The reality of virtual reality
A game of trade-offs
The Nintendo Switch 2 launched on June 5, 2025. The console has a similar form factor to the Switch 1 but much better specs. However, even with these improvements, the Switch 2’s hardware specs aren’t that much more powerful than the PS5, which was released 5 years ago. It’s also not that much cheaper: The Switch 2 costs $450, while the PS5 costs $550.
But this all doesn’t seem to matter as the Nintendo Switch 2’s launch was highly successful, with the console selling over 3.5 million units globally in its first four days. This is because Nintendo doesn’t compete on specs or price. It has never played this game since the 80s and always relied on its strengths:
- Ease of use
- Game play
- Portability
- Innovative form factors
- Exclusives (Mario, Zelda, Pokémon)
- Accessibility & Family Appeal
Nintendo doesn’t compete with Sony’s PlayStation or Microsoft’s Xbox when it comes to raw power. Instead, Nintendo takes a different approach, playing to its unique strengths.
In our household, we have both a Nintendo Switch and a Meta Quest 3s. The Quest 3s is almost half the price of the Switch 2, but
when you switch from the Switch to VR, it will blow your mind.
The immersive experience is on a completely different level than traditional games. Also, headsets and games are cheaper on the Quest than on traditional consoles. Honestly, though, after a while, the headset collects dust, and we all play Mario Kart again. Why can’t VR hold people’s attention?
The visual turing test
I think the VR/AR world can learn a lot from Nintendo. Especially the Nintendo of the 80s, because from the beginning they didn’t focused on the most realistic graphics.
We are still far from realistic graphics where we can’t distinguish from reality anymore. Meta calls this the “Visual Turing Test” (sometimes referred to as the “VR Turing Test”). The name is inspired by the original Turing Test proposed by Alan Turing in 1950. Turing’s idea was that a computer can converse with a human in such a way that the human cannot reliably tell whether it’s a machine or a person, then the machine can be said to “think.”
The visual turing test is thus a similar concept to describe
the “ultimate goal” of making virtual reality visuals so realistic that users cannot tell them apart from the real world.
But what are exactly the specs needed to pass the Visual Turing test? Let’s discuss a few of them: PPD, FOV, Nits and refresh rate. What are these specs? Why are they important to VR? And at what level do they pass the Visual Turing test?
PPD
PPD is a key metric to describe how sharp and detailed a VR display looks to your eyes.
- Pixels (display resolution): How many pixels the headset’s screens can show.
- Per Degree (field of view): How many degrees of your vision those pixels are spread across.
So PPD = horizontal display resolution ÷ field of view (in degrees). It tells you how many pixels your eye gets for every degree of vision.
~60 PPD is what you mostly see in news articles as human eye resolution. Though recent research and prototypes suggests it’s a lot higher, above 90 PPD even.
FOV
FOV stands for Field of View. It’s the angular extent of the observable world that you can see through the headset at once — basically, how wide your virtual “window” into the VR world is.
- Horizontal FOV — How wide your vision is from left to right.
- Vertical FOV — How tall your vision is from top to bottom.
Manufacturers often quote the diagonal FOV as a single number, but the actual experience depends on both horizontal and vertical. Human eyes sees about 210° horizontally and 135° vertically.
NITS
Nits is a unit of brightness.
- 1 nit = 1 candela per square meter (cd/m²)
The higher the nits, the brighter the display looks. Brighter displays can make virtual environments look more realistic (sunlight, outdoor scenes). Too many nits too close to your eyes can however cause eye strain.
Small note on black levels. Brightness alone doesn’t guarantee good visuals. OLED headsets, for example, may have lower peak nits than LCD, but offer true blacks which means a better perceived contrast.
Typical Ranges:
- VR headsets: ~100–200 nits (lower than TVs/monitors, since the screens are just centimeters from your eyes).
- Phones/monitors: ~300–600 nits.
- High-end HDR TVs: 1000–2000+ nits.
- Specialized outdoor displays: 2000–5000 nits for sunlight visibility.
If you look directly at the sun, its apparent luminescence is ~1.6 billion nits. However typical daylight is around ~10,000–25,000 nits and a bright summer day ~50,000–100,000 nits.
Refresh rate
The refresh rate shows how many times per second a display updates the image. It’s measured in hertz (Hz). For example: 90 Hz = 90 frames per second per eye. In VR, refresh rate isn’t just about smoothness, it directly affects how natural and comfortable the experience feels.
Thresholds & Human Perception:
- 60 Hz → Barely acceptable on flat screens, too low for VR (causes discomfort).
- 72–80 Hz → Minimum comfort level for VR.
- 90 Hz → Standard baseline today (Meta Quest, Vive, Index). Comfortable for most people.
- 120 Hz → Feels very smooth, noticeably more natural for fast motion.
- 144 Hz+ → Luxury level; diminishing returns begin.
However for true “reality matching” in VR, research suggests ~200–240 Hz would be needed, especially for fast eye/head movements.
There are other visual specs that are also important such Contrast Ratio and the Color Gamut, but the four described are the most common specs you will find in the VR world.
A game of trade offs
So how far are we in headsets on the market from graphics that we can’t distinguish from reality? Here is a table that shows the specs of the Oculus DK1 and the latest Meta Quest 3:
The table shows that we are far from passing the visual Turing test today. However, a lot of progress has been made in the last decade, and more progress is on the horizon. Meta Labs has created several prototypes that pass the Visual Turing Test for specific specifications, such as:
- Prototype 1: “Tiramisu”: 90 PPD
- Prototype 2: “Boba”: 200°×70° FOV
- Prototype 3: “Starbust”: 20,000 Nits
- Prototype 4: “Mirror Lake”: Apparently has a very high refresh rate, but specifics were undisclosed.
