able to easily substitute cameras. This will place a greater premium on differentiation of the front end of the camera: optics, sensor performance, camera size, functionality and reliability. Proprietary camera interfaces that locked users into a particular vendor will recede, creating a more vibrant marketplace of open standards competition.

Overall, 2002 promises steady progress for imager and camera technologies. The hype and oversold expectations in some pockets of the imaging industry over the past few years will die down as a more enduring suite of products emerges. This will offer users better choices and more flexibility to meet their imaging needs.

Meet the author

Dave Litwiller is vice president of corporate marketing and business development at Dalsa Corp. in Waterloo, Ontario, Canada.