Merge pull request 'v0.27.0 Decoder fixes, offset exploration and testing.' (#34) from dev into main

Reviewed-on: #34

CHANGELOG.md

@@ -8,6 +8,395 @@ All notable changes to seismo-relay are documented here.
 
 ---
 
+## v0.27.0 — 2026-08-28
+
+**Per-sample decoder verification at scale, plus the offset investigation.**
+The series-3 codec is now verified sample-by-sample against **14,338** preserved
+Blastware ASCII exports — 1,249 waveform and 13,089 histogram, spanning 45 units
+and files back to 2018. That is 11x the ground truth the production store
+carried, and it found one real codec bug (below).
+
+### Fixed
+- **Sub-minute histograms with a partial final block decoded to nothing**
+  (`histogram_codec.detect_multi_interval_stride`). The stride search confirmed
+  itself on a third block header whenever the body was long enough to hold one —
+  but a body can exceed two strides and still contain only two real blocks, because
+  a *partial* final block leaves trailing padding. BE18193 `T193L0XM.CI0H` (51
+  intervals at 2 s = one full 30-interval block plus a 21-interval remainder, in a
+  2787-byte body) therefore had its correct stride of 612 discarded and produced an
+  empty decode. A missing third header now means end-of-stream rather than
+  disqualification; the block-counter check, which is what actually prevents the
+  false positives that once mis-dispatched 9,082 files, is unchanged.
+
+  Found by decoding the full DL2 archive against its preserved Blastware ASCII
+  exports. Across **63,535 unique** histogram binaries the fix recovers **4 files** —
+  `K440HJCN.3C0H` and `K557IF1U.8K0H` (stride 252), `T191HVNP.0S0H` (92) and
+  `T193L0XM.CI0H` (612) — with **zero** files regressed. Verification over all
+  14,340 archive pairs goes 14,337 → 14,338 exact, the only remainder being two
+  series-4 IDF files that belong to a different codec.
+
+  (The DL2 export keeps a byte-identical `Sent/` mirror of its root, so a naive
+  walk double-counts every binary — 127,035 paths are 63,535 distinct files. The
+  ASCII exports are *not* mirrored, so the 14,340 pair count is already distinct.)
+
+  ⚠ Prod stores hold `.h5` files generated before this fix. Those 4 events stay
+  empty until `backfill_sidecars.py` is re-run — not worth a two-hour prod backfill
+  on its own; fold it into the next one.
+
+- **Histogram/waveform twin matching is now interval-based** (`find_twins`). A real
+  trigger is recorded twice — as a triggered waveform (stamped at the trigger instant)
+  and inside the scheduled histogram whose interval contains it (stamped at the 7am/7pm
+  interval start) — so the two twins can be **hours apart**. The old ±5-minute window
+  silently missed them, which broke review propagation (flagging one twin didn't flag its
+  twin). Twins are now matched by same serial + identical `peak_vector_sum` + opposite
+  record type + the waveform falling within the histogram's interval (bounded by the next
+  same-serial histogram). `window_seconds` is retained but ignored. Fixes terra-view #102
+  sub-task 2.
+
+- **`/health` reported a hard-coded `0.1.0`** instead of the real service version.
+  `sfm/server.py` now derives its version from `minimateplus.event_file_io.TOOL_VERSION`,
+  making that constant the single source of truth for the service version and the
+  sidecar stamp alike — one place to bump at release.
+
+- **`CLAUDE.md` had 793 NUL bytes appended** after its last line, which made `grep`
+  treat the file as binary and silently skip it. Present since at least v0.21.0.
+  Stripped.
+
+### Added
+- **`docs/offset_investigation.md`** — a dated journal of the "offset" hardware
+  fault: base rate, detector design, per-unit case files, ruled-out hypotheses
+  (each kept with the evidence that killed it), and Instantel's own autozero
+  procedure with its 2027–2069 acceptance window.
+- **`scratch/verify_against_ascii.py`** — decodes a corpus of BW binaries and
+  diffs every sample against the paired `_ASCII.TXT`. Includes a saturation
+  carve-out: BW clamps clipped events to the range maximum and writes `OORANGE`,
+  while the decoder faithfully reports counts past nominal full scale.
+- **`scratch/offset_scan3.py`** — offset detector. Measures the resting floor in
+  the *pre-trigger* window (definitionally quiet) and requires it to hold across
+  pre / middle / end. Result: **5 of 45 units (11%)**, stable across a 2x
+  threshold range. Supersedes `offset_scan.py` and `offset_scan2.py`, both kept
+  as the reasoning trail.
+
+### Verified
+- **19,244 healthy channel-events sit at a pre-trigger floor of exactly 0.000
+  (62.7%), 94.5% within ±1 quantisation unit, median +0.0000.** No systematic
+  zero-point bias in the decoder — an independent confirmation of the
+  32000-count geo full scale, arrived at from a different direction than the
+  ASCII sample comparisons.
+
+---
+
+## v0.26.0 — 2026-08-27
+
+**Series-3 decode correctness.**  Two body-model rewrites, a systematic
+scale error affecting every geophone reading ever produced, a recovered
+file format, and two artifact-hygiene bugs where stale files outlived the
+decodes that made them.  All 11,603 series-3 binaries in the production
+snapshot now pass every check.
+
+### Fixed
+
+- **Series-3 health sweep: 11,603 / 11,603 binaries now clean on every check.**
+  Swept every series-3 file with the live decoder against five independent
+  checks — decode exceptions, zero samples, unequal geo channel lengths, peaks
+  above range full scale, decoded peak vs the device-reported PPV, and waveform
+  length vs the declared record time. Three real defects surfaced and were
+  fixed:
+
+  - **`block[22]` is not a constant and must not be tested.** It was documented
+    as always `0x00` but carries data on loud blocks, and rejecting those threw
+    away the interval holding the event peak.
+    `BE18350/T350L7HR.NL0H` block 92 has `block[22]=0x26` and a Tran peak of
+    `0x0563` = 1379 counts = **6.895 in/s** — exactly the device-reported PPV —
+    while the file as a whole decoded to 0.015 in/s. `block[0]==0x00`,
+    `block[4]==0x0A` and the 4-byte tail are six bytes of constraint, which is
+    what keeps trailer content out.
+
+  - **Block-model dispatch now goes on signature strength, not on whichever
+    decoder returns first.** A multi-interval body also yields scattered
+    standard-tail blocks by coincidence; dispatching on "first non-empty"
+    handed 193 BE18193 files to the standard walker and produced peaks of
+    149 in/s against a 10 in/s full scale.
+
+  - **Multi-interval stride detection requires the block counter to increment
+    by exactly 1.** Without it the detector false-positives on ordinary
+    standard-block bodies: those carry a header every 32 bytes, and
+    `192 = 12 + 20×9` and `512 = 12 + 20×25` are both multiples of 32, so a
+    stride "fits" while actually skipping 6 or 16 real blocks. That misrouted
+    9,082 files.
+
+  Partial-block garbage is now trimmed within the final block only, stopping at
+  the first slot with a non-zero tail word or a geo peak above full scale.
+  Trimming purely from the end left garbage stranded behind one slot that
+  happened to have a zero tail word; trimming on the tail word alone truncated
+  four BE9440 files by up to 2,800 intervals.
+
+- **Sub-minute histogram intervals are packed several to a block — 415 files
+  recovered.** The device always writes one minute of data per block, so a
+  shorter interval just means more intervals in a longer block:
+
+  | interval | intervals/block | stride |
+  |---|---|---|
+  | 1 min | 1 | 32 (the standard block) |
+  | 15 s | 4 | 92 |
+  | 2 s | 30 | 612 |
+
+  `stride = 12 + n * 20`. Each 20-byte record carries 8 × uint16
+  **little**-endian values — peak and half-period per channel — plus a 2-word
+  tail whose first word is `0000` on every real interval (a session ending
+  mid-block leaves buffer garbage in the remaining slots, which decoded as
+  peaks thousands of times the real value until that check was added).
+  **The standard 32-byte block is big-endian; this variant is not.**
+
+  These 415 files (216 on BE18193 at 2 s intervals, 199 on BE9440 at 15 s)
+  previously decoded to nothing at all — and before that were being accepted
+  by the *waveform* codec, which returned garbage peaking up to 400× the
+  device-reported PPV.
+
+  Ground truth `BE9440/K440L3AQ.T70H` — 5,710 intervals — matches its
+  Blastware ASCII export on **17,130/17,130** geo peaks, **22,840/22,840**
+  frequencies and **5,710/5,710** mic dB(L) values. Across all 455 affected
+  files, **1,354/1,365 (99.2%)** channel peaks match the device-reported PPV;
+  the 11 that don't are under-reads on BE9440 where the walk stops early.
+
+- **`backfill_sidecars.py` now removes a stale `.h5` when nothing decodes.**
+  It previously skipped the write "so we don't replace whatever's there with an
+  empty placeholder", which silently preserved output from a superseded
+  decoder. After the record-chain fix, 415 histogram files stopped decoding (an
+  unmapped block variant on BE18193 and BE9440) but kept `.h5` files whose peaks
+  ran up to **400× the device's own reported PPV** — garbage feeding the charts
+  and the false-trigger detector with nothing marking it. Reports a
+  `stale_h5_removed` count.
+
+- **The series-3 waveform body is a RECORD CHAIN, not a tag stream — this
+  supersedes the segment-header model, including the fixes made earlier the
+  same day.**
+
+  Records are self-delimiting. `off+2` is a `uint16 BE` length and
+  `next_record = off + 2 + len`; the chain ends on a record whose `chan_id` is
+  `0x06`. `off+8` carries a 3-valued mode enum:
+
+  | mode | header | data section |
+  |---|---|---|
+  | `02 00` | 14 B | anchors, then **cumulative deltas** |
+  | `01 00` | 10 B | no anchors, **absolute** values |
+  | `00 03` | 10 B | **no tags at all** — raw 12-bit packed absolute |
+
+  **`40 NN` is an ordinary int16 BE data block** (`2*NN + 2`), never a segment
+  header. Reading it as a `2*NN + 16` header is what made walks drift — and the
+  "variable-prefix segment descriptors" reported earlier today were not a format
+  feature at all, just walker drift of exactly
+  `4 - (old_stop - true_record_start)` on all 25 affected files.
+
+  Measured against the production snapshot:
+
+  | | before | after |
+  |---|---|---|
+  | all four channels equal length | 156 / 1388 | **1388 / 1388** |
+  | ASCII sample-count exact | 72 / 75 | **75 / 75** |
+  | ASCII fully exact | 70 / 75 | **73 / 75** |
+  | device PPV, waveform (live decode) | 1288 / 1306 | **1306 / 1306** |
+  | device PPV, histogram (live decode) | 4434 / 4459 | **4458 / 4459** |
+
+  Mean absolute PPV ratio error on waveforms is now 0.00000. The 2 remaining
+  ASCII imperfections differ by exactly 1 LSB on samples sitting at the
+  ±10.000 in/s rail.
+
+  **This also eliminated the walker-over-read class.** 24 of those 35 files
+  were histograms that `read_blastware_file` fed to the *waveform* codec first;
+  the old walker accepted them and returned garbage (one yielded 98,923
+  "intervals"), while the record-chain decoder correctly returns `None` so they
+  fall through to `histogram_codec`.
+
+  `00 03` records are decoded rather than skipped. Skipping them does not merely
+  lose samples — it silently shifts the time base of everything after them on
+  that channel (observed on `BE9558/K558LOF2.820W`, MicL displaced by exactly
+  512 samples with nothing marking the gap).
+
+  Footer detection now prefers whichever `0e 08` candidate yields a chain
+  terminating on `0x06`, since the signature can occur inside a sample stream.
+  Blast radius: 1 file of 1,388.
+
+  The superseded model is retained as `decode_waveform_legacy` and pinned by
+  `micromate/idf_file.py`, whose Thor IDFW body-offset search trial-decodes
+  candidate offsets and keeps whichever yields the most samples — the new
+  decoder correctly returns `None` where the old one returned garbage, which
+  changes that heuristic's winner. Switching Thor over is deferred until that
+  search is reworked to use the record chain directly.
+
+- **Series-3 histogram block is uniformly big-endian, and the stream's final
+  block has its own tail — the codec was clipping large peaks and dropping the
+  last interval of nearly every histogram.**
+
+  - **Peaks and half-periods are `uint16` big-endian**, not `uint8` plus an
+    "annotation" byte: `T_peak` `[5:7]`, `T_halfperiod` `[7:9]`, `V_peak`
+    `[9:11]`, and so on.  Only `block_ctr` `[2:4]` is little-endian.  The old
+    model silently **clipped any peak above 1.275 in/s** — the final interval
+    of `BE18193/T193LQ9K.OE0H` reads 8.270 in/s in Blastware's own export and
+    decoded as 0.590.  The "annotation" byte was the half-period's high byte,
+    which is why it was non-zero exactly on the sub-Hz intervals BW renders
+    as `<1.0`.
+  - **The marker is `block[4]` alone.**  Testing `[4:6]` as a `uint16 LE`
+    marker forced `block[5] == 0` — which is what capped the peak at one byte
+    in the first place.
+  - **The last block of each stream carries tail `9c 06 00 42`** instead of
+    `1e 0a 00 00`, with arbitrary bytes at `[21:23]`.  Rejecting it dropped
+    the final interval of nearly every histogram, and that interval is
+    frequently the one holding the event peak — so the file's reported PPV
+    came out low.
+
+  Verified against **1211 production histograms** paired with their preserved
+  Blastware ASCII exports, which carry a full per-interval data table:
+  **1211/1211 now decode exactly** (interval count plus every per-interval
+  peak), and 842,442 per-interval frequency comparisons match with zero
+  mismatches.  Before this fix: **1 of 1196**.
+
+  `decode_histogram_body_full` records now expose `is_terminal` in place of
+  the removed `annotations` tuple.
+
+- **Geophone full scale is 32000 ADC counts, not 32768 — every geo reading was
+  2.3% low.** The verified body codec emits geo samples in 16-count units whose
+  documented LSB is exactly 0.005 in/s, and `decoded_to_adc_counts` multiplies
+  by 16, so one ADC count is `0.005/16` in/s and Normal range (10.000 in/s) is
+  `10.0 / (0.005/16)` = **32000** counts. Both `sfm/event_hdf5.py` and
+  `minimateplus/event_file_io.py` divided by 32768, scaling every geophone
+  sample and every derived peak down by `1 - 32000/32768` = **2.34%**.
+
+  Measured against 216 per-channel comparisons with preserved Blastware ASCII
+  exports: **32768 → 151/216 exact** (worst error 0.238 in/s on a 10 in/s
+  event); **32000 → 216/216 exact**, worst error 0.005 in/s (exactly 1 LSB —
+  pure quantization). The error scales with amplitude, so it was invisible on
+  quiet events and worst on the loud ones that matter for compliance.
+
+  The mic path is unaffected — it back-solves its own per-count factor from the
+  device-reported peak.
+
+  **Scope:** the scale lives in `_samples_to_float`, which every event passes
+  through regardless of which codec produced the samples — so this affected
+  **waveforms, histograms and series-4 (Thor IDF) alike**, not just waveforms.
+  Verified after regeneration: series-3 histogram peaks vs their ASCII reports
+  now sit at a median ratio of 1.0000 across 1,137 comparisons (0.9766 under
+  32768); series-4 peaks vs device peaks moved from a median 0.960 to 0.983
+  across 1,468 comparisons.  The four block-framing fixes below are
+  waveform-only — histograms decode via `histogram_codec.decode_histogram_body`,
+  which is untouched.
+
+- **Series-3 waveform codec: four block-framing cases caused silent channel
+  truncation.** `walk_body` hit its unknown-tag `break` mid-stream and every
+  channel decoded after that point came out short — typically Vert/Long/MicL,
+  sometimes at a third of their true length, with no error raised.
+
+  - **Wide-NN RLE `0X NN`** — the 12-bit NN encoding already handled for
+    `1X NN` / `2X NN` also applies to the `00 NN` RLE tag. Runs longer than
+    252 samples must use the wide form (e.g. `01 0c` = 268 repeats).
+  - **`30 NN` with NN > 0x10** — the `0 < NN <= 0x10` guard was arbitrary;
+    data-section `30 NN` blocks reach at least NN = 0x18. The length formula
+    (`NN × 1.5 + 2`) was already correct.
+  - **Variable-width `40 NN` segment headers** — NN is the *count of
+    previous-channel continuation deltas*, so the header is `2 × NN + 16`
+    bytes and every field after the deltas shifts by `2 × NN`. Only `40 02`
+    (20 bytes) was handled; `40 01` (18) and `40 03` (22) both occur.
+  - **Tagless segment headers** — a segment header can appear with no
+    `40 NN` tag at all: just the 14-byte tail
+    `[field2:2][len:2][channel_id:4][marker:2][anchors:4]`. This is the NN=0
+    case (no continuation deltas needed, so no tag and no delta bytes). It is
+    where the walk stopped in 7 of the 8 events still truncating after the
+    first three fixes.
+
+### Changed
+
+- **Segment channel now comes from the header's own channel-id byte** rather
+  than from rotation position. The field previously documented as a
+  "monotonic uint32 LE counter" is really `[channel][00][00][segment_index]`
+  with `0x46`=Tran `0x47`=Vert `0x48`=Long `0x49`=MicL — verified on
+  **1697 of 1697** segment headers across the ground-truth corpus with zero
+  disagreements. Rotation-by-position is kept only as a fallback for unknown
+  ids; it was fragile because a single missed or extra header (exactly what
+  tagless headers caused) desynced every channel after it.
+
+- **`parse_segment_header` return shape** — now `n_prev_deltas`,
+  `prev_deltas`, `marker`, `anchors`, `channel`, `segment_index` in place of
+  the fixed-offset `anchor_bytes` / `fixed_pattern` / `tail` keys. The old
+  `fixed_pattern` (`02 00 00 01`) conflated the 2-byte constant marker with
+  the first anchor. `counter` is retained as the raw uint32 of the id field.
+
+### Verification
+
+Against the 75 ground-truth events (BW binary paired with its preserved
+`_ASCII.TXT` export), decoding end-to-end through the production path:
+
+| | before | after |
+|---|---|---|
+| exact (full length, within 1 LSB) | 37 | **72** |
+| truncated | 23 | **3** |
+| full length, value error > 2 LSB | 15 | **0** |
+
+Worst remaining error among the 72: 0.0050 in/s = exactly 1 LSB.
+
+No regressions — the byte-exact fixture suite still passes, and the full-suite
+failure list is unchanged from baseline (16 pre-existing failures from
+gitignored fixtures).
+
+### Notes
+
+- **The "DC offset" symptom is _not_ a decode bug.** Events whose geo trace
+  sits at a constant level instead of oscillating around zero
+  (dominant-axis `|mean| / peak` >> 0) reproduce *exactly* in Blastware's own
+  ASCII export — e.g. `BE12599/N599LQD7.8E0W` Tran reads mean +0.345,
+  min +0.335, max +0.355 in both. It is a known recurring hardware fault (the
+  operators call it an "offset"): the affected channel's baseline exceeds the
+  unit's own geo trigger level, so the unit retriggers continuously and floods
+  the ACH queue with garbage events. Store-wide it affects 2 units of 21 across
+  6 episodes; see `scratch/offset_candidates.csv` and the project memory notes.
+
+- ~~**Still open:** 3 of 75 ground-truth events truncate at a segment-header
+  variant with a variable-width prefix.~~ **Resolved later the same day** — the
+  record-chain rewrite (above) showed there is no variable prefix; it was
+  walker drift. All 75 are now sample-count exact.
+
+- **Still open after this release:**
+  - **Series-4 (Thor / Micromate) is not verified** — UM-series sits at ~48%
+    against device peaks with a ~1.7% systematic bias and a near-zero tail.
+    Thor IDFW is pinned to `decode_waveform_legacy` deliberately.
+  - **14 sensitive-range files** show a decoded/truth ratio of exactly 8.0
+    (= 10.0/1.25) — a units bug, not a codec one. Never chased.
+  - **`backfill_sidecars.py --force` also inserts DB rows** for store files
+    that have none (1,286 on the snapshot; one-time per store), and the
+    dry-run does not report that count before you commit to it.
+  - **Verification is uneven:** per-sample proof on the 11% of files with a
+    preserved `_ASCII.TXT`, peak-and-structure consistency on the other 89%.
+
+---
+
+## v0.25.0 — 2026-08-25
+
+**reviewed_real 3-state review flag + twin review-propagation.** The
+`events` table and the `/db/events` feed now carry `reviewed_real`, a
+3-state review flag mutually exclusive with `false_trigger`, mirrored from
+the sidecar review block — plus histogram/waveform twin review-propagation
+(flagging one flags both, matched by serial + identical PVS + timestamp
+window).
+
+### Added
+
+- **`events` column** `reviewed_real` — `INTEGER NOT NULL DEFAULT 0`, added
+  via the existing incremental `_migrate` ADD COLUMN pass (auto-migrates on
+  `SeismoDb()` construction, no manual migration). `query_events` /
+  `get_event` (and thus `/db/events`) return it automatically (`SELECT *`).
+- **Mutual exclusivity with `false_trigger`** — setting `reviewed_real=1`
+  clears `false_trigger`, and vice versa, enforced on both review paths:
+  the sidecar review PATCH (`update_event_review`) and the quick
+  `PATCH /db/events/{id}/false_trigger` endpoint (`set_false_trigger`).
+- **`find_twins`** — matches an event's histogram/waveform twins by serial +
+  identical peak-vector-sum + a timestamp window.
+- **`propagate_review_to_twins`** — copies an event's `false_trigger`/
+  `reviewed_real` state onto its twins, wired into both the
+  `PATCH /db/events/{id}/sidecar` review path and the quick
+  `PATCH /db/events/{id}/false_trigger` path, so flagging one flags both
+  regardless of which endpoint made the change.
+
+---
+
 ## v0.24.0 — 2026-08-22
 
 **Waveform-shape metrics on events.** The `events` table and the `/db/events`

CLAUDE.md

@@ -2,7 +2,49 @@
 
 Ground-up Python replacement for **Blastware**, Instantel's Windows-only software for
 managing MiniMate Plus seismographs. Connects over direct RS-232 or cellular modem
-(Sierra Wireless RV50 / RV55). Current version: **v0.21.0**.
+(Sierra Wireless RV50 / RV55). Current version: **v0.27.0**.
+
+---
+
+## Where things stand (updated 2026-08-28)
+
+Read this first when picking the project back up.
+
+- **Series-3 decode is verified per-sample at scale (v0.27.0).** The full DL2
+  archive decodes **14,338 / 14,338** paired files exactly against their
+  preserved Blastware ASCII exports — 1,249 waveform + 13,089 histogram, 45
+  units, files back to 2018. That is 11x the ground truth the prod store
+  carried, and it supersedes the old "per-sample on 11%, peak-only on 89%"
+  caveat. Harness: `scratch/verify_against_ascii.py` (note its saturation
+  carve-out — BW clamps clipped events, the decoder reports true counts).
+  Independent corroboration of the 32000-count scale: 19,244 healthy
+  channel-events sit at a pre-trigger floor of exactly 0.000 (62.7%), 94.5%
+  within ±1 quantisation unit, median +0.0000 — no zero-point bias.
+- **Series-4 (Thor / Micromate) is NOT verified.** UM-series sits at ~48%
+  against device peaks with a ~1.7% systematic bias and a near-zero tail.
+  Thor IDFW is pinned to `decode_waveform_legacy` deliberately.
+- **Open, not blocking:** 14 sensitive-range files show an exact 8x
+  (= 10.0/1.25) units discrepancy; `scripts/backfill_sidecars.py --force` also
+  inserts DB rows for store files that have none (one-time per store) and the
+  dry-run does not report that count.
+- **After any codec change, regenerate the store** — `backfill_sidecars.py`
+  then `backfill_event_shape.py`, DB backup first. Stored `.h5` files do not
+  update themselves. No `--force` needed as long as `TOOL_VERSION` was bumped
+  (it gates regeneration). ⚠ On the office NAS this takes **~2 hours**
+  (~1.5 files/sec vs 85/sec on the dev box — gzip-4 in `sfm/event_hdf5.py`
+  against a Synology CPU). Budget it up front.
+  **v0.27.0 owes prod a backfill:** the partial-final-block fix recovers 4
+  histograms that are still empty in the store.
+- **The "offset" hardware fault has its own journal** --
+  `docs/offset_investigation.md`. **5 of 45 units (11%)**, and the fault is
+  **persistent** — it stays until the geophone is serviced. Detect it with
+  `scratch/offset_scan3.py`: the resting floor in the **pre-trigger** window,
+  required to hold across pre/middle/end. Never score only the dominant-peak
+  axis and never use the mean — both produce false recoveries (see the
+  retraction banner in the journal). Instantel's autozero procedure and its
+  2027-2069 acceptance window are recorded there too. Best open lead is
+  `SUB 0x0E` (unimplemented), which may carry those very numbers.
+
 
 When new information about the protocol is discovered, please update the instantel_protocol_reference.md with the findings in addition to this document
 
@@ -223,6 +265,118 @@ custom delta + RLE + variable-width codec.
   `NN + 2` for int8 blocks).  Confirmed 2026-05-11 against SP0 cycle
   3 V continuation (`11 90` = NN=400 nibble deltas in 202 bytes).
 
