The Vault

Cutting Edge 3D Display Technology Optical Hardware
White Paper / Jan 2018 / ar/vr

Check out this report detailing the State of Art (SoA) of optical hardware currently used in the field of 3D display technology, as well as emerging optical hardware technologies that feature Light Field (LF) and Holographic techniques.

We found some problems with the form:
To view this content please fill out the form below to tell us about you.
Privacy Policy
InterDigital.com Cutting Edge 3D Display Technology Optical Hardware Jukka-Tapani Mäkinen | 3/2017 VTT Technical Research Centre of Finland 3D Display and Rendering Technology will play a pivotal role in powering Augmented Reality, Mixed Reality, VR Devices and Experiences. The purpose of this report is to collect the State of Art (SoA) optical hardware currently used in the field of 3D display technology. The first part of the document presents some basic background information on the human visual perception and its effects to the 3D display technologies. The second part lists existing 3D display types and briefly explains the optical hardware used in them. The third part of the document is focused on emerging 3D display optical hardware technologies with main emphasis on Light Field (LF) and Holographic techniques. InterDigital | Goggleless AR & VR Project 2 About InterDigital lnterDigital, Inc., designs and develops advanced technologies that enable and enhance mobile communications and capabilities. Founded in 1972, our engineers have designed and developed a wide range of innovations used in mobile technology products and networks that include 2G, 3G, 4G and IEEE 802-related products and networks. For more than four decades, lnterDigital has been a pioneer in mobile technology and a key contributor to global wireless standards. Our team of more than 150 engineers-nearly 75 per cent of whom hold advanced degrees, including nearly 50 PhDs-has unparalleled expertise in major mobile connectivity and content delivery technologies. Since 2000, lnterDigital has spent close to $900 million on technology research and development. The company’s activities are organized around the concept of the Living Network: a future where intelligent networks self-optimize to deliver service tailored to the content, context and connectivity of the user, device or need. About Innovation Partners Innovation Partners is InterDigital’s technology sourcing initiative based on partnerships with leading technology firms, innovators and research organizations worldwide. Innovation Partners recognizes the importance of partnerships to capitalize on opportunities to expand the company’s research and development efforts in response to the evolution of the mobile technology industry. For more information, please visit www.innovation-partners.com. About VTT Technical Research Centre of Finland VTT Technical Research Centre of Finland Ltd is the leading research and technology company in the Nordic countries. The centre’s research and innovation services give their partners, both private and public, all over the world a competitive edge. VTT Technical Research Centre of Finland Ltd paves the way for the future by developing new smart technologies, profitable solutions and innovation services. For more information, please visit www.vttresearch.com. InterDigital | Goggleless AR & VR Project 3 Table of Contents 1. Introduction ............................................................................................................................................................................................... 5 1.1 Definition of a perfect 3D display ............................................................................................................................................................................ 5 1.2 Depth perception and depth ques .......................................................................................................................................................................... 6 1.3 Human visual perception ............................................................................................................................................................................................... 7 1.3.1 Vergence-accommodation conflict (VAC) .............................................................................................................................................. 8 1.3.2 Vestibular-visual conflict (motion sickness) ........................................................................................................................................... 9 1.4 Classification of 3D display solutions ..................................................................................................................................................................... 9 1.4.1 Differences between head mounted and goggleless 3D solutions ............................................................................................ 10 1.4.2 Single user vs. multiple users ........................................................................................................................................................................ 10 1.4.3 3D display types with different form-factors ......................................................................................................................................... 11 2. AR and VR systems with glasses ......................................................................................................................................................... 12 2.1 Head-mounted devices ................................................................................................................................................................................................ 12 2.1.1 Immersive VR systems ...................................................................................................................................................................................... 12 2.1.2 Smart glasses ......................................................................................................................................................................................................... 14 2.1.3 Retinal projection displays .............................................................................................................................................................................. 15 2.1.4 See-through displays ......................................................................................................................................................................................... 15 2.1.5 Advanced AR goggles: HoloLens and Magic Leap ............................................................................................................................ 17 2.1.6 Display based on sparse OLED matrix ..................................................................................................................................................... 18 2.1.7 Displays based on contact lenses ............................................................................................................................................................... 18 2.2 Multiplexed systems with single display ............................................................................................................................................................ 20 2.2.1 Spatially multiplexed systems ..................................................................................................................................................................... 20 2.2.2 Temporally multiplexed systems ................................................................................................................................................................. 21 3. Goggles-free 3D displays ..................................................................................................................................................................... 22 3.1 Parallax-based systems............................................................................................................................................................................................... 22 3.1.1 Parallax barriers ................................................................................................................................................................................................... 23 3.1.2 3D multiview displays based on moving parallax barriers ............................................................................................................. 23 3.1.3 Lenticular sheets ................................................................................................................................................................................................. 24 3.1.4 3D multiview displays made with multiple projectors ..................................................................................................................... 25 3.1.5 Moving lenticular and LC lenses ................................................................................................................................................................ 26 3.1.6 Directional backlight displays ...................................................................................................................................................................... 27 3.2 Volumetric displays ...................................................................................................................................................................................................... 28 3.2.1 Persistence of vision displays...................................................................................................................................................................... 28 3.2.2 Multi-planar systems ....................................................................................................................................................................................... 30 3.2.3 Pepper's Ghost and volumetric projection displays ....................................................................................................................... 30 3.2.4 Laser scanned volumetric displays .......................................................................................................................................................... 32 3.2.5 Some additional volumetric display types ............................................................................................................................................ 33 InterDigital | Goggleless AR & VR Project 4 4. Light Field and holographic display technologies ........................................................................................................................ 35 4.1 Light Field display technologies ............................................................................................................................................................................. 35 4.1.1 Definition of Light Field ................................................................................................................................................................................... 35 4.1.2 Difference between LF and regular 3D displays ................................................................................................................................ 36 4.1.3 Super Multi View displays .............................................................................................................................................................................. 36 4.1.4 Multiplexing of light fields .............................................................................................................................................................................. 37 4.1.5 Examples of existing light field 3D display technologies ............................................................................................................... 39 4.1.6 Companies developing LF systems ......................................................................................................................................................... 40 4.1.7 Technical challenges of LF displays .......................................................................................................................................................... 42 4.2 Holographic display technologies ........................................................................................................................................................................ 43 4.2.1 Definition of a hologram ................................................................................................................................................................................. 43 4.2.2 Technical challenges of holographic displays ..................................................................................................................................... 44 4.2.3 Static vs. dynamic holographic displays ................................................................................................................................................ 46 4.2.4 Example of an advanced CG holographic display ............................................................................................................................. 47 5. Discussion and summary ...................................................................................................................................................................... 48 5.1 Discussion ......................................................................................................................................................................................................................... 48 5.2 Comparison of depth ques and complexity of 3D display systems .................................................................................................... 49 5.3 Summary ........................................................................................................................................................................................................................... 50 Main references ............................................................................................................................................................................................ 51 Figure references .........................................................................................................................................................................................52 InterDigital | Goggleless AR & VR Project 5 1. Introduction The purpose of this report is to collect the State of Art (SoA) of optical hardware currently used in the field of 3D display technology. The first part of the document presents some basic background information on the human visual perception and its effects to the 3D display technologies. The visual system needs to be taken into account as a basis for any 3D display design in order to achieve a high quality viewing experience without side effects like headaches, nausea and motion sickness commonly associated with the currently used 3D display technologies. The second part lists existing 3D display types and briefly explains the optical hardware used in them. Currently available 3D display devices don’t have perfect image quality yet and there are multiple different, competing technical approaches that aim to improve the visual experience. The third part of the document is focused on emerging 3D display optical hardware technologies with the main emphasis on Light Field (LF) and Holographic techniques. The LF systems aim for creating high-resolution light fields both in spatial and angular domains in order to provide the users a much more natural viewing experience of 3D content. From the physical point-of-view Holographic dynamic displays would be the perfect optical solution for 3D viewing as they could provide the most natural scenes. Currently both LF and holographic 3D display systems require very complex designs with high-cost components, time-consuming calculations and huge bitrates. 