A preview of some of the most interesting papers appearing in the March 2009 issue of the Journal of the SID. To obtain access to these articles on-line, please go to www.sid.org
Edited by Aris Silzars
Uwe Vogel Daniel Kreye (SID Member) Bernd Richter, Gerd Bunk Sven Reckziegel, Rigo Herold Michael Scholles, Michael Törker Christiane Grillberger Jörg Amelung, Sven-Thomas Graupner Sebastian Pannasch, Michael Heubner, Boris Velichkovsky
Fraunhofer Institute for Photonic Microsystems
Abstract — First prototypes of bi-directional OLED microdisplay devices that combine both display and camera functionality on a single CMOS chip (OLED-on-CMOS) have been designed. The major goal of this integration is to provide capabilities for eye-tracking in see-through HMDs to achieve gaze-based human–display interaction, e.g., in augmented-reality applications. The development of the prototype was accompanied by user studies with a simulated bi-directional microdisplay consisting of a commercially available eye-tracker and a see-through HMD. These tests were aimed at providing basic minimum requirements in terms of temporal and spatial resolution of an eye-tracker to be implemented within the prototype, as well as to evaluate ergonomics of an appropriate user-interface design. A description of the current state of the hardware architecture and design aspects for bi-directional OLED microdisplays are also presented.
FIGURE 3 — Mock-up system consisting of a microdisplay and a head-mounted eye-tracker simulating the bi-directional see-through HMD.
FIGURE 6 — Detailed view of the used OLED structure sandwiched between metal electrodes (top) and a cross-sectional view of a CMOS device with an integrated OLED illustrate the functionality of a bi-directional OLED microdisplay (bottom).
Hiroshi Mukawa (SID Member) Katsuyuki Akutsu (SID Member) Ikuo Matsumura Satoshi Nakano Takuji Yoshida Mieko Kuwahara Kazuma Aiki
Sony Corp.
Abstract — A full-color eyewear display with over 85% see-through transmittance with a 16° horizontal field of view was developed. Very low color crosstalk, less than 0.008 Δu′v′ uniformity, and 120% NTSC color gamut were achieved. Waveguides with two in- and out-coupling reflection volume hologram elements enabled a simple configuration that has an optical engine beside the user's temples. The reflection volume hologram elements used on the waveguides realized a small thickness of 1.4 mm for each waveguide, and an out-coupling reflection volume hologram used as an optical combiner contributed a high see-through transmittance of 85% due to its wavelength selectivity. However, there are technical challenges in achieving a reasonable screen size and quality color images with optics that utilize holographic waveguides because holograms have large chromatic dispersions compared to conventional optical elements such as lenses and mirrors. Approaches to overcome these issues are described.
Figure 1 illustrates the basic structure of the holographic planar waveguide of the eyewear display. The waveguide has an in-coupling and an out-coupling reflection volume hologram which have exactly the same fringe pattern and a mirror symmetrically positioned. Reflection volume holograms were employed because their diffraction bandwidths are much smaller than those of transmission holograms and could potentially enlarge the field of view of the eyewear displays. Each of these holograms has red, green, and blue hologram layers to transmit full-color images through the waveguide.
FIGURE 1 — A basic structure of a holographic planar waveguide of the eyewear display.
Michael E. Miller (SID Member) Michael J. Murdoch
Eastman Kodak Co.
Abstract — Organic-light-emitting-diode (OLED) displays employing white-light-emitting OLEDs in combination with RGBW color filters can demand high peak currents to present images with bright, highly saturated colors. Image-processing methods that take advantage of a very highly efficient white subpixel in addition to filtered RGB subpixels to reduce the peak current and power of these displays are described. The image-quality impact of these algorithms are explored to develop a final image-processing algorithm.
FIGURE 2 — A pictorial example of the effect of the RGB-limiting algorithm with RGB values limited to 1.0, 0.67, 0.33, and 0.0 with a threshold of 0.75.
Taesung Kim (SID Member) Bo Ye Chung Phan Vu Nikhil Balram (SID Senior Member) Hugo Steemers
Marvell Semiconductor
Abstract — LCD motion blur is a well-known phenomenon, and several approaches have been developed to address it. This includes very-high-performance approaches based on motion-compensated frame-rate conversion (MC-FRC) and very-low-cost approaches based on impulsive driving. Impulsive-driving schemes are attractive because of their low cost, but suffer from two significant issues – loss of luminance and large-area flicker. A new impulsive-driving approach using motion-adaptive alternate gamma driving (MA-AGD), which removes motion blur and preserves the original luminance level without causing large-area flicker, is proposed.