+### ⚠ SUPERSEDED 2026-08-25 — the body is a RECORD CHAIN
+
+Everything in this section below about `40 NN` segment headers, tagless
+headers, variable header widths and channel rotation describes a model that
+is **wrong**.  The body is a chain of self-delimiting per-channel records:
+
+    off+2  len  uint16 BE  ->  next_record = off + 2 + len
+    off+4  chan_id  0x46 Tran / 0x47 Vert / 0x48 Long / 0x49 MicL / 0x06 = END
+    off+8  mode  02 00 = deltas+anchors (14B hdr)
+                 01 00 = ABSOLUTE values (10B hdr)
+                 00 03 = raw 12-bit absolute, NO TAGS (10B hdr)
+
+`40 NN` is an ordinary int16 BE data block (`2*NN + 2`), never a header.  The
+"variable prefix" of 0/2/4/6/8 bytes was walker drift, exactly
+`4 - (old_stop - true_record_start)`.
+
+All four channels now come out equal length in **1388/1388** files (was
+156/1388); ASCII sample-count exact **75/75**, fully exact **73/75**; device
+PPV on a live decode **1306/1306** waveform, **4458/4459** histogram.
+
+The old model survives as `decode_waveform_legacy` because
+`micromate/idf_file.py` pins it for Thor IDFW body-offset search.
+
+### Framing cases added 2026-05-11 → 2026-08-25
+
+Four more block-framing cases, each of which had been causing **silent
+channel truncation** — `walk_body` ends its loop on an unrecognised tag
+and `decode_waveform_v2` returns whatever channels it got, so an
+unhandled tag surfaces as short channels with no error raised.  Found by
+diffing 75 production events against their preserved Blastware ASCII
+exports (`<store>/<serial>/<file>_ASCII.TXT`).
+
+- **Wide-NN RLE `0X NN`** — the 12-bit NN encoding documented above for
+  `1X`/`2X` **also applies to the `00 NN` RLE tag**.  A narrow run maxes
+  out at NN=0xFC, so a quiet stretch longer than 252 samples must use
+  the wide form (e.g. `01 0c` = 268 repeats).
+- **`30 NN` is not capped at NN=0x10** — data-section blocks reach at
+  least NN=0x18.  The `NN × 1.5 + 2` length formula was already right;
+  only the guard was wrong.
+- **`40 NN` segment headers are variable width** — NN is the *count of
+  int16 BE continuation deltas for the PREVIOUS channel*, so the header
+  is `2*NN + 16` bytes and every field after the deltas shifts by
+  `2*NN`.  `40 01` (18 B) and `40 03` (22 B) both occur alongside the
+  common `40 02` (20 B).
+- **Tagless segment headers** — a header can appear with **no `40 NN`
+  tag at all**: just the 14-byte tail
+  `[field2:2][len:2][channel_id:4][marker:2][anchors:4]`.  This is the
+  NN=0 case (previous channel needed no continuation deltas).
+
+**The header "counter" is really a channel id.**  The 4-byte field long
+documented as a "monotonic uint32 LE counter" is
+`[channel_id][00][00][segment_index]`, with `0x46`=Tran `0x47`=Vert
+`0x48`=Long `0x49`=MicL — verified on **1697/1697** segment headers
+across the corpus, zero disagreements.  `decode_waveform_v2` now takes
+the channel from this field rather than from rotation position; a single
+missed or extra header (exactly what tagless headers caused) desyncs
+rotation and corrupts every channel after it.
+
+Corpus result, end to end through the production path:
+**exact 37 → 72, truncated 23 → 3, full-length value errors 15 → 0.**
+
+### Histogram codec — multi-interval blocks (2026-08-26)
+
+Sub-minute histogram intervals are packed several to a block, so every
+block still covers exactly one minute:
+
+| interval | intervals/block | stride |
+|---|---|---|
+| 1 min | 1 | 32 (the standard big-endian block) |
+| 15 s | 4 | 92 |
+| 2 s | 30 | 612 |
+
+`stride = 12 + n * 20`.  Block = `[00][segment][ctr uint16 LE][0a][00]`,
+then n x 20-byte records of 8 x uint16 **LITTLE**-endian values
+(`T_peak, T_halfp, V_peak, V_halfp, L_peak, L_halfp, M_peak, M_halfp`)
+plus a 2-word tail whose first word is `0000` on every real interval,
+then a 6-byte block trailer.
+
+⚠ The standard 32-byte block is BIG-endian; this variant is LITTLE-endian.
+
+Recovers **415 files** (216 on BE18193, 199 on BE9440) that decoded to
+nothing.  Ground truth `BE9440/K440L3AQ.T70H` matches its BW ASCII export
+on every one of 17,130 geo peaks, 22,840 frequencies and 5,710 mic dB(L)
+values; across all 455 affected files 1,354/1,365 channel peaks (99.2%)
+match the device-reported PPV.
+
+### Histogram codec — corrected 2026-08-25
+
+The histogram block is **uniformly big-endian**, and the stream's final
+block has its own tail signature.  Two long-standing errors:
+
+- **Peaks and half-periods are `uint16` big-endian**, not `uint8` +
+  an "annotation" byte.  `T_peak` is `[5:7]`, `T_halfperiod` `[7:9]`,
+  `V_peak` `[9:11]`, and so on; only `block_ctr` at `[2:4]` is LE.
+  The old model silently **clipped any peak above 1.275 in/s** — the
+  final interval of `BE18193/T193LQ9K.OE0H` reads 8.270 in/s in BW's
+  export and decoded as 0.590.  The "annotation" byte was the
+  half-period's high byte, which is why it was non-zero exactly on the
+  sub-Hz intervals BW renders as `<1.0`.
+- **The marker is `block[4]` alone.**  Testing `[4:6]` as a uint16 LE
+  marker forced `block[5] == 0`, which is what capped the peak at one
+  byte in the first place.
+- **The last block of the stream carries tail `9c 06 00 42`** instead of
+  `1e 0a 00 00`, with arbitrary bytes at `[21:23]`.  Rejecting it
+  dropped the final interval of nearly every histogram — frequently the
+  one holding the event peak, so the file's PPV read low.
+
+Verified against 1211 production histograms paired with their BW ASCII
+exports: **1211/1211 decode exactly** (interval count plus every
+per-interval peak), and 842,442 per-interval frequency comparisons match
+with zero mismatches.  Before: 1 of 1196.
+
 ### What's NOT solved
 
 - **MicL channel conversion to dB(L)** — the codec emits MicL as
@@ -230,9 +384,25 @@ custom delta + RLE + variable-width codec.
   shows mic in dB(L) with ~6 dB quantization steps.  Need to map
   ADC counts → dB(L) for direct comparison; likely
   `dB = 20*log10(|counts|) + offset` or similar.
-- **Walker edge cases** — SP0/SS0/SV0 don't walk the full event due
-  to block-length quirks past the first few segments.  Every sample
-  reached is correct; the walker just needs robustness improvements.
+- **Variable-prefix segment descriptors** — 3 of the 75 ground-truth
+  production events still truncate.  The walk reaches a segment header
+  whose channel-id field is preceded by a *variable-width* prefix (2, 4
+  or 6 bytes observed; the standard tagless form always has 4), carrying
+  an `01 00` marker instead of `02 00`.  The marker is **not** simply an
+  anchor count — `01 00` records appear with both 2- and 4-byte anchor
+  fields in the same file.  Examples: `BE12599/N599LPNB.JF0W` @1155,
+  `BE12599/N599LPWJ.980W` @849, `BE9558/K558LOF2.820W` @1485.
+(The series-3 histogram codec was fixed 2026-08-25 — see below.)
+
+- **Micromate (UM-series) IDF decode is ~1000× low** — e.g.
+  `UM11402_20260406130113.IDFW` gives a Tran peak of 0.0009 in/s against
+  a device-reported 1.1168.  The Thor IDF path decodes sanely, so this
+  is UM-specific.
+- **Thor IDF per-count LSB** — after the 32000 geo full-scale
+  correction, series-4 Thor peaks sit at a median 0.983 of the
+  device-reported peak (was 0.960 under 32768).  Closer but not exact;
+  Thor likely uses its own per-count LSB rather than the BW
+  16-count/0.005 in/s convention.
 
 ### Decoded sample counts (across the fixture bundle)
 
@@ -264,6 +434,14 @@ then `decoded_to_adc_counts()` to scale to int16 ADC counts (geos × 16;
 mic pass-through).  The `.h5` sidecars SFM produces now contain
 correct samples for any event without walker edge cases.
 
+**Geo full scale is 32000 ADC counts, NOT 32768** (fixed 2026-08-25).
+One decoder unit = 16 ADC counts = exactly 0.005 in/s, so
+`10.000 in/s / (0.005/16)` = 32000.  Consumers must use
+`sfm.event_hdf5._GEO_INT16_FS` / `event_file_io._GEO_INT16_FS` (both
+32000).  Dividing by 32768 reads every geophone sample 2.34% low —
+that was a live bug in both modules until 2026-08-25.  Mic is
+unaffected (it back-solves its scale from the device-reported peak).
+
 The original int16 LE decoder is preserved as
 `_decode_a5_waveform_LEGACY` for reference but is not called.
 
@@ -1640,4 +1818,4 @@ body) because writing a dial string may require DLE escaping for embedded contro
 
 To parse BW TX captures: use `bridges/captures/` scripts or adapt the `find_write_frames()` pattern
 in `/tmp/analyze_write_payload.py` — it correctly handles `0x10 0x03` DLE-escaped ETX bytes
-inside write frame data (the naive parser terminates early at the escaped `0x03`).
+inside write frame data (the naive parser terminates early at the escaped `0x03`).

README.md

@@ -1,4 +1,4 @@
-# seismo-relay  `v0.22.0`
+# seismo-relay  `v0.27.0`
 
 A ground-up replacement for **Blastware** — Instantel's aging Windows-only
 software for managing seismographs.  Supports both the **MiniMate Plus

docs/histogram_codec_re_status.md

@@ -1,3 +1,30 @@
+> ## SUPERSEDED 2026-08-25 — the block is uniformly BIG-ENDIAN
+>
+> The `uint8` peak / `annotation` byte model described below is wrong,
+> though it decoded quiet data correctly.  The real layout:
+>
+> - **Every per-channel field is `uint16` big-endian.**  `T_peak` is
+>   `[5:7]`, `T_halfperiod` `[7:9]`, `V_peak` `[9:11]`, and so on.
+>   Only `block_ctr` at `[2:4]` is little-endian.
+> - The **marker is `block[4]` alone**, not a `uint16 LE` at `[4:6]`.
+>   Testing `[4:6] == 10` forced `block[5] == 0`, which is exactly what
+>   capped every geo peak at one byte (255 counts = 1.275 in/s).
+> - The **"annotation" byte was never an annotation** — it is the high
+>   byte of the big-endian half-period.  That is why it was non-zero
+>   precisely on the sub-Hz intervals Blastware renders as `<1.0`.
+> - The **final block of the stream carries tail `9c 06 00 42`** instead
+>   of `1e 0a 00 00`, and arbitrary bytes at `[21:23]`.  Rejecting it
+>   dropped the last interval of nearly every histogram — often the one
+>   holding the event peak, so the file's PPV read low.
+>
+> Verified against 1211 production histograms paired with their Blastware
+> ASCII exports: **1211/1211 decode exactly** (interval count plus every
+> per-interval peak), and 842,442 per-interval frequency comparisons match
+> with zero mismatches.  The uint8 model scored 1204/1211 — the seven
+> failures are exactly the files containing a peak above 1.275 in/s.
+>
+> The section below is retained as the reasoning trail.
+
 # Histogram body codec — FULLY DECODED (2026-05-20)
 
 Clean working status doc for the MiniMate Plus histogram-mode event

docs/instantel_protocol_reference.md

@@ -11,6 +11,10 @@
 
 | Date | Section | Change |
 |---|---|---|
+| 2026-08-26 | S7.6.2, S15 | **MULTI-INTERVAL HISTOGRAM BLOCKS - sub-minute intervals pack several per block.** The device always writes one minute of data per block, so a shorter histogram interval means more intervals packed into a longer block: 1 min -> 1 (the standard 32-byte block), 15 s -> 4 (stride 92), 2 s -> 30 (stride 612), with `stride = 12 + n * 20`. Each 20-byte record holds 8 x uint16 **little**-endian values (peak/half-period per channel) plus a 2-word tail whose first word is `0000` on every real interval. The standard block is big-endian - the variant is not. This recovers **415 files** (216 on BE18193 at 2 s, 199 on BE9440 at 15 s) that previously decoded to nothing, and before that were being accepted by the WAVEFORM codec and returning garbage up to 400x the device-reported PPV. |
+| 2026-08-25 (3) | S7.6.1, S15 | **THE WAVEFORM BODY IS A RECORD CHAIN, NOT A TAG STREAM — supersedes the segment-header model entirely.** Records are self-delimiting: `off+2` is a uint16 BE length and `next = off + 2 + len`; the chain ends on a record whose chan_id is `0x06`. `off+8` holds a 3-valued mode enum - `02 00` (14-byte header, anchors, cumulative deltas), `01 00` (10-byte, no anchors, ABSOLUTE values), `00 03` (10-byte, no tags at all, raw 12-bit absolute). **`40 NN` is an ordinary int16 BE data block of length 2*NN+2**, never a header; reading it as a 2*NN+16 header is what made walks drift, and the 'variable prefix' of 0/2/4/6/8 bytes reported earlier the same day was walker drift, exactly `4 - (old_stop - true_record_start)`. Verified: chain terminates on `06` in 1387/1388 files; all four channels equal length in **1388/1388** (was 156/1388); ASCII sample-count exact 72/75 -> **75/75**, fully exact 70/75 -> **73/75**; device PPV on a live decode **1306/1306** waveform (mean abs ratio error 0.00000) and **4458/4459** histogram. Also eliminated the walker-over-read class: 24 of those files were histograms the waveform codec was wrongly accepting. The superseded model is retained as `decode_waveform_legacy` because `micromate/idf_file.py` pins it for Thor IDFW body-offset search. |
+| 2026-08-25 (2) | S7.6.2, S15 | **HISTOGRAM BLOCK IS BIG-ENDIAN + terminal block tail.** The 32-byte histogram block's per-channel fields are uint16 **big-endian** (`T_peak` [5:7], `T_halfperiod` [7:9], `V_peak` [9:11], `V_halfperiod` [11:13], `L_peak` [13:15], `L_halfperiod` [15:17], `M_peak` [17:19], `M_halfperiod` [19:21]); only `block_ctr` [2:4] is little-endian. The marker is `block[4]` alone - the previous uint16 LE marker test at [4:6] forced `block[5] == 0` and thereby capped every geo peak at 255 counts (1.275 in/s), silently clipping larger peaks. The byte previously documented as a per-channel "annotation" is the high byte of the big-endian half-period, which is why it was non-zero exactly on sub-Hz intervals. Separately, the **final block of each stream carries tail `9c 06 00 42`** rather than `1e 0a 00 00` and holds arbitrary bytes at [21:23]; rejecting it dropped the last interval of nearly every histogram. Verified against 1211 production histograms paired with their Blastware ASCII exports: 1211/1211 decode exactly, plus 842,442 per-interval frequency comparisons with zero mismatches (previously 1 of 1196 files fully correct). |
+| 2026-08-25 | §7.6.1, §15, Appendix E (NEW) | **BODY CODEC + SCALE PASS — five findings, all verified against 75 production events paired with their preserved Blastware ASCII exports.** (1) **Geo full scale is 32000 ADC counts, not 32768** — one decoder unit (16 counts) is exactly 0.005 in/s, so 10.000 in/s = 32000 counts. Consumers dividing by 32768 read every geophone sample and derived peak **2.34% low**; the error scales with amplitude so it was invisible on quiet events and worst on loud ones. 216 per-channel comparisons: 32768 → 151/216 exact (worst 0.238 in/s on a 10 in/s event); 32000 → 216/216 exact, worst 1 LSB. Affects waveforms, histograms and series-4 alike. (2) **Wide-NN RLE `0X NN`** — the 12-bit NN encoding already known for `1X`/`2X` also applies to the `00 NN` RLE tag (runs > 252 samples). (3) **`30 NN` is not capped at NN=0x10** — data-section blocks reach at least 0x18; the length formula was already right. (4) **`40 NN` segment headers are variable width** — NN counts the previous-channel continuation deltas, so the header is `2*NN + 16` bytes; `40 01` and `40 03` occur alongside `40 02`. A header can also appear **tagless** (the NN=0 case): just the 14-byte tail. (5) **The header field documented as a "monotonic uint32 LE counter" is really `[channel_id][00][00][segment_index]`** with 0x46=Tran 0x47=Vert 0x48=Long 0x49=MicL — verified on 1697/1697 segment headers, zero disagreements. Decoders should take the channel from this field, not from rotation position. Items (2)–(5) each caused **silent channel truncation**: an unhandled tag ends the walk and the decoder returns short channels with no error. Corpus result end-to-end: exact 37 → 72, truncated 23 → 3, full-length value errors 15 → 0. New Appendix E documents the field-observed "offset" device fault. |
 | 2026-05-20 | §2, §3, §4.2, §5.1, §5.3, §6, §7.5b, §7.6.1, §7.6.3, §7.6.4, §7.7.2, §7.7.3, §7.7.5, §7.8.4, §7.8.7, §7.9, §8, §11, §12, §13, §14, §15, Appendix D | **DOC AUDIT PASS — accuracy sweep against `CLAUDE.md` + `minimateplus/` code.** Fixed: (1) S3 frames terminate on bare ETX, not DLE+ETX — §2/§3 rewritten. (2) §3 payload layout corrected — byte[1]=flags, byte[2]=SUB (was wrongly labelled DLE/ADDR). (3) §4.2 — probe responses do NOT carry data length; lengths are hardcoded `DATA_LENGTHS` constants. (4) §5.1 — removed stale duplicate "SUB 1C = TRIGGER CONFIG READ" row; SUB 0A lengths corrected from `0x30/0x26` to `0x46/0x2C` (real event / boundary marker). (5) §5.3 — added missing write-frame format (BW_CMD-only doubling, DLE-aware checksum, offset formula, ack format, SUB 71 chunk parameters). (6) §6 — fixed "SUB 06 → channel config read" → event storage range. (7) §7.5b / §8 — added the 10-byte `sub_code=0x03` continuous-mode timestamp variant alongside the 9-byte single-shot layout; peak vector sum location corrected from "fixed offset 87" to `tran_pos − 12` (label-relative). (8) §7.6 / §7.6.1 / §7.6.3 / §7.6.4 — switched compliance-anchor convention from the 10-byte form to the canonical 6-byte `\xbe\x80\x00\x00\x00\x00`; recording_mode confirmed at anchor−8 in BOTH read and write (was wrongly listed as anchor−3 write / anchor−4 read); sample_rate at anchor−6, histogram_interval at anchor−4, record_time at anchor+6; geo_range row added at channel_label+33. (9) §7.7.2 — token byte position corrected from `params[6]` to `params[7]`. (10) §7.8.4 — fi==9 skip marked FIXED (already removed from code); chunk-count totals updated. (11) §7.8.7 — TODO replaced with current state of `_decode_a5_metadata_into`. (12) §7.9 — Histogram Interval upgraded ❓ → ✅. (13) §11 — POLL example wire bytes corrected; SUB 5A row added to checksum table. (14) §13 — device-under-test updated for current primary unit (BE11529 / S338.17). (15) §14 — TCP Idle Timeout fixed (0→2 min); Data Forwarding Timeout units clarified. (16) §15 (renumbered from second §14) — open-question items already resolved in CLAUDE.md closed out. (17) Appendix D — extension taxonomy rewritten to reflect the AB0T timestamp encoding (D.5.2/D.5.3); EXTENSION REFUTED warning replaced with the resolved encoding. |
 | 2026-05-08 | §7.6.1 (RETRACTION) | **❌ RETRACTED — "raw int16 LE 8 bytes/sample-set" body codec was never validated.** The original 4-2-26 confirmation was based on misreading broken-decoder output (full-scale ±32K noise) as evidence the signal had saturated. BW's own 0C peaks for that capture (Tran=0.420 / Vert=3.870 / Long=0.495 in/s) prove the signal was NOT saturated — none of those exceed 13K ADC counts. No event in the project's archive has ever come close to saturation, yet the decoder consistently produces ±32K noise on every event. Conclusion: the body codec is not raw int16 LE; the actual encoding is open. Body byte distribution is heavily skewed (24% `0x00`, 10.5% `0x10`, lots of `10 XX` pairs) — likely a delta encoding with `0x10` as escape, but unverified. Retraction box added at top of §7.6.1; "fully-saturating event" claim removed from channel-identification note. The histogram codec in §7.6.2 IS verified and decoded correctly (different recording mode, 32-byte blocks); use it as a structural hint when reverse-engineering the waveform codec. |
 | 2026-02-26 | Initial | Document created from first hex dump analysis |
@@ -1101,14 +1105,51 @@ Every block starts with a 2-byte tag.  Five tag types are confirmed:
 |-----------|-------------------------------------|-----------------------|
 | ``10 NN`` | Small-delta data block              | NN/2 + 2 bytes        |
 | ``20 NN`` | Literal data block (int8-shaped)    | NN + 2 bytes          |
-| ``00 NN`` | 2-byte marker between data blocks   | 2 bytes               |
+| ``00 NN`` | RLE zero-delta run                  | 2 bytes               |
 | ``30 NN`` | Trailer summary block               | NN × 4 bytes          |
-| ``40 02`` | Segment header                      | 20 bytes (fixed)      |
+| ``40 NN`` | Segment header                      | 2 × NN + 16 bytes     |
 
 NN is always a multiple of 4.  ``10 NN`` and ``20 NN`` data blocks
 alternate with ``00 NN`` markers — every ``10/20 NN`` block is
 followed by a ``00 NN`` marker before the next data block.
 
+###### Wide-NN forms — ``0X NN`` (CONFIRMED 2026-08-25)
+
+The 12-bit wide-NN encoding already documented for ``1X NN`` /
+``2X NN`` (low nibble of the tag byte carries the high nibble of NN,
+so effective ``NN = ((tag & 0x0F) << 8) | NN``) **also applies to the
+``00 NN`` RLE tag.**  A narrow RLE run maxes out at NN = 0xFC, so a
+quiet stretch longer than 252 samples must use the wide form.
+
+Confirmed against six production events, e.g. ``01 0c`` (NN = 268) in
+``BE9558/K558LKOF.460W``.  Before this was handled, the walker hit its
+unknown-tag break at the first long quiet run and silently truncated
+every channel decoded after that point.
+
+###### ``30 NN`` is not capped at NN = 0x10 (CONFIRMED 2026-08-25)
+
+Data-section ``30 NN`` blocks occur with NN up to at least 0x18 (24),
+e.g. ``30 18`` in ``BE18193/T193LQ45.NN0W`` and ``30 14`` in
+``BE18193/T193LQ9W.AF0W``.  The data-section length formula
+(``NN × 1.5 + 2``) holds for these; only the earlier ``NN ≤ 0x10``
+guard was wrong.
+
+###### ``40 NN`` segment headers are variable width (CONFIRMED 2026-08-25)
+
+``40 02`` is the common case, but **NN is the count of int16 BE
+continuation deltas the header carries for the *previous* channel**, so
+the header grows with NN and every field after the deltas shifts by
+``2 × NN``:
+
+```
+length = 2 (tag) + 2 × NN (prev-channel deltas) + 14 (fixed tail)
+```
+
+``40 01`` (18 bytes) and ``40 03`` (22 bytes) both occur in production
+files — see ``BE12599/N599LP1S.UO0W`` and ``BE18438/T438LO30.GA0W``.
+In each case the constant ``02 00`` marker sits at ``data[2×NN+8]`` and
+the following tag lands exactly on a valid block boundary.
+
 ##### Segments
 
 The body is divided into segments separated by ``40 02`` segment headers.
@@ -1129,32 +1170,179 @@ fit fewer.  Observed first-segment sizes in the bundled fixtures:
 based on incomplete walks; that figure is wrong.  Segments are
 flash-page-sized in bytes, not sample-count-sized.
 