1.1 Definition of a perfect 3D display A perfect three-dimensional (3D) display is commonly defined as a “window to the world”, through which viewers can look at a 3D scene as if the display screen were a transparent window to the real world outside. With good enough representation of the view, the viewer would not be able to tell the difference between artificial and real image. This definition may differ from the mental image many people have of a futuristic 3D display as the imagination has been affected by the numerous science fiction stories and visualizations (Figure 1). One well-known example, the Star Trek “holodeck”, is not a very realistic vision of near-future 3D displays as the pictured ultra-realistic, 360 degree Field-Of- View (FOV) for multiple users would require huge technical development and optical components not available in the foreseeable future. Another common example of a futuristic 3D display is the “Princess Leia” hologram shown in the Star Wars movies, which is unrealistic from the physical point of view as the image would be projected to thin air without any means to scatter the light towards the viewers. From these two science fiction examples, the projected hologram is actually closer to reality as there is already one optical hardware solution (Figure 40) available that is able to produce glowing dots of light inside a volume of air. Unfortunately this requires the generation of hot plasma with high-energy lasers that makes it a bit too dangerous technology outside the well-controlled laboratory environment. Figure 1. a) A realistic view of a current ”perfect” 3D display [P1] and b) a science fiction vision [P2] of futuristic 3D display that can be viewed from all sides. a) b) InterDigital | Goggleless AR & VR Project 6 Some of the currently used 3D display technologies actually challenge this “window” definition by presenting systems that can be viewed from more than one side, unlike a flat TV type of a display that can be viewed only from a specific angle from the front side of the device. For example the volumetric “crystal ball” type of a persistence-of-vision display shown in Figure 35 can be viewed from all angles around it and it may be hard to see it as a flat window. But whatever the case, a good 3D display device should be able to totally duplicate the entire visual sensation received by the viewer. A perfect 3D display should be able to offer to its viewers all necessary depth cues like motion parallax, binocular disparity, etc., commonly associated with human 3D visual perception [1]. 1.2 Depth perception and depth cues Depth perception can be defined as the visual ability to perceive the world in three dimensions. Depth perception arises from a variety of depth cues, which can be either physical or psychological (Figure 2) [1]. Physical depth cues can be introduced only by true objects or realistic 3D images, but psychological cues can be reproduced also with 2D images. There are big differences on how the different 3D display solutions are able to provide the depth cues. One of the most well-known examples of a not-so-perfect 3D display technology is the stereoscopic eyeglasses used in 3D cinema. For some people these movies cause a lot of eye fatigue or even nausea, which are connected to the fact that eyes focus to the surface of the screen and not to the 3D object, whereas the two eyes converge to the virtual object at another distance. Figure 2. Major physical and psychological depth ques used in human 3D perception. InterDigital | Goggleless AR & VR Project 7 Depth vision cues are processed in the brain and combined together to improve accuracy. Depth cue accuracy is highly dependent on viewing distance. Figure 3 shows the relation between depth cue accuracy and distance from the observer at three different spaces [2] defined as: 1) personal space (0 – 2 m), where humans perform most of their close up interactions and depth perception is highly refined, 2) action space (2 – 30 m), where users may interact reasonably accurately with other objects (e.g. throw a ball), but with less cues and accuracy and 3), vista space (>30m), where objects appear flat and distance estimations become poor. Figure 3. Accuracy of different depth ques in relation to distance from the observer. 1.3 Human visual perception Perception of a 3D scene is created subconsciously in the brain with an analysis of the various features of 3D objects and with sensory feedback from the visual system. All visual cues contribute to this dynamic and adaptive sensing process. The depth cues are emphasized differently at different viewing distances (Figure 3). In general, the effects of major physical depth cues decrease with the increasing distance from the objects, but the effects of psychological depth cues remain the same. Figure 4. A stereographic photograph from the year 1901 [P4]. InterDigital | Goggleless AR & VR Project 8 The most well-known technique for presenting three-dimensional images is stereoscopy. In this method two two- dimensional (2D) images are displayed separately to the left and right eye (Figure 4). Perception of depth is created when these images taken from slightly different viewing angles are combined in the brain. It is important to note that the presentation of two 2D images is perceptually not the same thing as displaying an image in full 3D. The biggest difference is the fact that head and eye movements will not give more information about the objects being displayed – the 2D images are able to present only the same two slightly different viewpoints. These types of systems are commonly called 3D displays, although stereoscopic displays would be the more accurate term. Most stereoscopic displays do not qualify as real 3D displays, but all real 3D displays are also stereoscopic, because they are able to present the image pairs to the two eyes of the viewer. It is quite difficult for any 3D display to provide all the physical and psychological depth cues at the same time. For example, the volumetric 3D display techniques presented in paragraph 3.2 may not be able to provide occlusion, shading or texture due to the transparency of displayed voxels and some technologies, such as the stereoscopic display, may even provide conflicting depth cues making the displays fall short of ideal. 1.3.1 Vergence-accommodation conflict (VAC) Current stereoscopic displays are an unnatural way of making 3D scenes. There is a neural connection between light sensitive cells on the retinas and the cells sensing eye muscle movement. The associated areas in the brain work together in order to create perception of depth. Autostereoscopic 3D displays lack the correct focus cues as the image information is limited to the plane of the display as illustrated in Figure 5. When the eyes focus to a different point than where they converge, physiological signals in the brain get mixed up. Depth cue mismatch of convergence and accommodation leads to e.g. eye strain, fatigue, nausea and slower eye accommodation to object distance. Tolerance for the comfortable depth mismatch ranges ~20-30% of distance between viewer and display. This phenomenon also requires a non-proportional depth squeezing in artificial images and allows only short-time viewing of scenes with large depth. Tolerance range for convergence mismatch is called the Panum’s fusion area and it is in the range of 15-30 arc min (0.25- 0.5 degrees) [3]. Larger convergence errors cause a breakdown in binocular fusion and perception of depth is disrupted. Retinal blur [4] visual cue caused by the limited depth of focus is also missing from the image of current 3D displays. This shortcoming is visualized at the bottom part of Figure 5a. Vergence-accommodation conflict is one of the main drivers for moving from the current stereoscopic 3D displays to the more advanced light field systems! Figure 5. Illustration of the different focal distances when looking a real world object and autostereoscopic 3D display. a) With natural scenes the eyes converge to the same point where they focus, but with 3D displays the focus plane is on the display surface and b) all objects are in focus [P3]. a) b) InterDigital | Goggleless AR & VR Project 9 1.3.2 Vestibular-visual conflict (motion sickness) The primary role of the vestibular system is to provide the brain with information to regulate posture and to coordinate eye and head movements. It consists of several neural pathways and sensory organs that integrate to form a complete mental picture (Figure 6). The vestibular system is responsible for: detecting linear and angular head movement and head position in space, assisting gaze stabilization of the visual field, maintaining balance and postural control, providing spatial orientation or perception of body movement. Motion sickness is caused by the conflict between visual and vestibular signals. This is the most disturbing side-effect of VR systems that totally block the view of surroundings and replaces it with a virtual world that doesn’t perfectly follow the head and eye movements of the user. Figure 6. Integration of neural pathways to one vestibular system [P5]. 1.4 Classification of 3D display solutions Figure 7 shows the taxonomy of different 3D display solutions [5]. This order is used as a basis for the classification of different 3D display technologies in the next chapters of this document. Each chapter describes the basic optical functioning principle or technology of the particular type of display and shows practical examples of devices that have been made either as products or prototypes. Figure 7. Taxonomy of 3D displays. InterDigital | Goggleless AR & VR Project 10 1.4.1 Differences between head mounted and goggleless 3D solutions The first division of 3D display solutions is made between systems utilizing glasses or goggles and systems that can be used without them. In both of these classes there are technologies that allow multiple users and technologies that work only for a single user. Goggleless 3D solutions can be technically much more challenging than systems with some kind of headgear. All visual information that a person can use enter the human perception system through the eye pupils. Head mounted devices have the great advantage of being very close to the eyes and they can cover the large Field-Of-View (FOV) with much more compact optical constructions than what is possible with any goggleless displays. They can also be vastly more efficient in producing the needed amount of light as the “target area” is well defined in a relatively fixed position and they can also be much more efficient in handling of the image data as the single user will not need more than one viewpoint to the 3D scene. The goggleless displays will need to be physically large if one wants to cover a significant portion of the viewer’s FOV and the systems become much more expensive to make than goggles. As the user position is not fixed to the display device, the light emitted needs to be spread over a large angular space in order to make the picture visible from multiple positions, which easily leads to a situation where most of the light is wasted. Furthermore, as the user without goggles is free to change the position around the 3D display, it will need to provide several different “views” of the same 3D scenery, which multiplies the amount of image information needed to be displayed with the number of views that one wants to create around the central position. One approach to ease the burden of heavy data handling with goggleless displays is to use specialized eye tracking systems in order to determine the position and line of sight of the user(s). In this case the image or the “light field” can be directed straight towards the pupils and not spread out to the whole surrounding space. By knowing the exact position of the eyes the “viewing window” size can be reduced enormously. This technique of course comes with the price of needing the eye tracking and projection systems that require their own hardware and processing power and also limit the number of possible viewers. 1.4.2 Single user vs. multiple users Smart glasses and virtual reality goggles are by their nature personal, single user devices. However, these can also be made for multiple users by creating a system where the same virtual/augmented digital environment is shared by several people all wearing their own headgear showing each individual their own point-of-view of the 3D scene. The synchronization between different user views and positions is done by using a set of positional and rotational sensors, either in the surroundings or in the head mounted device itself, that monitor the head movements of the participants. This way the experience can be shared and users interact inside the same virtual space (Figure 8). Number of needed viewing directions is dependent on the number of actual viewers and the image processing calculations can be done for each display (2 per user) separately. Figure 8. A shared virtual experience with multiple users wearing goggles [P6]. InterDigital | Goggleless AR & VR Project 11 1.4.3 3D display types with different form-factors The different 3D display types can also be classified on the basis of their form-factors as shown in Figure 10. The head- mounted devices naturally occupy less space than all the goggleless solutions. This also means that the devices can be made with smaller components and less material, which has an effect to the manufacturing cost. Volumetric displays take space from all three spatial directions and require a lot of physical material making these systems heavy, expensive to manufacture and more difficult to transport to different locations. Due to the heavy use of materials, the volumetric displays tend to have fairly small “windows” and limited FOV. Projection-based 3D displays typically have one large but flat component which is the screen and a system that projects the image(s) over free space from a distance. These systems can be made more compact for transportation and they also cover much larger FOVs than e.g. volumetric displays. Flat 3D displays may require a lot of space in the two spatial directions, but as the third direction is only virtual, they are fairly easy to transport and assemble to different environments. This last category is likely the most desirable goggleless 3D display form-factor for regular consumers as they are the easiest to assemble at home. Goggleless systems which are intended for a single user can generate the 3D image with a minimum of two different views – one for each eye. If the system is intended for multiple users, it will need to provide multiple, different viewing directions (Figure 9) and at least two images to each direction for the stereoscopic effect. In addition, if the system aims for realistic presentation, each viewing direction should also have image content specifically created for that particular point-of-view from the same 3D scene. This requirement for multiple directions and multiple image content will place hard requirements not only on the graphical processing and data transfer speeds, but also on the optical hardware that needs to handle the multiplexing task with a single display and by taking into account the “refresh rate” of the human visual system. Limitations of currently available optical hardware can be seen in the complex definitions and limitations on e.g. “optimal viewing distance” or “transition zones” seen in the currently available multi-view glasses-free flat 3D displays. The regular autostereoscopic 3D televisions and cinema have avoided the complexity needed in multi-view displays by making the individual users wear special glasses that work together with the display. a) b) Figure 9. a) Creation of multiple views from a single display [P7] and b) viewing directions [P8]. Figure 10. Four different types of 3D displays based on form factor [P9][P10][P11][P12]. InterDigital | Goggleless AR & VR Project 12 2. AR and VR systems with glasses The glasses-based systems can be divided into two main categories. In the first category, the whole system is built into a device that is mounted to the viewer’s head. The devices are generally called Head Mounted Displays (HUD), VR/AR glasses/goggles (Figure 12) or smart glasses (Figure 13) and they are based on microdisplays or projectors that scan the image directly to the eye retina. They are built for single user only. The second category consists of systems that have two optical parts - glasses that are mounted in front of the eyes and a display that is further away from the viewer (Figure 24). In this case there can also be several viewers at the same time as the glasses enable multiplexing of different display images to two eyes of each user, creating the 3D effect for all simultaneously. 2.1 Head-mounted devices The HMDs can also be divided into two distinctive categories: immersive and see through devices. The immersive goggles block the direct view of the surrounding world and replace it with an artificially created virtual view. For this reason they are called virtual reality (VR) glasses. Augmented Reality (AR) glasses use partially transparent optical components in front of the eye(s) making it possible to augment the natural view with digital content by superimposing images and graphics on top of the natural FOV. VR glasses typically cover both eyes, but AR systems can be made for just one eye only. If the head mounted device is made very small and integrated to a frame that resembles eye- or sunglasses, the devices are usually called smart glasses and not goggles. 