Figure 1. Popularity of the search terms (a) “Stereoscopy” and (b) “3D television” between January 2007 and October 2016, based on Google Trends (https://www.google.com/trends, accessed 29.10.2016).
About five years ago the popularity of stereoscopy for personal entertainment exploded. The majority of new TV models supported active or passive stereoscopy and the main manufacturers of computer equipment began selling monitors and notebooks with stereoscopic screens. Some producers of digital cameras issued compact stereoscopic models with two lenses, and even with autostereoscopic displays.
After several years the landscape is different. Stereoscopic three-dimensional (3D) monitors and laptops are rare, and none of the big camera manufacturers offer a stereoscopic model. Stereoscopic 3D TVs are still on sale, but 3D support is rather an additional gadget, irrelevant for most clients. Heavy investment in research and development of 3D certainly proved unprofitable, and further advancement of this technology has slowed down.
We analyze the main reasons for this, and present possible future solutions. We will try to find out if the 3D market collapsed because of imperfections of technology, or because there is no real demand for stereoscopy. The answer is important especially for manufacturers, because it will help to decide about further investments in research and development of 3D technology.
Stereoscopy is a technique for creating the illusion of depth based on binocular vision; it is almost as old as photography.
Stereoscopy is a technique for creating the illusion of depth based on binocular vision. It is almost as old as photography: the first stereoscope was built in the 1830s, only a few years after production of the first permanent photograph .1 Soon stereoscopes become popular gadgets for home entertainment, and by the end of the nineteenth century there were millions of images in circulation . In 180 years of the history of stereoscopy, there have been many waves of popularity, alternating with periods when it was considered obsolete. Similarly, 3D movies had several short periods of popularity, which all ended rapidly: the first commercial 3D movie premiered in the 1920s but was unsuccessful, and later attempts by Hollywood to use 3D technology in the 1950s and 1970s failed as well .
Several years ago another wave of stereoscopy popularity appeared, but this time the situation was different. Modern stereoscopy is based on digital images, displayed on screens of considerable size. The quality of both still and motion images is incomparably better. At the beginning of the 2010s it was difficult to find a new TV model without 3D technology, and many laptops and even mobile phones were equipped with stereoscopic displays. The era of 2D imaging seemed to have finished. But consumer interest in 3D disappeared as quickly as it emerged, aligning with the popularity of the Internet search term stereoscopy, which has fallen more than threefold since 2011, and for the phrase 3D television, where there was a tenfold fall (Fig. 1).
In this article we look for the main reasons for the rapid collapse of this emerging technology that seemed to offer such promise. The collapse was unforeseen, even by large and experienced manufacturers who invested heavily in the new technology. We try to answer the question of whether stereoscopy is not yet mature enough to meet users’ needs and will come back with more success in the future, or whether consumers are in fact satisfied with 2D images and any future 3D technology will only be a passing fashion.
Let us recap briefly the most common stereoscopic technologies that were available when the market for such devices started to grow around 2010. In fact all these technologies had been known for decades, but this was the first time they were in common use:
- Passive display, based on polarized light, where consecutive lines of the screen emit light with different polarization. Inexpensive binoculars are a pair of polarizing filters, which allow each eye to observe every second line. A disadvantage is a reduction of vertical resolution by half.2 Passive display can be based on circular or linear polarization, depending on the standard.
- Active display, based on an active shutter, where images for the right and for the left eye are presented consecutively. Glasses utilizing this method are more expensive and heavier than those used in the passive system, and they include a battery to operate active shutters synchronized with the display.
- Autostereoscopic display, which does not require glasses. Instead, the user must take a specific position, centrally in front of the display. The two most common autostereoscopic technologies are parallax barriers and lenticular lenses.
- Color-based anaglyphs – the simplest solution. These do not require any special equipment except colored glasses (among many standards, red-cyan is the most popular). However, with this method, representation of colors is very limited.