-The 18-byte ``40 02`` payload structure:
+The ``40 NN`` payload structure (offsets shown for the common NN=2 /
+18-byte-payload case; add ``2 × (NN − 2)`` to every offset from ``[4:6]``
+onward for other widths):
 
-| Offset    | Field                                       | Status      |
-|-----------|---------------------------------------------|-------------|
-| [0:2]     | T_delta at first sample of new segment      | ✅ confirmed|
-|           | (int16 BE, in 16-count units)               |             |
-| [2:4]     | Likely T_delta at sample seg_start+1        | 🟡 likely   |
-| [4:6]     | Unknown (varies; possibly a checksum)       | ❓ open     |
-| [6:8]     | Byte length to next segment header − 2      | ✅ confirmed|
-|           | (uint16 BE; useful for walker pre-scan)     |             |
-| [8:12]    | Monotonic uint32 LE counter                 | ✅ confirmed|
-|           | (starts ~0x47, increments by 1 per segment) |             |
-| [12:14]   | Constant ``02 00``                          | ✅ confirmed|
-| [14:18]   | Unknown 4-byte field                        | ❓ open     |
+| Offset (NN=2) | Generic          | Field                                  | Status      |
+|---------------|------------------|----------------------------------------|-------------|
+| [0:4]         | [0 : 2NN]        | NN × int16 BE continuation deltas for  | ✅ confirmed|
+|               |                  | the PREVIOUS channel (16-count units)  |             |
+| [4:6]         | [2NN : 2NN+2]    | Unknown (varies; possibly a checksum)  | ❓ open     |
+| [6:8]         | [2NN+2 : 2NN+4]  | Byte length to next segment header − 2 | 🟡 likely   |
+|               |                  | (uint16 BE; off by ±4 on some files)   |             |
+| [8:12]        | [2NN+4 : 2NN+8]  | Monotonic uint32 LE counter            | ✅ confirmed|
+|               |                  | (starts ~0x47, +1 per segment)         |             |
+| [12:14]       | [2NN+8 : 2NN+10] | Constant ``02 00``                     | ✅ confirmed|
+| [14:18]       | [2NN+10 : 2NN+14]| THIS channel's 2-sample anchor pair    | ✅ confirmed|
+|               |                  | (2 × int16 BE)                         |             |
 
-Examples from event-c (1 sec single-shot):
+⚠️ An earlier draft listed ``[14:18]`` as an "unknown 4-byte field" and
+``[12:16]`` as a constant ``02 00 00 01``.  Both were wrong: the constant
+is only the 2-byte ``02 00``, and the four bytes after it are the anchor
+pair the decoder needs.  Corrected 2026-08-25.
+
+###### Tagless segment headers (CONFIRMED 2026-08-25)
+
+A segment header can appear with **no ``40 NN`` tag at all** — just the
+14-byte tail:
 
 ```
-Segment header 1 (offset 235):
-  40 02 | 00 00 00 00 | 0a 4b 01 1e | 47 00 00 00 | 02 00 00 01 | 00 01
-                                                  ^counter=0x47
-Segment header 2 (offset 523):
-  40 02 | ff fe ff fe | 13 f5 01 06 | 48 00 00 00 | 02 00 00 01 | 00 02
-                                                  ^counter=0x48 (+1)
+[field2:2][len_to_next:2][channel_id:4][marker:2][anchors:4]
 ```
 
+This is the NN=0 case: the previous channel needed no continuation
+deltas, so there is no tag and no delta bytes.  Detect it by the six
+bytes at ``[4:10]`` — a known channel id, two zero bytes, a small
+segment index, then the ``01 00`` / ``02 00`` marker.
+
+It is where the walk stopped in 7 of the 8 events that still truncated
+after the wide-RLE / ``30 NN`` / variable-width-``40 NN`` fixes.
+
+###### The header "counter" is really a channel id (CONFIRMED 2026-08-25)
+
+The 4-byte field previously documented as a *"monotonic uint32 LE
+counter (starts ~0x47, increments by 1 per segment)"* is actually:
+
+```
+[channel_id:1][00][00][segment_index:1]
+```
+
+| channel_id | channel |
+|---|---|
+| ``0x46`` | Tran |
+| ``0x47`` | Vert |
+| ``0x48`` | Long |
+| ``0x49`` | MicL |
+
+Verified on **1697 of 1697** segment headers across the ground-truth
+corpus — every one agrees with the channel the rotation would assign,
+zero disagreements, no other id values observed.  The old reading was
+plausible because the id byte cycles 0x46→0x47→0x48→0x49 and the
+segment index increments, which *looks* monotonic in LE.
+
+Decoders should take the channel from this field rather than from
+rotation position: one missed or extra header (exactly what tagless
+headers used to cause) desyncs rotation and corrupts every channel
+after it.
+
+###### Geophone full scale is 32000 counts, not 32768 (CONFIRMED 2026-08-25)
+
+The body codec emits geo samples in 16-count units whose LSB is exactly
+**0.005 in/s**.  With the consumer-side ``×16`` to ADC counts, one ADC
+count is ``0.005 / 16`` in/s, so Normal range (10.000 in/s) is
+
+```
+10.0 / (0.005 / 16) = 32000 counts
+```
+
+Dividing by 32768 scales every geophone sample and every derived peak
+down by ``1 - 32000/32768`` = **2.34%**.  Measured on 216 per-channel
+comparisons against preserved Blastware ASCII exports: 32768 gave
+151/216 exact (worst error 0.238 in/s on a 10 in/s event); 32000 gives
+216/216 exact with a worst error of 0.005 in/s — exactly 1 LSB, i.e.
+pure quantization.
+
+This also explains why Blastware reports geo peaks slightly above
+nominal full scale (e.g. 10.14 in/s): the ADC has headroom past 32000.
+
+**Scope — this is not waveform-specific.**  The scale is applied where
+ADC counts become physical units, which every event passes through
+regardless of which codec produced the samples.  Verified after
+re-deriving the whole production store:
+
+| source | median ratio ours/device, 32768 | with 32000 |
+|---|---|---|
+| series-3 waveform (vs ASCII sample table) | 0.9766 | **1.0000** |
+| series-3 histogram (vs ASCII PPV, n=1137) | 0.9766 | **1.0000** |
+| series-4 Thor IDF (vs device peak, n=1468) | 0.960 | **0.983** |
+
+The series-4 figure is closer to correct but not exact — the Thor
+per-count LSB is its own open question (see §15).
+
+###### SUPERSEDED 2026-08-25 — the body is a RECORD CHAIN, not a tag stream
+
+Everything above about ``40 NN`` segment headers, tagless headers, variable
+header widths and channel rotation describes a model that is **wrong**.  It
+produced nearly-correct output because the block table happens to tile the
+data sections correctly, but the framing is not what the device writes.
+
+The body is a chain of **self-delimiting per-channel records**:
+
+```
+off+0   field2    uint16     purpose unknown (not a length, not a checksum)
+off+2   len       uint16 BE  next_record = off + 2 + len   <- authoritative
+off+4   chan_id   0x46 Tran / 0x47 Vert / 0x48 Long / 0x49 MicL
+                  0x06 = END OF WAVEFORM STREAM
+off+5   0x00
+off+6   0x00
+off+7   segment index
+off+8   mode      2 bytes, a 3-valued enum
+off+10  anchors   2 x int16 BE, ABSOLUTE  -- present ONLY when mode == 02 00
+```
+
+**Mode enum**, all three ground-truth verified:
+
+| mode | header | data section |
+|---|---|---|
+| ``02 00`` | 14 bytes | anchors emitted, then blocks are **cumulative deltas** |
+| ``01 00`` | 10 bytes | no anchors, blocks carry **absolute** sample values |
+| ``00 03`` | 10 bytes | **no tags at all** — raw 12-bit packed absolute samples |
+
+Census over the 1,388 production series-3 waveform binaries:
+``02 00`` x 32,617, ``01 00`` x 74, ``00 03`` x 71.
+
+**``40 NN`` is an ordinary int16 BE data block** of length ``2*NN + 2``
+(1 <= NN <= 8), never a header.  The superseded model read it as a header of
+length ``2*NN + 16``, which is exactly why walks drifted: the "variable prefix"
+of 0/2/4/6/8 bytes reported earlier was walker drift, precisely
+``4 - (old_stop - true_record_start)``, on all 25 affected files.
+
+Block table for data sections (``NN = ((tag_hi & 0x0F) << 8) | tag_lo``):
+
+| tag | length | samples |
+|---|---|---|
+| ``0X NN`` | 2 | NN (RLE hold — holds the previous value in BOTH delta and absolute modes) |
+| ``1X NN`` | NN/2 + 2 | NN (4-bit nibble) |
+| ``2X NN`` | NN + 2 | NN (int8) |
+| ``30 NN`` | NN*1.5 + 2 | NN (12-bit packed) |
+| ``40 NN`` | 2*NN + 2 | NN (int16 BE) |
+
+The ``30 NN`` "trailer length = NN*4" fallback must NOT be applied inside a
+record — it corrupts records whose ``30 NN`` sits near a boundary.
+
+**The preamble is segment 0's implicit Tran record.**  ``body[1:3]`` carries
+the same mode pair: ``00 02 00`` (1,387 of 1,388 files) means two int16 BE
+anchors at ``body[3:7]`` then delta blocks; ``00 00 03`` (1 file,
+``BE13121/O121L4L1.KF0W``) means raw 12-bit absolute from ``body[3]``, which
+cannot be block-walked — the first record must be located by scanning.
+
+**Verification.**  The length chain terminates on a ``0x06`` record in 1,387 of
+1,388 files (the exception has an ambiguous footer signature inside its sample
+stream).  All four channels come out at identical length in **1,388/1,388**,
+against 156/1,388 under the superseded model.  Against the 75 events with a
+preserved Blastware ASCII export: sample-count exact **72/75 -> 75/75**, fully
+exact **70/75 -> 73/75** (the 2 remaining differ by exactly 1 LSB on samples
+sitting at the +-10.000 in/s rail).  Against device-reported PPVs on a live
+decode: waveform **1306/1306** exact with mean absolute ratio error 0.00000;
+histogram **4458/4459**.
+
+This also eliminated the walker-over-read class entirely.  24 of those 35
+"histogram" over-reads were histogram files that ``read_blastware_file`` fed to
+the *waveform* codec first; the old walker accepted them and returned garbage
+(one produced 98,923 "intervals"), while the record-chain decoder correctly
+returns None so they fall through to ``histogram_codec``.
+
 ##### Trailer
 
 The trailer (after the last segment's data) is a sequence of 32-byte
@@ -2867,7 +3055,7 @@ The `.bin` files produced by `s3_bridge` are **not raw wire bytes**. The logger
 |---|---|---|---|
 | Timestamp 6-byte format byte[3] — purpose of the separator `0x00` byte | LOW | 2026-02-26 | Not blocking; 9-byte waveform record format (§8.2) fully confirmed without this byte. |
 | `trail[0]` in serial number response — unit-specific byte, derivation unknown. `trail[1]` resolved as firmware minor version. | MEDIUM | 2026-02-26 | |
-| Full channel ID mapping in SUB `5A` stream (01/02/03/04 → which sensor?) | MEDIUM | 2026-02-26 | |
+| ~~Full channel ID mapping in SUB `5A` stream~~ — **RESOLVED 2026-08-25 for the waveform body:** every `40 NN` segment header carries `[channel_id][00][00][segment_index]` at `data[2*NN+4]`, with `0x46`=Tran `0x47`=Vert `0x48`=Long `0x49`=MicL. Verified on 1697/1697 segment headers with zero disagreements against the decoded channel rotation. See §7.6.1. | RESOLVED | 2026-02-26 | Resolved 2026-08-25 |
 | ~~Exact byte boundaries of project string fields in SUB `71` write frame~~ — **RESOLVED 2026-05-05:** project/client/operator/seis-loc/extended-notes come from SUB 5A metadata pages at counter `0x1002` / `0x1004` (§7.8.7), NOT from the SUB 71 write payload. `_decode_a5_metadata_into` locates them via ASCII label scans. | RESOLVED | 2026-02-26 | Resolved 2026-05-05 |
 | Purpose of SUB `09` / response `F6` — 202-byte read block | MEDIUM | 2026-02-26 | |
 | Purpose of SUB `2E` / response `D1` — 26-byte read block | MEDIUM | 2026-02-26 | |
@@ -2894,6 +3082,11 @@ The `.bin` files produced by `s3_bridge` are **not raw wire bytes**. The logger
 | **ACH inbound server — RESOLVED.** `bridges/ach_server.py` implements full inbound ACH pipeline. `--clear-after-download` flag for delete-after-upload workflow. Post-erase key-reuse detection via `max_downloaded_key` high-water mark. | RESOLVED | 2026-04-11 | |
 | **Sensor Check dropdown byte location** — byte offset in 1A compliance config payload for the "Sensor Check: Before monitoring / After each event / Disabled" setting is NOT YET LOCATED. Confirmed: unit always runs with "Before monitoring" set. Need a capture with "Disabled" to diff. | MEDIUM | 2026-04-08 | Still open |
 | **RV55 DCD/DTR default** — newer Sierra Wireless RV55 firmware does not assert DCD/DTR by default, so the MiniMate Plus never detects TCP disconnect and stays idle instead of resuming monitoring. Root cause: RV55 ACEmanager `DCD Control` setting. Workaround not yet found. | MEDIUM | 2026-04-11 | Still open |
+| ~~**Variable-prefix segment descriptors**~~ - **RESOLVED 2026-08-25:** there is no variable prefix. The body is a chain of self-delimiting records (see S7.6.1); the 0/2/4/6/8-byte prefix was walker drift from reading `40 NN` as a segment header. All 25 affected files now chain cleanly to the `06` terminator. | RESOLVED | 2026-08-25 | Resolved 2026-08-25 |
+| ~~**Histogram codec drops trailing intervals (series-3)**~~ - **RESOLVED 2026-08-25.** Two errors, both in the block model. (1) The block is **uniformly big-endian**: peaks and half-periods are uint16 BE (`T_peak` [5:7], `T_halfperiod` [7:9], `V_peak` [9:11], ...); only `block_ctr` [2:4] is LE. The old uint8-peak + "annotation"-byte model silently clipped any peak above 1.275 in/s, and the "annotation" byte was really the half-period high byte - non-zero exactly on the sub-Hz intervals BW renders `<1.0`. The marker is `block[4]` alone; testing [4:6] as a uint16 LE marker forced `block[5] == 0`, which is what capped the peak at one byte. (2) The **final block of the stream carries tail `9c 06 00 42`** instead of `1e 0a 00 00`, with arbitrary bytes at [21:23]; rejecting it dropped the last interval of nearly every histogram - often the one holding the event peak. Verified on 1211 production histograms vs their BW ASCII exports: **1211/1211 exact** (interval count + every per-interval peak) and 842,442 frequency comparisons with zero mismatches; was 1/1196. | RESOLVED | 2026-08-25 | Resolved 2026-08-25 |
+| ~~**Unmapped histogram block variant (415 files, 2 units)**~~ - **RESOLVED 2026-08-26.** When the histogram interval is SHORTER than one minute the device packs several intervals into a single block so each block still covers exactly one minute: 1 min -> 1 interval (the standard 32-byte block), 15 s -> 4 (stride 92), 2 s -> 30 (stride 612). `stride = 12 + n * 20`. Block is `[00][segment][ctr uint16 LE][0a][00]` then n x 20-byte records of 8 x uint16 **LITTLE**-endian `T_peak,T_halfp,V_peak,V_halfp,L_peak,L_halfp,M_peak,M_halfp` plus a 2-word tail whose first word is `0000` on every real interval (a session ending mid-block leaves buffer garbage in the remaining slots), then a 6-byte block trailer. **Note the endianness flip** - the standard 32-byte block is big-endian. Ground truth `BE9440/K440L3AQ.T70H` (5,710 intervals at 15 s) decodes with 17,130/17,130 geo peaks, 22,840/22,840 frequencies and 5,710/5,710 mic dB(L) matching its Blastware ASCII export exactly; across all 455 affected files 1,354/1,365 channel peaks (99.2%) match the device-reported PPV. | RESOLVED | 2026-08-25 | Resolved 2026-08-26 |
+| **Micromate (UM-series) IDF decode is ~1000x low** — e.g. `UM11402_20260406130113.IDFW` decodes a Tran peak of 0.0009 in/s against a device-reported 1.1168. Distinct from the Thor IDF path, which decodes sanely. Suspect a different per-count LSB or a body offset that does not hold for UM-series files. | MEDIUM | 2026-08-25 | Still open |
+| **Thor IDF per-count LSB** — after the 32000 geo full-scale correction, series-4 Thor peaks sit at a median 0.983 of the device-reported peak (was 0.960 under 32768). Closer, but the residual ~1.7% suggests Thor uses its own per-count LSB rather than the BW 16-count/0.005 in/s convention. A code comment in `sfm/waveform_store.py` claims Thor's LSB is 0.0003 in/s, which would predict Thor reading *high* — the measurement shows the opposite, so that comment is unverified. | LOW | 2026-08-25 | Still open |
 
 ---
 
@@ -3320,3 +3513,141 @@ file[anc−3]: histogram_interval_LO
 *All findings reverse-engineered from live RS-232 bridge captures.*  
 *Cross-referenced from 2026-03-02 with Instantel MiniMate Plus Operator Manual (716U0101 Rev 15).*  
 *This is a living document — append changelog entries and timestamps as new findings are confirmed or corrected.*
+
+---
+
+## Appendix E — Known Device Faults (field-observed, 2026-08-25)
+
+This appendix records *device* behaviour, not protocol.  It exists
+because these signatures are easy to mistake for decoder bugs — and one
+of them was, for a while.
+
+### E.1 The "offset" fault
+
+> **The full investigation now lives in `docs/offset_investigation.md`** --
+> base rate, detector definition, per-unit case files, ruled-out hypotheses,
+> and Instantel's own autozero procedure with its 2027-2069 acceptance
+> window.  This appendix is kept as the protocol-side summary.
+
+**Symptom.**  One geophone channel's baseline steps away from zero and
+stays there.  The trace still carries the real AC signal, but it rides
+on a DC pedestal of a few tenths of an in/s.  Operators call this an
+"offset"; it is a known recurring hardware fault, historically resolved
+by returning the unit to Instantel, shelving it until calibration, or a
+per-channel re-zero in an advanced/professional Blastware build.
+
+**Why it floods the queue.**  The pedestal exceeds the unit's own geo
+trigger level, so the channel sits permanently above threshold and the
+unit retriggers as fast as it can rearm — one event every 3–7 minutes
+for as long as the fault persists.  In the surveyed snapshot the worst
+episode produced 193 events in 38 hours.
+
+| unit | window | events | channel | pedestal | that unit's trigger level |
+|---|---|---|---|---|---|
+| BE9558 | 2026-05-15 → 05-16 | 193 | Long | +0.335 | 0.200 |
+| BE18438 | 2026-02-25 → 02-26 | 64 | Vert | +0.327 | 0.200 |
+| BE18438 | 2025-11-15 | 12 | Vert | +0.312 | 0.200 |
+| BE9558 | 2026-04-14 / 04-29 / 05-04 | 5 total | Tran | +0.29…+0.44 | 0.200 |
+
+**It is NOT a decode artifact.**  These events reproduce *exactly* in
+Blastware's own ASCII export — e.g. `BE12599/N599LQD7.8E0W` Tran reads
+mean +0.345, min +0.335, max +0.355 in both our decode and BW's.  Any
+shape-based false-trigger detector must treat this as a real device
+condition, not corrupt data.
+
+**Detection rule that works:**
+
+```
+dominant geo axis:  |mean| / peak  > 0.7
+              AND:  |mean| >= 0.9 x that unit's geo trigger level
+group hits into episodes by serial with a 12 h gap
+```
+
+A bare `|mean|/peak` threshold is useless — a quiet trace sitting at the
+0.010 in/s noise floor clears any ratio test.  On the surveyed store the
+bare rule flagged 7,184 events; the rule above flags 274, all real.
+Candidate list: `scratch/offset_candidates.csv`.
+
+**What the data rules out.**
+
+- *Not the geophone.*  The on-device sensor check passes on every unit,
+  offset or not — test frequency 7.2–8.1 Hz, damping ratio 3.3–4.7,
+  zero failures fleet-wide, including mid-episode.  The coil and its
+  mechanical response are healthy.
+- *Not the battery.*  6.6–6.8 V on the affected units, same as the rest.
+- *Not environmental.*  Only 2 units of 21 ever show it.  BE17353 logged
+  395 waveform events over nine months with zero occurrences; BE7145,
+  295 events, zero.  Weather acts on all of them equally.
+- *Not condensation.*  BE18438's Vert held 0.179 → 0.189 for 34 hours
+  **including a 10-hour overnight gap**, then stepped to 0.33 at 13:24
+  in the afternoon.  Dew would peak overnight; it was flat overnight.
+
+**What the data shows.**  The pedestal is *piecewise constant* — it
+holds rock-steady, survives power-off gaps, and changes only in discrete
+steps.  A small common-mode diurnal wobble rides on top of it, but that
+wobble is present on the healthy channels too (it is the unit's normal
+thermal breathing) and is not the fault.
+
+**The fault is on the geophone side.**  Operator report (2026-08-25):
+*attaching a different geophone to an affected unit makes the offset go
+away.*  That is the channel-swap test, already run in the field many
+times, and it rules out the unit's analog front-end and any stored
+per-channel zero constant — a constant lives in the unit and would
+survive a sensor swap.
+
+The stored data agrees.  MicL — a separate transducer on its own cable —
+shows no offset during either episode (`|mean|/peak` = 0.17 and 0.02),
+while the geo channels on the same unit at the same moment are pinned.
+
+**Two distinct failure patterns**, both geophone-side:
+
+| unit | Tran | Vert | Long | MicL | reading |
+|---|---|---|---|---|---|
+| BE18438 | 0.16 | **0.97** | 0.18 | 0.02 | one element only |
+| BE9558 | 0.90 | 0.81 | **0.99** | 0.17 | all three, Long dominant |
+
+(values are `\|mean\|/peak` on the dominant axis)
+
+BE18438 has a single faulted conductor pair with the other two clean.
+BE9558 has all three geo channels offset by different amounts
+(Long +0.340, Tran +0.037, Vert −0.020) — that points at a shared return
+or ground path in the housing rather than one element.
+
+**Candidate mechanisms.**  A geophone coil is passive and cannot
+generate sustained DC on its own, so an offset originating on the sensor
+side has to come from one of:
+
+1. *Galvanic corrosion at a connector or splice* — dissimilar metals
+   plus moisture form a literal cell.  This generates DC directly, in
+   the tens-of-millivolts range that matches the measured offset
+   (0.29 in/s ÷ 6.206053 in/s per volt ≈ 46 mV referred to the ADC
+   input).  Fits the discrete steps, surviving power cycles, and being
+   cured by swapping the sensor.
+2. *A leakage path to shield or another conductor*, letting the unit's
+   own bias/reference divide into the signal pair.  Best fit for the
+   BE9558 all-three-channels pattern.
+3. *Changed coil DC resistance* (partial short or degraded winding)
+   interacting with the amplifier's input bias current —
+   `V_offset = I_bias × R_source`.  Best fit for the BE18438
+   single-element pattern.  Note the on-device sensor check would not
+   catch this: it measures ring-down frequency and damping, not DC.
+
+**Caveat on the swap test.**  Swapping a geophone requires stopping and
+restarting monitoring, and these units run Sensor Check "Before
+monitoring" — so the restart also re-zeros the channels.  The swap
+therefore does not cleanly separate *"new sensor"* from *"the restart
+re-zeroed it"*.  Two cheap controls settle it: restart monitoring
+**without** swapping (if the offset clears, it was the re-zero), or swap
+the **original** sensor back (if the offset returns, it is the sensor).
+The operator's remedy being a swap rather than a restart is weak
+evidence for the sensor, since a restart is the easier thing to try
+first.
+
+**Best single measurement when it recurs:** DC voltage across the
+suspect geophone's connector pins with the sensor **disconnected from
+the unit**.  A galvanic cell (mechanism 1) shows an open-circuit
+voltage; mechanisms 2 and 3 do not.
+
+**Note:** the surveyed population is subject to survivorship bias —
+flooded events were routinely deleted before this snapshot, so units
+that look clean here may have had episodes that were scrubbed.