While MC-FRC (motion-compensated frame-rate conversion) technology has been rapidly adopted as an ultimate solution for better motion performance, impulsive-driving technology, despite its simple and cheap implementation, has always been in the research mode due to its side effects. Flicker and luminance loss are two major drawbacks of impulsive-driving technologies, keeping them from being commercialized. In terms of luminance loss, there is a way to compensate the light loss by applying alternate gamma drive (AGD), which makes one frame darker to give the impulsive effect and another frame brighter to compensate for it by making the average brightness level even. However, flicker is almost impossible to eliminate completely because its generation principle conflicts with impulsive driving.
FIGURE 1 —The luminance variation of each driving scheme. (a) Native response of liquid-crystal molecules. (b) Boosted response by RTC.
Seongjong Yoo (SID Member) Yongjoo Song Jiwoon Jung Myunghee Lee
Samsung Electronics Co., Ltd.
Abstract — A 12-bit segmented R–C DAC to support a linear gamma curve has been proposed and fabricated in a 720-channel LCD source driver with a 16-V 1-poly 3-metal high-voltage CMOS process. The proposed DAC has a global resistor string and sample-and-hold buffers. A MSB voltage selected by the upper 6 bits of input data and a LSB voltage selected by the lower 6 bits of input data are summed by using a sample-and-hold operation with offset cancellation in the proposed DAC. The measured DNL was less than 0.3 LSB, and the output voltage deviation was less than 3 mV in all gray levels. Although two sample-and-hold buffers were adopted to operate alternatively, the die size was as small as 24.9 mm2, whichwas only an 8.3% increase compared to that of a conventional 8-bit 720-channel source driver. Because of its good performance with small area, the proposed DAC can be a good low-cost solution for a 10-bit TV system.
FIGURE 13 — Die photo with COF package.
TABLE 1 — Chip summary.
Fang-Cheng Lin (SID Student Member) Yi-Pai Huang (SID Member) Ching-Ming Wei (SID Student Member) Han-Ping D. Shieh (SID Fellow)
National Chiao Tung University
Abstract — Field-sequential color (FSC) is a potential technique for low-power liquid-crystal displays (LCDs). However, it still experiences a serious visual artifact, color break-up (CBU), which degrades image quality. Consequently, the "Stencil Field-Sequential-Color (Stencil-FSC)" method, which applies local color-backlight-dimming technology at a 240-Hz field rate to FSC-LCDs, is proposed. Using the Stencil-FSC method not only suppressed CBU efficiently but also enhanced the image contrast ratio by using low average power consumption. After backlight signal optimization, the Stencil-FSC method was demonstrated on a 32-in. FSC-LCD and effectively suppressed the CBU, which resulted in more than a 27,000:1 dynamic contrast ratio and less than 40-W average power consumption.
FIGURE 1 — (a) Target image_ Peter Pan (©Disney), each sub-frame image using the (b) conventional FSC-LCD and (c) Stencil-FSC method.
Michael E. Becker (SID Member)
Display-Metrology & Systems
Abstract — A method for evaluation of the contrast of moving step-grating patterns under smooth-pursuit eye-tracking conditions without imaging data acquisition and image analysis is introduced. Periodic optical responses of the display to a set of simple driving signals have been recorded at a fixed location, and the luminance vs. time data has been evaluated to obtain two types of contrast for characterization of the dynamical performance of the display under test: the frame-convoluted contrast and the frame-integrated contrast. The relation of this characterization with respect to modulation transfer functions from impulse responses and to thedynamic modulation transfer function from sine-gratings is explained and discussed. The approach described here provides a detailed and comprehensive characterization of the dynamical properties of electronic displays, including both extreme cases of step-responseand impulse-response with quantities that are related to visual perception. With this type of evaluation, the visual resolution of displays can be described by the same characteristics in the static and the dynamic case. The method is attractive due to limited instrumental efforts and the transparent method of evaluation.
Visual targets with initially sharp edges moving across certain electronic-display screens (e.g., LCDs and PDPs) are often perceived as blurred by human observers. This is caused by the hold-type characteristics of the temporal response of the display and by integration of the human visual system while smoothly following the movement of the target (i.e., smooth-pursuit eye tracking, SPET). Increased LCD response times, especially when switching between intermediate levels of gray, further deteriorate the visual quality of moving objects and thus contribute to motion blur, but they are not the actual cause.
FIGURE 1 — Intensity profile ( i.e., intensity vs. DUT pixel) of sine and step gratings shown in the inserts. The sine grating with wavelength λ is approximated by a series of steps. The step-grating has a minimum wavelength λ of two pixel (one pixel for each optical state).
