The Limits of VR

Every strength of virtual reality has a shadow, and honesty means naming them as clearly as the benefits - because the deepest limit of VR is inseparable from its deepest strength. VR works by *replacing your view entirely* with a computed world; that total substitution is exactly what lets you stand inside the unbuilt building convincingly. But the same substitution means total isolation from the real world and from other people. When you put on a VR headset you are, for practical purposes, gone: you cannot see your desk, your drawings, your colleagues, your own hands unless the system draws them, or the person who just walked in. You cannot glance at a plan and back at the model, sketch a note, point at a screen with someone beside you, or hold the quick, fluid, shoulder-to-shoulder conversation that is the ordinary texture of design work.

This isolation is a real cost, not a detail. Design is deeply collaborative and comparative - you constantly cross-reference drawings, calculations, precedents and each other - and VR walls you off from all of it for the duration. Classic single-user VR immerses one person while everyone else watches a flat spectator feed, which can make a review feel like an audience to someone's private experience and slows the fast back-and-forth of group decisions. Even multi-user VR, where people share the virtual space as avatars, cannot fully restore the ease of bodies in a real room with real drawings on a real table. And there is a plain safety dimension: a person in a headset is blind and deaf to the actual room, its furniture, its edges and its other occupants, so someone must keep them safe.

The honest reading is not that isolation disqualifies VR - the immersion it buys is genuinely valuable for the specific tasks earlier lessons defended - but that isolation is the *price* of immersion, and you should pay it deliberately, only when the immersive value is worth being cut off from the real world and the room. For a focused, well-run review or a client walkthrough, it is often worth it. For everyday design work - drawing, detailing, coordinating, the constant cross-referencing and quick conversation - being sealed inside a computed world is a serious handicap, and this is the first and most structural reason a screen so often wins: the screen keeps you *in* the room, with your drawings, your team and your other tools all in view at once. VR's power to take you somewhere else is also its inability to keep you here.

The second family of limits is bodily, and it is not a minor inconvenience. Motion sickness - VR sickness or cybersickness - affects a significant share of people, from mild unease to genuine nausea, and it comes from a fundamental sensory conflict: your eyes see motion (you glide down a corridor) while your inner ear feels stillness (you are standing in a room), and the brain, which evolved to treat that mismatch as a sign of poisoning, responds with queasiness. Some people are barely affected; others feel ill within seconds of smooth locomotion. It is not something users simply get over, and it is not their failing - it is a property of the medium, softened but not cured by teleport locomotion, comfort vignettes and modest speeds. For a designer this means you cannot assume everyone in the room can comfortably use VR, and you must design and plan around it.

Alongside sickness sits plain fatigue. Current headsets are heavy on the face, warm, and pressed against the head; the optics can strain the eyes; and a subtler issue, the vergence-accommodation conflict, means your eyes converge on virtual objects at various apparent distances while always focusing on a screen at a fixed distance, which tires them over time. The honest consequence is that comfortable VR use is measured in minutes, not hours - most people want to come out after ten to twenty minutes - and the dream of designing all day in a headset simply does not survive contact with the current hardware. All-day comfortable kit does not yet exist anywhere. So VR is a tool for punctuated, focused sessions, not a sustained working environment, and any plan that assumes otherwise is planning against the hardware.

These bodily limits interact with everything else in the module. They cap the length of a design review and a walkthrough, forcing you to plan tight, focused routes rather than open exploration. They constrain who can participate - a client prone to motion sickness may not be able to use the very tool meant to help them. They raise the accessibility question honestly: VR is not equally usable by everyone, and the health, comfort and safety of any headset are matters for the manufacturers' guidance and appropriate professionals, not for a designer's optimism. None of this cancels VR's value, but all of it means immersion is bought with real bodily cost, which is one more reason it must earn its place task by task rather than becoming a default - and one more reason the fast, comfortable, all-day screen remains the workhorse of design.