docs/offset_investigation.md

@@ -0,0 +1,514 @@
+# The "offset" fault — investigation journal
+
+> ## ⚠ CORRECTED 2026-08-28 (same day) — the v1 detector was wrong
+>
+> Brian pushed back on the finding that offsets "come and go": in the field,
+> once a unit develops one it stays broken until the geophone is replaced.
+> He was right, and the challenge exposed **two real flaws** in the v1 detector:
+>
+> 1. **It scored only the axis with the largest peak.** A real event on one axis
+>    hid a persistent pedestal on another. BE12599 on 2026-08-21 read "clean"
+>    solely because Long had a 1.065 in/s event — Tran was sitting at
+>    **+0.4732 in/s** at that moment and was never examined.
+> 2. **It used the MEAN**, which a real transient perturbs. The **median** is the
+>    resting baseline — most samples sit at it, so a blast does not move it.
+>    Same event, Long channel: mean **+0.0783** vs median **-0.0050**.
+>
+> Both flaws manufactured false recoveries. The corrected detector
+> (`scratch/offset_scan2.py`, per-channel median) shows the pedestal is
+> **persistent**, exactly as the field experience says. See §2b and §3b.
+>
+> **Then Brian proposed a better detector still** — measure the floor during
+> the *pre-trigger* window, and require it to hold across pre/middle/end.
+> That is now the detector of record (§2c). Final answer: **5 of 45 units
+> (11%)**, stable across a 2x threshold range.
+>
+> Sections below that were written against v1 are marked; v1 numbers are kept
+> for the reasoning trail, not as current fact.
+
+A running record of the **offset** hardware fault on Instantel Series III
+seismographs: a geophone channel whose trace sits displaced from zero rather
+than centred on it.
+
+This is a *journal*, not a spec. Findings are dated, dead ends are kept with
+the reason they died, and every number says where it came from. When something
+here is superseded, strike it and say why rather than deleting it — the point
+is that a future session can tell what was actually established from what was
+merely believed at the time.
+
+Companion material:
+- `scratch/offset_scan.py` — the detector
+- `scratch/verify_against_ascii.py` — decoder verification harness
+- `docs/instantel_protocol_reference.md` — wire protocol, incl. the
+  unimplemented `SUB 0x0E` this investigation now wants
+
+---
+
+## TL;DR (current state, 2026-08-28)
+
+- **It is real device data, not a decode bug.** Settled early and confirmed
+  against Blastware's own ASCII exports.
+- **Base rate: 5–6 of 45 units (11–13%)** across the full DL2 archive,
+  2018–2026. This *confirms* the earlier 2-of-21 (9.5%) estimate from the much
+  smaller Terra-View DB — survivorship bias from deleted events had **not**
+  concealed a wave of cases.
+- **The fault is bimodal, not a drift continuum.** A unit is either clean or
+  grossly off. Loosening the amplitude threshold 11× adds no new units.
+- **The unit's own sensor check cannot see it.** 102 offset events, zero
+  sensor-check failures. Do not try to use it as a screen.
+- **Cause is still unsettled.** Instantel's autozero fixes the minority of
+  cases; the rest are hardware. We cannot yet tell which is which remotely.
+- **Best open lead:** `SUB 0x0E` (channel sensor data, 8 channels × 10 bytes,
+  unimplemented) may carry the very numbers Instantel says to check against
+  **2027–2069**. Untested.
+
+---
+
+## 1. What the fault looks like
+
+A healthy geophone trace is centred on zero. An offset channel is parked away
+from zero, so the channel **mean approaches its own peak**. In Blastware the
+signature is "parallel lines above or below the zero line" (Instantel's own
+wording).
+
+Consequences observed in the field:
+- The unit can **self-trigger on its own offset** when the displacement exceeds
+  the geo trigger level, producing streams of junk events with no ground
+  motion. Instantel has a separate FAQ for this symptom (13-0-22, *"Unit
+  triggers continuously without activity"*).
+- Recorded PPV for that channel is meaningless while the fault persists.
+
+---
+
+## 2. The detector
+
+Implemented in `scratch/offset_scan.py`. Operates on raw BW binaries only — no
+DB, no sidecars.
+
+```
+for each series-3 waveform binary:
+    decode -> per-channel ADC counts
+    dominant axis = channel with the largest |peak|
+    flag when   |mean| / peak  >  0.70
+          and   |mean|         >= 0.90 x the unit's geo trigger level
+episodes = per-serial runs of flagged events, split on a >12 h gap
+```
+
+Why each term:
+
+| term | purpose |
+|---|---|
+| `\|mean\|/peak > 0.7` | the discriminator. A DC-parked trace has mean ≈ peak. |
+| `\|mean\| >= 0.9 × trigger` | amplitude floor — suppresses quiet traces where mean and peak are both tiny and the ratio is meaningless. |
+| dominant axis only | the fault is per-channel; scoring all three dilutes it. |
+| 12 h episode gap | separates deployments/visits rather than counting events. |
+
+Trigger level comes from a paired `_ASCII.TXT` when one exists, else the
+per-serial median learned from that unit's ASCII files, else 0.2 in/s.
+
+**Known limitation.** Event traces contain real ground motion, so this can only
+see offsets large enough to *dominate* the trace. A mild offset on a real blast
+is invisible. Instantel's A/D-mode check (§5) is the only thing that sees the
+mild end. Our base rate is therefore a **gross-offset** rate.
+
+### 2b. Detector v2 — per-channel median (CURRENT)
+
+`scratch/offset_scan2.py`. Supersedes the above.
+
+```
+for each series-3 waveform binary:
+    for each geo channel independently:
+        pedestal = median(samples)          # resting baseline, robust to blasts
+        flag the CHANNEL when |pedestal| >= 0.025 in/s   (5 A/D counts)
+a unit has a real fault when a channel is flagged on >=3 CONSECUTIVE events
+```
+
+Why median: a DC pedestal shifts every sample, so it moves the median. A real
+event moves only a minority of samples, so it does not. This removes the need
+for the `m/p` ratio guard entirely — that guard existed only to compensate for
+using the mean.
+
+Why per-channel: the fault is on one geophone axis. Scoring only the dominant
+axis means any event with motion elsewhere hides it.
+
+Why "3 consecutive": the 0.025 in/s floor is only ~2x a healthy channel's
+resting median (observed 0.010-0.015), so isolated flags are noise. Persistence
+is the discriminator — and it is what the field experience predicts.
+
+---
+
+## 3b. Archive results, corrected (v2)
+
+| | v1 (dominant axis, mean) | **v2 (per-channel median)** |
+|---|---|---|
+| units with any flagged event | 6 of 45 | 19 of 45 |
+| **units with a sustained pedestal (>=3 consecutive)** | — | **8 of 45 (18%)** |
+| runs of >=3 consecutive | — | 29 |
+| runs of 1-2 events (noise) | — | 69 |
+
+Units with a sustained pedestal: **BE9558, BE10895, BE11007, BE11529, BE12599,
+BE13117, BE18003, BE18438**. BE10895 and BE18003 were invisible to v1.
+
+**The affected channel is most often Vert**, which v1 got wrong — it named
+whichever axis had the largest peak. BE13117 and BE18438 are both Vert faults.
+
+Longest / clearest runs:
+
+| unit | ch | span | events | median in/s |
+|---|---|---|---|---|
+| BE13117 | Vert | 2023-05-03 → 05-04 | 194 | 0.035 → **1.915** |
+| BE18438 | Vert | 2026-02-25 → 02-26 | 75 | 0.180 → 0.370 |
+| BE9558 | Vert | 2020-02-11 (6 h) | 33 | 0.065 → 0.090 |
+| BE12599 | Tran | 2026-08-14 → 08-23 | 8 | 0.030 → **0.565** |
+| BE18003 | Vert | 2021-03-17 → 06-11 | 3 | 0.040 → 0.060 |
+
+BE12599 began **2026-08-14**, not 08-17 as v1 reported, and was still faulting
+at the last event in the archive.
+
+### 2c. Detector v3 — PRE-TRIGGER floor + constant-floor test (CURRENT)
+
+`scratch/offset_scan3.py`. Brian's method, and better than v2 for a reason
+worth naming: **the pre-trigger window is definitionally quiet** — it is the
+buffer captured before the trigger fired — whereas a whole-record median is
+merely *robust* to the event. `pretrig_samples` comes from the STRT record.
+
+```
+per channel:
+    pre    = median of the first pretrig_samples samples
+    mid    = median of the middle third
+    end    = median of the final third
+    spread = max(pre,mid,end) - min(pre,mid,end)
+
+    offset when  |pre| >= floor  AND  spread <= 0.02 in/s
+real fault when a channel is flagged on >=3 CONSECUTIVE events
+```
+
+A DC offset is a **constant floor** — present before the trigger, during, and
+after. The spread test rejects transients (settling, handling, a long event
+tail) that move one segment relative to the others, which is what v2's
+whole-record median could not do.
+
+**The empirical noise floor justifies the threshold.** Across 19,244
+non-flagged channel-events the pre-trigger floor distributes as:
+
+| floor | share |
+|---|---|
+| −1 unit (−0.005) | 18.4% |
+| **0.000** | **62.7%** |
+| +1 unit (+0.005) | 13.4% |
+
+**94.5% within ±1 quantisation unit; median exactly +0.0000, mean −0.0008.**
+So there is **no systematic zero-point bias in the decoder** — an independent
+confirmation of the 32000-count scale. A healthy channel really does read
+0.000, and "any constant floor that is not 0.000" is the right signal, with
+±1 unit of slack for quantisation.
+
+**The result is threshold-insensitive**, which is what distinguishes a real
+signal from a tuned one:
+
+| floor | units flagged | sustained units |
+|---|---|---|
+| 2 units (0.010) | 34 | 15 ← into the noise |
+| 3 units (0.015) | 26 | 8 |
+| **4 units (0.020)** | 17 | **5** |
+| **5 units (0.025)** — Instantel's | 12 | **5** |
+| **8 units (0.040)** | 8 | **5** |
+
+### FINAL RESULT: 5 of 45 units (11%)
+
+**BE9558, BE11529, BE12599, BE13117, BE18438.**
+
+Unchanged across a 2x threshold range. BE11007 and BE10895 drop out — the
+spread test identifies them as transients, not pedestals.
+
+The 11% headline happens to match v1's, but the reasoning and the unit list
+differ: v1 included BE11007 and named the wrong *channel* on most units.
+
+---
+
+## 3. Archive results (2026-08-28)
+
+Source: DL2 event export, 6,577 **unique** series-3 waveforms, 45 units.
+See [`dl2-archive`](#8-data-and-tooling) for the `Sent/` mirror trap.
+
+**283 suspect events, 15 episodes, 6 of 45 units (13.3%).**
+Excluding BE11007 (§4, likely not an offset at all): **5 of 45 = 11.1%**.
+
+### Threshold sensitivity — the bimodality result
+
+Re-scoring the same corpus at a range of amplitude floors, with two
+ratio cut-offs (1 A/D count = 0.005 in/s, see §5):
+
+| \|offset\| floor | m/p > 0.7 | m/p > 0.9 |
+|---|---|---|
+| 5 cts (0.025 in/s) — *Instantel's own* | 333 ev / 6 units | 279 ev / **5 units** |
+| 10 cts (0.050) | 294 / 6 | 274 / 5 |
+| 20 cts (0.100) | 250 / 5 | 244 / 4 |
+| 40 cts (0.200) | 209 / 5 | 203 / 4 |
+| 80 cts (0.400) | 152 / 4 | 148 / 2 |
+| 160 cts (0.800) | 144 / 2 | 141 / 1 |
+
+Relaxing the floor by 11× (0.27 → 0.025 in/s) adds ~14% more events and **no
+new units**. There is no population of mild offsets hiding below our threshold
+*in event data*. Either a unit is clean or it is grossly off.
+
+---
+
+## 4. Per-unit case files
+
+Ordered by severity. `m/p` medians are on the offending channel.
+
+### BE13117 — one violent day, never again
+`145 / 454 events (32%)`, **1 episode**, 2023-05-04, 6.8 h.
+Offset climbed **0.393 → 1.875 in/s within the episode**. `m/p` median
+**0.996** — the trace is almost pure DC. No recurrence in the rest of its 454
+events. No ASCII files in the archive, so no calibration history.
+
+### BE18438 — recurring, months apart
+`87 / 293 (30%)`, **2 episodes**: 2025-11-15 (1.2 h, n=12, 0.279 → 0.369) and
+2026-02-25 (**28.8 h**, n=75, 0.183 → 0.366). `m/p` median 0.967.
+Clean across all 196 events preceding its 2025-08-12 calibration.
+
+### BE9558 — six years apart
+`38 / 196 (19%)`, **4 episodes**: 2020-02-11 (6.3 h, n=33, but only
+0.068 → 0.086 — very mild), then 2026-04-14, 2026-04-29, 2026-05-04
+(0.28–0.45). `m/p` median 0.919. Calibrated 2026-06-26; 0/7 events flagged
+after, but n=7 is far too small to call it fixed.
+
+### BE12599 — the live case ⚠
+`6 / 77 (8%)`, **6 single-event episodes, one per day at exactly 05:00**,
+2026-08-17 → 2026-08-23. Offset rose 0.383 → 0.565 then fell back to 0.345.
+`m/p` ≈ 0.965, geo trigger 0.3 in/s — **the offset exceeds the trigger level,
+so the unit is triggering on its own fault**. Last calibrated 2025-08-12.
+
+This is the most recent and the most useful: a currently-faulting unit is the
+natural experiment for the re-zero-vs-repair question (§7).
+
+### BE11529 — marginal
+`4 / 99 (4%)`, 1 episode 2025-07-08, 0.4 h, offsets only 0.051 → 0.058 in/s.
+`m/p` median 0.959, so DC-dominated, but the magnitude is near the noise of
+this method. Treat as unconfirmed.
+
+### BE11007 — probably NOT an offset
+`3 / 70 (4%)`, 1 episode 2022-01-17, offsets 5.500 → 6.904 in/s — by far the
+largest. But `m/p` is only **0.719–0.738** against ≥0.9 for every other unit,
+and the peaks are 7.6–9.4 in/s on a 10 in/s range. That reads as a **large
+real blast with asymmetric ground motion**, not a parked trace. Excluded from
+the headline base rate.
+
+---
+
+## 5. Instantel's own procedure and thresholds
+
+From two Instantel technical-support FAQs supplied 2026-08-28
+(answers **13-0-21** *"How to determine offsets"* and **12-0-10** *"Removing
+offsets on an Instantel Series III monitor"*; created 2008/2007, last updated
+2009-03-06).
+
+### Identifying (13-0-21)
+
+1. Create or use an event with the **manual minimum trigger** set for the
+   connected geophone and microphone — i.e. an event that recorded no real data.
+2. Save it and open in Blastware.
+3. An offset shows as **parallel lines above or below the zero line**.
+4. Put the unit in **A/D mode** — on Series III, press and hold `OPTION`, then
+   press `START MONITOR`.
+5. **Display counts higher than 5**, with no vibration or overpressure present,
+   indicate an offset.
+
+### Removing — the autozero (12-0-10)
+
+1. Be in a **quiet area with low vibration**.
+2. Power on the Blastmate III / Minimate Plus.
+3. Connect the geophone and microphone — **LINEAR mic only**.
+   ⚠ *Do not connect an "A" weight microphone, regardless of what the monitor
+   displays.*
+4. Press `Test`.
+5. Wait for the **Sensor Check** results to appear.
+6. Press `OPTION` and `START MONITOR` **simultaneously**.
+7. `Performing Autozero` appears; press `Enter`.
+8. Confirm the sensors are properly connected; press `Enter`.
+9. Wait for the autozero to complete.
+10. Press `Enter` twice → Main Menu, *Ready To Monitor*, offset corrected.
+
+### The go/no-go number — 2027 to 2069
+
+> When you perform an Autozero on any Series III unit, the lists of numbers in
+> the **X1 and X8 gains should all be between 2027 and 2069**. If not, repeat
+> the Autozero. **If the numbers are extremely out of the specified range, then
+> the unit should be sent in for repair.**
+>
+> If this process does not remove the offset problem, return the unit **and
+> sensors** to Instantel for repair.
+
+This is the documented explanation for the field experience (Brian's dad,
+2026-08-28) that **a re-zero works maybe 10% of the time** — the autozero only
+recovers units whose zero reference is still near-correct.
+
+### Scale derivation (inference, well-supported — not proven)
+
+2048 is 12-bit midscale. Our codec's geo full scale is 32000 internal counts =
+10 in/s, with 1 decoder unit = 16 counts = exactly 0.005 in/s
+(`geo-full-scale-is-32000-counts`). ±2000 A/D counts about 2048 therefore maps
+to ±10 in/s at **0.005 in/s per A/D count**. That makes:
+
+- Instantel's ">5 counts" threshold ≈ **0.025 in/s**
+- the 2027–2069 window = **±21 counts = ±0.105 in/s** of tolerated zero error
+
+Consistent and mutually corroborating, but we have not confirmed the A/D-count
+scale directly from a device reading.
+
+---
+
+## 6. Ruled out — keep these dead
+
+### Condensation / humidity — DEAD (2026-08-25)
+Proposed, then killed by its own controls: BE18438 stayed flat across a 10-hour
+overnight gap, and only 2 of 21 units showed the fault while 19 sat in the same
+weather. The apparent "diurnal cycle" was an artifact of binning by hour-of-day
+across two days. See `waveform-dc-offset-is-real-device-data`.
+
+### Clipping as a false-positive source — RULED OUT (2026-08-28)
+A rail-hitting trace would fake an offset (mean → peak). It isn't happening:
+median suspect peak is only **10% of full scale**, p90 is 18.6%. Only BE11007's
+3 events exceed 50% FS, and none reach 98%.
+
+### The sensor check as a predictor — DOES NOT WORK (2026-08-28)
+Tested on 102 offset events across 4 units:
+
+| unit | state | n | failed | median ratio | median freq |
+|---|---|---|---|---|---|
+| BE11529 | offset | 4 | **0** | 3.90 | 7.6 |
+| BE11529 | clean | 14 | 0 | 3.80 | 7.5 |
+| BE12599 | offset | 6 | **0** | 4.00 | 7.4 |
+| BE12599 | clean | 13 | 0 | 4.00 | 7.6 |
+| BE18438 | offset | 87 | **0** | 3.70 | 7.6 |
+| BE18438 | clean | 25 | 0 | 3.80 | 7.5 |
+| BE9558 | offset | 5 | **0** | 3.90 | 7.8 |
+| BE9558 | clean | 44 | 0 | 3.80 | 7.5 |
+
+Zero failures on either side and indistinguishable ratios/frequencies. The
+swing test measures geophone frequency response and damping — it never examines
+DC zero. **A grossly offset unit passes its own self-check.** This is why the
+fault goes unnoticed until somebody looks at waveforms.
+
+### "Offsets are transient / come and go on their own" — RETRACTED 2026-08-28
+v1 reported episodes lasting hours that ended spontaneously. **This was an
+artifact of the v1 detector** (see the banner at the top). With the per-channel
+median, the pedestal persists. Every clear case reads clean again only after a
+multi-day-to-multi-month gap consistent with service: BE13117 6 days, BE18438
+24 days, BE9558 63 days **with a confirmed Instantel calibration inside the
+gap**. BE12599 never reads clean — it is still faulting at the end of the
+archive. This matches the operational experience: once a unit develops an
+offset it stays broken until the geophone is replaced.
+
+### "Offsets develop N months after calibration" — CONFOUNDED, NOT A FINDING
+Tempting, and it looked strong:
+
+| unit | suspect before latest cal | after |
+|---|---|---|
+| BE18438 | 0 / 196 | 87 / 97 |
+| BE12599 | 0 / 62 | 6 / 15 |
+| BE11529 | 0 / 82 | 4 / 17 |
+| BE9558 | 38 / 189 | 0 / 7 |
+
+But bucketing suspects by months-since-calibration gives **one unit per bucket**:
+`0–3mo={BE11529}`, `3–6 & 6–9mo={BE18438}`, `9–12mo={BE9558}`,
+`12–15mo={BE12599}`. The apparent "51% failure rate at 6–9 months" is entirely
+BE18438's single February 2026 episode. Five units with roughly one episode
+each cannot support a population trend. **Do not re-derive this.**
+
+Also note: all affected units are calibrated on a **~12–13 month cadence**, so
+"sent to Instantel" is the routine annual schedule, not evidence of a
+fault-driven return.
+
+---
+
+## 7. Open questions
+
+### Q1 — Is it a latched bad zero or analog degradation?
+The question that decides everything. A latched zero is correctable (possibly
+over the wire); degradation means a repair. Instantel's 2027–2069 rule implies
+*both* populations exist, with the split roughly 10/90 in the field.
+
+**BE12599 is the natural experiment** — faulting as of 2026-08-23. Read its
+values, run the autozero, read them again.
+
+### Q2 — Can we read the autozero numbers over the wire? (best lead)
+Instantel says to check *"the lists of numbers in the **X1 and X8 gains**"* —
+4 sensors × 2 gains = **8 channels**. The protocol reference already documents
+an unimplemented command with exactly that shape:
+
+```
+SUB 0x0E  ->  RSP 0xF1     "channel sensor data"
+  2-step read; channel selector in params[6:8] = 0x0000..0x0007
+  data length 0x0A (10 bytes) per channel
+```
+
+Blastware's *Unit Channel Test* sequence:
+`POLL×N → 0x15 → 0x01 → 0x08 → 0x01 → 0x0E×8 → 0x98×2 → 0x0E×8`
+— note the **second `0x0E` pass carries live ADC readings**.
+
+**Hypothesis (untested):** `0x0E` returns the numbers Instantel wants compared
+against 2027–2069. If true, SFM could diagnose an offset remotely *and* predict
+whether a re-zero will succeed — converting a 10%/90% shipping gamble into a
+decision made before packing a box.
+
+**How to test.** `bridges/ach_mitm.py` is a generic TCP proxy:
+
+```bash
+python bridges/ach_mitm.py --bw-host <MODEM_IP> --bw-port 9034 --listen-port 9999
+```
+
+Point Blastware at the proxy and run **Unit Channel Test**.
+⚠ In this topology the output filenames are reversed — the tool labels the
+*connecting* side "unit", so `raw_s3_*.bin` holds Blastware's bytes and
+`raw_bw_*.bin` the unit's.
+
+Capture priority: (1) BE12599 while faulting, (2) a known-good unit as control,
+(3) before/after an autozero on the same unit. Eight 10-byte payloads with an
+expected value near 2048 is a very constrained puzzle.
+
+### Q3 — What is the mild-offset rate?
+Unmeasurable from event files (§2). Only the A/D-mode check sees it. Would
+need a fleet sweep in A/D mode, or Q2 to succeed.
+
+### Q4 — Does an offset recur on the same unit after service?
+BE9558 shows episodes in 2020 and 2026; BE18438 twice in four months. Suggestive
+of recurrence, but service records aren't in the data — only calibration dates.
+
+---
+
+## 8. Data and tooling
+
+| what | where |
+|---|---|
+| detector | `scratch/offset_scan.py` |
+| current results | `/home/serversdown/dl2-archive/offset_archive.csv` |
+| earlier candidate list (Terra-View DB, 274 events) | `scratch/offset_candidates.csv` |
+| archive working copy | `/home/serversdown/dl2-archive/files/` |
+| archive source | NAS `DeathStar` 10.0.0.2, `/volume1/Uploads/TMI/DL2-Event-backup-8-25-26/Event/autocall home/` |
+
+⚠ **The DL2 export keeps a byte-identical `Sent/` mirror of its root.** 13,077
+waveform paths are 6,577 distinct files. Always dedupe by basename — this
+doubled two reported figures before it was caught.
+
+---
+
+## 9. Chronology
+
+| date | event |
+|---|---|
+| 2026-08-25 | Reported as a *waveform decode bug* — traces with a DC offset. Investigation shows the offset is **real device data**; the decoder is correct. |
+| 2026-08-25 | Brian relays his dad's description: a known hardware fault called an "offset"; usually sent to Instantel. |
+| 2026-08-25 | First detection pass over the Terra-View DB: **2 of 21 units**, 274 events, 11 months. Flagged as vulnerable to survivorship bias — flooded events were routinely deleted. |
+| 2026-08-25 | Condensation hypothesis proposed, then **killed by its own controls**. |
+| 2026-08-25 | Parked pending the multi-year archive. |
+| 2026-08-28 | DL2 archive pulled (33 GB, 546k files; 6.6 GB working set). |
+| 2026-08-28 | Archive scan: **6 of 45 units**, 283 events, 15 episodes. Prior base rate **confirmed**, not overturned. |
+| 2026-08-28 | Clipping ruled out; `m/p` established as the discriminator; BE11007 reclassified as probably a real blast. |
+| 2026-08-28 | Calibration-timing correlation attempted and **rejected as confounded**. |
+| 2026-08-28 | Instantel FAQs supplied: autozero procedure, the **2027–2069** window, the **>5 counts** threshold. Explains the ~10% re-zero success rate. |
+| 2026-08-28 | Bimodality established; sensor check proven **blind** to offsets; `SUB 0x0E` identified as the best open lead. |
+| 2026-08-28 | **v1 detector retracted.** Brian challenged the "come and go" finding against field experience. Two flaws found: dominant-axis-only scoring and mean-instead-of-median. Corrected detector shows persistent pedestals on **8 of 45 units**, and the gaps are service windows. |
+| 2026-08-28 | **Detector v3 (Brian's method):** pre-trigger floor + pre/mid/end consistency. Healthy channels proven to sit at 0.000 +/-1 unit (94.5%), confirming no decoder zero-point bias. Final: **5 of 45 units (11%)**, threshold-insensitive. |