2.1.1 Immersive VR systems Immersive VR systems have two separate displays (or one display physically divided into two separate sections) placed in front of each eye in order to create the stereoscopic effect. The user typically wears goggles that have small LCD (Liquid Crystal Display) or organic light-emitting device (OLED) displays with magnifying lenses in front of each eye. Freeform optical designs can somewhat improve the performance and reduce size, but the basic optical principles are the same for all VR systems. The optical designs tend to be similar as well due to the fact that the human eye is used as the basis for the design specifications and as the eyes of different individuals are basically the same in size and focal length, the optical designs converge to similar solutions. HMDs may also be coupled with head-tracking devices, allowing the user to see different views of the virtual objects by head movements. Performing this update quickly enough to avoid motion sickness requires very fast image processing and the display refresh rate is one of the most important parameters to optimize in a high-quality VR system. Figure 11. Illustration of the optical principle used in making of the large virtual image in VR goggles. InterDigital | Goggleless AR & VR Project 13 The functioning principle of the optics in VR goggles is very simple, as illustrated in Figure 11. A magnifying projection lens is positioned between the eye and display. The display size is magnified and the virtual image transformed from a very close range to a distance dependent on the power of the projection lens. The main reason to do this is the fact that normal eyes can’t focus to objects at very close range (<10cm) and the eye muscles are more relaxed when the focus accommodation is to a longer distance. VR glasses block the view of the outside world and there is no need for the eyes to accommodate to different distances as the images will always be focused to the virtual display surface. This means that there is no need other than eye relaxation to make the VR virtual image appear further away than ~1m. However, AR glasses allow the user to see real world objects at different distances making the eyes change their focus and careful consideration is needed as to the distance the virtual image is positioned in order to avoid unnecessary eye straining focus adjustments. One of the first big players in the VR market was Oculus Rift. Figure 12a shows the most recent hardware and optics of the head mounted device. A pair of magnifying lenses is used in making the virtual image (110 deg FOV) from a single high-resolution (2160 x 1200 pix) OLED display that is manufactured by Samsung. The device costs ~$600 and it is connected to a PC with a cable. The system uses positional tracking with IR-leds and external desktop tracking hardware. There are some indications that sales of the device have flatlined since the product’s release in March 2016 as the early adopter market dries up and interest from regular consumers has been much lower than expected. Figure 12. a) Oculus Rift [P13] and b) Google cardboard [P14] VR goggles. a) b) Other well-known companies making VR headsets are Sony (PlayStation VR), HTC (Vive) and Samsung. Samsung Gear VR has a different approach to the hardware as it uses a mobile phone as the display device and the head mounted piece is basically just a mount for holding the magnifying lenses and mobile phone display at fixed positions. A similar, but very low cost (~15$) product is the Google cardboard (Figure 12b), which shows that the requirements for the optical and optomechanical features in VR headgear can be quite easy to achieve with simple hardware. InterDigital | Goggleless AR & VR Project 14 2.1.2 Smart glasses Probably the most well-known example of a smart glass device is the Google Glass (Figure 13), which was first introduced as a consumer product. However, due to the lower than expected interest in the market, the business focus was later changed to professional users that can utilize the device in logistics centers or warehouses. Those smart glasses that have a see-through display element can be also classified as AR devices as the images are superimposed on top of the real world view. Some of these devices have optics that distort the real image heavily and the augmented experience quality is not good. In those cases the see-through feature is mainly for avoiding occlusion of a large part of the field of view, which can be hazardous for users that move a lot with the device. Figure 13. Google Glass smart glasses [P15] and patent picture related to the optics [P16]. All of the smart glasses are based on very small displays. The microdisplay type can be Liquid Crystal Display (LCD), Organic Light Emitting Diode (OLED) or Liquid Crystal on Silicon (LCoS). LCoS is currently the most common type and it is used together with a so-called “light engine” that contains illumination LEDs and polarizing beam splitter cubes. OLEDs don’t need a separate light engine as the display itself generates the light. For this reason, the designs based on OLEDs can be made smaller and lighter, and they will likely become the most common display components in smart glasses when the brightness levels become adequate for outside use. Figure 14. Optoelectronic components of the Vuzix M100 and the imaging prism InterDigital | Goggleless AR & VR Project 15 The image from the microdisplays is projected to the eye with a similar magnifying-lens principle to what is used in the VR headsets (Figure 11). However, in order to make the designs smaller for the glass frames, smaller displays and more complex optics are used. As an example, Figure 14 shows the optoelectronic components of the Vuzix M100 device. The LCD display used (and the whole WhitePearl display module) is made by Kopin, which is one of the large microdisplay manufacturers. The magnifying piece has three optical surfaces that act together as an imaging prism that also folds the optical path from the display to the eye in order to save space. This injection molded plastic element has integrated features for mounting and represents the desired minimalistic design approach. 2.1.3 Retinal projection displays MicroVision is the creator of PicoP® scanning technology, an ultra-miniature laser projection and imaging solution based on laser beam scanning. The company has previously worked on “retinal display” technology for military and industrial applications. All of the HMDs use retinal projection in some form, but this approach offers an alternative to the different miniature displays. In this technology the laser beams are scanned to the retina directly with a Micro-Electro-Mechanical- System (MEMS) mirror. The scanning principle has potential for enabling very small miniaturized designs (Figure 15a), but the laser packaging technology counteracts this benefit by making the systems quite complex and expensive to manufacture (Figure 15b). a) b) Figure 15. Patent images of MicroVision’s a) HMD design [P17] and b) retinal projection system [P18]. 2.1.4 See-through displays Head-mounted or wearable glasses may be used to view a see-through image imposed upon the real-world view, creating augmented reality (AR) experience. This is done by reflecting or projecting the digital images through partially reflective mirrors or lightguides positioned in front of the eye. The real-world view is seen through the partially transmitting surfaces and the display or projection system is positioned outside the FOV. InterDigital | Goggleless AR & VR Project 16 Figure 16 shows the optical structure of Lumus AR glasses. The image of the microdisplay is projected into the see- through lightguide element with the aid of an in-coupling structure that can be a mirror, prism, surface microstructure or a holographic grating. The lightguide transfers the image in front of the eye and an out-coupling structure projects it to the eye pupil. The partially transparent out-coupling structure can be a semi-transparent mirror or a diffractive/ holographic grating. Field-of-view of these devices is usually limited by the lightguide material and the type of out-coupling structure. The best systems with the largest FOV utilize higher refractive index materials and holographic gratings that are more expensive to manufacture than lower cost alternatives. Figure 16. Patent images of Lumus AR optical and optoelectronic components [P19]. Figure 17. Patented diffractive lightguide structures by a) TruLife Optics [P20] and b) Dispelix [P21]. TruLife Optics is a UK-based company selling the technology developed by Colour Holographic Ltd and National Physical Laboratory (UK). It claims to be the world’s leading optical component developer for wearable augmented reality devices. They provide holographic optical technology for head-up display and augmented reality systems. The flat lightguides (10 cm x 3 cm) have holographic in- and out-coupling structures that are created with RGB lasers on special material. Dispelix is a Finnish company offering a novel diffractive grating technology developed at VTT Technical Research Centre of Finland. The cost to manufacture the structures is relatively low and they are able to offer good quality out-coupling of colored images, which is a property difficult to achieve with most diffractive structures that naturally disperse the different colors. Example patent images from these companies are shown in Figure 17. a) b) InterDigital | Goggleless AR & VR Project 17 The big benefit of using a thin see-through element is the fact that the real world view can be seen without large distortions. The thicker light manipulating elements used in Google Glass distort the image and effectively block the user’s view from some part of the FOV. This makes it impractical to extend the virtual display size and for this reason the thin elements could be used for much larger FOV displays. A lot of criticism has also been placed on the look and intrusive nature of the more traditional HMD devices with cameras and a thin see-through element can be a convenient way to make the AR display structure more discreet. These facts make the see-through elements very desirable components for many product manufacturers and there are several patents utilizing this approach in one way or another as shown in Figure 18. Figure 18. Patent pictures with smart glass see-through element designs by a) Samsung [P22], b) Amazon [P23] and c) Google [P24]. 2.1.5 Advanced AR goggles: HoloLens and Magic Leap One of the most advanced, currently available, head mounted AR devices is the Microsoft HoloLens (Figure 19a). It is a stand-alone AR headset that was unveiled at a Windows 10 event at the beginning of 2015. Development Edition of the device ($3000) was launched on 30 March, 2016 and it uses “holographic lenses”, one for each eye, which are transparent lightguide pieces that have three layers of glass (RGB). The lightguides are most likely quite high in cost. FOV of the device is ~35 degrees and it has a depth camera and integrated speakers. There is some indication that the HoloLens device might be using “Light Field technology” (see chapter 4.1), but this might also be just a naming issue as many kinds of systems use this term without actually being real light field (LF) systems. According to Microsoft, the best holograms will appear to be about two meters away from the user, because the two virtual displays are fixed at that optical distance. According to some test user reviews, the strain of trying to focus each eye becomes too great, and discomfort sets in if the “hologram” moves too close or too far away. At an apparent distance of one meter, virtual objects rendered by the HoloLens will begin to fade out, and they’ll disappear as they close to 0.85 meters away. The fact that focal distance of the images is very limited points to a more traditional autostereoscopic system rather than to a true LF system. Figure 19. a) Visualization [P25] of Microsoft Hololens use case and b) patent images associated to the Magic Leap technology [P26][P27]. a) b) c) a) b) InterDigital | Goggleless AR & VR Project 18 Magic Leap is a company that is aiming for true AR light field goggles, but it has started to use the term “Mixed Reality” (MR) instead of AR. The first prototype device was made in 2011. The current device is described as: “a chunky pair of sunglasses wired to a square pack that fits in a pocket”. Figure 19b shows some patent images associated with the Magic Leap technology. According to the available information, the device will possibly be based on an array of very small projectors and a so called “photonic light field chip”, which probably means a lightguide element. It is likely that the diode laser array acting as a light engine is positioned in the outside package, which is connected to the headset with a fiber bundle. The head mounted device probably contains an array of fibers that are vibrated with piezoelectric actuators. The illuminating laser light is pulsed and as the end of a single fiber rotates through a spiral-shaped path a complete 2D picture is projected from each fiber separately [6]. An array of these projectors would be able to create the array of different views needed in a LF system. Unfortunately, single-mode fibers and lasers require highly accurate optoelectronic packaging, which is likely to lead to a high cost device. However, the use of extremely small laser projector arrays would enable a true light field design as several separate image sources could be integrated to a very small package in the goggles. 2.1.6 Display based on sparse OLED matrix LusoVU is a Portuguese company developing smart eyewear. The company has released a description of a concept design (Figure 20) for AR glasses that use a sparse grid of OLED pixels integrated to eyeglasses. The light emitting pixels are positioned on the inside-surface of the lenses and light is emitted to the viewing direction instead of the eyes directly. A holographic combiner element is used in front of the lens for reflecting the light back to the eyes. As the OLED pixels have a lot of space in between them, the display layer is also partially transparent and allows AR use. A monochromatic proof-of-concept demonstrator device has been presented, but there is no information on the possible product release date. The company’s current product is called Eyespeak, which are stand-alone communication glasses that read eye gestures and use them for typing with a virtual keyboard. Figure 20. AR glass product concept developed by LusoVU 2.1.7 Displays based on contact lenses Innovega is a company working on HMDs based on the use of contact lenses. The iOptik lenses (Figure 21) have polarizing filters and a secondary magnifying lens, which focuses the light of a nearby microdisplay to the eye. Light rays are transferred independently to the retina, allowing users to focus both far away and close range in full field of view. In other words, the special contact lenses make it possible to view images on a very close range HMD display while at the same time viewing distant objects. The main benefit of this approach is the large FOV that can be achieved by positioning the optics directly on top of the eye, coupled with the slightly smaller size of the required HMD optics. The contacts do away with the need for the “bulky” optics of HMDs – but this comes with the cost of needing to wear contact lenses. InterDigital | Goggleless AR & VR Project 19 The next step towards nearly invisible wearable virtual displays would be to integrate the light sources directly to contact lenses. First trials with LEDs embedded into flexible and curved contact lenses have already been made. For AR display, the light emitting element (or display matrix) needs to be partially transparent. In addition to light emitting elements, the display needs some optical structures like Fresnel lenses or microlenses that collect and direct the light into the eye pupil. Without the additional optics, the eye lens can’t focus the image to the retina. An example of such a microlens structure can be seen in the patent filed by Samsung in 2014 (Figure 22). If the contact lens display is used for showing 3D images, two lenses, one for each eye, are needed and their image content will need to be synchronized. Figure 21. Patent images of a contact lens as a part of an AR system developed by Innovega [P28]. Figure 22. Samsung contact lens display patent pictures [P29]. The main challenges of this technology approach are miniaturization of the display elements and its control electronics as well as powering of the display. Currently the powering can be done with inductive coils built into glasses that feed electric energy to the coils in the contact lenses. The biggest hurdle for the adaptation of this technology is likely to be the fact that the lens needs to be inserted into the eye. University of Washington has made some demonstrators that integrate conductors to the flexible lenses and have also inserted these into the eye of a rabbit [7]. Contact lens displays can be seen as possible future technology, but the products that would have only contact lenses as displays are not likely to emerge in the near future. InterDigital | Goggleless AR & VR Project 20 2.2 Multiplexed systems with single display Multiplexed systems based on glasses or goggles use a single display in the creation of the stereoscopic image. The division of images can be made either by spatially separating the projected light beams to the two eyes of a single user or by temporally alternating the images directed to the left and right eye. In both cases, the headgear acts as a filter or shutter that blocks the other view from being visible to the wrong eye. These systems are easy to make for multiple users simply by making multiple glasses. 