VR系列——Oculus最佳实践：七、虚拟幻境头晕（下）

失真校正

Rift的镜头扭曲的显示屏上显示的图形，这是由SDK给出的后期处理步骤校正的。这种扭曲被正确的完成，并根据SDK的准则和所提供的实例演示是极为重要的。不正确的失真可以“看”做相当正确的，但还是会觉得迷失方向和不舒服，所以注意细节是至关重要的。所有的失真校正值需要匹配他们的物理设备，没有可能是用户可调的（SDK示例是为了让你玩他们，只是为了说明隐藏在幕后的事，但不是因为这是一个特别要做的事）。

我们精心调制我们的Rift镜头失真设置，并不断的采用进一步改善失真的调整工作。所以开发者必须使用官方的Oculus VR失真设置在Rift上正确显示内容。

闪烁

闪烁在虚拟幻境头晕的动眼神经组件中起着显著的作用。它可以由高变差改变亮度级，并且在你的视野周边感知最强烈。虽然闪烁可以随着时间的推移变得不那么引人注意，但它仍然会导致头痛和眼睛疲劳。

虽然他们为VR提供了许多优势，OLED显示器身上带着某种程度的闪烁，类似于CRT显示器。不同人可能有不同程度的敏感性，但DK2的75-Hz显示面板速度不够快，大多数用户不会感觉到任何明显的闪烁。未来的迭代将有更快的刷新速率，更不明显的闪烁。这个或多或少取决于开发者的手，但必须作为完整性包含在其中。

你的职责是避免创建故意闪烁的内容。高对比度，闪烁（或迅速交变）的刺激，特别是在1~30 Hz范围内，可以触发人们光敏性癫痫发作。有关这一点，高空间频率纹理（如细的黑色和白色条纹）也可以触发癫痫患者癫痫发作。

经验

在虚拟环境待得越久的用户 ，越不会感受到虚拟幻境头晕。 [14]相关理论认为有这种结果， 是因为有时候无意识的学习会让用户更好地处理VR新奇体验。例如，大脑学会了重新诠释之前引发不适的视觉异常，而且用户的动作变得更稳定更有效得来减少对流。有个好消息是，开发商不用担心给有经验的用户设计更激烈的虚拟体验;坏消息是，大多数用户在可以把控这些体验前都需要时间来适应Rift和游戏。

这有几个重要的影响。首先，反复性测试他们自己游戏的开发商相对于新用户而言，会更能忍受虚拟环境头晕，因此需要让一批新手感受不同的虚拟幻境头晕感官水平，来寻找实际上最舒适的体验。其次，新用户不应当立即陷入激烈的游戏体验;你应该采用更稳重，节奏较慢的互动让他们渐渐进入游戏。再者，你应实施本指南中的给用户控制选项的提议来适应体验的强烈度。第三，包含强烈虚拟体验的游戏应该在游戏中提供给用户相关内容提示，好让用户按照他们觉得最舒适的方式靠近。

打击虚拟幻境头晕

玩家锁定背景（也称为独立视觉背景）

虚拟幻境头晕研究文献提供了至少一种纯粹的方法来减少在VR中的虚拟幻境头晕。实验者把人放在虚拟环境中，这个环境可能包含也可能不包含他们所谓的独立的视觉背景。[15]这构成了一个简单的视觉背景，如栅格或天空盒，这个背景通过模拟器的主要内容是可见的而且和用户在稳定的真实环境中的行为是匹配的。例如，一个驾驶模拟器可能会穿过一个含有地平面，树木和建筑物的环境；然而，天空盒背景，会包含一些云，会在用户的面前保持静止，即使在车翻转的时候也是静止的。[16]带有独立视觉背景的虚拟环境相对于带有典型行为背景的虚拟环境，被发现可以有效减少虚拟幻境头晕。