- Head-mounted displays, where the user wears a device with two small displays, one for each eye, and magnifying lenses. Good quality head-mounted displays are rather expensive, especially considering that each user needs a separate device.
Reasons for Collapse
Our discussion is related to applications in home entertainment, which is mostly based on technologies 1)–3). Here we try to answer the question of why the market for stereoscopic devices such as 3D TV sets, 3D laptops, and 3D computer screens shrunk rapidly after a short time of dynamic growth.
Inconvenience and Health Issues
Watching in 3D is never as convenient as in 2D. To watch in 3D, it is necessary to put on special glasses, and to charge the battery or use a cumbersome cord if they are active. Autostereoscopic devices, where a user must occupy a specific position, are even less convenient, especially if several people would like to watch the same material. Wearing special glasses is not a big issue in the cinema, and maybe this is why 3D cinema has not lost so much popularity, but glasses are hard to accept at home, where watching television is often one of several simultaneous activities. Moreover, users must keep their head upright, as tilting one’s head to the side introduces vertical disparity that causes visual discomfort or even makes watching impossible (this effect was studied in detail in ). For passive glasses with a linear polarizer, adverse effects include ghosting, where part of the left image is passed to the right eye and vice versa.
Some users complain that watching 3D content causes health problems such as eye strain, headache, nausea, and general fatigue , . One reason is the vergence-accommodation conflict , when the eyes must focus on points of the screen plane, while at the same time verging at points at different distances. Recent research ,  confirms that the vergence-accommodation conflict contributes to user fatigue, and proposes how to alleviate this when creating 3D content. Vergence-accommodation conflict can also be solved using special corrective lenses . However, the research  demonstrates that the mismatch between vergence and accommodation it is not the only factor responsible for adverse effects and its elimination can only alleviate the problem rather than solve it.
Another issue is excessive binocular disparity . This can be reduced when creating 3D content, but disparity also depends on the equipment used. This means that content well-suited for one display may cause adverse effects when watched, for example, on a screen of a different size. Surveys indicate that some viewers suffer headaches when using active shutter glasses . Finally, watching 3D content is not recommended for children under six, whose vision is still developing.
No Real Demand — Added Value is Less than It Appeared in the Beginning
Many people are enchanted when looking at 3D content for the first time. The stereoscopic experience brings the user much closer to the feeling of reality than a two-dimensional image, and the inconvenience related to wearing 3D glasses may seem irrelevant.
Why did the market for stereoscopic devices such as 3D TV sets, 3D laptops, and 3D computer screens shrink rapidly after a short time of dynamic growth?
Nevertheless, when compared to 2D content, a stereo image does not provide as much new information as we could expect. Our brain infers depth from many different clues, such as perspective, apparent size of objects, occlusions, shadows, etc. This is why in most cases we can get almost the same information from a 2D image as from a 3D image of the same scene. This is why consoles for the remote control of professional (e.g., military) drones are equipped with large 2D screens and do not include 3D displays. Stereoscopic systems are sometimes used in hobby drones for First Pilot View (FPV) flying but their popularity is decreasing (many FPV pilots use head-mounted displays but there is usually non-stereoscopic image from a single board camera). After getting used to stereoscopic view, when seeing depth is not a curiosity anymore, the difference between this and 2D is not as relevant. As Bell  argues, the main reason for the failure of 3D television is that there was never a demand for this technology. The long history of stereoscopy, with its waves of popularity that always ended after a short time, seems to confirm this thesis.
Insufficient Reality and Limitations when Creating 3D Content
Some users complain that watching 3D content causes health problems such as eye strain, headache, nausea, and general fatigue.