docs/superpowers/plans/2026-08-25-b2a-reviewed-real-twin-propagation.md

@@ -0,0 +1,264 @@
+# Phase B2-A (seismo-relay) — reviewed_real + 3-state mirror + twin propagation
+
+> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
+
+**Goal:** Give SFM a persisted, queryable `reviewed_real` flag alongside `false_trigger` (mirrored from the sidecar review block, mutually exclusive), and propagate a review to an event's histogram/waveform twin so flagging one flags both.
+
+**Architecture:** Mirror the existing `false_trigger` mechanism. New `events.reviewed_real` column (auto-migrate). `update_event_review` mirrors BOTH flags from the sidecar review block and enforces mutual exclusivity on the columns. A `find_twins` matcher (same serial + identical peak_vector_sum + timestamp within a window) drives `propagate_review_to_twins`, which the sidecar-PATCH endpoint calls after mirroring the primary. Terra-View reads `reviewed_real` from `/db/events` (SELECT *).
+
+**Tech Stack:** Python 3.10, raw sqlite3, FastAPI, pytest. Runner: `/home/serversdown/seismo-relay/.venv/bin/python3`.
+
+## Global Constraints
+
+- 3 review states are **mutually exclusive**: an event is `false_trigger=1` XOR `reviewed_real=1` XOR neither. Setting one true forces the other's column to 0.
+- The sidecar JSON stays the source of truth for full review state; the `false_trigger`/`reviewed_real` columns are derived indexes (like today). B2-A propagates twins at the **column** level (what the feed/peak/export read); twin sidecars are not rewritten (known limitation — noted).
+- Twin match = **same serial AND identical `peak_vector_sum` (exact equality) AND `timestamp` within ± window (default 300 s), excluding the event itself.** Identical PVS is the safety anchor.
+- Known pre-existing test failures (~16, missing gitignored fixtures under `tests/fixtures/`) are unrelated — confirm zero NEW failures, don't try to fix them.
+- Run tests with `/home/serversdown/seismo-relay/.venv/bin/python3 -m pytest`.
+
+---
+
+### Task 1: `reviewed_real` column on events
+
+**Files:** Modify `sfm/database.py` (`_SCHEMA` CREATE TABLE `events` + the Migration-1 rebuild `CREATE TABLE` + the `_migrate` ADD COLUMN loop). Test: `tests/test_reviewed_real_column.py`.
+
+**Interfaces:** Produces `events.reviewed_real INTEGER NOT NULL DEFAULT 0`.
+
+- [ ] **Step 1: Failing test**
+```python
+# tests/test_reviewed_real_column.py
+import sqlite3
+from sfm.database import SeismoDb
+def _cols(db):
+    with sqlite3.connect(db.db_path) as c:
+        return {r[1] for r in c.execute("PRAGMA table_info(events)")}
+def test_fresh_db_has_reviewed_real(tmp_path):
+    assert "reviewed_real" in _cols(SeismoDb(tmp_path/"s.db"))
+def test_existing_db_migrates_reviewed_real(tmp_path):
+    p = tmp_path/"s.db"; db = SeismoDb(p)
+    with sqlite3.connect(p) as c:
+        c.execute("ALTER TABLE events DROP COLUMN reviewed_real")
+    assert "reviewed_real" not in _cols(db)      # dropped (read via existing handle/connection)
+    SeismoDb(p)                                   # re-open migrates
+    assert "reviewed_real" in _cols(SeismoDb(p))
+```
+> If sqlite < 3.35 lacks DROP COLUMN, fall back to building a table without the column and asserting re-open adds it (same as the shape-columns test did).
+
+- [ ] **Step 2: Run → FAIL** (`pytest tests/test_reviewed_real_column.py -q`).
+- [ ] **Step 3: Implement**
+  - In `_SCHEMA` `events` CREATE TABLE, after the `false_trigger ... DEFAULT 0,` line add: `    reviewed_real           INTEGER NOT NULL DEFAULT 0,  -- 0=no, 1=operator-confirmed real (mutually exclusive with false_trigger)`
+  - In the `_migrate` ADD COLUMN loop tuple add: `("reviewed_real", "INTEGER NOT NULL DEFAULT 0"),`
+  - **Do NOT** add it to the Migration-1 rebuild `CREATE TABLE events (...)` block — that block uses a positional `INSERT ... SELECT * FROM events_old` and, by convention, contains only the columns that existed when Migration 1 was written; every later column is added by the ADD COLUMN loop only. Adding it there crashes `_migrate` on genuinely legacy (pre-Migration-1) DBs.
+- [ ] **Step 4: Run → PASS.**
+- [ ] **Step 5: Commit** `feat(db): reviewed_real column on events (+ auto-migrate)`
+
+---
+
+### Task 2: `update_event_review` mirrors both flags + mutual exclusivity
+
+**Files:** Modify `sfm/database.py` `update_event_review`. Test: `tests/test_update_event_review_reviewed_real.py`.
+
+**Interfaces:** Consumes a `review` dict that may carry `false_trigger` and/or `reviewed_real` (bools). Produces mutually-exclusive column state.
+
+- [ ] **Step 1: Failing test**
+```python
+# tests/test_update_event_review_reviewed_real.py
+from sfm.database import SeismoDb
+from minimateplus.models import Event
+def _ins(db, eid_key="01110000", serial="BE1"):
+    ev = Event(index=0); ev._waveform_key = bytes.fromhex(eid_key)
+    db.insert_events([ev], serial=serial)
+    return db.query_events(serial=serial)[0]["id"]
+def test_confirm_real_sets_and_clears_ft(tmp_path):
+    db = SeismoDb(tmp_path/"s.db"); eid = _ins(db)
+    db.update_event_review(eid, {"false_trigger": True})
+    assert db.get_event(eid)["false_trigger"] == 1
+    db.update_event_review(eid, {"reviewed_real": True})   # confirming real clears FT
+    row = db.get_event(eid)
+    assert row["reviewed_real"] == 1 and row["false_trigger"] == 0
+def test_flag_ft_clears_reviewed_real(tmp_path):
+    db = SeismoDb(tmp_path/"s.db"); eid = _ins(db)
+    db.update_event_review(eid, {"reviewed_real": True})
+    db.update_event_review(eid, {"false_trigger": True})
+    row = db.get_event(eid)
+    assert row["false_trigger"] == 1 and row["reviewed_real"] == 0
+```
+> Build the Event inline like `tests/test_zc_freq_columns.py` if the import differs.
+
+- [ ] **Step 2: Run → FAIL.**
+- [ ] **Step 3: Implement** — replace the body of `update_event_review` so it handles both keys:
+```python
+    if not isinstance(review, dict):
+        return False
+    has_ft = "false_trigger" in review
+    has_real = "reviewed_real" in review
+    if not has_ft and not has_real:
+        with self._connect() as conn:
+            row = conn.execute("SELECT 1 FROM events WHERE id=?", (event_id,)).fetchone()
+        return row is not None
+    sets = {}
+    if has_ft:
+        sets["false_trigger"] = 1 if review.get("false_trigger") else 0
+    if has_real:
+        sets["reviewed_real"] = 1 if review.get("reviewed_real") else 0
+    # mutual exclusivity: a true in one forces the other column to 0
+    if sets.get("false_trigger") == 1:
+        sets["reviewed_real"] = 0
+    if sets.get("reviewed_real") == 1:
+        sets["false_trigger"] = 0
+    assign = ", ".join(f"{k}=?" for k in sets)
+    params = list(sets.values()) + [event_id]
+    with self._connect() as conn:
+        cur = conn.execute(f"UPDATE events SET {assign} WHERE id=?", params)
+    return cur.rowcount > 0
+```
+- [ ] **Step 4: Run → PASS** (+ run `tests/test_*false_trigger*`/existing review tests to confirm no regression).
+- [ ] **Step 5: Commit** `feat(db): update_event_review mirrors reviewed_real + enforces 3-state exclusivity`
+
+---
+
+### Task 3: `find_twins` matcher
+
+**Files:** Modify `sfm/database.py` (add `find_twins`). Test: `tests/test_find_twins.py`.
+
+**Interfaces:** Produces `find_twins(event_id, *, window_seconds=300) -> list[dict]` — same serial, identical peak_vector_sum, timestamp within ±window, excluding self.
+
+- [ ] **Step 1: Failing test**
+```python
+# tests/test_find_twins.py
+import datetime
+from sfm.database import SeismoDb
+from minimateplus.models import Event, Timestamp
+def _ins(db, key, serial, pvs, ts):
+    ev = Event(index=0); ev._waveform_key = bytes.fromhex(key)
+    ev.timestamp = ts
+    # peak_vector_sum comes from peak_values; simplest: insert then UPDATE pvs directly
+    db.insert_events([ev], serial=serial)
+    row = [r for r in db.query_events(serial=serial) if r["waveform_key"] == key][0]
+    import sqlite3
+    with sqlite3.connect(db.db_path) as c:
+        c.execute("UPDATE events SET peak_vector_sum=? WHERE id=?", (pvs, row["id"]))
+    return row["id"]
+def test_find_twins_matches_same_serial_pvs_near_time(tmp_path):
+    db = SeismoDb(tmp_path/"s.db")
+    base = Timestamp(raw=b"", flag=0x10, year=2026, unknown_byte=0, month=2, day=25, hour=20, minute=19, second=5)
+    twin = Timestamp(raw=b"", flag=0x10, year=2026, unknown_byte=0, month=2, day=25, hour=20, minute=19, second=45)
+    far  = Timestamp(raw=b"", flag=0x10, year=2026, unknown_byte=0, month=2, day=25, hour=21, minute=0,  second=0)
+    a = _ins(db, "01110001", "BE1", 0.4763, base)
+    b = _ins(db, "01110002", "BE1", 0.4763, twin)   # twin: same serial+pvs, 40s apart
+    c = _ins(db, "01110003", "BE1", 0.4763, far)    # same pvs but >window away
+    d = _ins(db, "01110004", "BE1", 0.9999, twin)   # near time but different pvs
+    ids = {r["id"] for r in db.find_twins(a, window_seconds=300)}
+    assert ids == {b}
+```
+> Adjust the Event/Timestamp construction to match how `tests/test_waveform_store.py::_make_synthetic_event` builds them if fields differ.
+
+- [ ] **Step 2: Run → FAIL.**
+- [ ] **Step 3: Implement**
+```python
+    def find_twins(self, event_id: str, *, window_seconds: int = 300) -> list[dict]:
+        row = self.get_event(event_id)
+        if not row:
+            return []
+        serial = row.get("serial"); pvs = row.get("peak_vector_sum"); ts = row.get("timestamp")
+        if serial is None or pvs is None or not ts:
+            return []
+        try:
+            t = datetime.datetime.fromisoformat(ts.replace(" ", "T"))
+        except ValueError:
+            return []
+        lo = (t - datetime.timedelta(seconds=window_seconds)).isoformat()
+        hi = (t + datetime.timedelta(seconds=window_seconds)).isoformat()
+        with self._connect() as conn:
+            rows = conn.execute(
+                "SELECT * FROM events WHERE serial=? AND id!=? AND peak_vector_sum=? "
+                "AND timestamp BETWEEN ? AND ?",
+                (serial, event_id, pvs, lo, hi),
+            ).fetchall()
+        return [dict(r) for r in rows]
+```
+- [ ] **Step 4: Run → PASS.**
+- [ ] **Step 5: Commit** `feat(db): find_twins (serial + identical PVS + time window)`
+
+---
+
+### Task 4: propagate a review to twins + wire into the sidecar-PATCH endpoint
+
+**Files:** Modify `sfm/database.py` (add `propagate_review_to_twins`); `sfm/server.py` (`db_event_sidecar_patch`). Test: `tests/test_twin_propagation.py`.
+
+**Interfaces:** `propagate_review_to_twins(event_id, *, window_seconds=300) -> list[str]` copies the event's `false_trigger`/`reviewed_real` columns onto each twin; returns twin ids.
+
+- [ ] **Step 1: Failing test** (DB-level)
+```python
+# tests/test_twin_propagation.py — reuse the _ins helper pattern from test_find_twins
+def test_propagate_copies_flags_to_twins(tmp_path):
+    db = SeismoDb(tmp_path/"s.db")
+    # (build primary + twin via the _ins helper as in test_find_twins)
+    # flag the primary FT, then propagate:
+    db.update_event_review(primary_id, {"false_trigger": True})
+    moved = db.propagate_review_to_twins(primary_id)
+    assert twin_id in moved
+    assert db.get_event(twin_id)["false_trigger"] == 1
+```
+- [ ] **Step 2: Run → FAIL.**
+- [ ] **Step 3: Implement**
+  - `sfm/database.py`:
+```python
+    def propagate_review_to_twins(self, event_id: str, *, window_seconds: int = 300) -> list[str]:
+        row = self.get_event(event_id)
+        if not row:
+            return []
+        ft = 1 if row.get("false_trigger") else 0
+        real = 1 if row.get("reviewed_real") else 0
+        twins = self.find_twins(event_id, window_seconds=window_seconds)
+        moved = []
+        with self._connect() as conn:
+            for tw in twins:
+                conn.execute("UPDATE events SET false_trigger=?, reviewed_real=? WHERE id=?",
+                             (ft, real, tw["id"]))
+                moved.append(tw["id"])
+        return moved
+```
+  - `sfm/server.py` `db_event_sidecar_patch`: after the existing `_get_db().update_event_review(event_id, new_sidecar.get("review", {}))`, add:
+```python
+        # Propagate the review to the event's histogram/waveform twin(s) so
+        # flagging one flags both (column-level; twins share serial+PVS+near time).
+        try:
+            _get_db().propagate_review_to_twins(event_id)
+        except Exception as exc:
+            log.warning("twin review-propagation failed for %s: %s", event_id, exc)
+```
+  (Guard the `if body.review is not None:` block so propagation only runs when review changed.)
+- [ ] **Step 4: Run → PASS.**
+- [ ] **Step 5: Commit** `feat(review): propagate false_trigger/reviewed_real to twins on sidecar PATCH`
+
+---
+
+### Task 5: expose in feed guard + version bump
+
+**Files:** Test `tests/test_reviewed_real_in_feed.py`; `pyproject.toml`, `sfm/server.py` version, `CHANGELOG.md`.
+
+- [ ] **Step 1: Guard test** — a `query_events` row dict includes `reviewed_real` (SELECT * returns it).
+```python
+from sfm.database import SeismoDb
+from minimateplus.models import Event
+def test_query_events_includes_reviewed_real(tmp_path):
+    db = SeismoDb(tmp_path/"s.db")
+    ev = Event(index=0); ev._waveform_key = bytes.fromhex("01110000")
+    db.insert_events([ev], serial="BE1")
+    assert "reviewed_real" in db.query_events(serial="BE1")[0]
+```
+- [ ] **Step 2: Run → PASS** (columns already present from Task 1).
+- [ ] **Step 3: Bump** `pyproject.toml` 0.24.0 → 0.25.0; `sfm/server.py` version="0.25.0"; add `## v0.25.0` CHANGELOG entry ("reviewed_real 3-state review flag + histogram/waveform twin review-propagation").
+- [ ] **Step 4: Full suite** — confirm zero NEW failures beyond the ~16 pre-existing.
+- [ ] **Step 5: Commit** `chore(release): v0.25.0 — reviewed_real + twin review-propagation`
+
+---
+
+## Self-Review
+
+**Spec coverage:** reviewed_real column (T1) ✓; mirror + mutual exclusivity (T2) ✓; twin match serial+identical-PVS+window (T3) ✓; twin propagation wired into the review path (T4) ✓; feed exposure + version (T5) ✓. Twin **sidecar** rewrite intentionally deferred (column-level only — documented in Global Constraints); this is the SFM half — the 3-state UI + confirm-real PATCH is the separate **B2-B (terra-view)** plan.
+
+**Placeholder scan:** none — every step has real code (the two "adjust Event construction to match test_waveform_store" notes point at a concrete existing helper).
+
+**Type consistency:** `false_trigger`/`reviewed_real` INTEGER columns used identically across T1/T2/T4; `find_twins`→`propagate_review_to_twins` both key on the same match; server calls the DB methods by the exact names T2–T4 define.

docs/waveform_codec_re_status.md

@@ -1,3 +1,22 @@
+> ## SUPERSEDED 2026-08-25 — the body is a RECORD CHAIN
+>
+> The tag-dispatch model described in this document — `40 NN` segment headers,
+> tagless headers, channel rotation — is **wrong**.  It produced nearly-correct
+> output only because the block table happens to tile the data sections.
+>
+> The body is a chain of self-delimiting per-channel records: `off+2` is a
+> uint16 BE length, `next = off + 2 + len`, and the chain ends on a record whose
+> chan_id is `0x06`.  A 3-valued mode enum at `off+8` selects delta / absolute /
+> raw-12-bit semantics.  `40 NN` is an ordinary int16 BE data block.
+>
+> See the record-chain section of `docs/instantel_protocol_reference.md` §7.6.1
+> and the implementation in `minimateplus/waveform_codec.py`.
+>
+> Result: all four channels equal length in 1388/1388 files (was 156/1388);
+> ASCII sample-count exact 75/75, fully exact 73/75; device PPV 1306/1306.
+>
+> This document is retained as the reasoning trail.
+
 # Waveform body codec — FULLY DECODED (2026-05-11)
 
 This is the **clean working note** for the body-codec reverse-engineering
@@ -102,12 +121,24 @@ correct.
 |          |                       | nibble first; signed 0..7 / 8..F = -8..-1)|
 | `20 NN`  | NN + 2 bytes          | int8 signed deltas (1 per byte)          |
 | `00 NN`  | 2 bytes               | RLE: append NN copies of current value   |
-| `30 NN`  | NN*2 in data section, | Unknown content.  Only in loud-from-     |
-|          | NN*4 in trailer       | start events.                            |
-| `40 02`  | 20 bytes (fixed)      | Segment header                           |
+| `30 NN`  | NN*1.5 + 2 in data    | 12-bit signed deltas (see below).        |
+|          | section, NN*4 trailer |                                          |
+| `40 NN`  | 2*NN + 16 bytes       | Segment header (NN = prev-channel deltas)|
 
 NN is always a multiple of 4.
 
+**Wide-NN forms.**  `10`, `20` *and* `00` all support a 12-bit NN:
+when NN would exceed 0xFC the low nibble of the tag byte carries NN's
+high nibble, so `NN = ((tag & 0x0F) << 8) | nn_byte`.  Confirmed for
+`1X`/`2X` in 2026-05-11 and for `0X` (RLE) in 2026-08-25 — e.g.
+`01 0c` = a 268-sample zero-delta run.
+
+**`40 NN` is variable width.**  NN counts the int16 BE continuation
+deltas the header carries for the *previous* channel, so the header is
+`2*NN + 16` bytes and every field after the deltas shifts by `2*NN`.
+`40 01` (18 B) and `40 03` (22 B) both occur alongside the common
+`40 02` (20 B).  Confirmed 2026-08-25.
+
 Implementation: `walk_body()` in `minimateplus/waveform_codec.py`.
 
 ### 7-byte preamble
@@ -207,6 +238,61 @@ TL;DR table above are now locked in by pytest regression tests.
    still bails out partway through.  Lower priority since the other
    7 events walk cleanly.
 
+4. **Variable-prefix segment descriptors** (found 2026-08-25).
+   3 of 75 ground-truth production events still truncate.  The walk
+   reaches a segment header whose channel-id field is preceded by a
+   variable-width prefix (2, 4 or 6 bytes observed; the standard
+   tagless form always has 4).  These also carry an `01 00` marker
+   instead of `02 00`.  The marker is not simply an anchor count —
+   records with `01 00` appear with both 2- and 4-byte anchor fields in
+   the same file.  Examples: `BE12599/N599LPNB.JF0W` @1155,
+   `BE12599/N599LPWJ.980W` @849, `BE9558/K558LOF2.820W` @1485.
+
+## Segment header: channel id and tagless form — 2026-08-25
+
+The 4-byte field previously read as a "monotonic uint32 LE counter" is
+`[channel_id][00][00][segment_index]`, with `0x46`=Tran `0x47`=Vert
+`0x48`=Long `0x49`=MicL.  Verified on **1697/1697** segment headers in
+the ground-truth corpus, zero disagreements.  `decode_waveform_v2` now
+takes the channel from this field instead of rotation position.
+
+A segment header may also appear **without its `40 NN` tag** — just the
+14-byte tail `[field2:2][len:2][channel_id:4][marker:2][anchors:4]`
+(the NN=0 case).  `is_tagless_segment_header()` detects it from the six
+bytes at `[4:10]`.
+
+## Geo scale: full scale is 32000 counts — 2026-08-25
+
+One decoder unit (16 ADC counts) is exactly 0.005 in/s, so Normal range
+(10.000 in/s) is `10.0 / (0.005/16)` = **32000** ADC counts.  Consumers
+that divided by 32768 read every geophone sample 2.34% low.  Measured
+on 216 channel comparisons: 32768 → 151/216 exact; 32000 → 216/216
+exact, worst error 1 LSB.
+
+**Scope — not waveform-specific.**  The scale is applied where ADC
+counts become physical units, which every event passes through
+regardless of source codec:
+
+| source | median ratio ours/device, 32768 | with 32000 |
+|---|---|---|
+| series-3 waveform (vs ASCII sample table) | 0.9766 | **1.0000** |
+| series-3 histogram (vs ASCII PPV, n=1137) | 0.9766 | **1.0000** |
+| series-4 Thor IDF (vs device peak, n=1468) | 0.960 | **0.983** |
+
+The four block-framing fixes are waveform-only — `histogram_codec` is
+untouched by them.
+
+## Ground-truth corpus (2026-08-25)
+
+Beyond the bundled fixtures, the production waveform store keeps each
+event's original Blastware ASCII export at
+`<store>/<serial>/<filename>_ASCII.TXT`.  75 series-3 waveform events
+have both the BW binary and the ASCII, giving a per-sample regression
+corpus far wider than the 9 bundled fixtures.  Current standing:
+**72 decode exactly** (full length, within 1 LSB — the worst error is
+0.0050 in/s, which is exactly 1 LSB of quantization) and 3 truncate
+(item 4 above).  Zero events have full-length value errors.
+
 ## `30 NN` block format — CRACKED 2026-05-11 late
 
 The `30 NN` block carries `NN` 12-bit signed deltas, packed as `NN/4`

micromate/idf_file.py

@@ -47,7 +47,19 @@ from dataclasses import dataclass
 from pathlib import Path
 from typing import Optional, Union
 
-from minimateplus.waveform_codec import decode_waveform_v2
+# Thor IDFW bodies are pinned to the SUPERSEDED tag-dispatch decoder.
+#
+# _find_waveform_body_offset() trial-decodes every candidate offset and keeps
+# whichever yields the most samples.  The series-3 record-chain decoder
+# correctly returns None where the legacy walker returned garbage, which
+# changes that heuristic's winner on 33 of 577 files.  The net effect measured
+# 2026-08-25 was positive (all-channels-equal 8/577 -> 506/577, mean abs PPV
+# error 0.228 -> 0.173 in/s) but Thor has no ASCII ground truth in the corpus
+# and its geo scaling is separately suspect, so the switch is deferred until
+# the body-offset search is reworked to use the record chain directly.
+from minimateplus.waveform_codec import (
+    decode_waveform_legacy as decode_waveform_v2,
+)
 
 from .models import IdfEvent, IdfPeaks, IdfReport
 

minimateplus/event_file_io.py

@@ -27,6 +27,7 @@ from typing import Optional, Union
 from .models import Event, PeakValues, ProjectInfo, Timestamp
 from . import blastware_file as _bw  # avoid circular reference at module load
 from .bw_ascii_report import BwAsciiReport
+from . import waveform_codec as _wc
 from .waveform_codec import decode_waveform_v2, decoded_to_adc_counts
 from .histogram_codec import decode_histogram_body
 
