Technical Progress Should Not Overwhelm Common Sense
by Tom Fiske
Welcome to the metrology issue of Information Display magazine. Technical progress is inexorable. As time goes on, photon-catching detectors get more sensitive. Sensor arrays get more pixels. Computers process more data than ever before. What we often need more of, however, is common sense. That is not to say that detector sensitivity, numbers of sensors on a chip, or computational power are not important. But they are not the only important things. For these advancements to be useful, one has to know what to do with the extra sensitivity and the additional data. Our contributions this issue help us along the way of optimizing that extra data and computer power.
We have two articles this month describing advances in metrology. The first one, from NIST scientists John Penczek and Paul Boynton and consulting display-metrology expert Ed Kelley, recommends a standardized method for finding the optical performance of displays in any ambient lighting environment. They describe a common-sense extension of the principles in the Information Display Measurements Standard (IDMS) that cover emissive, reflective, and transparent displays. With straightforward characterization of the reflective and transmissive properties of a display, one can use these principles to account for how ambient lighting will affect the visual performance of almost any display. These authors show us how to use these principles to report the optical properties of transparent displays in a variety of common lighting environments.
The second article is by Đenan Konjhodžić, Peter Khrustalev, and Richard Young of Instrument Systems, GmbH. They report on a new technique for extending the usefulness and accuracy of an imaging colorimeter. There are a few different paths one can take to increase the accuracy in such a system: Optimize the accuracy of the “CIE” filters, increase the number of filters used for the colorimetric measurements, or use a set of accurate spectroradiometric measurements of a typical spectral power distribution (SPD) to construct a transformation matrix. There will always be some deviation in the CIE matching filters. Increasing filters adds time and cost. Matrices increase accuracy, but limit the system to measuring accurately only a narrow range of SPDs. Konjhodžić and colleagues propose a method that uses six filters and a range of training spectra. Optimization and appropriate choice and weighting of the training spectra
improve the transformation matrices and increase the accuracy of the system for a general range of SPDs.
I had the privilege of reporting on Display Week for Information Display magazine. Along with several others, we covered various aspects of the event by writing blogs (http://idmagazinedisplayweek2015.blogspot.com/) and articles for this edition of ID. Although not strictly part of the metrology issue, my article covers image quality and metrology for Display Week 2015. In it you will find my take on high-dynamic-range and extended-gamut displays as well as on recent offerings from display-measurement system providers. •