White Paper Method of Measuring and Quantifying the Amount of Sparkle in Display-Touch Panel Stacks
Journal of Information Display, 2014 http://dx.doi.org/10.1080/15980316.2014.962633 Method of measuring and quantifying the amount of sparkle in display-touch panel stacks C.R. Evans, M. Skibinski and D. Gullick Visteon Engineering Services Ltd, Visteon Innovation Centre, 1 Springfield Lyons Approach, Chelmsford, Essex CM2 5LB, UK (Received 2 July 2014; accepted 25 August 2014) Presented herein is a measurement method that can be used to quantify sparkle with a value that is independent of the display under test, and that does not require any reference state. A jury appraisal was detailed, which was used to assess the level of sparkle in a display stack and grades the sparkle performance as not noticeable, noticeable but acceptable, or unacceptable. The measurement results are presented using the described method for both a 4.2-inch (131.2 dpi) display and an 8.0-inch (116.8 dpi) display with several different anti-glare films giving a range of sparkle performances. The values have been normalized to the perceived luminance to make the measured values independent of the display luminance. Using this method, it was reported that a measured value of less than 0.6 will not be noticeable, a value between 0.6 and 1.4 will be noticeable but acceptable, and any value above 1.4 will be unacceptable. Keywords: speckle; sparkle; display metrology; anti-glare (AG); display Introduction Due to the increase in the complexity and expected functionality of electronics in vehicles, there is a strong trend towards adding touch panels and optical lenses. To reduce the overall reflections from the system, an anti-glare (AG) film is often placed on the top surface. This film has a rough surface to diffuse the reflected light. This improves the quality of the display image to the user under brightlight ambient conditions. Parallel to this, there is another trend towards increasing the display resolution to try and match the display resolutions seen in the consumer market. This has led to the pixel sizes of the displays becoming approximately the same size as the features of the AG surface. This can cause a visual sparkle effect on the display image, which can be distracting to the viewer. To reduce this effect, the AG film can be altered or optimized so that the size of the features does not match the size of the pixels, but this sometimes has a negative impact on the reflection, haze, fingerprint visibility, and image clarity [1]. In the automotive industry, there is no agreed standard method of measuring and quantifying the amount of sparkle, and most products rely on a jury appraisal to assess the overall optical performance of the display stack. The current published methods [2,3] of sparkle measurement often require a reference state with no sparkle, which is often not possible. In this work, the effects of using several different AG films on two displays were investigated. The first display was a 4.2-inch WVGA (480 272 pixels) display with a dpi of 131.2 pixels per inch. The second display was an 8-inch display with a resolution of 800 480 pixels, giving a dpi of 116.8 pixels per inch. For both displays, a capacitive touch panel was attached to the metal bezel of the display using a 0.5-mm-thick gasket. The thickness of the touch panel was 2.18 ± 0.2 mm, and thus, the AG film was located 2.68 ± 0.3 mm away from the display surface. Figure 1 shows the construction of the displaytouch panel stack. For the top surface of the touch panel, several AG polarizer films were investigated to create a range of display-touch panel stacks with varying sparkle performance. Measurement method In high-resolution displays with poorly chosen AG films, the display pixel size becomes small enough to be comparable to the feature size in the AG film. This causes interference and scattering issues, which mean that the viewer sees a sparkle effect in the image (Figure 2). This effect is more prominent in larger areas of uniform color, and is more noticeable to the eye in green areas. For the accurate measurement of the pixel luminance, it is important to ensure that a detector resolution greater than 10 camera pixels can be achieved for each display pixel. In this study, a detector resolution of 14 camera pixels was achieved for each display pixel. The camera was focused on the display pixels through the touch panel, with the display showing an all-green test screen. Figure 3 shows *Corresponding author. Email: cevans6@visteon.com 2014 The Korean Information Display Society
2 C.R. Evans et al. Figure 1. Construction of the device under test. Figure 4. sparkle. The left side has high sparkle and the right side, low Figure 2. sparkle. The left side has high sparkle and the right side, low samples from Figure 3. It was found that displays with high-sparkle contents have a larger variance in peak luminance compared to displays with low-sparkle content. The peak luminance was recorded for each individual display pixel, and the standard deviation (SD) was calculated. A jury appraisal was then performed, in which people were asked to rate the performance of a range of AG films in terms of sparkle. An all-white test screen was displayed on the screen to maximize the amount of sparkle, and the testers were asked to judge the performance of several different films that were placed 2.68 mm away from the surface of the display. The jury panel consisted of a total of 30 people, half of whom were managers and half, engineers. All had a technical background, and all were experienced with displays in automotive applications. From this study, a rating of perceived sparkle was obtained for each of the different films. The users were initially shown a setup with no sparkle and one with high sparkle to ensure their familiarity with the phenomenon. Subsequently, the users were shown the unit under test so that they could grade the sparkle effect from 1 to 10. A rating of 1 2 represented no noticeable sparkle; 3 5, noticeable but acceptable sparkle; 6 7, highly noticeable but still acceptable sparkle, and 8 10, unacceptable sparkle. Figure 3. The top image shows low-sparkle configuration and the bottom image, high-sparkle configuration. the image that was captured for two AG films: one with a high sparkle effect and one with a small sparkle effect. These images were analyzed by placing a line across the horizontal axis of the green pixels and plotting the horizontal luminance across the pixels. Figure 4 shows the luminance distribution for the high- and low-sparkle Results The perceived sparkle was plotted against the SD of pixel peak luminance, as shown in Figure 5. Here, one linear trend for the 4.2-inch display (triangles) and a separate linear trend for the 8-inch display (squares) can be seen. When one stack with high sparkle is taken for the 4.2- and 8.0-inch displays, and when the display luminance is reduced by lowering the display backlight (filled-in triangles and squares, respectively), it can be seen that the measured sparkle quickly decreases but the perceived sparkle decreases only marginally. This shows that the measured SD has a linear response to the perceived sparkle only if the end luminance to the user remains constant.
