The iPhone Duo's Missing FaceTime Feature Is a Quiet Lesson in How We Treat Vision
MacRumors has confirmed something quietly strange: FaceTime's Eye Contact setting is absent on the iPhone Duo. The feature arrived in iOS 14 and uses machine learning to subtly shift your eyes during a video call, so it looks like you're gazing into the lens instead of at the other person's face on your screen. A forum member noticed their eyes looked "a little off" during a call and started digging. Turns out the setting simply isn't there.
It's a small story. It's also a perfect illustration of how the tech industry relates to eyes: we build elegant software to fake what a well-coordinated visual system does for free.
I've been turning this over for a few days now, because I keep meeting the same assumption in different disguises. We treat vision like a camera problem. A lens focuses light, a sensor records it, a model cleans it up. If the output looks wrong, we patch the output. That framing works beautifully for screenshots. It falls apart the moment the bottleneck isn't the picture, it's the person.
Correcting the picture vs. training the system
Eye Contact is a cosmetic layer. It changes how you appear on a two-dimensional call. It doesn't touch why someone's eyes might genuinely be pointing at different targets — the underlying binocular coordination issue behind amblyopia, lazy eye, and strabismus.
For a lot of people, that gap is the whole story. They've spent years being told the window for training has closed, or that the only tools available are adhesive patches and red-blue glasses that feel like a punishment. Meanwhile the industry's answer to "my eyes don't work together" has been, increasingly, "here's a filter that hides it."
Filters are fine. Training is better. And the difference between them is the difference between a post-processing pass and actually changing what the input can do.
What it actually looks like to train both eyes at once
Amblyotube, built by Seven Sports, is a Meta Quest app that turns ordinary YouTube-style viewing into an active vision session.
You pick which eye is affected, set a severity preset (Mild, Moderate, or Severe) or run a LogMAR-style vision check, then start a video you'd have watched anyway — flat in a virtual room or in Room View AR passthrough.
From there, each eye gets a different job:
- Dominant eye: a gentle, adjustable digital blur softens its view so it stops doing all the work.
- Lazy eye: the AI Sharpener detects human figures and draws them with a bold red outline, a yellow-green highlight, and a steady pulsing flicker — a moving target the weaker eye has to track.
- Optional Wedge Occlusion: part of each eye's view is blacked out, so each eye sees unique portions of the scene. That cuts binocular rivalry instead of fighting it.
- Magenta Focus Cue: a bright magenta marker appears in the lazy eye's view while a subtle neutral grey dot appears in the dominant eye's. When the brain merges them, you get one cohesive 3D image — the actual skill being trained.
No adhesive patches. No red-blue glasses. 30–40 minutes a day, with content you already like.
Why VR is the right container
Vision training fails for boring reasons: it's tedious, it's isolating, and it's hard to do consistently. VR solves the logistics — a headset can deliver two genuinely different images, one per eye, which is exactly what binocular training requires. What it can't solve is content. That's why Amblyotube rides on YouTube-style video rather than a sterile exercise screen. The exercise is the show.
If you've spent any time in the vision-therapy world, you know the graveyard of good intentions here. Patches get lost. Atropine drops sting and blur the wrong eye for hours. Vision-therapy exercises live on paper worksheets that a nine-year-old tolerates for about two weeks. The adherence problem isn't a discipline problem. It's a design problem. Any tool that asks a person to do something hard, every day, for months, has to answer one question honestly: would you actually want to sit down and do this? If the answer is no, the science doesn't matter.
The iPhone Duo story is a reminder that convenience features are not the same as capability. If your visual system never learned to fuse two images, no camera setting will hand you depth perception.
The honest limits of a headset
I want to be careful here, because the temptation with any new tool is to overpromise. A Quest is not a clinic. It has a fixed focal distance, it can't measure the actual alignment of your eyes the way a cover test or a synoptophore can, and it can't tell you whether your specific binocular issue will respond to this kind of training. Amblyotube is a training and assistive tool, not a diagnostic instrument. It doesn't replace an orthoptist or a pediatric ophthalmologist, and it shouldn't try to.
What it can do is something the older tools mostly can't: make daily practice something a person will actually keep doing. That's not a small thing. Most of the outcome variance in vision training, as in most rehab, lives in adherence — not in the cleverness of the protocol. A perfectly designed exercise that gets abandoned in week two produces nothing. A simpler one that survives to month six produces everything. That's the whole engineering argument, and it's the same argument we make about every habit-forming product: reduce friction, ride existing behavior, and make the boring version of the right thing easier than the fun version of the wrong one.
Practical notes
- Recommended session length: 30–40 minutes daily, maximum 1 hour.
- Age: 13+.
- Amblyotube is a supplementary vision training and assistive tool for binocular coordination and depth perception. It is not a diagnosis, treatment, or cure. If you have amblyopia, lazy eye, or any binocular vision concern, talk to an eye care professional first.
Download Amblyotube on the Meta Quest App Store: https://www.meta.com/en-gb/experiences/amblyotube/25906906972338493/
Top comments (0)