Eike is co-founder and "creative re-director" of Curvature Games, the Germany-based independent VR studio behind The Amusement. Eike has a PhD in computer science and did a lot of research on perception and locomotion in virtual reality before he founded Curvature Games. Besides his activities at Curvature Games, he is Professor at Hamburg University of Applied Sciences where he teaches game development and researches XR interaction.
Locomotion is one of the most fundamental challenges in virtual reality, shaping not only how players move but how they perceive and experience entire worlds. Over the years, developers have explored a wide range of techniques — from teleportation to smooth movement — to balance comfort, immersion, and accessibility. In this article, we take a closer look at an alternative approach: designing VR experiences around real walking in extremely limited physical spaces, and the techniques we used to make that possible in The Amusement.
Locomotion in VR has been a challenge ever since. And there have been interesting and creative solutions ever since, such as Budget Cuts' portal teleportation or the climbing approach of The Climb. And though smooth locomotion is the preferred locomotion technique for some core enthusiasts, a broad range of people get motion sick by continuous movements. So, teleportation became the standard technique.
In academic research, there is a lot of effort to investigate how we can leverage real walking in VR environments. Real walking is theoretically possible when we have 6DOF tracking, so tracking of orientation AND position. In consumer devices, this is basically the case since Oculus DK2, however, HTC Vive really defined what room-scale means: Having a space of some square meters in your real room to walk around with a VR headset. Oculus Quest (1) perfected that by transitioning to wireless.
There have been some great games that really leveraged that space in a creative way, e.g. Eye of the Temple, Tea for God, Unseen Diplomacy or recently Hotel Infinity. All of these games inspired us for The Amusement! However, the origin of our locomotion approach is a bit different.
When I started my PhD work in 2015, redirected walking was already investigated for some years and several redirection techniques have been introduced. Translation gains, rotation gains, and curvature gains are core techniques in redirected walking that subtly manipulate how users perceive movement in VR. Translation gains scale the user’s physical movement, making them walk slightly more or less in the virtual world than they do in reality. Rotation gains adjust how much the virtual world rotates relative to the user’s head turns. Curvature gains go a step further by gently bending a user’s walking path in the real world while they perceive it as straight in VR, helping keep them within a limited physical space.
F. Steinicke, G. Bruder, J. Jerald, H. Frenz and M. Lappe, "Estimation of Detection Thresholds for Redirected Walking Techniques," in IEEE Transactions on Visualization and Computer Graphics, vol. 16, no. 1, pp. 17-27, Jan.-Feb. 2010, doi: 10.1109/TVCG.2009.62.
Another approach is impossible spaces that I would also count as a redirected walking technique. In this case, it's not the player's movement that is manipulated but the virtual environment itself. The environment is structured in ways that could not exist in the real world, allowing multiple areas to overlap or reuse the same physical space. By carefully controlling what the player sees and when, developers can seamlessly transition between these spaces without the user noticing the illusion.
Epplée, R., & Langbehn, E. (2021). “Overlapping Architecture: Implementation of Impossible Spaces in Virtual Reality Games”. in REALITIES, 37.
For me, the goal was to find a smart combination of these techniques, to make real walking possible in very small spaces. In 2018, I showed a very first concept of how this could work at SIGGRAPH. In this demo, impossible spaces, rotation gains and curvature gains were combined. The rotation gains were applied in a situation where the player was forced to turn their head a lot back and forth to solve a task. This way, we could turn the whole environment around them while they did not even notice it.
Eike Langbehn and Frank Steinicke. 2019. Space walk: a combination of subtle redirected walking techniques integrated with gameplay and narration. In ACM SIGGRAPH 2019 Emerging Technologies (SIGGRAPH '19). Association for Computing Machinery, New York, NY, USA, Article 24, 1–2. https://doi.org/10.1145/3305367.3327976
In another prototype from the same time, we could even show that it is possible to re-use physical proxy objects for passive haptic feedback when it is properly aligned with impossible spaces and redirections.
E. Langbehn, P. Lubos and F. Steinicke, "Redirected Spaces: Going Beyond Borders," 2018 IEEE Conference on Virtual Reality and 3D User Interfaces (VR), Tuebingen/Reutlingen, Germany, 2018, pp. 767-768, doi: 10.1109/VR.2018.8446167.
These demos were still designed for a space of 4x4 m. When we started Curvature Games in 2020 and kicked off the development of The Amusement, it was clear that the goal is to reduce this to 2x2 m. We decided to go with rotation gains and impossible spaces (and neglect translation gains and curvature gains for now).
What was really important to us, is that the environment felt natural and that it is not obvious to the player that they are tricked. This is difficult when using impossible spaces, but with proper level design we hope that we achieved it in the end. One fundamental was for example to always show a target in the distance that the player can use to orient themself (like the giant in our maze level).
Another important point is that we do not use too much overlapping geometry at once, but enrich these redirection techniques with a bunch of additional traversal techniques such as swinging, climbing, elevators and pulley platforms. This increases variety.
During the level design process, we used a 2x2m grid that was divided into 9 fields (so each field approx. 0.67x0.67 m). This is the minimum space for a player.
This way, we could always see which areas in a level are walkable and adapt our level geometry as well as our impossible spaces accordingly.
When we wanted to integrate a rotation gain task, for example, we put two grids in the level while one field of each grid needs to overlap. By applying the rotation gains during a head-turning task, the grids are rotated so they finally overlap completely.
Bringing natural-feeling locomotion into a space as small as 2x2 meters was a long and iterative journey, combining research, experimentation, and careful level design. While the techniques behind it are complex, the goal was always simple: create an experience that feels intuitive and seamless to the player. We’re excited about the possibilities this opens up and would love to hear your thoughts and feedback as we continue to explore what’s possible in VR locomotion.