Big VR Is Fragile — Small Vision Tools Are Where the Stack Wins

Schell Games — the veteran VR studio behind I Expect You To Die and Among Us 3D — confirmed it laid off 40% of its staff in an effort to stay afloat. That's the headline, and it's a hard one for anyone who's been in VR for a while.

It's also, quietly, a map of where VR is still working.

The shift nobody announced

When a studio that big trims that deep, the story people tell is "VR is struggling." The more accurate story is narrower: big-budget VR entertainment is struggling. The category that keeps growing is the unglamorous one — small, single-purpose tools that people open on a schedule, not a weekend.

As an engineer, that's the pattern I find interesting. Tentpole games are content pipelines: long production cycles, high burn, brittle revenue assumptions, and a team that has to stay whole until ship. A daily-use tool is a different shape entirely — a narrow surface, a compounding habit loop, and a feedback signal you can actually read from session telemetry instead of launch-week sales.

Vision training is one of those tools. And the reason it works in VR specifically is a hardware fact, not a content strategy.

Why this matters if you have lazy eye

Amblyopia and binocular coordination issues are usually treated in childhood, and a lot of adults are told, in effect, that the window has closed. Meanwhile the thing that actually helps — repeated, structured, engaging visual work — is boring enough that most people quit. That's a retention problem before it's ever a clinical one.

VR fixes the engagement problem almost by accident. A headset is already a two-channel display: two independent eyepieces, two independent feeds. That's a hardware property. Everything interesting in Amblyotube is built on top of it.

The architecture, briefly

Amblyotube, developed by Seven Sports, is recreational and educational software for visual coordination and attention exercises in VR, available on Meta Quest (Quest 2, 3, Pro, 3S). The core mechanism is dichoptic vision training: because the headset has two eyepieces, the app renders a different visual experience per eye while the user watches YouTube-style content. No rigid routine, no physical patch.

It's a nice example of software turning existing hardware asymmetry into a feature rather than a bug to correct.

What's actually in the render path

• Dominant Eye Shader (digital occlusion) — an adjustable blur, contrast, brightness, gamma, or opacity filter applied to the stronger eye, putting it in a "digital timeout." At 100% opacity it's full digital patching, but the primary goal is often partial transparency so both eyes learn to process simultaneously.
• Lazy Eye Sharpener (MFBF mode) — an AI engine identifies human figures (faces, bodies) in the video and applies selective sharpening specifically to the lazy eye's view. Segmentation-driven attention weighting, essentially.
• Controlled Flicker Stimulation — a controlled flicker or "jerk" added to human targets, giving the visual cortex a startle cue to pay attention to the ignored eye.
• Magenta Focus Cue (binocular fusion) — a moving circular cue rendered high-visibility magenta pink in the lazy eye and subtle grey in the dominant eye. Magenta stands out against natural video backgrounds (skin tones, greens, blues), pulling the weaker eye's focus and prompting the brain to merge both signals — practicing binocular fusion and 3D depth perception.
• Wedge Occlusion — a black wedge over parts of both visual fields that eliminates binocular rivalry by forcing each eye to view unique portions of the scene.
• Visual Accents — a soft yellow-green Highlight glow or a distinct red Outline tracing human silhouettes, each with optional "breathing" pulse controls (adjustable Pulse Hz and depth) to maintain attention.

Setup, presets, and a built-in check

On launch the user selects whether the affected eye is Left or Right. This is critical — the wrong choice trains the dominant eye instead, which is the kind of one-line config mistake that quietly inverts the whole therapy.

Condition presets are chosen while wearing corrective glasses if applicable: Mild (cannot read the last 2 lines of an eye chart), Moderate (cannot read the last 3–5 lines), Severe (cannot read 5 or more lines). Picking a preset tunes the default levels for the Dominant Eye Shader and Lazy Eye Sharpener.

There's also an integrated LogMAR vision check: a built-in screening tool that tests each eye individually and both together using standardized Landolt C optotypes at a virtual 3-meter distance across up to 30 trials per eye. A clinical staircase algorithm displays logMAR acuity scores and can automatically update the therapy profile to match the detected severity. From a systems standpoint that's the loop worth noticing — measure, then re-parameterize, without a clinician sitting in the loop for every session.

Interface-wise: a floating Control Panel in VR that can be resized or moved, working with both Meta Quest controllers and hand tracking. There's integrated YouTube search plus a preloaded Catalog of child-friendly videos that play offline without a network connection. Standard playback (play, pause, seek, volume, mute, language selection in English, German, French, Japanese, and more), and a Reset All button to restore default severity settings. Room View (AR mode) streams videos in augmented-reality pass-through, training overlaid on the user's real room, as an alternative to the default virtual room.

Why it beats patching, mechanically

Patching isolates one eye. Amblyotube trains both to work as a team, targeting depth perception and binocular vision. It also avoids the full-scene color distortion of red-blue anaglyph glasses and keeps natural visual colors. And it's designed with teenagers (13+) in mind, who tend to find traditional patching boring or hard to maintain — which is the retention problem every training app eventually runs into.

The practical part

• 30–40 minutes a day is the recommendation. Strict maximum of 1 hour per session to avoid VR eye strain.
• Ages 13+.
• No adhesive patches. No red-blue glasses.
• Amblyotube is a supplementary vision training and assistive tool for binocular coordination and depth perception. It is not a medical cure, a diagnosis, or a replacement for doctor-prescribed treatment — it's meant to complement professional medical guidance, not replace it.

The bigger point

If VR's future is tools instead of tentpoles, then the studios that survive won't be the ones chasing the next blockbuster. They'll be the ones building something a person uses on a Tuesday because it fits their life — and shipping a measurement loop tight enough to prove it's working.

Download Amblyotube on the Meta Quest App Store:
https://www.meta.com/en-gb/experiences/amblyotube/25906906972338493/