2.2.1 Spatially multiplexed systems The so-called anaglyph technology utilizes color filters for blocking the left and right eye views. It was used in the first multi-user 3D cinema systems as early as 1915 and is still in use today due to the simplistic and low cost hardware needed. Images directed to the two eyes are filtered with near complementary colors (red and green, red and cyan, or green and magenta) and the observer wears respective glasses with color filters (Figure 23). The human eye is sensitive to three primary colors and with clever selection of the color channels, the brain can be tricked to see a full color image as the two wrong-color images are combined. Negative features of the technology include color rivalry, reduction in image brightness and some unpleasant after-effects that have led to the development of alternative 3D cinema methods. Figure 23. A 3D-picture [P30] made with the anaglyph method and 3D-glasses [P31] with color filters. Stereoscopic display techniques based on light polarization are very well suited for video projection. Projector devices utilizing Liquid Crystal (LC) displays as the Spatial Light Modulators (SLM) emit polarized light by default. By using two projectors with orthogonal polarization orientation and polarized glasses with respective polarization directions, it is possible to filter out the two different images to each eye. In this use, the projection screens must preserve polarization well enough in order to avoid crosstalk. The best screens use aluminized surfaces for reflection or translucent acrylic sheets for back-illuminated systems. The latter option can be used for a rear-projection TV device, on which the front sheet consists of micropolarizers that do the actual polarization filtering on the pixel-by-pixel level. The big benefit of this technology is the possibility to use simple and low cost passive glasses that can be manufactured the same way as polarized sunglasses. InterDigital | Goggleless AR & VR Project 21 2.2.2 Temporally multiplexed systems The term “persistence of vision” (POV) refers to the optical illusion that occurs when visual perception of an object does not cease for some time after the light source has moved and image shifted away from the particular retinal position in the eye. This “memory effect” is exploited in cinema and video displays that have refresh rates faster than what the eye can see. It can also be used to advantage in the creation of time-multiplexed 3D displays. In these systems the left and right eye views are shown with a single display in rapid alternation and active shutters positioned in front of each eye open and close in synchronization with the display signal. The shutter glasses are based on LC technology and can be controlled with an infrared link. This is the most common technology used today in home 3D television systems (Figure 24). A big benefit of this technology is the possibility of using the same display device for both 2D and 3D images without any penalty to the resolution. The need to use active shutter glasses that are fairly heavy and require batteries can be seen as the negative side. Figure 24. Active glasses with LCD shutters [P32] intended to be used with a 3D television set [P12]. InterDigital | Goggleless AR & VR Project 22 3. Goggles-free 3D displays 3D displays that can be used without goggles can be divided into three main categories: 1) parallax-based systems that utilize 2D displays, 2) volumetric displays that use some optical method to generate and direct light rays inside a predetermined volume in space and 3) Light Field (LF) and holographic systems that aim to reconstruct the natural wavefronts with 2D elements. The first two categories contain solutions that are generally available today and they are explained in this document chapter. The third category contains solutions that can be considered as emerging technologies and there are not that many commercial players developing these systems yet in their purest form. There are some existing systems and products that claim to be based on “light field technology” or name their images as “holograms”, although the technology used can’t be considered to fall under these categories. In order to clarify the situation, the LF and holographic techniques are discussed in the next Chapter 4. 3.1 Parallax-based systems In autostereoscopic displays, the general idea is to divide a picture generated by a 2D display to at least two different images with two different views that are projected to the two eyes of a single user. Two of the most common types of optical structures used in these goggleless displays are parallax barriers that block the light and lenticular lens sheets that refract the light to different directions. The two options are illustrated in Figure 25a and a picture of an example lenticular sheet is shown in Figure 25b. Multiview 3D displays show more than two unique views that are usually spread only along the horizontal direction. The key issue is to properly separate the stereo image pair displayed on the same screen to deliver the left image to the left eye and the right image to the right eye for each user. Figure 25. a) Two of the most common types of optical structures used in glasses-free parallax-based 3D displays [P33] and b) lenticular lens sheet [P34]. a) b) InterDigital | Goggleless AR & VR Project 23 3.1.1 Parallax barriers Parallax barriers offer the simplest solution for dividing the picture of an autostereoscopic display. A light absorbing barrier, which can be a linear grating or a matrix of pinholes, is placed between two pixels occluding the left pixel from the left eye and right pixel from the right eye. One advantage of this approach is that 2D display content can be shown with the same display simply by activating both left and right pixels at the same time or by switching the barrier off (in the case of an active barrier). Unfortunately the parallax-barrier-based displays have several drawbacks that make them less than ideal solutions for consumer use. Image brightness is reduced as part of the light emitted from the pixels is absorbed by the barrier material. Spatial resolution is reduced as display pixels are used for the generation of multiple views. Combination of the pixel matrix and parallax barrier grating creates easily vertical dark lines (picket fence effect) to the image, which also limits the number of possible viewing zones and view angles. The image will also be flipped when the viewer’s eyes move from one viewing zone to another as the eyes see the images intended to the opposite side. If the barrier grating used has very narrow slits, some diffraction effects also appear, limiting the achievable resolution. Figure 26. BOE patent images of active parallax barrier technology [P35]. BOE Technology Group is the world’s leading supplier of display technologies. They have developed glasses-free 3D displays made with “Active Barrier technology”. This term refers to a type of parallax barrier system, where two barrier sheets can be actively aligned in order to allow free switching between 2D and 3D television modes as shown in their patent visualization on Figure 26. One example product is the Changhong Glasses-Free 65” UHD 3D TV that has a BOE Technology Group panel inside. The 2D mode has a full Ultra-High-Density (UHD) resolution (3840 x 2160 pixels), while 3D mode has regular HD resolution (1920 x 1080 pixels). The company has already presented some extremely high resolution 10K (10240 x 4320 pixels) displays in the size of 82”. Motivation for developing such high pixel count TV sets can partially come from the higher resolution needs of 3D displays. 3.1.2 3D multiview displays based on moving parallax barriers The field of view of an autostereoscopic display may be extended with reduced side effects by making some clever designs that utilize moving parallax barriers. This method is actually combining the benefits of spatial and temporal multiplexing in order to achieve a higher quality 3D display. One example of these solutions is shown in Figure 27a, where a parallax barrier plate is scanned across a display surface. By synchronizing the light emissions to the movement of the plate it is possible to create several different views in front of the display without the picket fence effect. InterDigital | Goggleless AR & VR Project 24 Figure 27b shows another example of a multiview display covering a 360° viewing area. This device is created with a rotating cylindrical parallax barrier and LED arrays. The physically limited number of LED arrays is compensated by rotating the array, while the multiview in the horizontal direction is created by the rotating parallax barriers. In each viewing direction, an appropriate view of a 3D image can be seen corresponding to that particular direction as the LED light emissions are synchronized to the 3D content and cylinder rotations. Key design parameters for such a system include the barrier interval, the aperture width, the width of the LEDs, and the distance between the LEDs and the barriers. This “outward viewing” design is for viewers surrounding the outside of the display device, but a similar design for “inward viewing” is also possible by locating the viewer inside a large rotating cylinder. 3.1.3 Lenticular sheets A lenticular lens sheet consists of a linear array of (usually) plano–convex cylindrical lenses (Figure 25b). The lenses project an image of the pixel row lying below the lens to different directions, making only one part of the row visible from one viewing direction. As a transparent component, the lenticular sheets have much better optical efficiency than the absorbing parallax barriers. One display may also have several lenticular sheets and other optical structures like prisms, Fresnel lenses and diffractive grooves stacked on top of each other that modify the multiview image fields. In addition to the high efficiency, the lenticular lens 3D displays have the benefit of using the same display components as what are used for normal 2D displays making them relatively simple to manufacture and low in cost. The main technical challenge in manufacturing is the highly accurate alignment required between the lenticular sheets and display pixel rows. a) b) Figure 27. Patents of 3D display structures based on a) scanning aperture plate [P36] and b) rotating parallax barrier cylinder [P37]. Figure 28. Patent images of a) lenticular lens sheet used for creation of multiple views [P38] and b) slanted lenticular array [P39]. a) b) InterDigital | Goggleless AR & VR Project 25 In a multiview 3D display based on lenticular sheets and spatial multiplexing, the resolution of the display screen is split among the multiple views. Figure 28a shows two patent images of such a display that has four pixels under one cylindrical lens generating four different viewing directions. When a viewer looks at the display from the right distance, two of the views cover the two eyes creating a stereoscopic image. If the viewer moves his or her head, the view zones change to the adjacent ones in both eyes simultaneously. As the stereo image is maintained and the overall viewpoint changes, perception of the three dimensions in the image is intensified. This enhanced 3D effect makes multiview displays more desirable for one user than the simpler displays relying on just two views. In the first design shown in Figure 28a, the cylindrical lenses are aligned with the vertical pixel columns on the 2D screen. One method to increase the resolution in the horizontal direction is to use slanted structures, as shown in Figure 28b, where the direction of the lenticular sheet is slightly tilted in respect to the display matrix. With this method, the pixels at different rows are visible only to different directions and more pixels are available for the multiple views in the horizontal direction - but this comes with the cost of lowered resolution in the vertical direction. Stream TV Networks is an example of a company that has developed glasses-free 3D TVs for consumers. Their Ultra-D TV has a stack of refractive and diffractive optical elements bonded to an LCD. According to the company’s white paper [8] on their display technology, individual sub-pixels are projected into space and generate “something similar to a light- field”. In this case, it is likely that light field refers to a multiview configuration rather than to a true light field design with multiple focal planes. 3.1.4 3D multiview displays made with multiple projectors One of the most used methods for creating a multiview 3D display is to utilize multiple projectors (Figure 29). Each of these creates images for a single view and projects it to a screen that combines the images in one spatial location. The lenticular or holographic screen needs to maintain the angular distribution for the multiple views. If the system is used only for horizontally placed views, the light is usually diffused in all directions in the vertical direction in order to cover different viewing heights. In the horizontal direction, the light is focused onto the reflective or transmissive diffuser screen and then projected backward or forward to the direction of the viewers. In the front projection system, the light passes through the vertical lenticular screen twice. In the first pass, the lenses focus the projector pixels onto the diffuse screen and in the second pass, the same lenticular lenses project the pixels back to the viewing direction. A transmissive screen can also be used, together with a double lenticular sheet [1]. Figure 29. Schematics of a) front and b) back projection systems with lenticular sheets. a) b) InterDigital | Goggleless AR & VR Project 26 A big benefit of the multi-projector approach is the fact that it can preserve the resolution of projected images for each view. Higher resolution and more viewing zones can be achieved simply by using more projectors with higher pixel counts. The size of the display screen can also be scaled without significantly increasing cost and complexity of the system. The main drawback of the multi-projector approach is the cost associated with using multiple projectors. But as the size and cost of individual projectors is continually decreasing, these systems can become more common in the near future. The main technical challenge in building such a system is the accurate alignment and image calibration of the several projectors that can have slightly different color co-ordinates and brightness values. The densest multiview projector systems are also called light field displays (discussed in Chapter 4.1) as they are able to provide a fairly natural viewing experience of 3D content. However, they often still lack the multiple focal planes associated with real light field technology. Some systems have been produced with more than 100 views and there are already companies (e.g. Holografika, Zebra imaging) offering very high resolution systems that are able to provide realistic 3D visual experience enhanced with 3D audio. In order to make these even more convincing 3D displays, some additional projectors could be used for generating images at different focal distances. This would mitigate the VAC (discussed in Chapter 1.3.1) and transform them into true light field systems. 3.1.5 Moving lenticular and LC lenses A lot of different variations of the integral imaging technique have been developed. What they all have in common is the use of refractive small-scale optical components that are usually in the sheet or panel format. Figure 30a illustrates a 3D display concept based on a moving lenticular sheet module that projects the different view images generated with a fast LCD at high speed to the viewing direction. At each position of the lenticular sheet module, the display screen produces an image corresponding to that particular viewing direction. With the back and forth motion of the lenticular sheet module, the multiview images can be scanned through a wide range of viewing angles. This design effectively combines the benefits of spatial and temporal multiplexing to one fairly compact system. The main technical challenge is how to create the necessary small scale back and forth motion of the large lenticular sheet module in such a way that the movement speed is constant. Especially with large display sizes the module design may become very challenging and costly [1]. Figure 30. Schematics of multiview display designs using a) moving lenticular sheet module and b) switchable liquid crystal lens used for a 2D/3D display. a) b) Another presented idea for generating a multiview 3D display is to use lenses made of Liquid Crystal (LC) material. The optical properties of the material can be controlled by applying voltage to electrodes connected to the structure and it can serve as a tunable lens in front of a Spatial Light Modulator (SLM), projecting the SLM images to multiple directions. This basic idea is illustrated in Figure 30b. When a voltage is applied between transparent Indium-Tin-Oxide (ITO) layers, the shape of the LC cells forms an array of lenses optically equivalent to lenticular lenses. The main advantage of using a LC lens is the fact that it can be electronically switched between 2D and 3D display modes without losing resolution and also without using any moving parts that add complexity and unreliability to the structure. The main technical challenge connected to this technology is the cross talk between different views caused by the irregularity of the LC alignment [1]. InterDigital | Goggleless AR & VR Project 27 3.1.6 Directional backlight displays One method of creating a multiview 3D display is to use an LCD screen with directional backlight structure. Two light sources, one for each eye’s view, can be used together with a lightguide. The lightguide has out-coupling structures that project the display back-illumination to two different directions according to which light source is used. By alternating the display image content in unison with the light sources, a stereoscopic view with two different pictures can be created. One company that has developed this method and the structures utilized in the out-coupling is 3M, which also holds a patent visualized in Figure 31. Figure 31. Directional backlight for the creation of stereoscopic view in a patent written by 3M [P40]. Figure 32. Directional backlight 3D display concept [P41] with diffractive gratings patented by LEIA [P42]. LEIA is a company that aims for goggleless, small form-factor 3D displays for mobile devices and automotive industry. Their technology is based on structured backlighting and ordinary LCDs [9]. LED light is collimated and coupled into a lightguide that has diffractive dots for light extraction. The special grating patterns are deposited or molded on the surface of a glass or plastic lightguide. The diffractive patterns scatter the light out of the lightguide to different directions determined by the pattern orientation. The patterns can be designed to generate multiple views and even different focal planes for light field effects. A prototype has been created with 64 different views and a 90° viewing angle. Figure 32 presents a concept picture and two patent images made by LEIA. InterDigital | Goggleless AR & VR Project 28 3.2 Volumetric displays Volumetric 3D displays generate three-dimensional images in true 3D space instead of just projecting virtual images from a 2D surface to different viewing positions like all the previously presented systems. Each voxel (volumetric pixel) of the image is located physically at the spatial position where it is supposed to be according to the image content and it reflects or emits light from that position toward all directions resulting in a real image in the eyes of viewers. Such 3D displays are able to provide both physiological and psychological depth cues to the human visual system in a more natural way and they can be considered to be more desirable devices than most existing flat displays. However, the different techniques used in creating the volumetric displays provide depth cues with varying quality and no perfect solutions exist today. 