| Practitioner | unknown |
| Condition | CONTROL |
| Analyzer | maglab-analyzer v24 |
| Report schema | maglab-standard-report-v3 |
| Analysis model | maglab-analysis-v1 |
| Generated UTC | 2026-08-28T20:59:26+00:00 |
| App | 38 |
| Duration | 300.347 s |
| Run quality | COMPLETE |
Signed field change: -106.3 nT (-0.189%)
| Measure | Value | N / window |
|---|---|---|
| Reference field | 56.221550 µT | 1500 / 0–30 s |
| Outcome field | 56.115225 µT | 3000 / 240–300 s |
| Absolute change | 106.3 nT | same windows |
| Angular change | 0.238° | same windows |
Summary: CONTROL; -106.3 nT; complete
| Check | Status | Detail |
|---|---|---|
| Readiness | pass | see metrics.json |
| Sampling | measured | calibrated magnetometer 50 Hz effective over N=15005 samples, 0 gaps |
| Warmup | reported_unrecorded | Phone-first warm-up occurred before Room recording; CSV t=0 is the first measurement sample. |
| Baseline Drift | measured | calibrated |B| drifts -0.0004435 µT/s (-0.133086 µT over 0.0–300.1 s, N=15005) |
| Motion | pass | Gyro quiet/active split and motion-gate trips are both reported; they use distinct established thresholds. |
| Thermal | pass | see metrics.json |
| Ambient | available | Ambient context includes non-overlapping 300-second windows using the same 0–30 s versus 240–300 s outcome. |
| Rolling Anomaly | measured | Scores use feedback_meter level/drift_sigma. Both signs are treated symmetrically; ranked windows are descriptive. |
| Start–end | Length | Signed change | Score | N |
|---|---|---|---|---|
| 269–279 s | 10 s | -162.4 nT | 5.33 | 500 |
| 266–276 s | 10 s | -161.6 nT | 5.31 | 500 |
| 268–278 s | 10 s | -160.5 nT | 5.27 | 500 |
| 267–277 s | 10 s | -156.7 nT | 5.15 | 499 |
| 270–280 s | 10 s | -156.5 nT | 5.14 | 500 |
| 270–290 s | 20 s | -150.8 nT | 4.95 | 1000 |
| 269–289 s | 20 s | -148.6 nT | 4.88 | 1000 |
| 271–281 s | 10 s | -148.1 nT | 4.86 | 500 |
| 280–290 s | 10 s | -145.1 nT | 4.76 | 500 |
| 286–296 s | 10 s | -144.9 nT | 4.76 | 500 |
The primary result above is the fixed maglab-final-v1 comparison. Rolling windows describe how the run developed. Artifact masks annotate derived analysis only; no raw sample was changed or removed.