Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

Measurements of the photometric characteristics of displays are important for many reasons, such as assessing performance for product development, manufacture, and quality control purposes, enabling purchasers to compare products on a consistent basis, and as inputs to software used to predict display legibility in specific installations. A range of measurements may be required, as summarized in Table 1. This chapter will provide a brief overview of these measurements, the types of instrumentation that are available, and the major potential sources of measurement error.

Table 1 Measurements for characterizing the optical performance of a visual display
Full size table

Luminance

Luminance is the most basic measurement for a display. Not only is it the foundation of many of the other measurements (as detailed in part 51, “Advanced Display Measurement Procedures”), but it is also often used within the display industry as a general rule of thumb regarding the suitability of the display for a particular application. For example, a display with a luminance of 500 cd m−2 will be on the borderline of being acceptable (“bright enough”) for use in daylight or other high illumination conditions, whereas a luminance of 1,500 cd m−2 would be considered to be very likely to be acceptable. Luminance is usually measured for both a black and white screen using a luminance meter (sometimes called a spot photometer) or a telespectroradiometer (see Part 50, “Standard Display Measurement Procedures” and chapter “Standards and Test Patterns” for more details). The measurements are usually performed with the measuring instrument perpendicular to the surface of the display, or at the angle at which the user would normally view the display (if this is not perpendicular to the surface). The size and location on the display of the measured area should be stated, since most displays show some nonuniformity in luminance over the surface. It is also important not to measure too small an area, since this can result in large differences in the measured results with small changes in size or position (at the extreme, if the measurement area is the same size as a single pixel, small movements can mean that the area is either located precisely over a pixel, giving a “high” reading, or only partially over a pixel, giving a “low” reading; neither reading will adequately represent the true display luminance). Precautions must also be taken to minimize the effect of stray light from areas of the screen other than the defined measurement area, and this generally involves the use of either a stray light tube on the measurement instrument or a mask on the surface of the display. Other major sources of error are the performance of the measurement instrument (see Sect. 10, “Mobile Displays, Microdisplays, Projection, and Headworn Displays” and chapter “Measurement Devices”) and external influences on the output of the display (e.g., CRTs are susceptible to external magnetic fields and LED displays can be affected by changes in ambient temperature). Results are expressed in terms of candela per meter squared (cd m−2).

Color

The range of colors that can be displayed (the “color gamut”) is evaluated by measuring the chromaticities of red (R), green (G), blue (B), and white (R = G = B) screens. The approach, precautions, and sources of error are similar to those for measurements of luminance, but in this case, the instrumentation used is a colorimeter (which is typically placed directly on the surface of the screen) or a telespectroradiometer (which is imaged onto the screen and provides measurements of the spectral radiance). Results are expressed in terms of CIE (x,y) or (u′,v′) chromaticity coordinates or using another specified color system. More details are given in Part 50, “Standard Display Measurement Procedures” and chapter “Measurement Devices.”

Gray-Level Step

These measurements are also performed in a similar manner to luminance measurements, but in this case, the white screen is varied over the full range of RGB drive levels (i.e., from 0 to 256), and the luminance is determined for each step (see Parts 50, “Standard Display Measurement Procedures” and 51, “Advanced Display Measurement Procedures”). A test pattern generator is usually used for this purpose (see chapter “Standards and Test Patterns”) to avoid problems that may arise if a personal computer (PC) is used to set the gray levels (a PC provides the means for altering brightness, contrast, and gray-level step in a way that is often hidden from the operator). Results are usually expressed graphically, as shown in Fig. 1.

Fig. 1
figure 1

Example of results from gray-level step measurements

Full size image

Contrast Ratio

Contrast ratio is defined as the ratio of the highest luminance to the lowest luminance that the display system is capable of producing. The larger the contrast ratio, the greater is the difference between the brightest whites and the darkest blacks that can be displayed. A high contrast ratio is a desirable aspect of any display, but it is not always possible to make a direct comparison between the contrast ratio values provided by different display manufacturers, due to differences in the measurement methodologies used. The most representative measure of contrast ratio for assessing overall display performance is static contrast ratio , which refers to the ratio between the luminances of the brightest white and the darkest black that can be displayed simultaneously. Dynamic contrast , on the other hand, refers to the ratio between the deepest blacks and the brightest whites that a display can display, but not at the same time, and generally results in higher contrast values. This is particularly true in the case of displays employing backlights (e.g., LCDs), where light can bleed through from the backlight into black areas when an image containing both white and black areas is being viewed (thus reducing contrast ratio), whereas the backlight can be reduced or even turned off if a fully black image is displayed.

A further complication is that, regardless of whether static or dynamic contrast is measured, the results obtained can depend significantly on the ambient lighting conditions. Contrast ratio is often quoted for dark room conditions, that is, with no ambient illumination present and minimal reflections from the surroundings. In most instances, this is not the environment in which displays are used, and different values will be obtained if ambient illumination is present, and/or the surrounding walls, floor, and ceiling can reflect light from the display back onto the screen.