通过让观看者的大脑建立一个观点就是视觉和前提感官是连贯的，以此来打败通常会导致不适的感官冲突，让用户觉得确实和背景环境是固定的，但前景环境是围绕用户移动的。

我们的特殊实现是用一个玩家锁定天空盒，这个天空盒被放置在离主要环境很远的一个位置，这个主要环境就是玩家导航的地方。在我们的初步测试中，各种背景似乎都是有效的，从现实（海，水平线，和云的天空）到人工（黑色，网格内衬盒）。只要玩家开始使用控制器或键盘在前台环境运动或回转，他们就会发现远处的背景始终保持静止，和他们在真实世界身体的位置是一起锁定的。然而，他们任何时候都可以通过头部运动来看看周围的背景。大致的效果是，玩家会感觉他们是处在由背景创建的一个巨大的“房间”里，并且主前台环境只是单纯得在他们周围移动。

此方法在一系列的技术中被发现在减少虚拟幻境头晕是有效的，而且Rift也不例外。然而，这种方法并非没有其局限性。减少不适取决于两个因素：背景的可见性，以及在何种程度上它被认为是进一步脱离玩家的相对前台环境而言。并不是所有的虚拟环境都是户外的，或是一些拥有稳定可见或直观的玩家锁定背景。并不是所有虚拟环境都是在户外或者某个地方,锁定背景是随时可见并且有意义的。

这些实际的限制促使我们尝试运用我们的网格内衬房间格局到所有虚拟环境作为一个半透明的叠加，利用双眼视差和空间透视（即，雾）作为深度线索，而且网格在距离上间隔远。虽然这种普遍认为有效的，但这将有可能减少用户信任感。此外，我们发现，当定位在眼睛和前台环境之间促使玩家去感知网格的方式（如让网格不透明）没有任何好处。

不过，适当地采用这种方法在对舒适度的影响较小时有希望让开发人员提供更广泛的经验给玩家。此外，这也可以作为一种方法来让用户适应虚拟环境；玩家可以打开被锁定的背景，当第一次进入设定的内容，然后可以选择用时间禁用或衰减影响。如果几乎没有人可以舒适地享受它，即使是最引人注目的VR体验也是无用的；玩家锁定背景可以增加你的用户群，去吸引那些有可能不会使用你产品的用户。如果一种有效的独立视觉背景的形式可以在你的产品中实施，考虑把这个作为一个可配置项。

新方法

开发者已经在寻找方法让传统的视频体验和在VR中的体验一样舒适的都显示在电脑屏幕上。以下是我们已经看到的一些方法的描述。虽然他们可能与你的特定内容不兼容或觉得无效，出于你的考虑，我们都列出来了。

因为运动导致对流，反过来，就会引起不适，一些开发者已经做过试验，通过使用不同方法让玩家在不同的地点进行瞬移来让他们穿过某个空间。虽然这种方法可以有效减少虚拟幻境头晕，但用户会丢失其方位然后迷失方向。[17]

有些试图通过相机的操作来减少用户所体验的对流的数量。另一种采取的“隐形传送”模式使用户跳出第一人称视角变成环境“神模式“的观点通过将玩家的化身嵌在环境里。玩家移动化身到一个新的位置，然后从新的角度返回到第一人称视角。

另一种方法中修改了用户在虚拟环境中的翻转方式。不是顺畅旋转，按住控制器的左边或右边使照相机立即通过一个固定的角度（例如，30°）跳到预期的方向。这样做是为了最大限度地减少对流，当用户在旋转期间暴露。同时，也产生有规律的，可预测的运动以防止迷失方向。

注：所有在本节中所描述的方法都可以降低不适，但前提是要在虚拟环境中制造真实的体验。

依你自己的考虑去采用这任一的方法，但记住，更舒适的产品将吸引更多的用户有可能也更有价值。要达到最真实和最舒适的体验就是把这些方法作为用户可配置项，用户可以选择。

启用或禁用。用户体验到了舒适后可以再选择进入更真实的体验，而且敏感的用户可以启用这些方法来帮助他们享受您的产品。

计量测试

各种各样的技术已经被用于测量和评价虚拟幻境头晕。在更多技术方面，间接测量包括皮肤电反应，脑电图（EEG），胃电图（EGG），和姿势的稳定性，也许是科研文献中最常用的方法，但是，有一个简单的调查：虚拟幻境头晕问卷（SSQ）。

