Source: ~/workspace/rf-detector/INTAKE-013-014-2026-10-31.md — the 60-minute paired run, Sat Oct 31 2026, both phones at the same post, simultaneous windows, both seals COMPLETE. For this demo, Intake #13 stands in as BASELINE and Intake #14 as SWEEP.
| Role in demo | Capture | Device | Observations | Unique addresses |
|---|---|---|---|---|
| BASELINE | Intake #13 | Moto g play 2024 | 42,825 | 409 |
| SWEEP | Intake #14 | Samsung SM-S901U1 | 42,872 | 255 |
Total sightings sit within 0.3% of each other (42,825 vs 42,872): both radios heard the same room at the same rate.
Union: 489 addresses — every distinct address either capture heard. Check the arithmetic: 175 + 234 = 409 (baseline ✓) · 175 + 80 = 255 (sweep ✓) · 409 + 255 − 175 = 489 (union ✓).
Address-resolution asymmetry is not a hearing verdict. In this demo the two sets differ mostly because the two phones *resolve and surface addresses differently* — the intake note says so in plain words, and draws no device verdict from the 409-vs-255 gap. The sighting rates prove the point: 42,825 vs 42,872 observations means both radios heard the same room at the same rate.
The same discipline binds a real hotel sweep: a “new” or “vanished” bucket is a fact about address sets, not proof something arrived or left. Method demo on sealed house data — not a hotel session. Full concept: SPEC-6.md.
Arrival capture = BASELINE. Later capture from the same spot = SWEEP. Diff = new / changed / vanished emitters. Passive receive only — the Detector listens; it never probes, joins, or transmits. The comparison rides the Scope importers that already produced this paired read. No new hardware, no new app.