physical problems
One of the biggest difficulties of virtual reality is making the user feel a sensation of immersion without feeling nausea, dizziness, etc. Experiencing these symptoms when using virtual reality is known as cybersickness or virtual reality sickness and is similar to classic motion sickness, or the motion sickness that pilots experience in simulators. The perception of these symptoms also depends on the person. For some, vomiting appears within a few minutes, while others can enjoy virtual reality for hours without any consequences.
The problem lies in a mismatch between the vestibular system (the liquids and fluids in the cavities inside the ear, which send information to the brain about direction, angles, etc.) and the visual system.[105].
These side effects of virtual reality have different causes. Developers try to perfect their systems to avoid or combat them in the best possible way, these being latency, image duplication and persistence among others.
Latency is the delay between the action performed by the user and its representation on the screen, producing imbalances between the vestibular and visual systems, in turn causing nausea and dizziness.[105].
The common latency in video games is the time interval between the user pressing a button and the pixels updating, generally being a minimum of 50 ms. It is important not to confuse this delay with the time between when a user presses a button and the action is carried out, being insufficient for virtual reality, which requires a latency of at least 20 ms so that the user does not experience a delay. In fact, most experts believe that the limit is even lower, at 15 or even 7 ms. Oculus Rift has a delay under optimal conditions of between 30 and 40 ms.
This is because the process of rendering the image requires the tracking system to determine the exact position and orientation of the HMD, the application rendering the scene, so that the hardware transfers the rendered scene to the HMD screen and it begins to emit photons for each pixel.
The first step, tracking takes between 10 and 15 ms when it comes to optical tracking, which in itself is too long. Accelerometer tracking is much faster with a latency of 1 ms or less, but it is inaccurate and drifts a lot. One of the main problems is that 60 Hz screens, for example, already introduce a delay of about 15 or 16 ms in rendering. This value is dependent on the CPU and GPU, but is usually in that range except for older applications that require primitive rendering.
Finally, the hardware transfers the rendered scene to the HMD screen. For most frequency scanning based systems, this means a delay of about 16 ms in the worst case (assuming 60 Hz displays are used). If the image is transmitted immediately, that is, the photons begin to appear instantly upon arrival, the sum of the latencies mentioned above is much higher than 20 ms and is an abysmal distance from the desired 7 ms.
Another major drawback is image judder or duplication. It is a combination of two phenomena, image blurring and stroboscopy. Blurring or smearing is a motion blur present in virtual reality. Strobing or stroboscopy, on the other hand, consists of the perception of multiple copies of an image at the same time, making it appear that there is no movement between them. The union of these two phenomena constitute image duplications.
Judder normally produces dizziness and all related symptoms, so you should try to avoid it. One cause of judder is the fact that pixels move across the retina while they are on (causing smearing). The obvious solution to image duplication is an increase in frame rate. The problem is that to avoid it completely, a frame rate of between 300 and 1000 FPS would be necessary, something too far from reality. Therefore, although the solution is obvious, it is also totally impossible due to technological limitations.
The other solution has to do with persistence. Most displays have full persistence, so the pixels always stay on. The level of blur does not depend on what fraction of a frame the pixels are on, but on the total time that they are on. This is why a frame rate of around 1000 FPS would be ideal with full persistence, as the time would be just 1 ms.
Since this frame rate is currently unattainable, low persistence must be used to achieve the same result. With zero (or almost zero) persistence, the movement of lit pixels across the retina is eliminated, since they remain lit for a very short time. Thus, the blur component in image duplication is eliminated. However, low persistence also has disadvantages, as it can increase stroboscopy. In fact, the smearing itself obscures the stroboscopy quite a bit. By reducing the first using low persistence screens, the second is manifested more clearly. However, this problem is not that serious. The reason is that in the image that the eye is focusing on, stroboscopy will not occur, since the eye itself will avoid it by following it, because the same pixels will go to the same point on the retina in each frame. Although this effect will occur in the rest of the image, it will not be as noticeable since it will be out of focus.[106][107][108].
Privacy and data protection
As new technologies advance, concerns about data protection grow considerably. It is not only important who has access to our data, but also the great unknown of the purposes for which these data will be used.[109] This therefore generates an ethical debate that must be addressed.
It is important to consider that virtual reality not only makes use of our personal data, but also goes beyond the traditional limits of social networks. This means that these types of technologies go one step further and involve new categories of individual recognition with greater granularity and precision. Examples include the glasses used, which extract information from the variations of the iris, or the controls, which act as an interface with the virtual world and analyze postural changes to immediately analyze the emotional response.
The analysis of the relative position of the avatars in said world, the times and the way of reaction, generate a biomechanical study of the individual. If neural interfaces are also added, the result is knowledge of the individual at levels that were never intended to be reached with social networks.[110].
Thus, virtual reality generates a new paradigm in terms of data protection, covering not only the privacy and security of the individual's information, but also their physical and neural integrity.
Other problems
In addition to these technological impediments, virtual reality faces other problems. Firstly, although the short-term effects do not go beyond dizziness and vomiting, no one knows for sure how continued use of virtual reality can affect a person, either physically or mentally.
On the other hand, the costs of the necessary equipment are still too high for the average user. A high-quality virtual reality headset is around €600, and you must also take into account the price of a device (computer or console) capable of running the applications satisfactorily.
Finally, virtual reality needs to generate profits to be viable. Currently, most of the interested public are gamers, but it is necessary to attract more sectors in a broader way to survive economically.[112][113].