Sylvain Tourancheau (SID Member) Kjell Brunnström (SID Member) Börje Andrén (SID Member) Patrick Le Callet
University of Nantes
Abstract — The primary goal of this study was to find a measurement method for motion blur which is easy to carry out and gives results that can be reproduced from one lab to another. This method should be able to also take into account methods for reduction of motion blur such as backlight flashing. Two methods have been compared. The first method uses a high-speed camera that permits us to directly picture the blurred-edge profile. The second one exploits the mathematical analysis of the motion-blur formation to construct the blurred-edge profile from the temporal step response. Measurement results and method proposals are given and discussed.
Four recent monitor displays have been tested in this work. They were all TFT AMLCDs with a refresh frequency of 60 Hz, with different types of panel, sizes, and resolutions as depicted in Table 1. In the following, they are identified with letters from A to D. Both C and D were using backlight flashing (BF). The response time given by the manu-facturers is also mentioned.
Wen Song (SID Student Member) Kees Teunissen (SID Member) Xiaohua Li (SID Member) Yuning Zhang (SID Student Member) Ingrid Heynderickx (SID Fellow)
Southeast University
Abstract — The scanning-backlight technique to improve the motion performance of LCDs is introduced. This technique, however, has some drawbacks such as double edges and color aberration, which may become visible in moving patterns. A method combining accurate measurements of temporal luminance transitions with the simulation of human-eye tracking and spatiotemporal integration is used to model the motion-induced profile of an edge moving on a scanning-backlight LCD-TV panel that exhibits the two drawbacks mentioned above. The model results are validated with a perception experiment including different refresh rates, and a high correspondence is found between the simulated apparent edge and the one that is perceived during actual motion. Apart from the motion-induced edge blur, the perception of a moving line or square-wave grating can also be predicted by the same method starting from the temporal impulse and frame-sequential response curves, respectively. Motion-induced image degradation is evaluated for both a scanning- and continuous-backlight mode based on three different characteristics: edge blur, line spreading, and modulation depth of square-wave grating. The results indicate that the scanning-backlight mode results in better motion performance.
FIGURE 4 — The ten stimuli used in the perception experiment.
Hiroki Kikuchi (SID Member) Shigeki Hashimoto, Shinichiro Tajiri Tsuneo Hayashi, Yutaka Sugawara Michio Oka, Yoshiyuki Akiyama Akira Nakamura Naoya Eguchi (SID Member)
Abstract — A high-pixel-rate high-contrast (30,000:1) wide-color-gamut grating-light-valve laser projector is reported. A new optical engine enabling high-frame-rate (240 Hz) scan projection is employed. Panoramic wide-angle-scan projection with a 64:9 aspect ratio was also developed. Speckle noise is eliminated using a simple but highly efficient technique. The optical throughput efficiency of the grating-light-valve laser projector is reviewed.
The main concept behind the speckle reduction is to introduce the averaging of independent speckle patterns within the human eye's temporal and spatial resolutions. The averaging of N independent speckle patterns suppresses the measured speckle contrast by a factor of 1/ÖN. To maximize the effect of speckle reduction, the projector was designed to introduce the following diversities in speckle patterns: (a) Polarization diversity. (b) Diversity by line scan. (c) Wavelength diversity. (d) Temporal variation diversity.
FIGURE 1 — The GLV laser projector demonstrated at 2005 World Exposition, held in Aichi, Japan. The screen size is 50 x 10 m. Twelve 5000-lm laser projectors were used.
Vikram Bhatia (SID Member) Steven J. Gregorski Dragan Pikula Satish C. Chaparala David A. S. Loeber (SID Member) Jacques Gollier (SID Member) Joan D. Gregorski Martin Hempstead Yukihiro Ozeki, Yoshiaki Hata Kazuhiro Shibatani Fumio Nagai, Nobuyoshi Mori Yukinobu Nakabayashi Naoki Mitsugi, Satoshi Nakano
Corning Incorporated
Abstract — Efficient and compact green lasers are keystone components for micro-projector applications in mobile devices. An architecture that consists of an infrared-producing DBR (distributed Bragg reflector) laser with a frequency-doubling crystal is used to synthesize a green laser that has high electrical-to-optical conversion efficiency and can be modulated at speeds required for scanner-based projectors. The design and performance of a green-laser package that uses adaptive optics to overcome the challenge of maintaining alignment between the waveguides of the DBR laser and the frequency-doubling crystal over temperature and lifetime is described. The adaptive optics technology that is employed uses the piezo-based smooth impact drive mechanism (SIDM) actuators that offer a very small step size and a range of travel adequate for the alignment operation. The laser is shown to be compact (0.7 cm3 in volume) and capable of a wall-plug efficiency approaching 10% (at 100-mW green power). It was demonstrated that the adaptive optics enables operation over a wide temperature range (10–60°C) and provides the capability for low-cost assembly of the device.