@@ -49,7 +50,7 @@ SIDECAR_KIND   = "sfm.event"
 # bumped without a `pip install` re-run — leading to confusing stale
 # version stamps in sidecars.  Bump this constant and CHANGELOG.md
 # together at release time.
-TOOL_VERSION = "0.21.1"
+TOOL_VERSION = "0.27.0"
 
 try:
     # Best-effort: prefer the installed metadata when it's NEWER than the
@@ -659,6 +660,11 @@ def file_sha256(path: Union[str, Path], chunk_size: int = 65536) -> str:
 _GEO_NORMAL_FS_INS  = 10.0
 _GEO_SENSITIVE_FS_INS = 1.250
 _INT16_FS = 32768.0
+# Geophone full-scale count — 32000, not 32768.  One decoder unit (16 ADC
+# counts) is exactly 0.005 in/s, so 10.000 in/s = 32000 counts.  Must match
+# sfm.event_hdf5._GEO_INT16_FS or sidecar peaks disagree with the plotted
+# waveform by 2.3%.  Confirmed 2026-08-25 against the BW ASCII corpus.
+_GEO_INT16_FS = 32000.0
 
 # Microphone scale factor, psi per ADC count.  Approximate — exact factor
 # depends on the geophone-vs-mic ADC scaling and the firmware reference.
@@ -728,7 +734,7 @@ def _peaks_from_samples(samples: dict[str, list[int]]) -> PeakValues:
         if not ch:
             return 0.0
         m = max(abs(int(v)) for v in ch)
-        return m / _INT16_FS * _GEO_NORMAL_FS_INS
+        return m / _GEO_INT16_FS * _GEO_NORMAL_FS_INS
 
     tran = _peak_ins(samples.get("Tran", []))
     vert = _peak_ins(samples.get("Vert", []))
@@ -742,7 +748,7 @@ def _peaks_from_samples(samples: dict[str, list[int]]) -> PeakValues:
     pvs = 0.0
     n = min(len(samples.get("Tran", [])), len(samples.get("Vert", [])), len(samples.get("Long", [])))
     if n:
-        scale = _GEO_NORMAL_FS_INS / _INT16_FS
+        scale = _GEO_NORMAL_FS_INS / _GEO_INT16_FS
         T = samples["Tran"]; V = samples["Vert"]; L = samples["Long"]
         for i in range(n):
             t = T[i] * scale
@@ -838,7 +844,13 @@ def read_blastware_file(path: Union[str, Path]) -> Event:
 
     # Footer: locate the 0e 08 marker, validating the year is in a sane range.
     body_start = _bw._WAVEFORM_HEADER_SIZE + 21
-    footer_pos = -1
+    # The 0e 08 + plausible-year footer signature can occur inside the sample
+    # stream.  Collect every candidate and prefer the first whose body yields a
+    # waveform record chain terminating on the 0x06 marker; fall back to the
+    # first candidate otherwise.  Blast radius measured 2026-08-25: changes the
+    # chosen footer on exactly 1 of 1,388 series-3 waveform files
+    # (BE17353/S353L4O5.OX0W, false positive at 3800, real footer at 8576).
+    footer_candidates = []
     pos = body_start
     while True:
         pos = raw.find(b"\x0e\x08", pos)
@@ -846,10 +858,24 @@ def read_blastware_file(path: Union[str, Path]) -> Event:
             break
         yr = (raw[pos + 4] << 8) | raw[pos + 5]
         if 2015 <= yr <= 2050:
-            footer_pos = pos
-            break
+            footer_candidates.append(pos)
         pos += 1
 
+    footer_pos = -1
+    for cand in footer_candidates:
+        cand_body = raw[body_start:cand]
+        try:
+            recs = _wc.walk_records(cand_body)
+        except Exception:
+            recs = []
+        if recs:
+            tail = recs[-1]["end"]
+            if tail + 5 <= len(cand_body) and cand_body[tail + 4] == _wc.STREAM_END_ID:
+                footer_pos = cand
+                break
+    if footer_pos < 0 and footer_candidates:
+        footer_pos = footer_candidates[0]
+
     if footer_pos < 0 and len(raw) >= 26:
         footer_pos = len(raw) - 26
     if footer_pos < body_start:

minimateplus/histogram_codec.py

@@ -25,38 +25,60 @@ iterate 32-stride and stop before the tail.
 ────────────────────────────────────────────────────────────────────────────
 
     [0]    0x00                      always-zero tag
-    [1]    segment_id  (uint8)       0x00..0x03 — 256 blocks per segment
-    [2:4]  block_ctr  (uint16 LE)    resets each segment (0x0100, 0x0101, …)
-    [4:6]  0x000a (uint16 LE)        constant marker (= 10)
-    [6]    T_peak_count   uint8      Tran peak (count × 0.005 → in/s, max 1.275 in/s)
-    [7]    T_annotation   uint8      empirically non-zero on intervals with sub-Hz
-                                     or unmeasurable Tran freq; meaning not fully RE'd
-    [8:10] T_halfperiod   uint16 LE  Tran half-period in samples (freq = 512 / halfp Hz)
-    [10]   V_peak_count   uint8
-    [11]   V_annotation   uint8
-    [12:14] V_halfperiod  uint16 LE
-    [14]   L_peak_count   uint8
-    [15]   L_annotation   uint8
-    [16:18] L_halfperiod  uint16 LE
-    [18]   M_peak_count   uint8      MicL peak (count → dB via mic_count_to_db)
-    [19]   M_annotation   uint8
-    [20:22] M_halfperiod  uint16 LE  MicL half-period in samples (freq = 512 / halfp Hz)
-    [22:24] 0x00 0x00                constant
+    [1]    segment_id  (uint8)       0x00..0x03 - 256 blocks per segment
+    [2:4]  block_ctr  (uint16 LE)    resets each segment (0x0100, 0x0101, ...)
+    [4]    0x0a (uint8)              constant marker (= 10)
+    [5:7]  T_peak_count   uint16 BE  Tran peak (count x 0.005 -> in/s)
+    [7:9]  T_halfperiod   uint16 BE  Tran half-period in samples (freq = 512 / halfp)
+    [9:11] V_peak_count   uint16 BE
+    [11:13] V_halfperiod  uint16 BE
+    [13:15] L_peak_count  uint16 BE
+    [15:17] L_halfperiod  uint16 BE
+    [17:19] M_peak_count  uint16 BE  MicL peak (count -> dB via mic_count_to_db)
+    [19:21] M_halfperiod  uint16 BE  MicL half-period in samples
+    [21:23] 0x00 0x00                constant on standard blocks
     [24:28] 4-byte variable          purpose unknown (possibly CRC or timestamp delta)
-    [28:32] 0x1e 0x0a 0x00 0x00      constant block-end signature
+    [28:32] block-end signature      see "Two block tails" below
 
-NOTE on peak-count width: an earlier interpretation treated the peak
-fields as uint16 LE spanning [6:8] / [10:12] / [14:16] / [18:20].
-That happened to be byte-exact against the N844 fixture corpus only
-because every annotation byte in those fixtures was zero, making
-``uint16 LE == uint8``.  Cross-correlating BE9558 (K558) Tran-drift
-and BE18003 (T003) Histogram+Continuous events against the BW ASCII
-export proved peak is uint8 alone — see test_histogram_codec.py
-and docs/histogram_codec_re_status.md.
+**Every per-channel field is uint16 BIG-endian** (confirmed 2026-08-25).
+Only ``block_ctr`` at [2:4] is little-endian.
 
-Block-identification anchor: ``block[22:24] == b"\\x00\\x00"`` AND
-``block[28:32] == b"\\x1e\\x0a\\x00\\x00"``.  This is the reliable
-distinguisher from non-block content in the file.
+HISTORY - two earlier readings of this block were wrong in ways that
+cancelled out on quiet data:
+
+  1. *peak as uint16 LE at [6:8]* - produced 268 in/s peaks on any
+     interval whose next byte was non-zero.
+  2. *peak as uint8 at [6] with an "annotation" byte at [7]* - correct
+     for every peak below 256 counts (1.275 in/s), but it silently
+     **clipped larger peaks**: the final interval of
+     BE18193/T193LQ9K.OE0H reads 8.270 in/s in BW's export
+     (1654 counts = 0x0676) and decoded as 0x76 = 118 = 0.590 in/s.
+     The "annotation" byte was never an annotation - it is the high
+     byte of the big-endian half-period, which is why it was non-zero
+     exactly on the sub-Hz intervals BW renders as "<1.0".
+
+Both readings also forced ``block[5] == 0`` via a bogus ``uint16 LE``
+marker check at [4:6], which is what capped the peak at one byte.
+The marker is ``block[4]`` alone.
+
+Verified 2026-08-25 against 1211 production histograms paired with
+their Blastware ASCII exports: **1211/1211 decode exactly** (interval
+count plus every per-interval peak), and 842,442 per-interval
+frequency comparisons match with **zero** mismatches.
+
+Two block tails
+---------------
+Standard blocks end with ``1e 0a 00 00``.  The **final block of the
+stream** ends with ``9c 06 00 42`` instead, and carries arbitrary bytes
+at [21:23].  Rejecting it dropped the last interval of nearly every
+histogram - and the last interval is frequently the one holding the
+event peak, so the file's reported PPV came out low.  Observed in 1206
+of 1211 production histograms, always positioned after every
+standard-tail block.
+
+Block-identification anchor: ``block[0] == 0x00`` AND
+``block[4] == 0x0A`` AND the tail is one of the two signatures above;
+standard-tail blocks additionally require ``block[22] == 0x00``.
 
 ────────────────────────────────────────────────────────────────────────────
 Per-channel encoding
@@ -109,6 +131,12 @@ from typing import List, Optional, Tuple
 # real data block.  More distinctive than the byte-22 `00 00` (which
 # matches many false positives), so we anchor on this.
 _BLOCK_TAIL = b"\x1e\x0a\x00\x00"
+
+# The final block of a histogram stream ends with this instead.  It is a
+# real data block - same layout - and holds the last interval.  See the
+# module docstring, "Two block tails".
+_BLOCK_TAIL_TERMINAL = b"\x9c\x06\x00\x42"
+
 _BLOCK_SIZE = 32
 
 # Marker byte at block[4:6] of every histogram data block.  Used as
@@ -127,19 +155,29 @@ _FREQ_NUMERATOR = 512
 
 
 def _is_data_block(block: bytes) -> bool:
-    """Tight identification of a histogram data block."""
+    """Tight identification of a histogram data block.
+
+    Accepts both tail signatures.  ``block[4]`` alone is the marker -
+    ``block[5]`` is the high byte of the Tran peak and is non-zero on any
+    interval above 1.275 in/s, so it must not be part of the marker test.
+    The ``block[22] == 0`` constraint is what keeps trailer content out,
+    but it applies only to standard-tail blocks: terminal blocks carry
+    arbitrary bytes there.
+    """
     if len(block) < _BLOCK_SIZE:
         return False
-    if block[28:32] != _BLOCK_TAIL:
-        return False
-    if block[22:24] != b"\x00\x00":
-        return False
     if block[0] != 0x00:
         return False
-    marker = block[4] | (block[5] << 8)
-    if marker != _BLOCK_MARKER:
+    if block[4] != _BLOCK_MARKER:
         return False
-    return True
+    # The 4-byte tail plus block[0]==0 and block[4]==0x0A is already six bytes
+    # of constraint — enough to keep trailer content out.  There is NO extra
+    # test on block[22]: it was documented as a constant 0x00 but carries data
+    # on loud blocks, and rejecting those threw away the interval holding the
+    # event peak.  BE18350/T350L7HR.NL0H is the proof: its block 92 has
+    # block[22]=0x26 and a Tran peak of 0x0563 = 1379 counts = 6.895 in/s,
+    # exactly the device-reported PPV, while the file decoded to 0.015 in/s.
+    return block[28:32] in (_BLOCK_TAIL, _BLOCK_TAIL_TERMINAL)
 
 
 def _decode_block(block: bytes) -> Optional[dict]:
@@ -149,33 +187,23 @@ def _decode_block(block: bytes) -> Optional[dict]:
     Returns a record with per-channel peak counts (uint8) and
     half-periods (uint16 LE).
     """
-    # Peak counts are uint8 at bytes [6] / [10] / [14] / [18].  The
-    # adjacent bytes [7] / [11] / [15] / [19] hold an annotation field
-    # whose meaning isn't fully understood (empirically non-zero in
-    # intervals with sub-Hz or unmeasurable geo frequencies, mostly
-    # zero otherwise — see test fixtures from BE9558/BE18003 corpora).
-    # Crucially, those annotation bytes are NOT the high byte of the
-    # peak count: cross-correlating against BW's per-interval ASCII
-    # export proves the peak is uint8 alone.
-    #
-    # Reading the peak as uint16 LE (the original interpretation) was
-    # accidentally correct only because every block in the N844 fixture
-    # corpus had a zero annotation byte; non-N844 events with non-zero
-    # annotation bytes decoded to physically impossible peaks (e.g.
-    # 268 in/s per channel) and produced 35× inflated PVS sums when
-    # first run against prod data.  See histogram_codec_re_status.md.
-    t_peak = block[6]
-    v_peak = block[10]
-    l_peak = block[14]
-    m_peak = block[18]
-    t_halfp = block[8]  | (block[9]  << 8)
-    v_halfp = block[12] | (block[13] << 8)
-    l_halfp = block[16] | (block[17] << 8)
-    m_halfp = block[20] | (block[21] << 8)
+    # Every per-channel field is uint16 BIG-endian; only block_ctr is LE.
+    # See the module docstring for the two superseded readings and why
+    # each looked correct on quiet data.
+    def _be16(i: int) -> int:
+        return (block[i] << 8) | block[i + 1]
+
+    t_peak  = _be16(5)
+    t_halfp = _be16(7)
+    v_peak  = _be16(9)
+    v_halfp = _be16(11)
+    l_peak  = _be16(13)
+    l_halfp = _be16(15)
+    m_peak  = _be16(17)
+    m_halfp = _be16(19)
     segment_id = block[1]
     block_ctr  = block[2] | (block[3] << 8)
     var_meta   = bytes(block[24:28])
-    annotations = (block[7], block[11], block[15], block[19])
     return {
         "segment_id":  segment_id,
         "block_ctr":   block_ctr,
@@ -188,7 +216,7 @@ def _decode_block(block: bytes) -> Optional[dict]:
         "m_peak":      m_peak,
         "m_halfp":     m_halfp,
         "meta_var":    var_meta,
-        "annotations": annotations,
+        "is_terminal": block[28:32] == _BLOCK_TAIL_TERMINAL,
     }
 
 
@@ -224,6 +252,23 @@ def walk_body(body: bytes) -> List[dict]:
     return records
 
 
+def _walk_auto(body: bytes) -> List[dict]:
+    """Pick the block model by signature strength, not by which returns first.
+
+    The multi-interval variant announces itself with consecutive block headers
+    at an exact ``12 + 20*n`` stride — far stronger evidence than a handful of
+    scattered standard-tail blocks, which a multi-interval body will also yield
+    by coincidence.  Dispatching on "whichever decoder returns something"
+    handed 193 BE18193 files to the standard walker and produced peaks of
+    149 in/s against a 10 in/s full scale.
+    """
+    if detect_multi_interval_stride(body):
+        recs = walk_multi_interval_blocks(body)
+        if recs:
+            return recs
+    return walk_body(body)
+
+
 def decode_histogram_body(body: bytes) -> Optional[dict]:
     """Decode a histogram-mode body into per-channel peak-sample arrays.
 
@@ -239,7 +284,7 @@ def decode_histogram_body(body: bytes) -> Optional[dict]:
          to get 1-count ADC values, then ``count / 32767 * 10.0`` for in/s)
       - Mic channel:  use ``waveform_codec.mic_count_to_db(count)``
     """
-    records = walk_body(body)
+    records = _walk_auto(body)
     if not records:
         return None
     return {
@@ -261,7 +306,7 @@ def decode_histogram_body_full(body: bytes) -> Optional[List[dict]]:
 
     Returns ``None`` if the body has no valid blocks.
     """
-    records = walk_body(body)
+    records = _walk_auto(body)
     return records if records else None
 
 
@@ -281,3 +326,162 @@ def half_period_to_hz(halfp: int) -> Optional[float]:
 def geo_count_to_ins(count: int) -> float:
     """Convert a histogram geo peak count to in/s at Normal range."""
     return count * _GEO_LSB_INS
+
+
+# ── Multi-interval block variant (CONFIRMED 2026-08-26) ─────────────────────
+#
+# When the histogram interval is SHORTER than one minute, the device packs
+# several intervals into a single block so that every block still covers
+# exactly one minute of data:
+#
+#     interval size   intervals/block   stride
+#     1 minute        1                 32   <- the standard block above
+#     15 seconds      4                 92
+#     2 seconds       30                612
+#
+#     stride = 12 + n_intervals * 20
+#
+# Block layout:
+#     [0]      0x00
+#     [1]      segment_id (256 blocks per segment, same as the standard block)
+#     [2:4]    block_ctr uint16 LE (0x0100.., resets each segment)
+#     [4]      0x0a marker
+#     [5]      0x00
+#     [6 ...]  n x 20-byte interval records, each carrying 8 x uint16
+#              LITTLE-endian values:
+#                  T_peak, T_halfperiod, V_peak, V_halfperiod,
+#                  L_peak, L_halfperiod, M_peak, M_halfperiod
+#              then 2 more words; the first is 0x0000 on every real interval.
+#     [-6:]    6-byte block trailer
+#
+# ⚠ ENDIANNESS: the standard 32-byte block is BIG-endian.  This variant is
+# LITTLE-endian.  Do not share the accessor.
+#
+# These files previously decoded to nothing at all — 415 of them in the
+# production snapshot, 216 on BE18193 (2 s intervals) and 199 on BE9440
+# (15 s).  Before that they were being accepted by the WAVEFORM codec, which
+# returned garbage peaking up to 400x the device-reported PPV.
+#
+# Ground truth: BE9440/K440L3AQ.T70H (15 s intervals, 5,710 of them) decodes
+# against its Blastware ASCII export with 17,130/17,130 geo peak counts,
+# 22,840/22,840 frequencies and 5,710/5,710 mic dB(L) values matching exactly.
+
+_MULTI_HEADER_LEN   = 6
+_MULTI_RECORD_LEN   = 20
+_MULTI_TRAILER_LEN  = 6
+# At least 2 records: a 1-record block would have stride 12 + 20 = 32, which
+# collides with the standard big-endian block and mis-decodes it.
+_MULTI_MIN_RECORDS  = 2
+_MULTI_MAX_RECORDS  = 64
+
+# Geo full scale in 16-count units: 10.000 in/s / 0.005 = 2000.  A peak above
+# this is physically impossible and marks buffer garbage in a partial block.
+_GEO_MAX_COUNTS = 2000
+
+
+def _is_multi_header(body: bytes, off: int) -> bool:
+    return (off + _MULTI_HEADER_LEN <= len(body)
+            and body[off] == 0x00
+            and body[off + 4] == 0x0A
+            and body[off + 5] == 0x00)
+
+
+def detect_multi_interval_stride(body: bytes) -> Optional[int]:
+    """Block stride of a multi-interval histogram body, or None.
+
+    Found by locating the second block header; validated against
+    ``stride = 12 + n * 20`` and confirmed on a third block where present.
+    """
+    if not _is_multi_header(body, 0):
+        return None
+    lo = _MULTI_HEADER_LEN + _MULTI_TRAILER_LEN + _MULTI_RECORD_LEN * _MULTI_MIN_RECORDS
+    hi = _MULTI_HEADER_LEN + _MULTI_TRAILER_LEN + _MULTI_RECORD_LEN * _MULTI_MAX_RECORDS
+    for stride in range(lo, min(hi, len(body)) + 1, 2):
+        if (stride - 12) % _MULTI_RECORD_LEN:
+            continue
+        if not _is_multi_header(body, stride):
+            continue
+        # DECISIVE CHECK: consecutive blocks differ by exactly 1 in block_ctr.
+        # Without it this false-positives on ordinary standard-block bodies:
+        # those carry a header every 32 bytes, and 192 = 12 + 20*9 and
+        # 512 = 12 + 20*25 are both multiples of 32, so a stride "fits" while
+        # actually skipping 6 or 16 real blocks.  Sampling a standard body at
+        # stride 192 handed 9,082 files to the wrong decoder and produced peaks
+        # of 149 in/s against a 10 in/s full scale.
+        def _ctr(o: int) -> int:
+            return body[o + 2] | (body[o + 3] << 8)
+
+        if (_ctr(stride) - _ctr(0)) & 0xFFFF != 1:
+            continue
+        # Confirm on a third block WHEN ONE IS ACTUALLY PRESENT.  A body can
+        # be longer than two strides and still hold only two real blocks: a
+        # final *partial* block leaves trailing padding.  E.g. 51 intervals at
+        # 2 s = one full 30-interval block + a 21-interval remainder, in a
+        # 2787-byte body — long enough to demand a third header at 1224 that
+        # does not exist.  Requiring it unconditionally threw away the correct
+        # stride and the file decoded to nothing (BE18193 T193L0XM.CI0H).
+        # The block-counter check above is the decisive anti-false-positive
+        # test; this one is corroboration, so a missing third header means
+        # end-of-stream, not disqualification.
+        if (2 * stride + _MULTI_HEADER_LEN <= len(body)
+                and _is_multi_header(body, 2 * stride)):
+            if (_ctr(2 * stride) - _ctr(stride)) & 0xFFFF != 1:
+                continue
+        return stride
+    return None
+
+
+def walk_multi_interval_blocks(body: bytes,
+                               stride: Optional[int] = None) -> List[dict]:
+    """Decode a multi-interval histogram body into per-interval records."""
+    if stride is None:
+        stride = detect_multi_interval_stride(body)
+    if not stride:
+        return []
+    n_per_block = (stride - _MULTI_HEADER_LEN - _MULTI_TRAILER_LEN) // _MULTI_RECORD_LEN
+    if n_per_block < 1:
+        return []
+
+    def u16le(p: int) -> int:
+        return body[p] | (body[p + 1] << 8)
+
+    out: List[dict] = []
+    for off in range(0, len(body) - stride + 1, stride):
+        if not _is_multi_header(body, off):
+            break                      # end of the block run; trailer follows
+        for k in range(n_per_block):
+            q = off + _MULTI_HEADER_LEN + _MULTI_RECORD_LEN * k
+            out.append({
+                "_tail0": u16le(q + 16),
+                "segment_id": body[off + 1],
+                "block_ctr":  u16le(off + 2),
+                "t_peak": u16le(q),      "t_halfp": u16le(q + 2),
+                "v_peak": u16le(q + 4),  "v_halfp": u16le(q + 6),
+                "l_peak": u16le(q + 8),  "l_halfp": u16le(q + 10),
+                "m_peak": u16le(q + 12), "m_halfp": u16le(q + 14),
+                "meta_var": bytes(body[q + 16:q + 20]),
+                "is_terminal": False,
+            })
+    # A session ending mid-block leaves the remaining slots of the FINAL block
+    # filled with whatever was in the buffer.  Those decoded as peaks thousands
+    # of times the device-reported PPV, so they have to go — but only from the
+    # final block: a non-zero tail word occurs mid-file on real intervals, and
+    # trimming on that alone truncated four BE9440 files by up to 2,800
+    # intervals, while trimming purely from the end left garbage stranded
+    # behind one slot that happened to have a zero tail word.
+    #
+    # Within the final block, stop at the first slot that is not plausibly
+    # real: a non-zero tail word, or a geo peak above full scale.  16-count
+    # units put Normal-range full scale (10.000 in/s) at 2000 counts, so
+    # anything beyond that is physically impossible.
+    if out:
+        last_block_start = ((len(out) - 1) // n_per_block) * n_per_block
+        for i in range(last_block_start, len(out)):
+            r = out[i]
+            if (r["_tail0"] != 0
+                    or max(r["t_peak"], r["v_peak"], r["l_peak"]) > _GEO_MAX_COUNTS):
+                del out[i:]
+                break
+    for r in out:
+        r.pop("_tail0", None)
+    return out

minimateplus/waveform_codec.py

@@ -166,8 +166,14 @@ def find_data_start(body: bytes) -> int:
     # Try fixed offset 7 first (canonical preamble length).
     if len(body) >= 9:
         b, nn = body[7], body[8]
-        if (b in (0x00, 0x10, 0x20, 0x30) and nn % 4 == 0 and 0 < nn <= 0xFC) \
-                or (b == 0x40 and nn == 0x02):
+        # Accept the same tag vocabulary ``walk_body`` accepts, including the
+        # wide-NN forms (``0X``/``1X``/``2X``) and the variable-width ``40 NN``
+        # segment header.
+        if ((b & 0xF0) in (0x00, 0x10, 0x20) and nn % 4 == 0
+                and ((b & 0x0F) != 0 or 0 < nn <= 0xFC)) \
+                or (b == 0x30 and nn % 4 == 0 and 0 < nn <= 0xFC) \
+                or (b == 0x40 and 0 < nn <= 0x08) \
+                or is_tagless_segment_header(body, 7):
             return 7
     # Fall back to scanning the first 20 bytes.
     for i in range(min(20, len(body) - 1)):
@@ -178,6 +184,31 @@ def find_data_start(body: bytes) -> int:
     return -1
 
 
+# Channel-id byte carried in every segment header.  Previously mis-read as a
+# "monotonic uint32 LE counter"; it is really ``[channel][00][00][segment]``.
+# Verified 2026-08-25 on 1697/1697 segment headers across the ground-truth
+# corpus with zero disagreements against the decoded channel rotation.
+SEGMENT_CHANNEL_IDS = {0x46: "Tran", 0x47: "Vert", 0x48: "Long", 0x49: "MicL"}
+
+# A tagless segment header: the 14-byte tail of a ``40 NN`` header with no tag
+# and no previous-channel continuation deltas (the NN=0 case).
+_TAGLESS_HEADER_LEN = 14
+
+
+def is_tagless_segment_header(body: bytes, i: int) -> bool:
+    """True if a bare 14-byte segment header starts at *i*.
+
+    Layout ``[field2:2][len_to_next:2][channel_id:4][marker:2][anchors:4]``.
+    The discriminator is the 6 bytes at ``[4:10]``: a known channel id, two
+    zero bytes, a small segment index, and the ``01 00`` / ``02 00`` marker.
+    """
+    if i + _TAGLESS_HEADER_LEN > len(body):
+        return False
+    return (body[i + 4] in SEGMENT_CHANNEL_IDS
+            and body[i + 5] == 0x00 and body[i + 6] == 0x00
+            and body[i + 8] in (0x01, 0x02) and body[i + 9] == 0x00)
+
+
 def walk_body(body: bytes, start: Optional[int] = None) -> List[WaveformBlock]:
     """Walk the tagged-block sequence starting at *start* (auto-detected by default).
 