Stereoscopy can be regarded as a step towards giving the user a feeling of watching the real world. But the appearance of 3D content is not identical to reality. First, the geometry of a stereoscopic system is different from the geometry of the human vision system. Our eyes converge at the object we are viewing, while convergence of a pair of cameras is fixed. There are a number of rules about how to set a stereoscopic system, relating, for example, to the distance from cameras to the closest object in the scene (in relation to the base), the maximum parallax in relation to screen size, etc. . These rules, apart from imposing limitations when creating 3D content and making on-line 3D broadcasts difficult, imply that 3D content is not scalable. This means that the same stereoscopic image looks different when displayed on screens of different sizes. The relationship between the shooting distance and perceived distance in stereogram is nonlinear (except in a system with parallel optical axes) and the form of nonlinearity depends on system geometry (see [15, ch. 21.1]). There are a number of undesired effects, for example:
- Difficulty with focusing eyes on an object that is too close. A single object at too close distance may hamper watching the whole 3D image.
- Cardboard effect, which is an impression that the image consists of flat objects and a flat background. This effect can be avoided by applying a set of rules regarding the geometry of the stereoscopic system , .
- Dwarfism: A scene on a large scale, such as a mountainous landscape or city panorama, requires a long base to be used in order to produce the stereoscopic image. But the human brain is accustomed to interpreting a spatial image when the distance between the eyes is around 65 mm, and it follows the trick only partially. The impression of depth is perceived but the brain interprets the scene as a mock-up rather than as a real scene.
Sharing photos through social network services, email, MMS, etc. is often the main motivation for taking them. Content using 3D can only be shared with those who have devices capable of displaying it, and this makes it much less attractive. We consider this problem least important, because it is valid for many new technologies, which become attractive only when the group of users is large enough. This issue could have hampered the popularity of 3D in its initial phase, but it cannot be the main reason for the retail slump after initial dynamic growth.
Social Rather than Technical Explanation
A question we try to answer is whether 3D failed because technology is yet not mature enough to meet users’ needs, and therefore could be successful in the future, or whether lack of demand results from different reasons, for example psychological, which would make investment in the development of 3D technology pointless. Some of the problems listed above are related to technological limitations, and they can be solved in the future. For example, an autostereoscopic display with a head-tracking system capable of directing the image to the user’s eye, even for multiple viewers as proposed in , , may provide a convenient glasses-free solution. However, there are also non-technological barriers to the deployment of stereoscopy. In Table 1 we indicate prospects for future solutions to the issues indicated in this article.
Let us emphasize that we are considering applications in home entertainment and that in some other areas stereoscopy has proved to be useful regardless of temporal vogues.
Table 1. Prospects for future solutions to the issues raised.
For example, perception of depth is very important in telesurgery, and the console of the most common surgery robot DaVinci manufactured by Intuitive Surgical Inc. is equipped with a stereoscopic display .
As we can see from the table, there is a chance that solutions to most technological problems will be found, and in order to make any prediction it is crucial to estimate the truth of the thesis about lack of demand. Are users really indifferent as to whether they look at a flat or a spatial image? The explanation could be more social than technical: merging into a virtual world makes the user temporally lose contact with the real world. This is why virtual reality, although popular among some game players, is not common as a form of home entertainment: most people prefer to socialize in front of a 2D screen, rather than put on a wearable device that carries them to a virtual world but cuts off reality. A stereoscopic image, with or without glasses, is a substitute for virtual reality. Unlike 2D images, which are certainly part of our homes, the content of a stereoscopic screen apparently crosses our walls and draws our attention out of the room in which we sit. If this is the reason why even people fascinated by stereoscopic imaging abandon it after a short time, technological progress will never introduce 3D technology to homes.
Many people are enchanted when looking at 3D content for the first time as it brings the user closer to the feeling of reality than a two-dimensional image.
From the history of stereoscopy we can learn that the advantages of 3D technology make it apparently attractive, and these appearances mislead even experienced market players. It was not the first time that technological advances led big manufacturers to invest in 3D washouts. When other technological innovations bring a new wave of popularity for 3D, we can expect that such companies will be more careful before investing.
Pawel Rotter is assistant professor at the Department of Automatics and Biomedical Engineering of AGH-University of Science and Technology, Mickiewicza 30, 30–059 Kraków, Poland.