Display Measurement and the Development and Manufacturing Process
by Thomas G. Fiske
Welcome to the 2007 Display Metrology issue of Information Display magazine, for which I was proud to serve as Guest Editor. Display measurement is a critical part of any display-device development and manufacturing endeavor. Measurement, monitoring, and feedback are important for any successful process, but display optical evaluation is unique in that almost everyone in the enterprise – from the CEO to the mailroom clerk – can form an opinion about the optical quality of a display. After all, everyone's got eyes, so everyone's a display expert, right? Well, no. While it is true that anyone can form an opinion about display image quality, the validity and relevance of a given opinion is going to depend on the nature of a particular display application, the image-quality requirements for that application, the viewing environment, the fitness of an individual's vision, and the knowledge and experience of the observer.
Displays are subject to all sorts of undesirable optical artifacts: blemishes, spots, streaks, flicker, motion blur, contouring, and noise, among other things. These can be caused by problems in the display-manufacturing process, inferior materials, electronic-signal integrity issues, or substandard image capture, just to name a few. The causes of many artifacts are obvious and easily solved, but some require careful measurement, analysis, and several iterations to find the root cause — as a colleague likes to say, "If you can't measure it, you can't fix it." To be useful, however, the measurements have to be relevant and well planned, performed carefully and with the proper methodology. The results have to be analyzed by personnel with the experience to properly interpret the results, find the problem, and recommend corrective action. If this is not done with enough care and foresight, project schedules slip, manufacturing lines shut down, and profits decline. This is not necessarily a plea for hiring more display-measurement engineers, but it does highlight the importance of the ready availability of an effective portfolio of display-measurement expertise.
We offer a wide range of display-measurement topics in this issue, from human perception, to spatial and temporal analysis, to measurements for daylight-readable displays. The articles cover some of the important and recent advances in display-measurement methodology, techniques, and analysis. The authors for this issue have years of experience in the display industry, and we are fortunate to reap the benefits of their expertise.
The industry is busy creating displays that can render truly stunning images for all kinds of applications: modules for mobile devices, large displays for home entertainment and business use, and specialized displays for industrial and military uses. As diverse as these applications are, all of them are performing one essential task: conveying information to a human observer. One of the primary goals of the display-measurement task in a manufacturing environment is to determine if a particular display will succeed in performing its designated task. Since the ultimate customer is the human visual system (HVS), it behooves the manufacturer to choose inspection success criteria that are relevant to this ultimate customer. This is precisely the goal of several of the articles presented in this special issue.
Andrew B. Watson of the NASA Ames Research Center in Mountain View, California, incorporates a model of the HVS in the Spatial Standard Observer (SSO), a new tool for estimating the visibility of spatial patterns in displays. Watson brings some human sensibility to the world of display optical specification and measurement by applying this validated model of the HVS to common display-measurement tasks – mura detection, image blur, and projection-screen evaluation, for example. Tasks that are currently being done by trained human evaluators can now be done with a carefully set-up and calibrated machine-vision system that analyzes captured display images and characterizes and analyzes them across the spatial, temporal, and color domain. The results of the SSO model analysis are reported in units that are directly relevant to human vision. One still must set thresholds that are appro-priate for the particular application, but the reliability of the results and the cost savings of implementing such a scenario are evident.
Kees Teunissen at Philips Consumer Electronics and his colleagues also make a signifi-cant contribution with their article on how to characterize LCD motion blur. This is an important topic for LCD-TV applications, and much effort has been expended by manufacturers to optimize display systems to reduce this artifact. Tuenissen, et al., do a nice job implementing a motion-blur measurement scheme proposed by Klompenhouwer in 2005 by utilizing the temporal-impulse response of the display and, in so doing, bypass the complexity of alternate methods. They validate their method with a perception experiment.
Louis Silverstein and I present an article on determining the grating modulation of displays via photopic image capture and Fourier analysis. Modulation is an important parameter that describes the ability of a display to render spatial detail. We show our method to be better than the traditional method in that a more robust and perceptually relevant result is obtained. A comparison is made to the method described in the VESA Flat Panel Display Measurements (FPDM) Standard.
Edward Kelley of NIST has been a driving force behind the VESA FPDM and a champion of straightforward and robust display measurements for many years. Here, he presents a common-sense methodology for characterizing the daylight readability of displays.
We hope that you find the information in this issue useful and relevant to your pursuit of excellence in displays.
Thomas G. Fiske is a Principle Systems Engineer at Rockwell Collins Display Systems, 2701 Orchard Parkway, MS-40, San Jose, CA 95134; telephone 408/532-4986, fax –4105, e-mail: email@example.com. His primary experience is in display technology development and metrology.