3.2.1 Persistence of Vision displays Persistence of Vision (POV) displays rely on the optical illusion where multiple discrete points located in space are perceived as a collection of 3D points connected to a 3D geometry when the points are activated sequentially faster than what the eye can resolve at the same time when they sweep through a volume. One well-known example of a very simple volumetric POV display is the “floating” LED-display, which can be created from a single row of components simply by rotating or sweeping it fast through space and by activating the LEDs in synchronization with the propagation speed. A few different LED POV display system geometries are shown in Figure 33. Figure 33. Collection of persistence of vision LED display patent images [P43]. The earliest volumetric POV 3D displays were designed several decades ago. As an example of such a system, one patent image from the year 1977 is shown in Figure 34a. It consisted of a rotating panel with an embedded high-speed light emitter array. As the array rotated through the volume, a cylindrical 3D image space was created. A “multiplanar” volumetric POV 3D display system has also been developed by using a high-speed Digital Light Projector (DLP) and a rotating double helix screen as shown in Figure 34b. The system was capable of very high speeds and high spatial resolution. InterDigital | Goggleless AR & VR Project 29 Figure 34. a) Patent image of a POV display based on a rotating array of light sources [P44] and b) a volumetric POV display patent based on projection system and a rotating helix [P45]. One company commercializing the volumetric technology was Perspecta with their Spatial 3D product shown in Figure 35. They developed a 100-million-voxel swept screen volumetric display that was able to generate 10” (25 cm) diameter 3D image with 360° FOV. The system utilizes a DLP, which is used in projection of 2D images to a diffuser disc rotating with a speed of ~900 rpm. The display sweeps the spherical volume twice for every disc revolution, resulting in a refresh rate of 30 Hz, which can be considered as adequate for the POV effect. The device was directed to some special applications like medical visualization and molecular modelling. University of Southern California has studied the same concept further with spinning mirrors, anisotropic diffusers and reflecting color filters that can be used to create a 3D volumetric color image [10]. a) b) Figure 35. Volumetric 3D POV display patent [P46] based on light projection to a rotating vertical disk and b) photograph of the Perspecta volumetric display (courtesy Actuality Systems, Inc) [P10]. The main technical challenges of volumetric 3D displays are connected to the required high rotational speeds of the display element. The device can be a real safety hazard without shielding and if the shielding is made from high quality material like glass it adds weight, size and cost to the system. Without tight shielding, the rotating parts can also cause disturbing air currents and noise around the device. One of the major depth cues, occlusion, is also missing from these types of displays. The images seem translucent, making them non-realistic - but at the same time partially transparent shapes can be useful for some applications that require analysis of 3D structures. a) b) InterDigital | Goggleless AR & VR Project 30 3.2.2 Multi-planar systems One possible method of creating a volumetric 3D display is to stack several 2D display elements on top of each other, thus creating a display volume with 2D slices. As an example, Figure 36a shows a schematic presentation of a patented multilayer 3D display structure. The volumetric multilayer screen includes multiple liquid crystal display layers that are stacked, a back-illuminating light source and some polarization controlling optical components like polarizers and a quarter wave plate. Each LCD is activated separately and sequentially with its own image content presenting one slice of the 3D image. Figure 36. Schematic presentations of patented multi-planar 3D displays that utilize a) LCD screens [P47] and b) an image projector and electrically switchable screens [P48]. In another patented system, the display volume consists of a stack of LC sheets whose optical transmission rates are switchable by the voltage applied on them. The LC sheets stay optically clear when there is no voltage over the elements, but become scattered when the voltage is turned on. By synchronizing the timing of image projection from a high-speed projector and the ON/OFF state of each LC sheet, 2D slices of a 3D image can be displayed in proper 3D locations. LightSpace Technologies is one example of a company that commercialized such a system into a product named DepthCube [11]. The main advantage of such a system is the simplistic design that requires no moving parts. The system based on LC sheets is also fairly easy to control as the image signal needs to be processed only for the projector and not for the whole stack of displays. The main technical challenge is the required high switching speed of the LC display or screen that is difficult to achieve with current technology. Low image brightness due to short exposure times and optical transmission losses in the LC stack are also limiting the performance of these display types. 3.2.3 Pepper's Ghost and volumetric projection displays Pepper's Ghost is an illusion technique used in theaters, amusement parks, museums, television and concerts. It has a long history, dating back to the 19th century, and remains widely performed today. Some well-known examples of the illusion are the appearance of Tupac Shakur onstage with Dr. Dre and Snoop Dogg at the 2012 Coachella Music and Arts Festival and the appearance of “ghosts” at the Haunted Mansion at Disneyland. Teleprompters are a modern implementation of Pepper's Ghost. a) b) InterDigital | Goggleless AR & VR Project 31 a) b) c) Figure 37. a) A patent image of a Pepper's Ghost illusion system used in the creation of virtual presence [P49], b) the system in use [P50] and c) a 3D display utilizing the same optical principle [P51]. One commonly used large scale Pepper's Ghost arrangement is made by utilizing a set of projectors and a semi- transparent screen that is placed between the audience and live performers as shown in Figure 37a. The bright image generated with the projector reflects from the nearly invisible screen and the virtual image of a person appears to be located next to the real people standing at a distance from the screen (Figure 37b). Illumination levels behind the screen are adjusted to fit the virtual image brightness making the appearance more convincing. The same Pepper's Ghost principle based on tilted transparent plates, bright displays and dark surfaces is used in several smaller scale point-of- sales displays that appear to create images floating in midair. One example of these is shown in Figure 37b. A volumetric 3D display can also be made simply by using several projectors that illuminate a volumetric, mostly transparent and light scattering screen that can be constructed from fog, water mist or acrylic panels. With several well- aligned projectors, the image can be presented from different angles allowing the viewer(s) to see the same object from different projection angles making the 3D perception more convincing. The effect is similar to a “hologram” that floats in the volume and people can walk around or even through the screen. Some of the companies developing these systems call the displays as holographic even though the optical functioning principle is clearly not holography. Looking Glass is one example of a company that has commercialized such a “holographic” display. Their volumetric displays are made with a simple arrangement of an image projector and a stack of acrylic panels that are illuminated from below with the projector. With careful alignment of the image source, panels and 3D image content, a volumetric 3D image can be formed with the scatter points integrated to the acrylic plates. Figure 38. Patent images of a volumetric projection display made with microdroplets [P52]. InterDigital | Goggleless AR & VR Project 32 The Pepper's Ghost principle is often used together with multiview configuration. By hiding the projectors from direct view and by mixing the virtual and real 3D image content, it is possible to enhance the 3D experience and create “holograms” that seem to float in the air and interact with real life objects. From the optics point of view these systems are not very complex and the main challenge is in how to create the volumetric screen in such a way that the system remains low cost and elegant. For example, microdroplet formation in the patent drafted in Figure 38 is probably the most technically demanding part of the system that can use just a commercially available projector device for the image generation. 3.2.4 Laser scanned volumetric displays The basic idea in all volumetric 3D display systems is to have a volume filled with voxels that can be selectively excited at any desired location. One way to achieve this goal is to have two independently controlled radiation beams that activate a voxel only when they intersect. Figure 39 shows some patent images describing a 3D display system based on a process known as two-photon upconversion. The process utilizes the energy of two infrared laser beams to pump a special material into an excited energy level, from which it can make a visible light emission by fluorescence transition. The substrate materials need to be suitable for two photon absorption with two different wavelengths. By scanning the two laser beams across the volume, selected voxels can be turned on and off sequentially. Figure 39. Schematic presentation of a patented [P53] two-photon up-conversion system, energy states associated with the fluorescence transition and a table of some materials available for the conversion. Some prototypes have been built that show the feasibility of the two-photon upconversion concept for a 3D display. The main benefit of the technique is in the fact that there are no moving parts inside the display volume, although they may be required for the laser scanning system. Constructions based on solid materials are quite difficult to scale up to larger sizes, but similar systems can be developed also for atomic vapor or liquid phase materials that fill the required volume. One major challenge of this technique is the creation of a full-color image as the different colors require different materials and excitation lasers. InterDigital | Goggleless AR & VR Project 33 Figure 40. Schematics of the laser induced plasma system developed by Burton Inc. One Japanese company, Burton Inc., has developed a laser scanning 3D display technique that can produce visible voxels in the air from the plasma generated by a single focused high intensity laser beam [12]. They noticed that when laser beams are strongly focused, air plasma emission can be induced only near the focal point of the beam. Currently they have used femtosecond lasers which emit bursts that last only some tens of femtoseconds and have succeeded in experiments to display 2D images in the air. The images were constructed from sparse dot arrays produced using a combination of an IR laser and galvanometric scanning mirrors. Later on they extended the scanning and focusing mechanism in 3D space, and produced 3D images in the air with a system drafted in Figure 40. This technology is capable of actually creating the “Princess Leia” type of “holograms” in thin air! However, the use of very high intensity laser beams also makes the technique quite dangerous especially for the eyes and it is likely that the technology can’t be easily commercialized. 3.2.5 Some additional volumetric display types One research group at the University of Texas has developed a volumetric 3D display based on a transparent glass cube embedded with optical fibers. As each fiber can be separately illuminated with the help of a SLM, the fiber ends functioning as voxels inside the cube can be used for the creation of 3D geometry. The created prototype had only 48 x 48 x 12 voxels. The main advantage of this approach is the simplicity of the hardware, but it is at the same time very hard to scale for larger sizes without high costs associated with the manufacturing of the coherent fiber bundles needed in the system. Figure 41. Picture of a patented volumetric LED display structure [P54]. Many research groups have attempted to build volumetric 3D display devices using 3D matrices of LEDs. For example, an 8 x 8 x 8 voxel display prototype was developed at MIT. The concept of these LED array displays is straightforward, but implementation is quite challenging if the goal is to develop high-resolution systems. The main technical challenge in this approach lies in the voxel self-occlusion caused by the opaque LEDs themselves. Cross illumination among LEDs is also a concern as the active components also illuminate the off-state components creating additional light scatter points. However, some companies have already developed and patented their own low resolution solutions, one example of which is shown in Figure 41. InterDigital | Goggleless AR & VR Project 34 a) b) Figure 42. a) Schematic presentation and b) draft of a patented vibrating mirror 3D display system [P55]. One additional method for creating a volumetric 3D display uses a varifocal mirror system that creates a virtual image in front of a viewer. One example of such a system is shown in Figure 42 and it consists of a vibrating circular beam- splitter mirror and a high-speed image projector device. The projector is synchronized to a loudspeaker that makes the flexible mirror vibrate with sound waves. As the mirror bends it also changes focal length and the different images that are projected to it seem to appear at different physical locations in space. Another similar system has been presented that uses membranes (e.g. soap bubbles) as the reflecting surface and it was named “colloidal display”. This display type was used in the creation of virtual surface 3D textures. Technical limitations of such systems come from limited size and flexibility of the vibrating mirror and refresh rate of the projection system. InterDigital | Goggleless AR & VR Project 35 4. Light Field and holographic display technologies Light Field (LF) display technology aims to improve 3D visual perception by eliminating the vergence-accommodation conflict experienced with current autostereoscopic displays. It does this by introducing multiple views and correct focus cues to the eyes. LF systems are designed to create these so-called light fields both in spatial and angular domains, unlike the conventional 3D displays that can basically only control the spatial domain with high pixel densities. Holographic dynamic displays would be the perfect optical solution for 3D viewing. They would be able to produce naturally the extremely dense light fields that resemble the natural viewing conditions in which the light is diffracted from surrounding objects to all directions. From this point of view, the holographic displays can be seen as the ultimate goal for LF technology development. Both of these 3D display types require very complex designs with high-cost components, time-consuming calculations and huge bitrates. There are multiple different and competing technical approaches to both LF and holographic displays. Unfortunately some companies developing 3D display techniques also use the terms quite loosely when describing and naming their products or prototypes, which clearly adds confusion when the different technical approaches are analyzed. Some companies refer to their systems as “holographic” even though they may use simple volumetric projection techniques that don’t have any connection to holography - other than the mental image of what the general public thinks is a “hologram”. Some systems that use holographic components are also showing “holograms”, although a more accurate term for the approach would be “dense light field images” as the number of views or focal planes is clearly limited by some parts of the hardware. This varying use of terminology combined with the complex (and partly confidential) technical details makes it difficult to classify the different existing products and prototypes clearly as light field or real holographic systems. With the strictest definition based on holography physical principles alone, very few holographic commercial video displays have been demonstrated as most of the solutions are restricted to a discrete number of views making them, at best, light field displays. 