@@ -210,9 +241,15 @@ def walk_body(body: bytes, start: Optional[int] = None) -> List[WaveformBlock]:
             # Wide-NN int8 block: ``2X NN`` extends NN to 12 bits the same way.
             wide_nn = ((t0 & 0x0F) << 8) | t1
             length = wide_nn + 2
-        elif t0 == 0x00 and t1 % 4 == 0:
+        elif (t0 & 0xF0) == 0x00 and t1 % 4 == 0:
+            # ``00 NN`` RLE zero-delta run, plus its wide form ``0X NN``
+            # (X != 0) which extends NN to 12 bits exactly like ``1X``/``2X``:
+            # NN = ((t0 & 0x0F) << 8) | t1.  A narrow run maxes out at
+            # NN=0xFC, so quiet stretches longer than 252 samples must use
+            # the wide form.  Confirmed 2026-08-25 against six production
+            # events (e.g. ``01 0c`` = 268 repeats in K558LKOF.460W).
             length = 2
-        elif t0 == 0x30 and t1 % 4 == 0 and 0 < t1 <= 0x10:
+        elif t0 == 0x30 and t1 % 4 == 0 and 0 < t1 <= 0xFC:
             # Data-section ``30 NN`` blocks carry NN 12-bit signed deltas packed
             # as NN/4 groups of (2-byte high-nibble field + 4 × int8 low byte).
             # Length = NN/4 × 6 + 2 = NN × 1.5 + 2 (= 8 for NN=4, 14 for NN=8,
@@ -229,8 +266,28 @@ def walk_body(body: bytes, start: Optional[int] = None) -> List[WaveformBlock]:
                 length = cand_data
             else:
                 length = cand_trailer
-        elif t0 == 0x40 and t1 == 0x02:
-            length = 20
+        elif t0 == 0x40 and 0 < t1 <= 0x08:
+            # ``40 NN`` segment header.  NN is the number of int16 BE
+            # continuation deltas the header carries for the PREVIOUS
+            # channel, so the header grows with NN:
+            #     length = 2 (tag) + 2*NN (deltas) + 14 (fixed tail)
+            # ``40 02`` (20 bytes) dominates, but ``40 01`` (18) and
+            # ``40 03`` (22) both occur in production files.  Confirmed
+            # 2026-08-25; the constant ``02 00`` marker moves with NN too
+            # (see :func:`parse_segment_header`).
+            length = 2 * t1 + 16
+        elif is_tagless_segment_header(body, i):
+            # Segment header with no ``40 NN`` tag (NN=0 — the previous channel
+            # needed no continuation deltas).  Emit it as a synthetic ``40 00``
+            # block whose ``data`` is the whole 14-byte record, so the nd=0
+            # offsets in :func:`decode_waveform_v2` line up unchanged.
+            blocks.append(WaveformBlock(
+                offset=i, tag_hi=0x40, tag_lo=0x00,
+                data=bytes(body[i : i + _TAGLESS_HEADER_LEN]),
+                length=_TAGLESS_HEADER_LEN,
+            ))
+            i += _TAGLESS_HEADER_LEN
+            continue
         else:
             # Unknown tag; stop.  Caller can inspect ``i`` to see where.
             break
@@ -256,7 +313,7 @@ def split_segments(blocks: List[WaveformBlock]) -> List[List[WaveformBlock]]:
     segments: List[List[WaveformBlock]] = []
     current: List[WaveformBlock] = []
     for b in blocks:
-        if b.tag_hi == 0x40 and b.tag_lo == 0x02:
+        if b.tag_hi == 0x40:
             if current:
                 segments.append(current)
             current = [b]
@@ -268,23 +325,40 @@ def split_segments(blocks: List[WaveformBlock]) -> List[List[WaveformBlock]]:
 
 
 def parse_segment_header(block: WaveformBlock) -> Optional[dict]:
-    """Decode the 18-byte payload of a ``40 02`` segment header.
+    """Decode the payload of a ``40 NN`` segment header.
 
-    Returns a dict with the labelled fields, or None if *block* is not
-    a ``40 02`` header.
+    NN (the tag's low byte) is the number of int16 BE continuation deltas
+    the header carries for the PREVIOUS channel, so every field after
+    those deltas shifts by ``2 * NN``.  The payload is ``2 * NN + 14``
+    bytes.  ``40 02`` is the common case; ``40 01`` and ``40 03`` also
+    occur in production files (confirmed 2026-08-25).
+
+    Returns a dict with the labelled fields, or None if *block* is not a
+    segment header or is too short.
     """
-    if not (block.tag_hi == 0x40 and block.tag_lo == 0x02):
+    if block.tag_hi != 0x40 or block.tag_lo > 0x08:
         return None
-    if len(block.data) < 18:
+    nd = block.tag_lo
+    if len(block.data) < 2 * nd + 14:
         return None
     p = block.data
-    counter = int.from_bytes(p[8:12], "little", signed=False)
+    counter = int.from_bytes(p[2 * nd + 4 : 2 * nd + 8], "little", signed=False)
     return {
-        "anchor_bytes": p[0:4],          # 4-byte field, role unconfirmed
-        "field2": p[4:8],                # 4-byte field, role unconfirmed
-        "counter": counter,              # uint32 LE — increments by 1 per segment
-        "fixed_pattern": p[12:16],       # always b"\x02\x00\x00\x01"
-        "tail": p[16:18],                # last 2 bytes
+        "n_prev_deltas": nd,
+        # ``nd`` int16 BE deltas extending the previous channel.
+        "prev_deltas": [
+            int.from_bytes(p[2 * k : 2 * k + 2], "big", signed=True)
+            for k in range(nd)
+        ],
+        "field2": p[2 * nd : 2 * nd + 4],   # 4-byte field, role unconfirmed
+        "counter": counter,                 # legacy: raw uint32 LE of the id field
+        "channel": SEGMENT_CHANNEL_IDS.get(p[2 * nd + 4]),
+        "segment_index": p[2 * nd + 7],
+        "marker": p[2 * nd + 8 : 2 * nd + 10],      # always b"\x02\x00"
+        "anchors": [
+            int.from_bytes(p[2 * nd + 10 : 2 * nd + 12], "big", signed=True),
+            int.from_bytes(p[2 * nd + 12 : 2 * nd + 14], "big", signed=True),
+        ],
     }
 
 
@@ -367,8 +441,16 @@ def decode_tran_initial(body: bytes) -> Optional[List[int]]:
     return out
 
 
-def decode_waveform_v2(body: bytes) -> Optional[dict]:
+def decode_waveform_legacy(body: bytes) -> Optional[dict]:
     """
+    SUPERSEDED 2026-08-25 — the tag-dispatch / segment-header model.
+
+    Retained because ``micromate/idf_file.py`` trial-decodes Thor IDFW bodies
+    at many candidate offsets and keeps whichever yields the most samples;
+    the record-chain decoder returns None where this one returned garbage,
+    which shifts that heuristic's winner.  Thor is pinned here until its own
+    body-offset search is reworked.  Do not use for series-3.
+
     Decode the body into per-channel sample arrays.
 
     Status (2026-05-11 evening — channel-rotation hypothesis CONFIRMED):
@@ -420,8 +502,11 @@ def decode_waveform_v2(body: bytes) -> Optional[dict]:
                 for byte in blk.data:
                     cur += _i8(byte)
                     out[channel].append(cur)
-            elif blk.tag_hi == 0x00:
-                for _ in range(blk.tag_lo):
+            elif (blk.tag_hi & 0xF0) == 0x00:
+                # RLE zero-delta run.  Wide form ``0X NN`` carries the high
+                # nibble of a 12-bit NN in the tag byte, same as ``1X``/``2X``.
+                run = ((blk.tag_hi & 0x0F) << 8) | blk.tag_lo
+                for _ in range(run):
                     out[channel].append(cur)
             elif blk.tag_hi == 0x30:
                 # 12-bit signed deltas, packed as NN/4 groups of 6 bytes each:
@@ -461,34 +546,54 @@ def decode_waveform_v2(body: bytes) -> Optional[dict]:
     # previous-channel extension deltas at every segment boundary.
     last_value = {"Tran": last_tran_value, "Vert": None, "Long": None, "MicL": None}
 
+    prev_channel = "Tran"
     for k, hi in enumerate(seg_idx):
-        channel = rotation[k % 4]
-        prev_channel = "Tran" if k == 0 else rotation[(k - 1) % 4]
         header = blocks[hi]
-        if len(header.data) < 18:
+        # Channel comes from the header's own id byte, which is authoritative.
+        # The old rotation-by-position fallback is kept for headers whose id
+        # byte isn't one of the four known values — but a single missed or
+        # extra header would desync rotation and corrupt every later channel,
+        # which is exactly what tagless headers used to cause.
+        _nd = header.tag_lo
+        channel = None
+        if len(header.data) >= 2 * _nd + 8:
+            channel = SEGMENT_CHANNEL_IDS.get(header.data[2 * _nd + 4])
+        if channel is None:
+            channel = rotation[k % 4]
+        # ``40 NN``: NN int16 BE continuation deltas for the previous channel
+        # come first, so every later field shifts by 2*NN.  NN is usually 2
+        # but 1 and 3 both occur (confirmed 2026-08-25).
+        nd = header.tag_lo
+        if len(header.data) < 2 * nd + 14:
             continue
-        # Validate: real segment headers have bytes [12:14] = `02 00`.
-        # Trailer/footer "40 02" markers contain ASCII serial bytes or other
-        # non-header data there and would otherwise be mis-interpreted as
-        # segment headers, adding spurious samples at the tail.
-        if header.data[12:14] != b"\x02\x00":
+        # Validate: real segment headers have the constant `02 00` marker
+        # right after the counter.  Trailer/footer "40 NN" markers contain
+        # ASCII serial bytes or other non-header data there and would
+        # otherwise be mis-read as segment headers, adding spurious tail
+        # samples.
+        if header.data[2 * nd + 8 : 2 * nd + 10] != b"\x02\x00":
             break
-        # Extend the PREVIOUS channel by 2 more samples (deltas in bytes [0:4]).
-        prev_d0 = int.from_bytes(header.data[0:2], "big", signed=True)
-        prev_d1 = int.from_bytes(header.data[2:4], "big", signed=True)
+        # Extend the PREVIOUS channel by NN more samples.
         if last_value[prev_channel] is not None:
-            v = last_value[prev_channel] + prev_d0
-            out[prev_channel].append(v)
-            v += prev_d1
-            out[prev_channel].append(v)
+            v = last_value[prev_channel]
+            for d in range(nd):   # NB: not `k` — that's the segment index
+                v += int.from_bytes(
+                    header.data[2 * d : 2 * d + 2], "big", signed=True
+                )
+                out[prev_channel].append(v)
             last_value[prev_channel] = v
         # Anchor pair for THIS segment's channel.
-        c0 = int.from_bytes(header.data[14:16], "big", signed=True)
-        c1 = int.from_bytes(header.data[16:18], "big", signed=True)
+        c0 = int.from_bytes(
+            header.data[2 * nd + 10 : 2 * nd + 12], "big", signed=True
+        )
+        c1 = int.from_bytes(
+            header.data[2 * nd + 12 : 2 * nd + 14], "big", signed=True
+        )
         out[channel].extend([c0, c1])
         # Apply delta blocks for this segment.
         next_hi = seg_idx[k + 1] if k + 1 < len(seg_idx) else len(blocks)
         last_value[channel] = apply_blocks(channel, c1, hi + 1, next_hi)
+        prev_channel = channel
 
     return out
 
@@ -576,3 +681,231 @@ def decode_a5_frames(a5_frames) -> Optional[dict]:
     if decoded is None:
         return None
     return decoded_to_adc_counts(decoded)
+
+
+# ── Record-chain body model (CONFIRMED 2026-08-25) ──────────────────────────
+#
+# The body is NOT a flat tag-dispatch stream with ``40 NN`` segment headers.
+# It is a chain of self-delimiting per-channel RECORDS:
+#
+#     off+0   field2    uint16     purpose unknown (not a length, not a checksum)
+#     off+2   len       uint16 BE  next_record = off + 2 + len   <- authoritative
+#     off+4   chan_id   0x46 Tran / 0x47 Vert / 0x48 Long / 0x49 MicL
+#                       0x06 = end of waveform stream
+#     off+5   0x00
+#     off+6   0x00
+#     off+7   segment index
+#     off+8   mode      2 bytes, a 3-valued enum (see below)
+#     off+10  anchors   2 x int16 BE, ABSOLUTE — present only when mode is 02 00
+#
+# Mode semantics, all ground-truth verified:
+#     02 00   14-byte header; emit the 2 anchors, then blocks are CUMULATIVE deltas
+#     01 00   10-byte header; no anchors; blocks carry ABSOLUTE sample values
+#     00 03   10-byte header; NO TAGS AT ALL — the data section is raw 12-bit
+#             packed ABSOLUTE samples (6 bytes -> 4 samples)
+#
+# ``40 NN`` is an ordinary int16 BE DATA block (length 2*NN + 2), never a header.
+# The previous model read it as a variable-width segment header of length
+# 2*NN + 16, which is why walks drifted and channels came out unequal.
+#
+# Verified over the 1,388 series-3 waveform binaries in the production
+# snapshot: the length chain terminates on a 0x06 record in 1,387 of them (the
+# exception has an ambiguous footer, handled by the caller), and all four
+# channels come out at identical length in 1,388/1,388 — against 156/1,388
+# under the superseded model.  Against the 75 events with a preserved
+# Blastware ASCII export: sample-count exact 72/75 -> 75/75, fully exact
+# 70/75 -> 73/75.
+
+CHANNEL_IDS = {0x46: "Tran", 0x47: "Vert", 0x48: "Long", 0x49: "MicL"}
+STREAM_END_ID = 0x06
+
+MODE_DELTA    = (0x02, 0x00)
+MODE_ABSOLUTE = (0x01, 0x00)
+MODE_RAW12    = (0x00, 0x03)
+_MODES = (MODE_DELTA, MODE_ABSOLUTE, MODE_RAW12)
+
+
+def _u16(b: bytes, p: int) -> int:
+    return (b[p] << 8) | b[p + 1]
+
+
+def _i16(b: bytes, p: int) -> int:
+    v = _u16(b, p)
+    return v - 0x10000 if v >= 0x8000 else v
+
+
+def data_block_len(body: bytes, p: int) -> Tuple[Optional[int], Optional[int]]:
+    """``(byte_length, n_samples)`` of the data block at *p*, or ``(None, None)``.
+
+    Data-section blocks only — there is no segment-header tag in this model.
+    ``30 NN`` has no trailer-length fallback here; that fallback corrupted
+    records whose ``30 NN`` sat near a record boundary.
+    """
+    if p + 2 > len(body):
+        return None, None
+    t0, t1 = body[p], body[p + 1]
+    hi = t0 & 0xF0
+    nn = ((t0 & 0x0F) << 8) | t1
+    if hi == 0x40:                          # int16 BE data block
+        return (None, None) if (nn == 0 or nn > 0x08) else (2 * nn + 2, nn)
+    if nn == 0 or nn % 4:
+        return None, None
+    if hi == 0x00:
+        return 2, nn                        # RLE hold
+    if hi == 0x10:
+        return nn // 2 + 2, nn              # 4-bit nibble
+    if hi == 0x20:
+        return nn + 2, nn                   # int8
+    if hi == 0x30:
+        return nn * 3 // 2 + 2, nn          # 12-bit packed
+    return None, None
+
+
+def unpack12(data: bytes) -> List[int]:
+    """Raw 12-bit packed samples: 6 bytes -> 4 signed values."""
+    out: List[int] = []
+    for g in range(len(data) // 6):
+        hi = (data[6 * g] << 8) | data[6 * g + 1]
+        for k in range(4):
+            x = (((hi >> (12 - 4 * k)) & 0xF) << 8) | data[6 * g + 2 + k]
+            out.append(x - 0x1000 if x >= 0x800 else x)
+    return out
+
+
+def is_record(body: bytes, p: int) -> bool:
+    """True if a per-channel record header starts at *p*."""
+    return (p + 10 <= len(body)
+            and body[p + 4] in CHANNEL_IDS
+            and body[p + 5] == 0x00 and body[p + 6] == 0x00
+            and 8 <= _u16(body, p + 2) <= len(body) - p
+            and (body[p + 8], body[p + 9]) in _MODES)
+
+
+def find_first_record(body: bytes) -> Optional[int]:
+    """Offset of the first record, or None.
+
+    Under the normal ``00 02 00`` preamble the leading bytes are segment-0's
+    Tran blocks, so walk them.  Under the ``00 00 03`` preamble that data is
+    raw 12-bit with no tags at all and cannot be block-walked — scan instead.
+    """
+    if len(body) >= 3 and (body[1], body[2]) == MODE_RAW12:
+        scan_from = 3
+    else:
+        i = 7
+        while i < len(body):
+            if is_record(body, i):
+                nxt = i + 2 + _u16(body, i + 2)
+                if nxt + 5 <= len(body) and (is_record(body, nxt)
+                                             or body[nxt + 4] == STREAM_END_ID):
+                    return i
+            length, _ = data_block_len(body, i)
+            if length is None:
+                return None
+            i += length
+        return None
+    for i in range(scan_from, max(scan_from, len(body) - 10)):
+        if is_record(body, i):
+            nxt = i + 2 + _u16(body, i + 2)
+            if nxt + 5 <= len(body) and (is_record(body, nxt)
+                                         or body[nxt + 4] == STREAM_END_ID):
+                return i
+    return None
+
+
+def walk_records(body: bytes, first: Optional[int] = None) -> List[dict]:
+    """Follow the length chain from *first* to the ``0x06`` terminator."""
+    if first is None:
+        first = find_first_record(body)
+    out: List[dict] = []
+    if first is None:
+        return out
+    p, seen = first, set()
+    while p is not None and p + 10 <= len(body):
+        if p in seen:
+            break
+        seen.add(p)
+        cid = body[p + 4]
+        if cid == STREAM_END_ID or cid not in CHANNEL_IDS:
+            break
+        length = _u16(body, p + 2)
+        if length < 8 or p + 2 + length > len(body):
+            break
+        out.append({"offset": p, "channel": CHANNEL_IDS[cid],
+                    "segment_index": body[p + 7],
+                    "mode": (body[p + 8], body[p + 9]),
+                    "end": p + 2 + length})
+        p += 2 + length
+    return out
+
+
+def decode_waveform_v2(body: bytes) -> Optional[dict]:
+    """Decode a Blastware waveform body into per-channel sample arrays.
+
+    Returns ``{"Tran": [...], "Vert": [...], "Long": [...], "MicL": [...]}``
+    in 16-count units (LSB = 0.005 in/s at Normal range), or None if *body*
+    is not a decodable waveform body.
+
+    Implements the record-chain model documented above.
+    """
+    if len(body) < 8 or body[0] != 0x00:
+        return None
+    preamble = (body[1], body[2])
+    if preamble not in (MODE_DELTA, MODE_RAW12):
+        return None
+    first = find_first_record(body)
+    if first is None:
+        return None
+
+    out: dict = {c: [] for c in ("Tran", "Vert", "Long", "MicL")}
+
+    def run(channel: str, start: int, end: int, absolute: bool) -> None:
+        cur = out[channel][-1] if out[channel] else 0
+        i = start
+        while i < end:
+            length, nn = data_block_len(body, i)
+            if length is None or i + length > end:
+                return          # stop this record; the chain resyncs at end
+            hi = body[i] & 0xF0
+            if hi == 0x00:
+                vals = [None] * nn
+            elif hi == 0x10:
+                vals = []
+                for k in range(nn):
+                    byte = body[i + 2 + k // 2]
+                    v = (byte >> 4) if k % 2 == 0 else (byte & 0xF)
+                    vals.append(v - 16 if v >= 8 else v)
+            elif hi == 0x20:
+                vals = [v - 256 if v >= 128 else v
+                        for v in body[i + 2:i + 2 + nn]]
+            elif hi == 0x30:
+                vals = unpack12(body[i + 2:i + length])
+            else:
+                vals = [_i16(body, i + 2 + 2 * k) for k in range(nn)]
+            for v in vals:
+                if v is None:
+                    pass                    # RLE hold, in delta AND absolute modes
+                elif absolute:
+                    cur = v
+                else:
+                    cur += v
+                out[channel].append(cur)
+            i += length
+
+    # Segment 0 is an implicit Tran record carried in the preamble.
+    if preamble == MODE_DELTA:
+        out["Tran"].extend([_i16(body, 3), _i16(body, 5)])
+        run("Tran", 7, first, absolute=False)
+    else:
+        out["Tran"].extend(unpack12(body[3:first]))
+
+    for rec in walk_records(body, first):
+        ch, off, mode, end = (rec["channel"], rec["offset"],
+                              rec["mode"], rec["end"])
+        if mode == MODE_DELTA:
+            out[ch].extend([_i16(body, off + 10), _i16(body, off + 12)])
+            run(ch, off + 14, end, absolute=False)
+        elif mode == MODE_ABSOLUTE:
+            run(ch, off + 10, end, absolute=True)
+        elif mode == MODE_RAW12:
+            out[ch].extend(unpack12(body[off + 10:end]))
+    return out