The specs of these prototypes are impressive, but they come at a price. Not only are the prototypes too expensive to be sold as consumer products, but they also sacrifice other specs. A large field of view (FOV) means a lower pixel density (PPD), and vice versa. For example, the Tiramisu prototype had an FOV of only 33°. VR reporter David Heaney writes the following about the Starbust prototype:
“The Starbust delivered stunning 20,000-nit brightness, but it was suspended from above with cables because it was so comically big and heavy, and it was so hot that it had enormous cooling fans on top.”
The Boba prototype had an impressive 200° horizontal field of view but only a 70° vertical field of view and a resolution of just 6 pixels per degree (PPD).
It’s not only specs that have to compete with each other, but specs itself need to compete with other aspects, such as form factor (weight and size) and computing power.
With all these trade-offs in the VR market, many product plans disappear in this virtual Bermuda Triangle.
The trade off that tech companies like Meta and Apple need to make show how they are caught in a game of trade-offs.
For example, the Meta Quest has a modest mobile processor. This limits its specs and makes the headset heavier due to the necessary battery. Meta’s competitors are facing similar issues, as evidenced by Apple’s Vision Pro headset (much to Mark Zuckerberg’s relief it seemed). Apple decided to give the headset more computing power and higher specs, such as 35 PPD (compared to the Quest 3’s 25 PPD). However, the M1 processor needs more power, so the battery must be large enough. The headset weighs 650 grams (135 grams more than the Quest 3) and still requires an additional 350-gram external battery.
In 2018, the Chinese company Pimax shipped a headset with a 150° horizontal field of view, but it had a large, heavy form factor. The Bigscreen Beyond 2, released in 2025, weighs just 107 grams but requires an external computer with a powerful GPU. It also requires external trackers. Furthermore, both the Pimax and the Bigscreen Beyond are tethered to a PC.
And last but not least all these choice have a literal cost, the price:
- Meta Quest 3: Starting at $499 USD for the 512GB model.
- Bigscreen Beyond 2: Starting at $1,019 USD for the base model. But needs to add additional trackers and a powerful PC.
- Apple Vision Pro: Starting at $3,499 USD.
Those seeking realistic graphics are bound to be disappointed
All VR and AR headset vendors constantly engage in this game of trade-offs. They are getting better at it over the years though. For example, the Beyond 1 was criticized for its limited field of view, but the Beyond 2 improved in this regard while being lighter and costing the same. Perhaps in the future, Bigscreen will offer higher PPD or built-in inside-out trackers.
Vivo, a Chinese competitor and copycat of Apple’s Vision Pro, seems to offer a similar headset that is 40% lighter than the Apple Vision Pro and costs one-third as much. However, this headset is not yet available and may have a much less powerful processor. Apple is also moving forward. Rumor has it that its updated headset will have more computing power with the M5 processor and be lighter and cheaper.
Meanwhile, Meta is improving its prototypes. They are creating new prototypes that make fewer trade-offs so they can finally be turned into products. The second iteration of the Tiramisu lowered its PPD from 90 to 60 and its nits brightness from 1400 to 700. However, it improved the field of view from 33° to 90°.
The Boba 3 prototype offers a more realistic field of view (FOV) of 180°×120° (instead of 200°×70° (6 PPD) of the first iteration), while maintaining a 30 PPD and the same form factor as the Quest 3. Note that the Boba 3’s weight of 840 grams is notably higher than the Quest 3’s weight (even with a strap) and that it requires a PC for computing.
Many appreciate the boundaries that companies like Meta, Apple, and Bigscreen are pushing. On the other hand, passing the visual Turing test can take years, if not decades.
If price weren’t a factor, they could create a $75,000 headset for big companies, universities, and the military, just as they did with the very first headsets in the early ’90s. Until then, I think most new users are impressed by headsets at first but disappointed later on.
A new game
It’s not all about graphics.
VR headset makers, like Meta and Apple, could take an example from Nintendo instead of being obsessed with the visual Turing test.
Even if Meta and Apple maintain their focus on computing and graphics, Google’s return to the VR arena with its new Android XR operating system may prompt a change. Overall, I applaud the increased emphasis on operating systems, as seen with the introduction of the Vision Pro. Apple had an interesting approach by bringing the 2D world of screens and apps into XR.
However, the Vision Pro missed what makes VR special: immersive experiences. No company has figured out what general computing in VR looks like yet. A new, Nintendo like player in VR could maybe settle down with lower specs, and less compute to focus on:
- Easy of use
- Comfort
- Less friction
- Productive apps
- Entertainment apps
- Protocols
- Less Motion Sickness
- Easier Setup
- Content
- A less fragmented ecosystem
- And so on
Hardware is easy; software is hard. Modems were critical for the internet, but HTML turned it into the web. Phones needed 3G/4G and touchscreens, but apps are what make them engaging.
I guess the big players should focus on protocols, interoperability, and operating systems. That way, smaller players can innovate in terms of ease of use and useful apps. Only then can a true XR (VR/AR) ecosystem exist. The current Horizon OS, for example, has too many quirks and bugs to be accepted by the general public.
Even with perfect specs, VR still struggles with being comfortable, useful, affordable, and socially acceptable in everyday life.
Until these issues are addressed, it will remain a niche product compared to smartphones and consoles. A new player needs to emerge that steps out of the game of trade-offs and creates a new, user-oriented game. There’s no play like it.