The third family of limits is about *fidelity and effort*, and it punctures a specific overconfidence: the belief that because VR looks immersive, what you see in it is reality. It is not. Materials, light and finishes in VR are approximations, sometimes good ones, but not the real thing. Real-time rendering - the technology that keeps the image updating fast enough to feel present - trades accuracy for speed; it simplifies how light bounces, how materials scatter it, how colours read under real conditions. So judgements that depend on true material and light behaviour - exact colour, the play of daylight through the day, glare, the subtle quality of a finish - are indicative at best in VR and must be verified with real samples, mock-ups, physically accurate offline renders and, ultimately, reality. Feeling that a space is the right *shape* in VR is trustworthy; concluding that a marble is the right *colour* from VR is not.

And the immersion that makes VR compelling is not free: preparing a model for good VR is real, skilled work. The coordinated BIM model that holds the design is usually too heavy and too detailed to run smoothly in real time, so it must be exported, cleaned, optimised, re-scaled correctly, given real-time materials and lighting, and tested for comfort - a pipeline this course devotes a later module to. Skip or skimp that work and the result is the opposite of helpful: a laggy, mis-scaled, ugly, nausea-inducing experience that actively harms the design conversation and the client's confidence. The effortless-seeming walkthrough is the product of unglamorous preparation, and a practice that underestimates it will be disappointed.

Taken together, these fidelity-and-effort limits reframe what VR is honestly for. It is an excellent instrument for *spatial* questions - size, proportion, height, sightline, sequence, enclosure - because those survive the approximation; the shape of the space is what VR renders faithfully. It is a poor and misleading instrument for *material and photometric* questions - exact colour, real light, finish quality - because those are exactly what real-time rendering approximates away. And it always carries a preparation cost that must be weighed against the value of the immersive session. Knowing this keeps you honest in two directions at once: trusting VR for the spatial judgements it is genuinely good at, and refusing to trust it for the material, photometric and dimensional facts that live with samples, mock-ups, accurate renders, the verified model and the specialists - never in the felt vividness of the headset.

Put the limits together and a clear, usable rule emerges for choosing the medium: immersion must earn its place against a screen and a physical model, task by task - and often it does not win. Reach for VR when the design question is genuinely about experiencing space at true scale - a review of felt spatial quality, an immersive walkthrough of a sequence, an experiential decision about height or sightline, communicating a whole space to a client who cannot read drawings. In those cases the immersion is worth the isolation, the bodily cost, the approximation and the preparation. Reach for a screen for the enormous majority of design work: drawing, detailing, documenting, quick edits, calculations, precise measured questions, fast collaborative back-and-forth, and anything where you need your drawings, tools and team all in view at once. The screen keeps you in the room, comfortable, all day, precise and shared - which is why it remains the workhorse and VR the specialist.

And do not forget the physical model, which for some purposes still beats both. A real scale model is shared effortlessly - a whole team and client stand around it, point, and discuss together with none of VR's one-at-a-time isolation - it is instantly graspable, needs no hardware or preparation, causes no motion sickness, and is superb for reading massing, urban context and the overall composition of a scheme from the outside. Where VR excels at putting one body *inside* a space at a time, the physical model excels at letting many people read a whole object *from outside* together. Neither replaces the other, and a thoughtful designer keeps all three - screen, physical model, VR - and chooses by the question.

That is the honest close of the module. VR is genuinely powerful for a specific, defensible set of design tasks, and this module has defended them without flinching - the review, the walkthrough, the felt decision. But it is isolating, bodily taxing, materially approximate, effortful to prepare, and comfortable only in short bursts, so it is a specialist instrument, not a new default, and the field's hype that it will replace the screen or the studio has repeatedly failed to arrive. The competent, spatially literate stance is neither evangelism nor scorn but disciplined choice: know exactly what VR is superb at, know exactly what it is bad at, keep the screen and the physical model in easy reach, make immersion earn its place every time - and never, in any medium, mistake what you feel in a headset for the verified truth that lives in the model, the drawings, the samples, the engineers and the codes.