There are two commonly used methods of measuring contrast ratio. The “full on/off” method compares the luminance of a white screen (R = G = B = max) with that of a black screen (R = G = B = 0) and has the advantage that it largely cancels out the effect of the external environment (equal proportions of light are reflected from the display to the room and back for both the “black” and “white” measurements, as long as the room stays the same). This method is generally suited only to dynamic contrast measurements, unless it is possible to control the display such that the backlight is fully on even when displaying a black image. The second method is to use a checkerboard pattern, in which the luminance values of all the white squares (or rectangles) are measured and averaged, and similarly the luminance values of the black squares (or rectangles) are measured and averaged. The ratio of the averaged white readings to the averaged black readings is the contrast ratio. This method provides static contrast values. However, accurate measurement of contrast using this checkerboard approach requires the use of a well-controlled dark room, with all walls, floors, ceilings, etc., totally black and nonreflective; this can be difficult and expensive to achieve.

Whatever method is used and regardless of whether static or dynamic contrast is being measured, results are expressed as the ratio between the luminances of the “white” and the “black” conditions, for example, 1,000:1. More details of contrast ratio definitions and measurement methods are given in chapters “Luminance, Contrast Ratio, and Gray Scale” and “Standards and Test Patterns.”

Spatial Luminance and Color Uniformity

The visual appearance and effectiveness of a display can be significantly degraded if there are perceptible variations in the luminance and/or color over the active area of the screen. Such nonuniformities may appear as a gradual variation from one part of the screen to another or as localized variations due, for example, to the structure within the backlight. Prior to the development of imaging photometers/colorimeters, display nonuniformity was assessed by making measurements at a large number of discrete points across the full area of the screen using a spot photometer, colorimeter, or telespectroradiometer (see chapter “Spatial Effects”). The problem with this approach lies in achieving sufficient spatial resolution and conducting a sufficient number of measurements to characterize fully the performance of the display. In practice, the scientist/engineer performing the measurement generally identifies the brightest and dimmest locations on the display by visual inspection and then performs several measurements around these areas. The nonuniformity can be calculated as a contrast ratio between the areas of highest and lowest luminance. The main problem with this approach is that it does not fully represent the overall nonuniformity of the display but gives only two specific worst case points. The low spatial resolution of the measurements can also mask small area nonuniformities. More recently, therefore, this point-by-point approach to measurements of display nonuniformity has been largely superseded by the use of imaging photometers and colorimeters, which produce two-dimensional maps of the variations in luminance (or chromaticity) over the full screen surface, as illustrated in Fig. 2.

Fig. 2
figure 2

Example of results from an imaging colorimeter, showing contrast ratio values in false color

Full size image

Angular Variations in Luminance and Color

Measurements of the angular variation in luminance and color provide the characteristics of the display over the whole forward hemisphere and can be used not only to determine the angular field of view of a display but also to provide a more comprehensive understanding of display legibility under a range of conditions (see chapter “Viewing Angle”). The highest angular resolution and measurement sensitivity is achieved through goniometric methods, in which the luminance or color distribution is mapped as a function of angle. However, these methods require long setup and measurement times and are consequently often too expensive to fulfill the needs of the display industry. Other methods have therefore been developed, based on the use of the latest imaging technologies (Rykowski et al. 2006), but a detailed description of these is beyond the scope of this chapter (see chapter “Measurement Devices” for more information).

Reflectance Measurements

Light reflected from the display surface into the user’s line of sight is superimposed on the displayed image and results in a degradation of the legibility of the displayed image. Conventionally, two types of reflection have been considered within the display community: specular reflection and diffuse reflection. However, with the increased use of antiglare and touch screen coatings, a third type of reflection, termed “haze,” is now also considered and can be a significant contributor to degraded visual performance.

Diffuse Reflectance

Diffuse reflectance measurements (see chapter “Ambient Light” for more details) are intended to quantify the amount of reflected light that will be superimposed on the displayed image from a uniformly distributed diffuse light source. This diffuse light source provides a reasonable approximation of the illumination environments in which displays are often used. For example, the illumination from the sky is diffuse in nature, so this is an appropriate condition to use for displays used out-of-doors, and although indoor environments generally have a somewhat complicated illumination distribution, even these are often adequately represented by a diffuse source to a first degree (e.g., office lighting is often designed to produce good uniformity across the working plane with no visible “bright spots”).

