UNNS LAB · CHAMBER LXV-C2
Ridgecrest 2019 — tenv3 Native Replay schema-locked to LXV-A
ENGINE v1.0.0 · SPEC v1.0.0
EVENTRIDGECREST_CAL_USA_2019
BREAK2019-07-06
PRE_FIT60d
PRE_STEP20d
POST_STEP20d
GUARD2d
SMOOTHING[1,3,7,14,21]
SNR_MIN5.0
RANK_METRIChoriz_mm
N_STATIONS6
⊡
Drop 6 tenv3 fixture files here
CCCC.tenv3.txt · P580.tenv3.txt · P593.tenv3.txt · P594.tenv3.txt · P595.tenv3.txt · P616.tenv3.txt
Waiting for fixture files…
VERDICT
REGIME INTERPRETATION
REGIME A — CSV ARTIFACT
Rank fully stable
Cosine ≈ 1.0 across all w
→ Previous LXV-C failure was adapter/CSV artifact
REGIME B — AMPLITUDE TIE
Rank fragile
Cosine ≈ 1.0
→ Near-fault gradient flattening (degenerate tie)
REGIME C — MULTI-LOBE
Cosine < 0.95
→ Genuine multi-vector rupture field → LXV-D justified
PROVENANCE
STEP ESTIMATES BY STATION
EN DISPLACEMENT VECTORS (w = 1, detrended)
RANK ORDER BY SMOOTHING WINDOW
SMOOTHING INVARIANCE METRICS (vs baseline w=1)
DIRECTION STABILITY