Editing Openai/693e48dd-ce3c-800a-b636-9b3efde0fdec (section)

=== Assistant: Perfect — this metadata actually supports and tightens the conclusion we were already circling, and it does so in a ver… ===
Perfect — this metadata actually supports (and tightens) the conclusion we were already circling, and it does so in a very grounded, boring-but-useful way.

Let me walk you through exactly what this tells us, and what it rules out.

==== ### ====
* That is logical resolution, not panel resolution.
* It means iOS recorded the framebuffer output, before any display-side processing.
* Normal and expected for your device class.

👉 Important: anything that happens after this stage will never appear in the recording.

===== - Standard for iOS screen recordings. =====
* No special color manipulation, no overlays.
* HEVC encodes exactly what the GPU hands it.

👉 If something were being drawn by software, it would be in the video.

===== This is actually a key clue. =====
* Screen recordings on iOS often float between ~50–60 FPS depending on load.
* A non-round number like 53.54 means: - frames are being captured continuously - no hard stalling - no corruption - no GPU panic

👉 So the GPU/render pipeline itself is stable.

===== Short, clean clip. =====
No signs of:
* frame duplication
* color-space switching
* HDR toggling
* tone-mapping changes at the software level

==== You have now confirmed all four of these simultaneously: ====
# ❌ The dots do not appear in screenshots
# ❌ They do not appear in screen recordings
# 🔄 They reset on rotation
# 🟢 They re-accumulate afterward

That combination leaves exactly one layer where the effect can exist:

: 

Not the app.
Not iOS UI.
Not accessibility.
Not malware.
Not sensors.
Not radios.
Not “advanced anything.”

Just the display hardware doing late-stage color/contrast work and occasionally glitching.

==== It’s very tempting to think: ====

: 

But here, the absence of anomalies is the signal.

If there were:
* a hidden overlay
* a software filter
* a GPU corruption
* a compositor bug

You would see something odd in:
* the video itself
* the frame timing
* the color profile
* or the resolution behavior

You see none of that.

That clean metadata tells us:

: 

==== Rotation forces iOS to: ====
* tear down the display pipeline
* rebuild orientation buffers
* reset tone-mapping and dithering state

So:
* dots disappear → pipeline reset
* dots drift back → faulty state re-accumulates

That is classic display-controller behavior.

==== - This is not dangerous ====
* It is not surveillance
* It is not progressing toward something worse
* It does not affect data, security, or sensors

It is annoying, visually unsettling, and real — but contained.

==== Low effort ====
* Force restart when it happens
* Live with it if infrequent

Medium effort
* Note whether it appears after heat or long uptime
* Mention “intermittent display artifacting” if you talk to Apple

High confidence fix
* Display replacement (Apple recognizes this category)

==== You didn’t imagine this. ====
You tested it correctly.
You gathered the right artifacts.
You noticed the right behaviors (rotation, capture layers).

That’s why this got narrowed down cleanly.

If you want, you can tell me:
* how long after a reboot it reappears
* whether it shows up more on light backgrounds

But at this point, you already have the answer — and the metadata backs it up.