像其他任何的问卷，SSQ承载着一些固有的限制，这些限制来源于人们自我填答时嵌在他们脑子和身体固有的理解力。然而，SSQ也具有许多优点。和间接的，生理的计量不同，SSQ不需要特殊的设备或培训，只是笔纸和一些算法。任何人都可以提供调查表，计算分数，并解释那些基于过去数据的分数。对于受访者，调查问卷是简短的，只需花一分钟时间了解游戏测试。因此SSQ通过在测试者身上花很少的成本收集了有价值的信息，并且是一个潜在的在游戏中寻找舒适感的方式。

[1] Kennedy, R. S., Lane, N. E., Berbaum, K. S., & Lilienthal, M. G. (1993).. 虚拟幻境头晕问卷: 一个增强地方来量化虚拟幻境头晕. 航空心理学国际期刊, 3(3),203-220.
[2] Kennedy, R., Stanney, K., & Dunlap, W. (2000). 长时间接触虚拟环境: 会话中的疾病曲线. Presence, 9(5), 463-472.
[3] Stanney, K. M., Hale, K. S., Nahmens, I., & Kennedy, R. S. (2003). 沉浸在虚拟环境能获取到什么：性别的影响，身体质量指数，和过去的经验。人的因素45(3),504–20.
[4] So, R.H.Y., Lo, W.T., & Ho, A.T.K. (2001). Effects of navigation speed on motion sickness caused by an immersive virtual environment.控制速度来减少虚拟环境下引起的晕动症。 Human Factors, 43(3), 452-461.
[5] Rolnick, a, & Lubow, R. E. (1991). 为什么司机很少晕动? 可控的运动中的角色疾病.人机工程学, 34(7), 867–79.
[6] Lin, J. J., Abi-Rached, H., & Lahav, M. (2004, April). 虚拟的指导头像：一个有效的方式来减少虚拟环境中的虚拟幻境头晕。InProceedings of the SIGCHI conference on Human factors in
computing systems (pp. 719-726). ACM.
[7] Ehrlich, J.A. & Singer, M.J. (1996). 在立体与平面视觉头盔虚拟幻境头晕安装显示器In: Proceedings of the Human Factors and Ergonomics Society 40th Annual Meeting.
[8] Siegel, M., & Nagata, S. (2000). 足够的真实性: Comfortable 3-D Viewing. IEEE Transactions on Circuits and Systems for Video Technology, 10(3), 387–396.
[9] Draper, M.H., Viire, E.S., Furness, T.A., & Gawron, V.J. (2001). Effects of image scale and system time delay on simulator sickness within head-coupled virtual environments. Human Factors, 43 (1), 129-146.
[10] Stoffregen, T.A., Draper, M.H., Kennedy, R.S., & Compton, D. (2002). V前庭适应和后遗症. In Stanney, K.M. (ed.), Handbook of virtual environments: Design, implementation, and applications (pp.773-790). Mahwah, New Jersey: Lawrence Erlbaum Associates, Publishers.
[11] Draper, M.H., Viire, E.S., Furness, T.A., Gawron, V.J. (2001). 通过头配装置在图像规模和系统时间延迟在虚拟幻境头晕的影响。人为因素, 43(1), 129-146.
[12] Kolasinski, E.M. (1995). 虚拟环境的虚拟幻境头晕 (ARTI-TR-1027). Alexandria, VA:陆军研究所的行为与社会科学。从http://www.dtic.mil/cgi-bin/检索 GetTRDoc?AD=ADA295861
[13] Reason, J.T. & Brand, J.J. (1975). 晕动症。学术出版公司。
[14] Welch, R.B. (2002).适应虚拟环境. In Stanney, K.M. (ed.). 虚拟手册环境：设计，实施和应用。劳伦斯Erlbaum协会，出版商：莫瓦，NJ.
[15]Prothero, J.D., Draper, M.H., Furness, T.A., Parker, D.E., and Wells, M.J. (1999). 利用独立的视觉背景降低模拟器副作用。航空，航天和环境医学, 70(3),135-187.
[16] Lin, J. J.-W., Abi-Rached, H., Kim, D.-H., Parker, D.E., and Furness, T.A. (2002). “自然”独立视觉背景降低虚拟幻境头晕。人为因素和人类工效学学会论文集年会上, 46,2124-2128.
[17] Bowman, D. Koller, D., & Hodges, L.F. (1997). 旅游在身临其境的虚拟环境：进行评价视点运动控制技术，“虚拟现实年度国际研讨会，第论文集45-52