This green-laser module is a small-form-factor implementation of the frequency-doubled architecture using the SIDM-based adaptive optics. The optical components, including the DBR laser and SHG, are assembled on a ceramic base that provides electrical interconnections. The high-temperature co-fired ceramic material provides a low-cost base on which the optical components are integrated. The multi-layered ceramic offers the options for running traces between the layers for high-speed electrical interconnection.
FIGURE 4 —Illustration of the layout of components inside a green-laser package. The SHG is mechanically angled.
Thomas D. Pawlik (SID Member) Marina E. Kondakova (SID Member) David J. Giesen (SID Member) Joseph C. Deaton (SID Member) Denis Y. Kondakov (SID Member)
Abstract — Organic light-emitting diodes (OLEDs) were investigated by an electron para-magnetic resonance (EPR) technique that uses the effective device conductance as the detec-tion channel. This technique enables us to identify and study charge carriers and triplet excitons with high sensitivity. By using a series of model devices, it was demonstrated that this type of spectroscopy provides information regarding triplet energy transfer and the location of the recombination zone. The fundamental understanding about the extent of the recombinationzone in various OLED architectures helps in the design of devices with improved performance.
The chemical identity of the triplet is reflected in its characteristic EPR spectrum. This enables us to identify the location of the recombination zone in an OLED device. Furthermore, the triplet distribution among the constituents in a mixed layer and the effects of blocking layers can be studied. The fundamental understanding about the extent of the recombination zone in various OLED architectures helps us design the devices with improved performance.
FIGURE 1 — Experimental EDEPR spectrum (bottom trace) and calculatedtriplet spectrum (top trace) of an OLED device of the structure ITO | CFx | NPB (75 nm) | TBADN (40 nm) | ALQ (35 nm) | LiF | Al, measured at 15 K. The schematic energy level diagram of a triplet as a function of an applied magnetic field is shown on the right side. The ΔmS = 1 and ΔmS = 2 EPR transitions are indicated by arrows.
Y. S. Liao G. M. Yang Y. S. Hsu
National Taiwan University
Abstract — Scribing followed by a breaking method has often been applied to separate the individual liquid-crystal-display (LCD) glass panels from the mother cells. In some special occasions, especially for compact mobile devices, the bending strength of the glass substrates after breakage is the most important concern. In this study, the effect of the geometrical characteristics of the scribing wheel on the bending strength of LCD glass substrates was investigated. A volume-crack effective coefficient was defined to estimate the effect of different geometrical characteristics of the scribing wheels. It was found that there is a strong correlation between the bending strength and the coefficient before chipping appears as the scribing load is increased. Thus, the coefficient can be used for the assessment of the geometrical effect of scribing wheels on the bending strength of glass substrates.
The four-bar bending test as shown in Fig. 4 was used to estimate the effect of the geometrical characteristics of the scribing wheels on the bending strength of the active-matrix liquid-crystal display (AMLCD) glass substrates. The bending force can be obtained by directly measuring the bending stress during the four-bar bending test. Under the conditions for the same glass-sample size and same fixture, the relationship between the bending force and bending strength will be linear.
FIGURE 4 — Schematic illustration of a four-bar bending test device. All of the units are in mm.
Nobuyuki Matsuki Ryoichi Ishihara Kees Beenakker
Delft University of Technology
Abstract — Scanning spread resistance microscopy (SSRM) was used to investigate the electrical activity of coincidence-site lattice (CSL) boundaries in location-controlled silicon islands fabricated using the μ-Czochralski (μ-CZ) process. Using SSRM, the electrical activity of random and Σ3 and Σ9 CSL boundaries, which are determined by electron backscattered diffraction (EBSD) analysis, were observed. Quantitative evaluation of the microscopic current mapping by SSRM revealed that Σ3 and Σ9 CSL boundaries, of which most of them are coherent, have much less electrical activity than the random grain boundaries. Some of the Σ3 and Σ9 CLS boundaries seemed to be incoherent, while the number of such incoherent CSL boundaries are very much limited according to previous TEM investigation and showed increased activities; however, their activities are still lower than that of the random boundaries.
Crystal-line growth in the μ-CZ process is illustrated schematically in Fig. 1. A significant reduction in the number of random boundaries in the channel region resulted in attaining a field-effect mobility of 597 cm2/V-sec on average. The mobility decreased by approximately 10% when the coincidence-site-lattice (CSL) boundary was perpendicular to the channel. Although the influence of the CSL boundaries is much less than that of random boundaries, it is very important to investigate their electrical properties because they inhibit further improvement of TFT performance.
FIGURE 1 — Schematic description of the μ-CZ process. (a) Formation of the grain filter. (b) Deposition of amorphous silicon. (c) Excimer-laser annealing and crystallization.