The baseline measurement technique for diffuse reflectance is to place the display inside a large integrating sphere, with a lamp (with baffle) placed behind the display such that this provides diffuse illumination onto the display. Measurements are made of the screen luminance for a measured level of diffuse illuminance and compared with the measured luminance under the same conditions for a calibrated reflectance standard (Kelley 2006). However, this method requires access to an integrating sphere that is large enough to accommodate the display (the diameter of the sphere should be at least ten times the diagonal of the display), and it is therefore not widely used. An alternative approach is the “sampling sphere method ” (Kelley 2006), in which the display is placed against the sample port of an integrating sphere rather than inside it. A lamp is placed inside the sphere, close to the wall, and baffled to prevent direct illumination of the display. The luminance of the display surface is measured using a luminance meter and compared with that measured under identical conditions but with the display replaced by a calibrated diffuse reflectance standard (see Fig. 3). The diffuse reflectance of the display, ρ dis, is given by

Fig. 3
figure 3

Schematic of sampling sphere measurement of diffuse reflectance

Full size image
$$ {\rho}_{\mathrm{dis}}=\frac{\rho_{\mathrm{std}}\;{L}_{\mathrm{dis}}}{L_{\mathrm{std}}} $$

where ρ std is the reflectance of the calibrated standard, L dis is the luminance measured when the display is in position, and L std is the luminance measured with the reflectance standard in position. If the display emits light, then the luminance of the display must be subtracted from the luminance measured under reflection to obtain the net reflected luminance. Measurements are usually made for a range of display conditions (white and black as a minimum) since the reflectance may vary depending on the display settings.

Specular Reflectance

Measurements of specular reflectance are made to determine the degree of “mirrorlike” reflection from a display (see chapter “Ambient Light” for further details). Specular reflections are generally several orders of magnitude larger than diffuse reflections and can be a major source of discomfort if a display is incorrectly positioned within a lit environment. Measurements are usually made by comparing the luminance of display when viewed at angle θ to the normal and illuminated by a point source at − θ to the normal with that for a calibrated specular reflectance standard (typically a piece of black glass) that is illuminated and viewed under identical conditions.

Angular Reflectance

Measurements of angular reflectance provide the reflectance characteristics of the display over the whole forward hemisphere and can therefore be used to determine the details of how the display reflectance will impact on its legibility under any given conditions (see chapter “Ambient Light” for further details). The importance of these measurements has grown in recent years due to the increasing use of display screens with antiglare or touch screen coatings, both of which introduce nontrivial “haze” reflections, that is, reflections that are intermediate between diffuse and specular in nature. The most serious effect of haze is to “broaden out” the specular reflection, making it less easy for an observer to avoid the reflection by changing their viewing location and thus resulting in a display that is less legible than would be the case if only specular reflection were present.

As in the case of measurements of angular luminance and color variations, the most comprehensive and accurate measurements of the angular reflectance properties of a display are obtained using goniometric methods, yielding the bidirectional reflectance distribution function (BRDF) (Kelley et al. 1998). As for other angular measurements, simplified approaches are under development based on imaging technologies, but these have not yet gained international acceptance and are outside the scope of this chapter (see chapter “Measurement Devices” for further information).

Temporal Performance (Motion Blur)

Measurements of the temporal performance of a display relate to the degree of image persistence for a dynamic image (see chapters “Temporal Effects” and “Standards and Test Patterns”). These measurements are particularly important for video images containing fast-moving, high-contrast targets, such as television coverage of football and tennis (where motion blur can cause the fast-moving ball to be hard to distinguish). The issue of motion blur has become more important with the advent of new displays, such as LCD screens. Unlike CRT displays, where the image is displayed only for a short time during each refresh cycle and is blank between each image, in an LCD display, the image is held on the screen during the entire refresh period. This means that for a fast-moving object in the image, the object position is correct for only a fraction of the time, and the eye interprets this as the object being blurred. In practice, there are two contributors to motion blur: the rise and decay time of the pixels and the hold time. The former can be measured using a fast photodiode and the latter is dictated by the display drive electronics.

Basic Measurement Instrumentation

As described in chapters “Measurement Instrumentation and Calibration Standards” and “Measurement Devices,” most measurements of a display are made using either a spectroradiometer (Commission International de lÉclairage 1984), which provides measurement results as a function of wavelength, or a filtered broadband detector (Commission International de lÉclairage 1982), which is designed to give an approximation to one or more of the CIE standard observer functions. The different characteristics of these instruments can lead to different, but in each case significant, measurement errors, as summarized in Table 2.

Table 2 Major potential sources of error with spectroradiometer (SR) and filtered broadband detector (BB) systems
Full size table

Both types of instrument are typically calibrated using a stable and reproducible reference source, such as a luminance gauge . The reference source may be calibrated in terms of its luminance, its chromaticity, or its spectral output (usually absolute spectral radiance) as a function of wavelength by a laboratory that is traceable to national standards. The instrument is calibrated by comparing the measured values for the reference light source with the calibration data, yielding a correction factor or factors. All instruments will show some drift in calibration with time, so it is important to check the calibration at regular intervals. Furthermore, reference sources also drift, both with time and with usage, so it is important that these are recalibrated at regular intervals by a laboratory providing measurements that are traceable to national standards.

Summary

The key measurements required in order to characterize the performance of a display are luminance, color, spatial uniformity, angular distribution, reflectance, and temporal characteristics. These provide basic, underpinning information relating to the quality, usability, and legibility of the display under different conditions of use. This section has provided an overview of the methods, instrumentation, and major sources of potential error and uncertainty associated with these measurements; more details on all these aspects are provided in Sect. 11, “Display Metrology.”