J. Inf. Disp. 3 Figure 5. Perceived sparkle against the SD of peak pixel luminance. The squares represent the 8-inch display and the triangles represent the 4.2-inch display. The filled-in icons represent one stack with high sparkle as the luminance is dimmed. Figure 6. Response of the human eye to luminance using an exponent of 0.5. This variation of the perceived sparkle with luminance can be explained by the human eyes response to luminance. The human eye has a greater sensitivity to light at lower luminance levels than at higher luminance levels. Due to the increased sensitivity of the eye to changes in luminance at lower luminance levels, the deviation in individual pixel luminance will be more noticeable to the viewer at a lower luminance level than at a higher luminance level. The relationship between the perceived luminance and the measured luminance is described by Stevens power law [4]: L p = L a m, (1) where L p is the perceived luminance, L m is the measured luminance, and a is the exponent that describes the stimulus used. For the human eye, the exponent is 0.5 [4]. Figure 6 shows the relationship between the perceived luminance and the measured luminance using an exponent of 0.5.
4 C.R. Evans et al. Figure 7. Perceived sparkle against SD* L s. The squares represent the 8-inch display and the triangles represent the 4.2-inch display. The filled-in icons represent one stack with high sparkle as the luminance is dimmed. Equation (2) calculates the gradient or sensitivity of the eye to change at a particular luminance L s = L p L m = 1 2 L1/2 1 m, (2) where L s is the luminance sensitivity for the human eye. If this perceived luminance factor is multiplied by the SD of pixel peak luminance, a measurement value of sparkle that is normalized to the sensitivity of the human eye to light can be calculated, using the following equation: Sparkle = SD L 0.5 m 2 = SD L s. (3) Using Equation (3), the perceived sparkle can be plotted against the SD of the peak pixel luminance multiplied by L s. Figure 7 shows that by normalizing the data to the sensitivity of the eye at the operating luminance, a unified response can be obtained for both the 4.2- and 8.0-inch displays for the measurements taken at different display luminance levels. This normalization allows the sparkle to be measured independent of the display under test or of the display luminance. Conclusion Presented in this paper is a measurement method that can give a sparkle value that is independent of both the display under test and the display luminance. A jury appraisal was detailed that grades the amount of sparkle in a display and in a potential touch panel stack based on a Likert scale from not noticeable to noticeable but acceptable and to unacceptable. The results of the measurement using the described method are presented for both a 4.2-inch display and an 8.0-inch display, with several different AG films giving a range of sparkle performance. Using this measurement method, it can be seen that if an SD* L s value of less than 0.6 is obtained, the sparkle will not be noticeable to the user. For a measurement value between 0.6 and 1.4, it can be said that the sparkle will be noticeable but will be generally acceptable, and any value above 1.4 will be deemed unacceptable. It is also worth noting that the jury appraisal was done with a test image of an all-white screen; thus, with more realistic display graphics, this effect will be less noticeable. References [1] P. Weindorf and B. Hayden, Anti-Glare Sharpness Measurement Investigations, SID Vehicle Displays & Interfaces, University of Michigan, Dearborn, MI, 2012. [2] J. Gollier, G.A. Piech, S.D. Hart, J.A. West, H. Hovagimian, E.M. Kosik Williams, A. Stillwell, and J. Ferwerda, SID Symp. Dig. Tech. Pap. 44, 295 (2013). [3] M.E. Becker and J. Neumeier, SID Symp. Dig. Tech. Pap. 42, 1038 (2011). [4] S.S. Stevens, Psychol. Rev. 64, 153 (1957).
White Paper About Visteon Visteon is a global company that designs, engineers and manufactures innovative cockpit electronics products and connected car solutions for most of the world s major vehicle manufacturers. Visteon is a leading provider of driver information and controls, audio and infotainment, and domain controllers; its brands include Lightscape, OpenAir and SmartCore. With corporate offices in Van Buren Township, Michigan, (U.S.); Shanghai, China; and Chelmsford, UK; Visteon has 50 facilities in 21 countries. Learn more at www.visteon.com. Visteon Corporation One Village Center Dr. Van Buren Township, MI 48188 1-800-VISTEON www.visteon.com Copyright 2015 Visteon Corporation Secure Virtualization for an Ever Increasing Vehicle Complexity 2