The cameras were everywhere.
That wasn't unusual.
What mattered was how they were everywhere.
---
Each lens had a purpose beyond storytelling.
Framing choices dictated movement.
Focus pulls tracked distance.
Wide shots mapped crowd density.
"They don't just record scenes," Noah said in her ear. "They record behavior."
Aria adjusted her mark by half a step.
"Every camera is a sensor," she said. "Most people forget that."
---
She watched the monitors between takes.
Live feeds split across angles.
Metadata scrolling beneath the image—timestamps, lens IDs, location tags.
Different vendors.
Different formats.
Same logic.
"They're optimizing coverage," Noah said. "Not for the film."
"For the environment," Aria replied.
---
A steadicam operator followed her through a hallway rehearsal.
Too smoothly.
She slowed.
The operator compensated instantly.
Latency: near zero.
"Not human reflex," Noah muttered.
"Assisted," Aria agreed.
---
She improvised.
Paused unexpectedly.
Turned before cue.
Changed blocking mid-step.
The camera corrected—barely.
Just enough.
The system adapted.
---
Behind the scenes, software flagged the anomalies.
Not as errors.
As data.
"They're learning your movement," Noah said.
Aria smiled faintly. "Then let's teach them something wrong."
---
She repeated the scene.
This time, she followed the blocking perfectly.
No deviation.
No surprise.
The system relaxed.
Confidence rose.
---
Later, she moved again—subtly different.
A fraction slower.
A fraction closer.
Unregistered.
Unmodeled.
Unnoticed.
Cameras were sensors.
But sensors needed calibration.
And Aria was quietly changing the baseline.
One frame at a time.