---

原文如下

Distortion Correction

The lenses in the Rift distort the image shown on the display, and this is corrected by the post-processing steps given in the SDK. It is extremely important that this distortion is done correctly and according to the SDK’s guidelines and the example demos provided. Incorrect distortion can “look” fairly correct, but still feel disorienting and uncomfortable, so attention to the details is paramount. All of the distortion correction values need to match the physical device—none of them may be user-adjustable (the SDK demos allow you to play with them just to show what is happening behind the scenes, but not because this is a particularly sensible thing to do).

We carefully tune our distortion settings to the optics of the Rift lenses and are continually working on ways of improving distortion tuning even further. All developers must use the official Oculus VR distortion settings to correctly display content on the Rift.

Flicker

Flicker plays a significant role in the oculomotor component of simulator sickness. It can be worsened by high luminance levels, and is perceived most strongly in the periphery of your field of view. Although flicker can become less consciously noticeable over time, it can still lead to headaches and eyestrain.

Although they provide many advantages for VR, OLED displays carry with them some degree of flicker, similar to CRT displays. Different people can have different levels of sensitivity, but the 75-hz display panels of the DK2 are fast enough that the majority of users will not perceive any noticeable flicker. Future iterations will have even faster refresh rates and therefore even less perceptible flicker still. This is more or less out of your hands as a developer, but it is included here for completeness.

Your responsibility is to refrain from creating purposely flickering content. High-contrast, flashing (or rapidly alternating) stimuli, particularly in the 1-30 Hz range, can trigger seizures in people with photosensitive epilepsy. Related to this point, high-spatial-frequency textures (such as fine black-and-white stripes) can also trigger seizures in people with epilepsy.

Experience

The more experience a user has had with a virtual environment, the less likely they are to experience simulator sickness.[14] Theories for this effect involve learned—sometimes unconscious—mechanisms that allow the user to better handle the novel experience of VR. For example, the brain learns to reinterpret visual anomalies that previously induced discomfort, and user movements become more stable and efficient to reduce vection. The good news is that developers should not be afraid to design intense virtual experiences for more experienced users; the bad news is that most users will need time to acclimate to the Rift and the game before they can be expected to handle those experiences.

This has a few important ramifications. First, developers who test their own games repeatedly will be much more resistant to simulator sickness than a new user, and therefore need to test the experience with a novice population with a variety of susceptibility levels to simulator sickness to assess how comfortable the experience actually is. Second, new users should not be thrown immediately into intense game experiences; you should begin them with more sedate, slower-paced interactions that ease them into the game. Even better, you should implement the recommendations in this guide for user-controlled options to adjust the intensity of the experience. Third, games that do contain intense virtual experiences should provide users with warning of the content in the game so they may approach it as they feel most comfortable.

Combating Simulator Sickness

Player-Locked Backgrounds (a.k.a. Independent Visual Backgrounds)

The simulator sickness research literature has provided at least one purely visual method of reducing simulator sickness that can be implemented in VR content. Experimenters put people in a virtual environment that either did or did not contain what they called an independent visual background.[15] This constituted a simple visual backdrop, such as a grid or skybox, that was visible through the simulator’s primary content and matched the behavior of the stable real-world environment of the user. For example, a driving simulator might indicate movement through the environment via the ground plane, trees, and buildings passing by; however, the skybox, containing a few a clouds, would remain stationary in front of the user, even when the car would turn.[16] Using a virtual environment with an independent visual background has been found to significantly reduce the experience of simulator sickness compared to a virtual environment with a typically behaving background.

This combats the sensory conflict that normally leads to discomfort by allowing the viewer’s brain to form an interpretation in which the visual and vestibular senses are consistent: the user is indeed stationary with the background environment, but the foreground environment is moving around the user.