4.1 Light Field display technologies 4.1.1 Definition of Light Field Light Field (LF) can be defined as a vector function that describes the amount of light flowing in every direction through every point in space. The direction of each ray is given by the 5D plenoptic function, and the magnitude of each ray is given by the radiance. The plenoptic illumination function is an idealized function used in computer vision and computer graphics to express the image of a scene from any viewing position at any viewing angle at any point in time. In practice, the LF has been reduced to a 4D parameterized field, as there is some redundancy in the idealized function. Each ray in space can be defined with two points on two planes as shown on the right in Figure 43. There are also some other possible ways to define the dimensions of the LF (e.g. by adding light ray intensity and color to the four co-ordinates and thus creating a “6D light field”). Figure 43. Some alternative parameterizations of the 4D light field [P56]. InterDigital | Goggleless AR & VR Project 36 4.1.2 Difference between LF and regular 3D displays In natural viewing conditions, the eyes converge and focus to the same object point. The current autostereoscopic 3D displays have only a crude control over the angle of light emitted from the display, which is adequate for showing two different images to the two eyes. However, the image focus points are always on the same flat display surface. This applies also to a multiview display that is able to present different projections of the content to the different viewing directions without the true convergence effects associated with a real scene. In contrast, true light field displays aim to create the whole field of rays in both spatial and angular domains. This requires a much finer control over the light angles emitted from the display and also considerably higher spatial resolution as more than two views are needed just for one user. Figure 44. Schematic presentation of the focal planes with natural viewing, regular autostereoscopic 3D display and light field display. The main reason for adapting the light field technology is to be able to provide natural cues for eye accommodation and retinal blur. This would lead to a better user experience as the eye fatigue, etc., associated with the vergence- accomodation conflict are removed. With high enough 3D resolution, the virtual image could be indistinguishable from real life. Differences between the eye focal planes associated with natural viewing condition, regular autostereoscopic display and a light field display are pictured in Figure 44. Both autostereoscopic and LF displays can create virtual object points on both sides of a flat display, but only the LF display can create the angular distribution necessary for correct focus cues where some parts of the scene are blurred and some parts are in sharp focus (Figure 5). 4.1.3 Super Multi View displays There are two fundamentally different ways to create light fields that contain the necessary focus cues. In the first approach, the object image is projected at the appropriate focal plane corresponding to its 3D location by providing multiple focal planes. In the second approach, parallax is created across each eye that produces the correct retinal blur corresponding to the 3D location of the object being viewed. This can be done by presenting multiple views per single eye with a very dense light field. This technique is sometimes referred to as Super Multi View (SMV). All real light field displays use one of these two approaches. Naturally the second approach is more suitable to a head-mounted, single- user device as the location of the pupils are much easier to determine and the eyes are closer to the display making it much easier to generate and direct the required dense field of light rays to the eye pupils. The first approach is better suited for displays that are located at a distance from the viewer and could be used without headgear. Figure 45. Different occlusion effects of three different light fields directed inside the eye pupil. InterDigital | Goggleless AR & VR Project 37 With current low density multiview displays, the views change in a stepwise fashion as the viewer moves in front of the device. This feature lowers the quality of 3D experience and can even cause a breakup of the 3D perception. In order to mitigate this problem (together with the VAC), some SMV techniques have been tested with as many as 512 views. The basic idea is to generate an extremely large number of views that make the transition between two viewpoints very smooth. If the light from at least two images from slightly different viewpoints enters the eye pupil simultaneously, a much more realistic visual experience follows [13]. In this case, motion parallax effects resemble the natural conditions better (Figure 45) as the brain unconsciously predicts the image change due to motion. The SMV condition can be met by simply reducing the interval between two views at the correct viewing distance to a smaller value than the size of the eye pupil. At normal illumination conditions the human pupil is generally estimated to be ~4 mm in diameter. If the ambient light levels are high (Sunlight), the diameter can be as small as 1.5 mm and in dark conditions as large as 8 mm. 4.1.4 Multiplexing of light fields Light field displays require a high amount of multiplexing from the optical hardware as all the different viewing directions and focal planes need to be presented through a single display surface. There are two basic ways of creating spatially multiplexed flat displays: an array of small sub-displays or a single larger display with sub-pixels (Figure 46). From the practical point-of-view, the utilization of one large display is more convenient as the system will be much thinner and simpler. If the display is stationary and placed on the wall, the light fields can be diffused in the vertical direction as the two eyes are positioned horizontally and 3D effect is adequate for the user. The distance between the viewer and display can’t be very long as light angles become too small if the SMV condition is targeted and more than one image needs to be projected inside the eye pupils. Figure 46. Two optional ways of realizing the required optical spatial multiplexing of LF displays. Stanford University has tested an arrangement based on the combination of Pepper's Ghost type of optics and a single LCD display. Spatial multiplexing was realized by placing three image planes at 0.67 diopter intervals to the viewing space with partially reflective panels. Maximum depth resolution of the eye is estimated to be ~1/7 diopters, which meant that in practice it was possible to position the display focal planes farther apart when the distance from the eye was increased. The panels reflected three different parts of the single display into the same optical path as shown in Figure 47. Arrangement was used for testing the eyes' ability to focus to virtual objects positioned between the panels. Results showed that the focus cues also worked in practice and the 3D volume can be divided to discrete focal planes without breaking the perception of a 3D image [14]. InterDigital | Goggleless AR & VR Project 38 Temporally multiplexed displays use the persistence of vision property as an advantage as images are created to different focal planes or viewpoints faster than what the eye can resolve. One way to realize different focal distances is to use tunable or birefringent lenses, on which focal length can be changed by voltage difference or by polarization [15][16]. As an example, a schematic of one experimental light field projection display system realized with a tunable lens is presented in Figure 48. The main limiting factor in temporally multiplexed systems is usually SLM component switching speed or refresh-rate. Due to the complex structure of tunable optical components, the lens-based systems are also relatively large and expensive to manufacture. They could be suitable for multiple user 3D displays as parts of LF projection systems or for single users as parts of head mounted or table-top displays. Figure 47. Schematics of a single-user volumetric LF display test system. Figure 48. Schematics of a temporally multiplexed LF display system that creates different focal planes with the help of a tunable liquid lens. InterDigital | Goggleless AR & VR Project 39 4.1.5 Examples of existing light field 3D display technologies The most commonly used goggleless multiuser light field 3D display technology utilizes multiple projectors and a holographic screen to generate a sufficient number of views (Figure 49a). These systems don’t necessarily contain one of the key elements of real light field systems, which is the ability to offer correct focal cues either by providing different focal planes or by directing more than one view to the eye pupils of the viewers. The projector systems would simply become too expensive to build if they had enough projectors to cover all the necessary focal planes or the extremely dense fields required for the SMV condition. One of the key elements used for making these advanced 3D displays is the holographic screen. The screen usually maintains the angular distribution in the horizontal direction, but acts as a diffuser in the vertical direction (Figure 49b). However, if the screen is used as a board display that can be viewed from all sides (Figure 52), the projector array needs to be two dimensional and also the screen needs to maintain the angular distributions in both directions. Each point of the holographic screen needs to emit light beams of different color and intensity to the various directions in a controlled manner. It also has to combine the different projected beams into a continuous 3D view without large discrepancies between different viewpoints. Figure 49. a) Holographic screen used together with an array or projectors in a LF system and b) description of light diffusing screen in a 3D projection display patent [P58]. Another example of an advanced LF display design developed at MIT Media Lab is shown in Figure 50. In this “tensor display” design type, LCD panels are used as switchable parallax barrier masks in front of the actual displays. With this arrangement, the light field can be generated by temporal multiplexing and requirements for high resolution can be balanced with a higher display refresh rate – a combination of spatial and temporal multiplexing. It is also possible to take into consideration the image content and adapt the mask accordingly with the benefit of better brightness and lower refresh rate requirement [17]. The structure presented in the picture also includes a rear LCD layer that is used in the making of directional backlight, which helps in the creation of the multiple light fields necessary for multiple users. The same basic structure has been utilized both in a HMD [18] and projection based [19] light field system. Figure 50. Pictures from an advanced light field “tensor display” design patented by MIT Media Lab showing the multilayer structure and example 3D image decomposition to different layers [P57]. a) b) InterDigital | Goggleless AR & VR Project 40 4.1.6 Companies developing LF systems The HoloVizio displays developed by a Hungarian company Holografika use both front and back projection in their displays. In the horizontal direction, a viewer can see only one very thin slit of images generated by one projector in a larger row of projector devices. The company has used as many as 80 projectors in some prototypes and with mirrors on both sides, the system has been able to generate as many as 200 views. Currently Holografika offers glasses free 26” to 140” 3D displays with very high viewing angles for multiple users. Their own definition is that the devices are “not purely holographic systems” although in some texts these may be also referred to as “holographic displays”. Most sources have clearly defined the company’s products as light field displays, although they could also be referred to as high field count multiview displays. In addition to the projection based systems, the company has also a patent (Figure 51) on light field display structures that are based on a dense matrix of RGB LEDs and lenticular sheets. Figure 51. Images from a patent made by Tibor Balogh, founder of Holografika [P59]. InterDigital | Goggleless AR & VR Project 41 Zebra Imaging is a company which has developed a high-resolution 3D display system that, according to the company, was the world’s first, real-time, full-parallax, self-contained, interactive, color holographic display. This “holographic” 3D display is not based on traditional holographic principles that rely upon interference of light in recreation of the 3D view and a better description for it would probably be “light field display”. The whole system can be scaled from small to large screen sizes and it supports simultaneous collaboration of several participants that can view the same image content from 360 degrees around the device. Figure 52 shows two images from two different patents filed by the company. The first image shows the technical functioning principle of the system that uses a 2D array of projectors that simultaneously create the different images that are combined in the holographic screen. The use of multiple projectors or SMLs requires demanding calculations for “hogels” (holographic pixels) and several parallel computational modules are required for the 3D image content creation. The second picture shows one of the applications intended for the system. Interactive battle field simulation can be seen as one potential special-use case where the added value of a high resolution goggle-free multiuser display is worth the cost associated with the complex system. Figure 52. Images from patents filed by Zebra Imaging [P60][P61]. Holoxica is a developer and supplier of 3D holographic solutions based in Scotland UK. The company is offering an interactive holographic 3D display, which they are calling a “2nd gen” prototype. The Holographic Optical Element (HOE) measures approximately 200 x 300mm. The system has a “holographic” laser projector and a holographic diffusion screen (Figure 53). Images are formed in mid-air 200mm out from the hologram plane. Image sizes up to 70x70 mm can be refreshed at video rates. The images are formed in three distinct planes, corresponding to the colors of the lasers in the laser projector R, G and B. The company has announced that it is also working on a volumetric “3rd gen” system. Figure 53. Images from a patent filed by Javid Khan, founder of Holoxica [P62]. InterDigital | Goggleless AR & VR Project 42 Yet another company offering 3D “holographic” display solutions is Fovi3D. According to their web-pages the company is owned by and working with the same optical hardware technology as Zebra Imaging. Zebra is also co-operating with the company Holoxica. Both Zebra and Holoxica make static holographic prints that they also market as “displays”. It is possible that all three companies, Zebra Imaging, Holoxica and Fovi3D, are developing the same light field display technology - based on the announced co-operation as well as similar naming of the different prototype systems. 