The Frame Before Unlock hero image: a monochrome optical bench scene with a red 656.46 nanometer beam, sensor pipeline frames, and Silicon Gremlin branding.

Silicon Gremlin · Show Me The Seams

The Frame Before Unlock

Deterministic Optical Stress Test

CASE STUDY / 2025.12.13 → 2026.01.10

I was testing whether a narrow-band optical stimulus could produce a predictable projection and color response through an astrophotography filter. The unexpected part was not the red light. It was the preserved screen recording and the frames that appear before recognizable on-device content.

AUTHOR / 0x6A656666 640–660 NM LOCKDOWN MODE SCREEN RECORDING OE1104440734889
00

INTENT

This started as an optics experiment.

The experiment was not “make the phone do something weird.” I was trying to observe whether a 640–660 nm source passed through an Optolong L-eXtreme could produce a repeatable projection behavior and a predictable color response. The optical phenomenon was the point. The recording artifact came afterward.

The hero image uses the Hα reference at 656.46 nm in vacuum as a visual shorthand for the admissible red band. It is there to model the wavelength reference correctly, not to claim that every capture used exactly that value.

The red dots were the experiment. The recording artifact was the part that refused to stay quiet.
Repeated red optical projection captures from the experiment.
Experimental projection captures.
01

OPTICAL PATH

Source → filter → projection → sensor.

At the highest level, the setup was simple: introduce a narrow-band optical source, pass it through the filter path, observe the projection, and see what the sensor preserved. The repeated red-point captures and split projections were the intended outputs. They provide experimental context for what happened later in the recording.

01

Source

A narrow-band red optical input in the 640–660 nm range.

02

Filter

Observed through the Optolong L-eXtreme to constrain the bandpass and projection response.

03

Projection

Looking for repeatable visible structure, including the split projection behavior.

04

Sensor

What the phone ultimately preserved — including the part that became more interesting than the test itself.

02

REPORTED OBSERVATION

The claim was about what the recording preserved.

The observation I cared about was not simply “there was a laser.” It was that the preserved screen recording captured a transition that appears to move from non-standard imagery, through a dark boundary-line state, into recognizable on-device content. That is why the source video matters more than any single still.

REPORT:: OE1104440734889
MODE:: Lockdown Mode
STIMULUS:: 640–660 nm optical input
RECORD:: HEVC screen recording, 04:46, 59.6 FPS
QUESTION:: what capture path produced the pre-unlock imagery?

I am preserving the classification as given: Apple was unable to identify a security issue. This page is not trying to relitigate the label. It is documenting why the recording remains interesting as a privacy and observability question.

03

THE LITERAL FRAMES BEFORE

The transition is the observation.

These are three points from the same preserved 4:46 screen recording. Read them as a sequence, not as isolated stills: a diffuse bright frame, the dark line artifact, and then recognizable screen content. The continuity between them is why the video is more important than the screenshot.

04:35.33

Diffuse / environment-like frame

Frame at 04:35.33 showing a bright diffuse image inside the recorded phone-shaped region.
04:38.75

Dark boundary-line frame

Frame at 04:38.75 showing a mostly dark image with a horizontal line across the recorded region.
04:40.00

Recognizable screen content

Frame at 04:40.00 showing recognizable text and interface content in the screen recording.
CONTINUITY

Same preserved recording

The important part is not that one still looks odd. The important part is that the recording preserves the transition between these states without a frame break.

04

WHY THIS BECAME A PRIVACY QUESTION

If It Wasn’t an Issue, Why Is My Ceiling Fan In The Recording Before Unlock?

The line is intentionally tongue-in-cheek. The concern is not.

This is not a post about being annoyed that an experiment received the wrong label. It is a post about a recording that appears to preserve environmental imagery before the device visibly reaches an unlocked state. That is why I think the artifact is worth discussing publicly even with the original classification left intact.

The mechanism can remain unresolved and the observation can still matter. A privacy question does not become uninteresting just because the final explanation is unfinished.

05

THE RECORD

Preserve the source, not just the reaction.

Press play. Watch the transition.
Then show me the seams.