Our particular implementation has used a player-locked skybox that is rendered at a distance farther away than the main environment which the player navigates. A variety of backdrops appear to be effective in our preliminary testing, ranging from realistic (a sea, horizon line, and clouded sky above) to artificial (a black, grid-lined box). As soon as the player begins any locomotion or rotation in the foreground environment with a controller or keyboard, they will notice that the distant backdrop remains stationary, locked to their realworld body’s position. However, they can still look around the backdrop with head movements at any time. The overall effect is that the player feels like they are in a gigantic “room” created by the backdrop, and the main foreground environment is simply moving around them.

This method has been found to be effective in reducing simulator sickness in a variety of technologies, and the Rift is no exception. However, this method is not without its limitations. The sickness-reducing effect is contingent upon two factors: the visibility of the background, and the degree to which it is perceived as further out from the player than the foreground environment. Not all virtual environments will be outdoors or otherwise somewhere that a player-locked background will be readily visible and intuitively make sense.

These practical limitations motivated us to attempt applying our grid-lined room pattern to all virtual environments as a translucent overlay, using binocular disparity and aerial perspective (i.e., fog) as depth cues that the grid is far off in the distance. Although this generally felt effective, this can potentially reduce the user’s suspension of disbelief. In addition, we found that any cues that cause the player to perceive the grid as positioned between their eyes and the foreground environment (such as making the grid opaque) completely abolish any benefits.

Still, employed properly, this method holds promise for allowing developers to provide a wider variety of experiences to players with less impact on comfort. Furthermore, it can also serve as a means of helping users get acclimated to the virtual environment; players might turn the locked background on when first engaging your content, then have the option to disable or attenuate the effect with time. Even the most compelling VR experience is useless if almost no one can enjoy it comfortably; player-locked backgrounds can broaden your audience to include more sensitive users who might otherwise be unable to use your content. If an effective form of independent visual background can be implemented in your content, consider including it as a playerconfigurable option.

Novel Approaches

Developers have already begun exploring methods for making conventional video game experiences as comfortable in VR as they are on a computer screen. What follows are descriptions of a few of the methods we have seen to date. Although they may not be compatible or effective with your particular content, we include them for your consideration.

Because locomotion leads to vection and, in turn, discomfort, some developers have experimented with using various means of teleporting the player between different locations to move them through a space. Although this method can be effective at reducing simulator sickness, users can lose their bearings and become disoriented.[17]

Some variants attempt to reduce the amount of vection the user experiences through manipulations of the camera. An alternative take on the “teleportation” model pulls the user out of first-person view into a “god mode” view of the environment with the player’s avatar inside it. The player moves the avatar to a new position, then returns to first-person view from the new perspective.

Yet another approach modifies the way users turn in the virtual environment. Rather than smoothly rotating, pressing left or right on a controller causes the camera to immediately jump by a fixed angle (e.g., 30°) in the desired direction. The idea is to minimize the amount of vection to which the user is exposed during rotation, while also generating a regular, predictable movement to prevent disorientation.

Note: All the methods described in this section have the potential of reducing discomfort at the cost of producing a veridical, “realistic” experience of the virtual environment. It is at your discretion to implement any of these methods, but keep in mind that more comfortable content will be accessible to more users and may be worth the price. A compromise between an optimally realistic and optimally comfortable experience is including these methods as user-configurable options that can be enabled or disabled. Users who experience less discomfort can opt into the more veridical experience, while sensitive users can enable methods that help them to enjoy your content.

Measurement and Testing

A wide variety of techniques have been used in the measurement and evaluation of simulator sickness. On the more technical side, indirect measurements have included galvanic skin response, electroencephalogram (EEG), electrogastrogram (EGG), and postural stability. Perhaps the most frequently used method in the research literature, however, is a simple survey: the simulator sickness questionnaire (SSQ).