4.1.7 Technical challenges of LF displays At minimum, light field images are two-dimensional collections of 2D pictures that represent the scenery at different depths and horizontal viewing angles (Figure 54a). The matrix of images can also be three-dimensional, if the 3D effect needs to also be visible vertically like in the case of a multi-user display that lies horizontally on a table and a group of people can view it from many different directions (Figure 52). With computer generated scenes this means complex computations (Fourier transforms) that are used for calculating the right level of focus blur and light intensity at each focal plane. Realistic systems need to use only coarse sampling and approximations as the amount of computational power and data transfer (Figure 54b) required becomes easily overwhelming for a video image that has to present over 30 FL images each second in order to be perceived as continuous. The first major technical challenge associated to high quality LF displays is the huge amount of data needed for the content creation and display control. Figure 54. a) Different images needed for different viewpoints [P63] and b) an example calculation made for the data rate needed for a 3D video display system [P64]. With natural objects, the very dense light fields are generated when an illuminating beam hits the object and the light is diffracted to all directions from surface features that are in the sub-µm-scale. Regular autostereoscopic displays need to generate only two viewing directions and they are able to do that with optical structures that are in the sub-mm-scale. Pixel sizes of current displays are feasible for the autostereoscopic technique - for example, a 50” HDTV has a pixel size of ~0.5mm. Light field displays try to approximate the diffraction angles generated by natural objects and they require optical structures (and pixels) in the µm-scale making them much more demanding to manufacture and design than the current autostereoscopic hardware due to the smaller sizes. Schematic presentation of this difference is shown in Figure 55. The second major technical challenge in LF displays is to find the right very small-scale optoelectronic components that are able to recreate simultaneously or quickly enough the angular and spatial distributions required from a dense light field image. Figure 55. Creation of light angular distributions in a natural scene, with regular autostereoscopic 3D display and LF display. a) b) InterDigital | Goggleless AR & VR Project 43 One nice example of a light field display structure that could have the capability of modulating light to the different directions and depths can be found from a patent granted to Paul Lapstun in 2014. Figure 56 shows some of the images included in the patent document. The first image shows the structure of a light field display element, which acts as a single pixel in the device. This element already contains light sources, collimating optics and a galvanometric scanner for projecting the light beams to different directions. By combining an array of these display elements, it is possible to create virtual focal points in the space both behind and in front of the display surface. The patent also presents the possibility of using tunable liquid lenses in order to generate separate beams that focus to different distances. Such a display system also requires control for the radiance and color for each display element (pixel), effectively making it six-dimensional. The block diagram on the right shows the different stages needed in the system for the control of a single 6D light field element. Figure 56. Images from an example light field display patent by Paul Lapstun [P65]. 4.2 Holographic display technologies Real holographic 3D systems can be considered as the “holy grail” of 3D display technology. They would be able to present, with extreme realism, a window to the virtual 3D world and high-resolution holographic images could be indistinguishable from their real-world counter parts. In practice, there are still numerous technical challenges to be solved before the systems can be considered ready for commercialization. Many promising real holographic 3D display technologies are being developed and several different demonstrators have been built – mostly by research groups at universities and research institutes like MIT, IMEC and University of Arizona, but also by some pioneering companies like Zebra Imaging, QinetiQ and SeeReal. By making some compromises on the pure holographic conditions, components available today can be used in creation of feasibility demonstrators that are able to show the future potential of Computer Generated Hologram (CGH) displays [1]. 4.2.1 Definition of a hologram Holography is a 3D display technique that allows all the characteristics of light wavefronts (with information on phase, amplitude and wavelength) scattered from an object to be recorded and later reconstructed in a way that the 3D image becomes visible to the naked eye. A schematic presentation of the traditional way of recording and presenting holograms is shown in Figure 57. The system uses coherent laser light to illuminate a real object that scatters (diffracts) the light to a holographic plate, which is able to record the light field with both spatial and angular accuracy that is in the same scale as the light wavelength used. When the plate is later illuminated with a similar wavefront as what was used as a reference beam during recording, an extremely high density light field is generated. Ideally there should be no difference between seeing a natural object and seeing a hologram of it as the eye would not be able to tell the difference. A real holographic 3D display would have to be able to manipulate light with structures that are in the same scale as the light wavelength, which is between 0.4 µm – 0.7 µm for the light visible to the human eye. InterDigital | Goggleless AR & VR Project 44 In classic holography, the 3D scene is encoded into the entire hologram and all regions of the recording contribute to each image point. If the holographic plate is broken into pieces, each piece will reconstruct the original scene but with less resolution and in a smaller size. When the hologram is illuminated, the combination of all of its regions reproduces the complete recorded scene by multiple interferences. A classic film hologram has a large diffraction angle, which means that the viewing zone from which the reconstructed object can be seen is large and both eyes of the viewer can fit into this zone easily creating the perception of a 3D image. The reconstructed light field is so dense that it also naturally fulfills the SMV condition discussed in Chapter 4.1.3. Furthermore, the viewer can also move around and see different perspectives of the scene making the 3D illusion even more convincing and making the hologram seem like an extremely dense static multiview display. Figure 57. Schematic presentations of the traditional way of a) recording [P66] and b) reconstructing [P67] holograms with coherent light source (laser) and recording medium. a) b) 4.2.2 Technical challenges of holographic displays In a CGH system (Figure 58a), one of the key components is the Spatial Light Modulator (SLM), which is used to manipulate light waves. In traditional holography, the recording medium is able to capture and reproduce the extremely small scale features of the wavefront due to the special material properties. With computer generated images, the recording part is simulated digitally and the output wavefront needs to be reproduced with a SLM component that is able to diffract light into large viewing angles. Unlike the holographic materials, the digitally addressed SLMs are discrete components and only sampled wavefronts can be created. The amount of information that can be presented in the holographic image is directly related to the spatial resolution and size of the SLM. This means that in order to be able to create a high resolution and artifact free holographic image, one would need to have very small scale pixels in the SLM component that also needs to be fairly large in size. What follows from this condition is a huge number of pixels, and on the system level, the huge number of pixels requires complex control electronics and extensive calculations for the image content creation. Figure 58. a) Schematic presentation of the basic building blocks used in a holographic display system and b) some existing Spatial Light Modulator (SLM) components and their diffraction angles [1]. a) b) InterDigital | Goggleless AR & VR Project 45 The diffraction angle of a holographic display system is proportional to the size of SLM pixels. Pixel size close to or below the wavelength of the visible light used is necessary in order to achieve high diffraction efficiency and wide viewing angles. Figure 58b lists some types of existing SLMs that are used in building 3D holographic displays. The current CGH systems are able to produce only relatively low quality images and a very narrow viewing angle, because they are typically based on LCD, LCOS or micromirror components that limit the FOV to less than 10° [1]. The Interuniversity Microelectronics Centre (IMEC) from Belgium has developed a very small scale MEMS chip that has over 1 billion diffractive nano-devices (DNDs). The size of a single DND ranges from 250 nm to 2400 nm. As the size of the pixels on the chip are close to the wavelength of light, the SLM chip becomes a diffractive element that can be used for CGH 3D displays. IMEC has also developed a laboratory prototype (HoloDis system) demonstrating holographic video with a FOV just below 40 degrees. Figure 59a shows two pictures from a patent filed by IMEC on this technology. Figure 59. Patented structures of a) diffractive MEMS SLM chip developed by IMEC [P68] and b) SLM and c) CGH prototype developed by MIT Media Lab [P69]. MIT Media Lab has developed a diffractive SLM based on anisotropic leaky-mode coupling [20]. Light is coupled as a guided wave to the surface of a special material (Figure 59b). An acoustic pattern, encoded with holographic information, couples the guided light into a leaky mode of orthogonal polarization, which results in light exiting the waveguide in a controlled manner. By multiplexing the acoustic wave, different colors can be coupled out of the waveguide selectively. The controlled diffraction of light out of the SLM can be used in a 3D holographic display by projecting the diffracted fields to a 3D space with a scanner system. The research group has also developed a demonstrator system capable of showing RGB holograms (Figure 59c). In addition to the limitations of available optical SLM hardware, computational requirements and available recording systems present a great challenge to the development of CGH technology. Ideally, all the information of the light wavefronts of a scene should be recorded, but there is no commercial technology available to do this. Another option would be to perform full electromagnetic wave simulations in order to produce the wavefronts scattered from all objects at all points in space. Computationally, this is practically impossible with current technology. For this reason a lot of approximations need to be done, which lower the quality of 3D images made with the prototype systems. It also helps if the position of a single viewer is known as this makes it possible to calculate, with Fast Fourier Transform (FFT) methods, only the wavefronts and associated control signals that are needed for reconstruction of the 3D image at the exact position of the viewer. However, this leads to a system which is single user only and the number of feasible applications becomes much more limited. a) b) c) InterDigital | Goggleless AR & VR Project 46 4.2.3 Static vs. dynamic holographic displays Traditional holographic recordings are static as they utilize a recording medium that can only be used once and as the setups required for the hologram creation are usually based on static components. However, there are now some new materials available that can be used several times over by erasing the information of the previous recording with one light beam and by recording another image with the standard recording and reference beam arrangement. These new types of materials are called sensitized photorefractive polymers (PRPs) and they are blurring the boundary between static and dynamic holographic displays. There are also some companies that have developed faster methods for producing large scale digital holographic prints making the generation of image content faster and more efficient. It should be noted that sometimes these prints are also called “displays”, even though they are static recordings and the image can’t be changed like in the case of a video image. University of Arizona has one of the leading research teams developing PRPs. They have made a holographic recording material that can be refreshed inside a few seconds. They have also built a prototype display system shown in Figure 60, which is capable of rewriting and reproducing a new holographic image at two-second intervals. The system’s object beam is modulated with a computer controlled SLM. Holographic information is recorded onto the polymer using a laser system producing light pulses that last around one nanosecond. The entire recording of a 10 cm x 10 cm holographic screen with 1 mm resolution takes 2 seconds. The 3D image is viewed by using a LED as the incoherent light source and the image can be seen under normal room lighting conditions. The hologram fades away after a couple of minutes by natural dark decay, or it can be erased by a new 3D image recording. [21] Figure 60. Schematics of the dynamic holographic display system developed by University of Arizona. InterDigital | Goggleless AR & VR Project 47 4.2.4 Example of an advanced CGH display SeeReal Technologies is a German company developing Computer Generated Hologram (CGH) systems. The fundamental difference between other holographic displays and SeeReal’s approach is in the extent of FOV where the holograms are projected. While most systems are aimed at producing images that can be viewed from inside as large of a viewing angle as possible, the company’s system relies on eye tracking and reconstruction of the image only in the relatively small virtual Viewing Windows (VW) positioned in front of the eyes. This approach has been visualized in Figure 61. Figure 61. Images from SeeReal’s holographic projection method patents [P70][P71]. Additionally, the optical setup of SeeReal’s system differs from other approaches. A VW is the Fourier transform of the hologram and it is located in the hologram’s Fourier plane. As the other display methods reconstruct the scene and provide viewing regions behind the Fourier plane, SeeReal’s displays provide VWs for observers in the Fourier plane and 3D scene is located between the Fourier plane and the SLM or behind the SLM. In the image reconstruction, each of the 3D scene points is drawn with a sub-hologram that is created with a 30 x 30 pixel array [22]. SeeReal’s approach significantly reduces the amount of information that needs to be processed making it possible to use currently available hardware with reasonable cost. As the VWs are small and far away from the SLM component, only small diffraction angles are needed in the creation of a high-quality holographic picture. As the size of the VW is limited to one diffraction order, a 20 mm VW at a 2 m distance can be created with pixel size of 50 µm. This approach makes large-size holographic displays feasible today, given that the eye tracking and hardware multiplexing for more than one user can be realized in a cost effective way. The company has already proven the feasibility of their approach with a holographic display prototype called Visio 20. InterDigital | Goggleless AR & VR Project 48 5. Discussion and summary 5.1 Discussion The field of 3D display technology is still quite young compared to its 2D counterpart, which has developed over several decades with multibillion dollar investments into research and product development. Many of the most advanced 3D display technologies are still in the stages of laboratory demonstrators or product prototypes. In order to move the 3D display technologies forward to mass market, one needs to find a few “killer applications” first where the added value of the three dimensional image is worth the added cost of the more complex display system. There are definitely some high-end market segments like defense, medical, space and science where the users would benefit greatly from the more natural way of presenting complex visual information. The mass market segments of 3D display technologies could include 3D TV, movies, 3D user interfaces and games, but they all require wider development of 3D recording technology and content creation in order to become interesting enough for the general public. 3D televisions were a big thing a few years back, but as the consumers were disappointed to the less than perfect technology and lack of proper content, the interest has clearly faded. Any 3D display technology that can break into these markets may find widespread adoption and high-volume production opportunities that make it possible to reduce the cost of the systems to a level affordable by the average consumer. The numerous head-mounted VR and AR devices are currently leading the way for consumer adaptation of the 3D technology and the need for goggleless hardware solutions will likely increase in the near future as people are getting used to the idea of the digitally enhanced visual world around them. Virtual presence and people collaborating in the virtual world still requires a lot of hardware development as the currently available consumer 3D display systems lack the natural feel of the real world environment. The Light Field and Computer Generated Hologram displays are still in the very early stages of development and there are several different competing technologies that try to fulfill the hard demands of a perfect 3D display. However, the many goggleless systems already created have proven the feasibility of these technologies and we can expect some break-through products that are able to deliver the wanted "Wow!" effect in the near future. Light field display technology aims to improve 3D visual perception by eliminating the vergence-accommodation conflict experienced with current autostereoscopic displays. It does this by introducing correct focus cues to the eyes with the help of light fields. This area of display technology is closely related to the holographic technology as the aim in both is to generate the extremely dense light fields that would resemble most natural viewing conditions and the holographic displays can be seen as the ultimate goal for LF technology development. Although both of these technological areas are developed simultaneously, the LF technology is much further along than CGH technology. This is visible from the amount of LF systems already offered by some companies and also from the maturity of prototypes shown by research groups at universities and research institutes. Display resolution, connected to the whole image generation pipeline, is one of the clear areas where technology development in LF and CGH systems is needed. Viewers today are already used to seeing very high quality 2D video images as ordinary home TV resolutions continue to climb from the current Full HD to 4K and even 8K. The image quality of future 3D displays should be targeted to a level comparable with existing 2D displays - or they should be able to deliver some additional visual effects that can compensate for the lower resolution. Holographic “displays” based on digital recordings on holographic material can generate very high-quality 3D still images, which can serve as another reference point for the expected image quality for 3D presentations. However, it is likely to take many years of basic hardware development before computer generated dynamic holographic 3D display systems can deliver performance anywhere close to these examples. Before that, any technology that is able to fulfill the Super Multi View condition has a chance to succeed and take the place of industry-standard 3D display technology. InterDigital | Goggleless AR & VR Project 49 5.2 Comparison of depth cues and complexity of 3D display systems There are very few published studies that compare the various 3D display technologies to each other. This is probably due to the challenging task of analyzing and matching a large number of different designs with different optical properties and evaluating their impact on the quality of viewers’ perception, comfort level, etc. It is virtually impossible to perform rigorous quantitative or empirical comparisons of optoelectromechanical design details and performance among a large variety of display techniques without getting into the situation of having to compare apples to oranges. The most comprehensive analysis that the author has found was made by Jason Geng in 2013 [1]. Dr. Geng’s paper contains a detailed table listing the pros and cons of the different 3D display systems existing at that time. He also provided a table that summarized the different depth ques that can be achieved with the competing approaches. The contents of the comparison table complemented with the author’s estimates on the relative cost and complexity of each system is recreated in Table 1. P R O P E R T Y P e rs p e c ti v e O c c lu si o n S h a d in g T e x tu re P ri o r K n o w le d g e A c c o m a d a ti o n C o n v e rg a n c e B in o c u la r d is p a ri ty M o ti o n p a ra ll a x C o m p le x it y & C o st 2D/ Pseudo 3D displays Yes Yes Yes Yes Yes No No No No Low Stereo 3D displays Yes Yes Yes Yes Yes No Yes Yes No Fairly low Multiview 3D displays Yes Yes Yes Yes Yes No Yes Yes Limited Fairly high Super multiview Yes Yes Yes Yes Yes Yes Yes Yes Yes High Volumetric Yes Limited Limited Limited Yes Yes Yes Yes Yes Very high LF displays/ Holographic Yes Yes Yes Yes Yes Yes Yes Yes Yes Very high Table 1. Comparison of depth que properties and relative complexity associated to different 3D display technologies. InterDigital | Goggleless AR & VR Project 50 5.3 Summary The human visual system needs to be taken into account as a basis for any 3D display design in order to achieve a high quality experience without side effects. Currently available 3D display devices don’t have perfect image quality yet and there is a lot of room for further technical development in this field. Due to the fact that head-mounted AR and VR systems are close to the eyes, they are much easier to make and can be more cost effective than goggleless 3D displays. It follows that goggle-based solutions will likely be more popular than goggleless displays for the consumer markets in the very near future. However, the need for goggleless 3D display solutions will likely increase as people get more acquainted with virtual visual content and seek more natural ways to interact with it. Light Field (LF) technology aims to improve 3D visual experience by eliminating the vergence-accommodation conflict by introducing correct focus cues to the eyes. LF systems are geared for creating high-resolution light fields both in spatial and angular domains that approximate the natural light fields generated by real objects. There are multiple different and competing technical approaches on how to realize this. Several different methods are under investigation by research groups in universities, research institutes and companies around the world. Some of the solutions have already reached commercial maturity, but are still highly complex, expensive and large systems outside the reach of consumers. Holographic dynamic displays would be the perfect optical solution for 3D viewing, but they have even greater technical development challenges than LF systems. There are no real holographic digital imaging devices available that would record the content, and the spatial light modulator components necessary for holographic accuracy, with small enough feature sizes, are still very much in the basic research stage. Research efforts have produced some promising results, but a practical, large display size, high-resolution, full-color, and dynamic holographic display still remains a significant challenge. Commercialization of this technology will take at least several years of basic research and intense technical development. Both light field and holographic 3D display systems require very complex designs with high-cost components, time-consuming calculations and extremely large bitrates. In order to reach full commercial maturity they require the development of not only the display hardware technology, but of the whole system with image acquisition, data storage, calculation of digital content and data transfer. InterDigital | Goggleless AR & VR Project 51 Main references [1] Jason Geng; Three-dimensional display technologies; Advances in Optics and Photonics, Vol. 5, Issue 4, pp. 456-535; (2013) [2] James Cutting, Peter Vishton; The integration, relative potency, and contextual use of different information about depth; Perception of space and motion, (pp. 69-117). San Diego, CA, US: Academic Press; (1995) [3] David M. Hoffman, Ahna R. Girshick, Kurt Akeley, Martin S. Banks; Vergence–accommodation conflicts hinder visual performance and cause visual fatigue; Journal of Vision, Vol.8, 33.; (2008) [4] Dhanraj Vishwanath, Erik Blaser; Retinal blur and the perception of egocentric distance; Journal of Vision, 10(10), pp. 1-16; (2010) [5] Matthew Hirsch; Presentation slides: Stereo and 3D Displays, Published on October 3 (2013) [6] Brian T. Schowengerdt, Richard S. Johnston, C. David Melville and Eric J. Seibel; 3D Displays using Scanning Laser Projection; SID Symposium Digest of Technical Papers, Volume 43, Issue 1; (2012) [7] University of Washington news archive; Contact lenses with circuits, lights a possible platform for superhuman vision; January 17 (2008) [8] Stream TV Networks; White paper: The Ultra-D 3D Displays address human depth perception in a more natural way giving the viewer the sense of being there. (Accessed March 2017) [9] David Fattal, Zhen Peng, Tho Tran, Sonny Vo, Marco Fiorentino, Jim Brug and Raymond G. Beausoleil; A multi-directional backlight for a wide-angle, glasses-free three-dimensional display; Nature 495, 348–351; (2013) [10] Andrew Jones, Ian McDowall, Hideshi Yamada, Mark Bolas, Paul Debevec; Rendering for an interactive 360° light field display; SIGGRAPH 2007 papers, Article No. 40; (2007) [11] Alan Sullivan; A solid-state multi-planar volumetric display; SID Symposium Digest of Technical Papers Volume 34, Issue 1; (2012) [12] Hidei Kimura, Taro Uchiyama, Hiroyuki Yoshikawa; Laser produced 3D display in the air; Proceeding SIGGRAPH ‘06, Emerging technologies Article No. 20; (2006) [13] Yasuhiro Takaki; High-Density Directional Display for Generating Natural Three-Dimensional Images; Proceedings of the IEEE, VOL. 94, NO. 3; (2006) [14] Kurt Akeley, Simon J. Watt, Ahna Reza Girshick, Martin S. Banks; A stereo display prototype with multiple focal distances; SIGGRAPH 2004 Papers, Pages 804-813; (2004) [15] Gordon Love, David Hoffman, Philip Hands, James Gao, Andrew Kirby, Martin Banks; High-speed switchable lens enables the development of a volumetric stereoscopic display; Optics Express, Vol. 17, No. 18, 31.08.2009, p. 15716-15725; (2009) [16] Nikhil Balram and Ivana Tošić; Light-Field Imaging and Display Systems; Information Display 4/16, pp. 6-13; (2016) [17] Douglas Lanman, Matthew Hirsch, Yunhee Kim, Ramesh Raskar; Content-Adaptive Parallax Barriers: Optimizing Dual-Layer 3D Displays using Low-Rank Light Field Factorization; MIT Media Lab - Camera Culture Group; (2011) [18] F. Huang, K. Chen, G. Wetzstein. The Light Field Stereoscope: Immersive Computer Graphics via Factored Near-Eye Light Field Displays with Focus Cues, ACM SIGGRAPH (Transactions on Graphics 33, 5); (2015) [19] Matthew Hirsch, Gordon Wetzstein, Ramesh Raskar; A Compressive Light Field Projection System; SIGGRAPH 2014. ACM Transactions on Graphics 33(4); (2014) [20] D. E. Smalley, Q. Y. J. Smithwick, V. M. Bove, J. Barabas and S. Jolly; Anisotropic leaky-mode modulator for holographic video displays; Nature 498, pp. 313–317; (2013) [21] P.-A. Blanche, A. Bablumian, R. Voorakaranam, et.al.; Holographic three-dimensional telepresence using large-area photorefractive polymer; Nature 468, 80–83; (2010) [22] Stephan Reichelt, Ralf Häussler, Norbert Leister, Gerald Fütterer, Hagen Stolle and Armin Schwerdtner; Holographic 3-D Displays - Electro-holography within the Grasp of Commercialization; Advances in Lasers and Electro Optics, book edited by Nelson Costa and Adolfo Cartaxo; (2010) InterDigital | Goggleless AR & VR Project 52 Figure references [P1] Example of a simple home theatre system; picture by Hrbrhauntr, licensed under CC BY-SA 4.0 [P2] Artwork done by David Revoy, licensed under CC BY 3.0 [P3] Example depth of field pictures by Alex1ruff, licensed under CC BY-SA 4.0 [P4] Old stereographic photograph; picture from US Library of Congress, licensed under public domain [P5] Schematics of the vestibule-ocular reflex; picture by Mikael Häggström, licensed under CC BY-SA 3.0 [P6] Picture of Oculus Rift booth; picture by Marco Verch, licensed under CC BY 2.0 [P7] Visualization of multiple 3D display views; picture by Franciscoceba, licensed under CC BY-SA 3.0 [P8] Patent US20120105954 filed by GRilli3D (2011); pictures licensed under UPSTO public domain [P9] Jeff Gattis (HTC) wears the HTC RE Vive; picture by Maurizio Pesce, licensed under CC BY 2.0 [P10] Perspecta display (Actuality Systems); picture by Gregg Favalora, licensed under CC BY-SA 3.0 [P11] Projection screen in a home theater; picture by Derek Jensen, licensed under public domain [P12] LG 3D TV display; picture by LG, licensed under CC BY 2.0 [P13] Oculus Rift CV1: picture by Samwalton9, licensed under CC BY-SA 4.0 [P14] Google Cardboard VR headset; picture by Evan-Amos, licensed under public domain [P15] Google Glass smart glasses; picture by Mikepanhu, licensed under CC BY-SA 3.0 [P16] Patent US20130070338 filed by Google (2011); pictures licensed under UPSTO public domain [P17] Patent US20120188148 filed by Microvision (2011); pictures licensed under UPSTO public domain [P18] Patent US5467104 filed by Furness et.al. (1992); pictures licensed under UPSTO public domain [P19] Patent US7391573 filed by Lumus (2004); pictures licensed under UPSTO public domain [P20] Patent WO2016142707 filed by Colour Holographic (2016); pictures licensed under WIPO public domain [P21] Patent US20150253570 filed by VTT (2013); pictures licensed under UPSTO public domain [P22] Patent USD768627 filed by Samsung (2015); pictures licensed under UPSTO public domain [P23] Patent US9158115 filed by Amazon (2013); pictures licensed under UPSTO public domain [P24] Patent US8773599 filed by Google (2011); pictures licensed under UPSTO public domain [P25] Microsoft Hololens visualization; picture by Microsoft Sweden, licensed under CC BY 2.0 [P26] Patent US20160110920 filed by Magic Leap (2015); pictures licensed under UPSTO public domain [P27] Patent US20160110912 filed by Magic Leap (2015); pictures licensed under UPSTO public domain [P28] Patent US8520309 filed by Innovega (2008); pictures licensed under UPSTO public domain [P29] Patent US20160091737 filed by Samsung (2015); pictures licensed under UPSTO public domain [P30] Example stereogram image by U.S. Department of the interior, licensed under U.S. public domain [P31] Anaglyph 3D glasses; image by 3dnatureguy, licensed under public domain (by author) [P32] Principle of active shutter glasses by Locafox.de; picture licensed under CC BY-SA 4.0 [P33] Parallax barrier and lenticular systems; picture by Cmglee, licensed under CC BY-SA 3.0 [P34] Lenticular panel; picture by Bernard Soulier, licensed under CC BY-SA 3.0 [P35] Patent US20160116752 filed by BOE (2013); pictures licensed under UPSTO public domain [P36] Patent US7573491 filed by Hartkop & Dunn (2005); pictures licensed under UPSTO public domain [P37] Patent US7961182 filed by JSTA (2005); pictures licensed under UPSTO public domain [P38] Patent US6118584 filed by Philips (1996); pictures licensed under UPSTO public domain InterDigital | Goggleless AR & VR Project 53 [P39] Patent US6064424 filed by Philips (1997); pictures licensed under UPSTO public domain [P40] Patent WO2011149641 filed by 3M (2011); pictures licensed under WIPO public domain [P41] Principle of directional backlight LF display; picture by Leia 3D, licensed under CC BY 4.0 [P42] Patent US9298168 filed by Leia (2014); pictures licensed under UPSTO public domain [P43] Patent US6265984 filed by Molinaroli (1999); pictures licensed under UPSTO public domain [P44] Patent US4160973 filed by MIT (1977); pictures licensed under UPSTO public domain [P45] Patent US6177913 filed by US Navy (1998); pictures licensed under UPSTO public domain [P46] Patent US20030176214 filed by Burak & Fiden (2003); pictures licensed under UPSTO public domain [P47] Patent US5745197 filed by Aerospace Corporation (1995); pictures licensed under UPSTO public domain [P48] Patent US20020163482 filed by Sullivan (2002); pictures licensed under UPSTO public domain [P49] Patent US5865519 filed by Maass (1996); pictures licensed under UPSTO public domain [P50] Pepper's Ghost technology used; picture by Riksutställningar, licensed under CC BY-SA 3.0 [P51] Pepper's Ghost 3D display; picture by Greg Goebel, picture licensed under CC BY-SA 2.0 [P52] Patent US8289274 filed by Sliwa & Tosaya (2008); pictures licensed under UPSTO public domain [P53] Patent US5684621 filed by Downing (1995); pictures licensed under UPSTO public domain [P54] Patent US8872420 filed by Brindisi (2014); pictures licensed under UPSTO public domain [P55] Patent US8823702 filed by Disney Enterprises (2012); pictures licensed under UPSTO public domain [P56] Alternative parameterizations of 4D light field; picture by MarcLevoy, licensed under CC BY-SA 3.0 [P57] Patent US8848006 filed by MIT (2013); pictures licensed under UPSTO public domain [P58] Patent US8328360 filed by Seiko Epson (2009); pictures licensed under UPSTO public domain [P59] Patent US6999071 filed by Balogh (2001); pictures licensed under UPSTO public domain [P60] Patent US6665100 filed by Zebra Imaging (2000); pictures licensed under UPSTO public domain [P61] Patent US7190496 filed by Zebra Imaging (2004); pictures licensed under UPSTO public domain [P62] Patent US8625183 filed by Khan (2009); pictures licensed under UPSTO public domain [P63] Multi-Viewer 3D display principle; picture by Silke Schermann, licensed under CC0 1.0 [P64] Patent WO2008138986 filed by SeeReal (2008); pictures licensed under WIPO public domain [P65] Patent US9456116 filed by Lapstun (2014); pictures licensed under UPSTO public domain [P66] Schematics of recording a hologram; picture by Bob Mellish, licensed under CC BY-SA 3.0 [P67] Schematics of reconstructing a hologram; picture by Bob Mellish, licensed under CC BY-SA 3.0 [P68] Patent US8643937 filed by IMEC (2011); pictures licensed under UPSTO public domain [P69] Patent US20140300695 filed by MIT (2014); pictures licensed under UPSTO public domain [P70] Patent US20130022222 filed by SeeReal (2011); pictures licensed under UPSTO public domain [P71] Patent US20150036199 filed by SeeReal (2014); pictures licensed under UPSTO public domain