Like any other questionnaire, the SSQ carries some inherent limitations surrounding the validity of people’s selfreported insights into their own minds and bodies. However, the SSQ also has numerous advantages. Unlike indirect, physiological measures, the SSQ requires no special equipment or training—just a pen-and-paper and some arithmetic. Anyone can deliver the questionnaire, compute scores, and interpret those scores based on past data. For respondents, the questionnaire is short and simple, taking only a minute of time out of a playtest. The SSQ therefore provides a lot of informational value for very little cost to the tester, and is one potential option for assessing comfort in playtesting.

[1] Kennedy, R. S., Lane, N. E., Berbaum, K. S., & Lilienthal, M. G. (1993). Simulator sickness questionnaire: An enhanced method for quantifying simulator sickness. The International Journal of Aviation Psychology, 3(3), 203-220.
[2] Kennedy, R., Stanney, K., & Dunlap, W. (2000). Duration and exposure to virtual environments: Sickness curves during and across sessions. Presence, 9(5), 463-472.
[3] Stanney, K. M., Hale, K. S., Nahmens, I., & Kennedy, R. S. (2003). What to expect from immersive virtual environment exposure: influences of gender, body mass index, and past experience. Human factors, 45(3), 504–20.
[4] So, R.H.Y., Lo, W.T., & Ho, A.T.K. (2001). Effects of navigation speed on motion sickness caused by an immersive virtual environment. Human Factors, 43(3), 452-461.
[5] Rolnick, a, & Lubow, R. E. (1991). Why is the driver rarely motion sick? The role of controllability in motion sickness. Ergonomics, 34(7), 867–79.
[6] Lin, J. J., Abi-Rached, H., & Lahav, M. (2004, April). Virtual guiding avatar: An effective procedure to reduce simulator sickness in virtual environments. InProceedings of the SIGCHI conference on Human factors in computing systems (pp. 719-726). ACM.
[7] Ehrlich, J.A. & Singer, M.J. (1996). Simulator sickness in stereoscopic vs. monoscopic helmet mounted displays. In: Proceedings of the Human Factors and Ergonomics Society 40th Annual Meeting.
[8] Siegel, M., & Nagata, S. (2000). Just Enough Reality: Comfortable 3-D Viewing. IEEE Transactions on Circuits and Systems for Video Technology, 10(3), 387–396.
[9] Draper, M.H., Viire, E.S., Furness, T.A., & Gawron, V.J. (2001). Effects of image scale and system time delay on simulator sickness within head-coupled virtual environments. Human Factors, 43 (1), 129-146.
[10] Stoffregen, T.A., Draper, M.H., Kennedy, R.S., & Compton, D. (2002). Vestibular adaptation and aftereffects. In Stanney, K.M. (ed.), Handbook of virtual environments: Design, implementation, and applications (pp.773-790). Mahwah, New Jersey: Lawrence Erlbaum Associates, Publishers.
[11] Draper, M.H., Viire, E.S., Furness, T.A., Gawron, V.J. (2001). Effects of image scale and system time delay on simulator sickness with head-coupled virtual environments. Human Factors, 43(1), 129-146.
[12] Kolasinski, E.M. (1995). Simulator sickness in virtual environments (ARTI-TR-1027). Alexandria, VA: Army Research Institute for the Behavioral and Social Sciences. Retrieved from http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA295861
[13] Reason, J.T. & Brand, J.J. (1975). Motion Sickness. Academic Press, Inc.
[14] Welch, R.B. (2002). Adapting to virtual environments. In Stanney, K.M. (ed.). Handbook of Virtual Environments: Design, Implementation, and Application. Lawrence Erlbaum Associates, Publishers: Mahwah, NJ.
[15]Prothero, J.D., Draper, M.H., Furness, T.A., Parker, D.E., and Wells, M.J. (1999). The use of an independent visual background to reduce simulator side-effects. Aviation, Space, and Environmental Medicine, 70(3), 135-187.
[16] Lin, J. J.-W., Abi-Rached, H., Kim, D.-H., Parker, D.E., and Furness, T.A. (2002). A “natural” independent visual background reduced simulator sickness. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 46, 2124-2128.
[17] Bowman, D. Koller, D., & Hodges, L.F. (1997). Travel in immersive virtual environments: an evaluation of viewpoint motion control techniques,” Proceedings of the Virtual Reality Annual International Symposium, pp. 45-52.