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "seismo-relay"
-version = "0.24.0"
+version = "0.27.0"
 description = "Python client and REST server for MiniMate Plus seismographs"
 requires-python = ">=3.10"
 dependencies = [

scratch/offset_candidates.csv

@@ -0,0 +1,275 @@
+event_id,serial,timestamp,filename,channel,offset_ips,peak_ips,mean_over_peak,trigger_level_ips,already_flagged_ft
+1e7a7808-034c-423b-80e9-d6788da80993,BE18438,2025-11-15T08:57:40,T438LBX4.W40W,Vert,0.2722,0.2930,0.929,0.2,0
+346e383f-c5a1-41d1-b972-3d652e599d39,BE18438,2025-11-15T09:13:15,T438LBX5.M30W,Vert,0.2747,0.2930,0.938,0.2,0
+45ef3023-22c5-44a8-a606-a0900bc17861,BE18438,2025-11-15T09:16:13,T438LBX5.R10W,Vert,0.2929,0.3027,0.967,0.2,0
+1cbcd2aa-ec55-4e2f-af42-7178f26cc6a1,BE18438,2025-11-15T09:19:29,T438LBX5.WH0W,Vert,0.2816,0.2979,0.945,0.2,0
+caf43634-4528-4fd5-9674-5f4f563661c0,BE18438,2025-11-15T09:22:32,T438LBX6.1K0W,Vert,0.2883,0.2930,0.984,0.2,0
+456ae646-c834-4e5c-82f1-7df18b3440a1,BE18438,2025-11-15T09:25:31,T438LBX6.6J0W,Vert,0.3189,0.3223,0.989,0.2,0
+afb91ba2-351b-42a0-b33f-acebd5f1829d,BE18438,2025-11-15T09:28:29,T438LBX6.BH0W,Vert,0.3057,0.3076,0.994,0.2,0
+39847146-3537-4300-943d-01c0c771e9cd,BE18438,2025-11-15T09:31:26,T438LBX6.GE0W,Vert,0.3225,0.3271,0.986,0.2,0
+e47335c1-fc3f-4c90-88c6-8edd0f296ce4,BE18438,2025-11-15T09:34:24,T438LBX6.LC0W,Vert,0.3275,0.3320,0.986,0.2,0
+25b6eda6-ba98-422f-8834-94b6cbe34971,BE18438,2025-11-15T09:37:22,T438LBX6.QA0W,Vert,0.3398,0.3418,0.994,0.2,0
+7af99377-e47f-4914-8e91-ac7f7f50b8c6,BE18438,2025-11-15T09:40:20,T438LBX6.V80W,Vert,0.3515,0.3564,0.986,0.2,0
+c4330605-c257-49ef-aede-20deb641e7d7,BE18438,2025-11-15T10:09:48,T438LBX8.8C0W,Vert,0.3605,0.3955,0.912,0.2,0
+b665caf9-64ee-4352-a391-60c1ddeebc1e,BE18438,2026-02-25T10:58:04,T438LH66.GS0W,Vert,0.1803,0.1953,0.923,0.2,0
+3b0c32c3-fd16-4b83-9917-942a56dea168,BE18438,2026-02-25T18:12:04,T438LH6Q.K40W,Vert,0.1869,0.1953,0.957,0.2,0
+ca66e601-d10f-4424-9c26-08a688e7b08c,BE18438,2026-02-25T18:17:16,T438LH6Q.SS0W,Vert,0.1870,0.1953,0.958,0.2,0
+31aa30ba-eb0a-49a9-98ca-c77d2cc31275,BE18438,2026-02-25T18:21:11,T438LH6Q.ZB0W,Vert,0.1868,0.1953,0.956,0.2,0
+fdb55a93-fc91-4fdf-a7f4-fdd1a6b48f59,BE18438,2026-02-25T18:26:27,T438LH6R.830W,Vert,0.1893,0.1953,0.969,0.2,0
+cbd718d7-2378-4d74-8656-842f229d19a8,BE18438,2026-02-25T18:33:25,T438LH6R.JP0W,Vert,0.1887,0.1953,0.966,0.2,0
+bfaf58ec-6163-47eb-9d2a-70da880bd948,BE18438,2026-02-25T18:37:48,T438LH6R.R00W,Vert,0.1872,0.1953,0.959,0.2,0
+ddaa530b-886e-4415-b93a-f4ebf4c4a882,BE18438,2026-02-25T18:59:54,T438LH6S.RU0W,Vert,0.1874,0.1953,0.959,0.2,0
+57f15284-ab69-48fc-9660-8c347c80c681,BE18438,2026-02-25T19:15:58,T438LH6T.IM0W,Vert,0.1873,0.1953,0.959,0.2,0
+03715718-7aa8-4e1d-9b0b-c2ab75fc1975,BE18438,2026-02-25T19:19:58,T438LH6T.PA0W,Vert,0.1891,0.1953,0.968,0.2,0
+9f921cd5-1bf9-4713-aef6-41e7f780eff1,BE18438,2026-02-25T19:24:02,T438LH6T.W20W,Vert,0.1871,0.1953,0.958,0.2,0
+44994949-58d4-44fa-bd78-ed75b8acc667,BE18438,2026-02-25T19:44:57,T438LH6U.UX0W,Vert,0.1877,0.1953,0.961,0.2,0
+c4be01c3-1001-4e17-aecd-d6d26dbb09f3,BE18438,2026-02-25T19:49:19,T438LH6V.270W,Vert,0.1884,0.1953,0.964,0.2,0
+3db004e6-ee34-43c5-ae8e-f22894ead5d0,BE18438,2026-02-25T20:08:22,T438LH6V.XY0W,Vert,0.1878,0.1953,0.962,0.2,0
+d276ae48-0434-41e5-af43-e0fb366acf11,BE18438,2026-02-25T20:11:33,T438LH6W.390W,Vert,0.1880,0.1953,0.963,0.2,0
+c027dc12-88fd-4a45-872b-a6a945f2008e,BE18438,2026-02-25T20:14:35,T438LH6W.8B0W,Vert,0.1874,0.1953,0.959,0.2,0
+9665bc3a-3d08-4852-b008-37ef8bfd0023,BE18438,2026-02-25T20:18:51,T438LH6W.FF0W,Vert,0.1884,0.1953,0.965,0.2,0
+5fb5a05e-d7e2-4268-9b85-cfe8d82aa6c9,BE18438,2026-02-25T20:29:50,T438LH6W.XQ0W,Vert,0.1890,0.1953,0.968,0.2,0
+57ec0b10-9324-4459-a07e-0c8532b8928f,BE18438,2026-02-25T20:35:39,T438LH6X.7F0W,Vert,0.1889,0.1953,0.967,0.2,0
+ac86f73f-1174-4fae-a206-ecdcb73e1ffc,BE18438,2026-02-25T20:39:12,T438LH6X.DC0W,Vert,0.1887,0.1953,0.966,0.2,0
+dcf25ddd-9cf0-41c1-b109-f3ef2e490ce5,BE18438,2026-02-25T20:43:41,T438LH6X.KT0W,Vert,0.1871,0.1953,0.958,0.2,0
+aa79bbdd-2392-4ef6-8a5c-873a9f72746d,BE18438,2026-02-25T20:46:51,T438LH6X.Q30W,Vert,0.1877,0.1953,0.961,0.2,0
+ad46d1ce-b417-4c8c-b029-bb3b008ba809,BE18438,2026-02-25T20:53:33,T438LH6Y.190W,Vert,0.1882,0.1953,0.964,0.2,0
+3d3a8f34-86a7-4d70-a7e8-fa768a723325,BE18438,2026-02-26T07:03:39,T438LH7Q.A30W,Vert,0.1809,0.1953,0.926,0.2,0
+7817525f-dc3d-4f5e-b2cb-5af9ca12ede8,BE18438,2026-02-26T07:09:42,T438LH7Q.K60W,Vert,0.1814,0.1953,0.929,0.2,0
+a2091302-0b34-46c1-ab83-174be2b59bcf,BE18438,2026-02-26T13:12:35,T438LH87.CZ0W,Vert,0.1896,0.1953,0.971,0.2,0
+e1c0c9de-a900-479f-9572-9675b13296a9,BE18438,2026-02-26T13:24:46,T438LH87.XA0W,Vert,0.2237,0.2588,0.864,0.2,0
+76a6dc6b-88e1-434b-a137-f5e9b63f248e,BE18438,2026-02-26T13:27:43,T438LH88.270W,Vert,0.3185,0.3223,0.988,0.2,0
+281b508b-7982-4668-86af-9b0b28be608a,BE18438,2026-02-26T13:30:36,T438LH88.700W,Vert,0.3263,0.3320,0.983,0.2,0
+68bbaf77-bb23-40fd-94f9-f928728d90dc,BE18438,2026-02-26T13:33:28,T438LH88.BS0W,Vert,0.3207,0.3271,0.980,0.2,0
+5c975640-6ff1-441a-aa11-41642f6a6b31,BE18438,2026-02-26T13:36:21,T438LH88.GL0W,Vert,0.3261,0.3320,0.982,0.2,0
+dea248ac-1124-4bba-b054-953f86c852b2,BE18438,2026-02-26T13:40:34,T438LH88.NM0W,Vert,0.3271,0.3320,0.985,0.2,0
+4e463201-f008-41ad-9c52-11fcc3d93095,BE18438,2026-02-26T13:43:29,T438LH88.SH0W,Vert,0.3357,0.3418,0.982,0.2,0
+507262c5-d58f-4a05-a6a3-a6cbb5b4de20,BE18438,2026-02-26T13:46:24,T438LH88.XC0W,Vert,0.3372,0.3418,0.987,0.2,0
+4c2d8ee8-842f-4e44-9e03-ff34d26e4fdb,BE18438,2026-02-26T13:49:17,T438LH89.250W,Vert,0.3253,0.3320,0.980,0.2,0
+b184b112-603f-44f4-ae97-087493fbd8ce,BE18438,2026-02-26T13:52:10,T438LH89.6Y0W,Vert,0.3321,0.3369,0.986,0.2,0
+2ff1a552-5a2e-4396-9083-b9c8d334eed7,BE18438,2026-02-26T13:55:04,T438LH89.BS0W,Vert,0.3378,0.3418,0.988,0.2,0
+928e21a7-1990-4792-803e-58014d8140ce,BE18438,2026-02-26T13:57:57,T438LH89.GL0W,Vert,0.3341,0.3369,0.992,0.2,0
+f521b5db-6100-41e6-954e-0abdbf9660c1,BE18438,2026-02-26T14:00:51,T438LH89.LF0W,Vert,0.3421,0.3467,0.987,0.2,0
+efba962f-e511-41ef-9fc5-313f202a9d96,BE18438,2026-02-26T14:03:46,T438LH89.QA0W,Vert,0.3506,0.3564,0.984,0.2,0
+0b51e273-8d10-4494-bda6-8c75eaeec66f,BE18438,2026-02-26T14:06:41,T438LH89.V50W,Vert,0.3448,0.3516,0.981,0.2,0
+6d46e4ee-1647-4169-b6b2-33da32240eb9,BE18438,2026-02-26T14:09:35,T438LH89.ZZ0W,Vert,0.3463,0.3516,0.985,0.2,0
+1ca294cd-4dc2-41fb-9374-e277d4550761,BE18438,2026-02-26T14:12:28,T438LH8A.4S0W,Vert,0.3442,0.3467,0.993,0.2,0
+7b683dcf-1aed-4d9b-be1e-b10f92374398,BE18438,2026-02-26T14:15:23,T438LH8A.9N0W,Vert,0.3486,0.3516,0.992,0.2,0
+2b5cb8e7-c947-4914-bcb5-504f74603547,BE18438,2026-02-26T14:18:21,T438LH8A.EL0W,Vert,0.3519,0.3613,0.974,0.2,0
+5ea3a6bf-10de-4911-aedc-66daecf86b76,BE18438,2026-02-26T14:21:16,T438LH8A.JG0W,Vert,0.3460,0.3564,0.971,0.2,0
+d93be7f5-39ec-45fd-9949-fa531f2f9d34,BE18438,2026-02-26T14:24:13,T438LH8A.OD0W,Vert,0.3457,0.3516,0.983,0.2,0
+cfa113f2-8cee-4a1d-b1e7-0e936fa771e2,BE18438,2026-02-26T14:28:21,T438LH8A.V90W,Vert,0.3485,0.3564,0.978,0.2,0
+46d322ac-38cb-4c99-a4c2-9e9d7ad06171,BE18438,2026-02-26T14:31:17,T438LH8B.050W,Vert,0.3367,0.3467,0.971,0.2,0
+42500a83-bd67-4deb-8bb2-963ee2b8a7ce,BE18438,2026-02-26T14:34:11,T438LH8B.4Z0W,Vert,0.3365,0.3516,0.957,0.2,0
+3ccf6bda-2e88-473c-951d-9ea86d0d9363,BE18438,2026-02-26T14:37:05,T438LH8B.9T0W,Vert,0.3472,0.3564,0.974,0.2,0
+9050487d-d238-4631-b8e1-5a693a026542,BE18438,2026-02-26T14:39:58,T438LH8B.EM0W,Vert,0.3420,0.3564,0.959,0.2,0
+e62989e4-a326-4bae-8d47-a8becee96462,BE18438,2026-02-26T14:42:52,T438LH8B.JG0W,Vert,0.3430,0.3564,0.962,0.2,0
+f0e30fab-dc57-4189-8439-48df17007a2c,BE18438,2026-02-26T14:45:46,T438LH8B.OA0W,Vert,0.3389,0.3516,0.964,0.2,0
+03e1ca09-78df-4c6f-9915-0b6e4f2141d2,BE18438,2026-02-26T14:48:43,T438LH8B.T70W,Vert,0.3449,0.3564,0.968,0.2,0
+945d39f7-1b98-4a1f-80a8-6a18fd9df25f,BE18438,2026-02-26T14:51:37,T438LH8B.Y10W,Vert,0.3522,0.3613,0.975,0.2,0
+d27eb2e3-e313-4125-98f4-f9d6d3737d28,BE18438,2026-02-26T14:54:32,T438LH8C.2W0W,Vert,0.3502,0.3613,0.969,0.2,0
+916901cb-ebea-441e-a2a1-40f4e49f32f4,BE18438,2026-02-26T14:57:25,T438LH8C.7P0W,Vert,0.3524,0.3662,0.962,0.2,0
+5d8d7e29-fdbb-4fa3-bd25-506c9352a354,BE18438,2026-02-26T15:01:35,T438LH8C.EN0W,Vert,0.3530,0.3613,0.977,0.2,0
+c6c7a3af-b03b-43dc-97de-024f8e09b62f,BE18438,2026-02-26T15:04:27,T438LH8C.JF0W,Vert,0.3514,0.3613,0.973,0.2,0
+b6856898-3d10-4967-9577-7f17902f77ea,BE18438,2026-02-26T15:11:34,T438LH8C.VA0W,Vert,0.3566,0.3613,0.987,0.2,0
+7a4802f6-621e-4230-8d8c-d7f0c9a65bf2,BE18438,2026-02-26T15:14:24,T438LH8D.000W,Vert,0.3594,0.3711,0.968,0.2,0
+79f38b7e-ec56-417e-943a-8c944252c2b8,BE18438,2026-02-26T15:18:32,T438LH8D.6W0W,Vert,0.3550,0.3613,0.983,0.2,0
+1b4190c4-0d54-48ba-8ff5-8ddb0eb91e50,BE18438,2026-02-26T15:22:22,T438LH8D.DA0W,Vert,0.3571,0.3613,0.988,0.2,0
+de57b5ae-a3a1-4ede-9d2b-3e87bfb3fd19,BE9558,2026-04-14T11:16:32,K558LJN3.BK0W,Tran,0.3448,0.3662,0.942,0.2,0
+43afeaf5-02ec-41f6-9e23-0c9772821ed4,BE9558,2026-04-14T11:27:15,K558LJN3.TF0W,Tran,0.3094,0.3223,0.960,0.2,0
+8123c0ef-84c9-4f6c-8d82-9dc32e2e470d,BE9558,2026-04-14T14:45:30,K558LJNC.ZU0W,Tran,0.2721,0.3564,0.763,0.2,0
+8c787af3-596e-411b-9e10-29485fa5114f,BE9558,2026-04-29T16:18:47,K558LKF9.BB0W,Tran,0.2943,0.3027,0.972,0.2,0
+431928ff-b4c4-4caa-b933-acc917f3717c,BE9558,2026-05-04T15:02:30,K558LKOF.460W,Tran,0.4364,0.5225,0.835,0.2,0
+6c3c07f8-c36a-4493-acce-7441c9d22cec,BE9558,2026-05-15T08:50:06,K558LL8B.7I0W,Long,0.2892,0.2930,0.987,0.2,0
+13c268df-b652-422e-a642-6453f6a314aa,BE9558,2026-05-15T10:18:34,K558LL8F.AY0W,Long,0.2907,0.2979,0.976,0.2,0
+9b0d0871-8810-467d-806a-5bbfa4e667fe,BE9558,2026-05-15T15:52:00,K558LL8U.QO0W,Long,0.2659,0.2979,0.893,0.2,0
+092640f9-a944-48b7-b872-364d75c2c5e7,BE9558,2026-05-15T16:13:12,K558LL8V.Q00W,Long,0.2428,0.2979,0.815,0.2,0
+0573741a-96ab-4b36-af7a-b5bc11a79009,BE9558,2026-05-16T03:23:26,K558LL9Q.R20W,Long,0.2861,0.2930,0.976,0.2,0
+321f03ea-6696-47de-ad46-2be43d4d6cae,BE9558,2026-05-16T03:30:49,K558LL9R.3D0W,Long,0.2886,0.2930,0.985,0.2,0
+cc423a6a-3e3c-466a-b39c-6146ca7f34c8,BE9558,2026-05-16T03:33:55,K558LL9R.8J0W,Long,0.2880,0.2881,1.000,0.2,0
+efa52a14-cc68-4bea-aab4-4ee3a3cbff3a,BE9558,2026-05-16T03:36:55,K558LL9R.DJ0W,Long,0.2890,0.2930,0.986,0.2,0
+92895bfe-122f-4cdc-bd3d-40609633d278,BE9558,2026-05-16T03:43:41,K558LL9R.OT0W,Long,0.2880,0.2881,1.000,0.2,0
+2a4c81db-2307-44e5-9b27-d5154d98e38d,BE9558,2026-05-16T03:46:36,K558LL9R.TO0W,Long,0.2921,0.2979,0.981,0.2,0
+656b0fdc-e275-4c9c-8936-9eee267ea04e,BE9558,2026-05-16T03:49:28,K558LL9R.YG0W,Long,0.2966,0.2979,0.996,0.2,0
+3b9989ae-e0f2-4652-9724-c145603543d2,BE9558,2026-05-16T03:52:29,K558LL9S.3H0W,Long,0.2880,0.2930,0.983,0.2,0
+837d0d84-d609-4f4a-b542-b7e2b064ea22,BE9558,2026-05-16T03:56:32,K558LL9S.A80W,Long,0.2880,0.2930,0.983,0.2,0
+affcb008-ab99-4bc5-8f35-83dbe620499d,BE9558,2026-05-16T04:02:50,K558LL9S.KQ0W,Long,0.2896,0.2930,0.989,0.2,0
+e1f1c6d2-545c-451f-9018-714d52e20a05,BE9558,2026-05-16T04:05:47,K558LL9S.PN0W,Long,0.3024,0.3076,0.983,0.2,0
+93ef80f1-7e83-4703-a0f9-89bbfeeae6c3,BE9558,2026-05-16T04:08:41,K558LL9S.UH0W,Long,0.2976,0.2979,0.999,0.2,0
+5c118ca9-fc37-4080-876d-b3f57ce0b121,BE9558,2026-05-16T04:11:42,K558LL9S.ZI0W,Long,0.2881,0.2930,0.983,0.2,0
+d98b2eaa-7763-4c78-88fb-c21195cb9978,BE9558,2026-05-16T04:14:38,K558LL9T.4E0W,Long,0.2898,0.2930,0.989,0.2,0
+6a146a01-0d02-4413-b231-ecb4a6fde776,BE9558,2026-05-16T04:17:35,K558LL9T.9B0W,Long,0.2880,0.2930,0.983,0.2,0
+f6587a55-27e3-454f-997e-57e7516a296c,BE9558,2026-05-16T04:20:32,K558LL9T.E80W,Long,0.3046,0.3076,0.990,0.2,0
+1fb48250-2214-41e6-92cb-362aff1ba62b,BE9558,2026-05-16T04:26:05,K558LL9T.NH0W,Long,0.2876,0.2930,0.982,0.2,0
+2f4f3d06-1189-4e28-a1ac-5454ea586a0c,BE9558,2026-05-16T04:59:47,K558LL9V.7N0W,Long,0.2961,0.3027,0.978,0.2,0
+62b7bdab-b744-41e4-8d0e-c9141b9d4821,BE9558,2026-05-16T05:02:46,K558LL9V.CM0W,Long,0.2903,0.2930,0.991,0.2,0
+cadcb6dc-74c0-48b5-974c-87bb7a57816c,BE9558,2026-05-16T05:05:43,K558LL9V.HJ0W,Long,0.3090,0.3125,0.989,0.2,0
+a0235177-be5a-4f11-b741-c65744698e5b,BE9558,2026-05-16T05:08:35,K558LL9V.MB0W,Long,0.3167,0.3223,0.983,0.2,0
+1968c2f6-f52c-4544-a0ed-f79b91ce0da0,BE9558,2026-05-16T05:11:28,K558LL9V.R40W,Long,0.3446,0.3516,0.980,0.2,0
+caa06b94-8813-4949-b47b-61ffb16e61c4,BE9558,2026-05-16T05:14:22,K558LL9V.VY0W,Long,0.3318,0.3320,0.999,0.2,0
+fdaa8101-f033-47e6-9c65-abd76dea1870,BE9558,2026-05-16T05:17:16,K558LL9W.0S0W,Long,0.3407,0.3418,0.997,0.2,0
+29bd16bd-81b9-4ccd-a981-f1a80858536c,BE9558,2026-05-16T05:23:05,K558LL9W.AH0W,Long,0.3450,0.3516,0.981,0.2,0
+1f8bc862-d5fd-4cce-abc0-e2bd8c2f11fa,BE9558,2026-05-16T05:26:00,K558LL9W.FC0W,Long,0.3417,0.3467,0.986,0.2,0
+20220068-d498-4fca-9a2a-a7fd2ee52ab1,BE9558,2026-05-16T05:28:53,K558LL9W.K50W,Long,0.3516,0.3564,0.986,0.2,0
+9816a261-3118-417a-8334-ace36687ad8f,BE9558,2026-05-16T05:31:47,K558LL9W.OZ0W,Long,0.3416,0.3467,0.985,0.2,0
+5d2c76d7-543d-4557-9caf-996a69ab00bd,BE9558,2026-05-16T05:34:42,K558LL9W.TU0W,Long,0.3537,0.3564,0.992,0.2,0
+c7b5ae00-52e7-4920-aa27-49d3b3daff92,BE9558,2026-05-16T05:37:36,K558LL9W.YO0W,Long,0.3517,0.3564,0.987,0.2,0
+07db8fd8-ccef-4a5a-9d97-9f318f73a478,BE9558,2026-05-16T05:40:31,K558LL9X.3J0W,Long,0.3574,0.3613,0.989,0.2,0
+790fcff9-e266-4a12-87f0-a072990d2533,BE9558,2026-05-16T05:43:26,K558LL9X.8E0W,Long,0.3650,0.3662,0.997,0.2,0
+f4245f61-1dde-43dc-a121-98e102f6a193,BE9558,2026-05-16T05:50:28,K558LL9X.K40W,Long,0.3856,0.3906,0.987,0.2,0
+5a3bce7b-358e-4260-876f-3f853163d7cf,BE9558,2026-05-16T05:53:22,K558LL9X.OY0W,Long,0.3847,0.3857,0.997,0.2,0
+37c2b4b1-7a1b-43e7-b513-2c16f582109c,BE9558,2026-05-16T05:56:15,K558LL9X.TR0W,Long,0.3906,0.3955,0.988,0.2,0
+57fde109-1d8a-45a1-b398-5bd28c3e47f1,BE9558,2026-05-16T05:59:10,K558LL9X.YM0W,Long,0.4000,0.4004,0.999,0.2,0
+1e41ab40-93be-405c-a464-f7a57524d3b3,BE9558,2026-05-16T06:02:02,K558LL9Y.3E0W,Long,0.3825,0.3857,0.992,0.2,0
+e9d61eb5-fae8-4902-8fab-4828ffa8e6da,BE9558,2026-05-16T06:04:57,K558LL9Y.890W,Long,0.3802,0.3809,0.998,0.2,0
+c0e964e0-7788-4e41-bc41-6e479c3dc81b,BE9558,2026-05-16T06:07:51,K558LL9Y.D30W,Long,0.3954,0.4004,0.987,0.2,0
+0b2df8e5-1689-4fe6-abad-30970ba10aa6,BE9558,2026-05-16T06:16:06,K558LL9Y.QU0W,Long,0.4051,0.4053,1.000,0.2,0
+2bc6ce73-464b-469c-9e8f-3bf6f6cfae1d,BE9558,2026-05-16T06:21:51,K558LL9Z.0F0W,Long,0.3910,0.3955,0.989,0.2,0
+b8603ee1-616a-451e-a2cd-61f8f2634cd0,BE9558,2026-05-16T06:24:44,K558LL9Z.580W,Long,0.3969,0.4004,0.991,0.2,0
+1aecf582-66a2-4ebb-80f9-04b3aeba331a,BE9558,2026-05-16T06:27:39,K558LL9Z.A30W,Long,0.3915,0.3955,0.990,0.2,0
+885aedfd-7a1a-4120-8176-7c344dd3d8fe,BE9558,2026-05-16T06:30:30,K558LL9Z.EU0W,Long,0.4025,0.4053,0.993,0.2,0
+ec15930f-47f5-4752-afc2-427ee156ac95,BE9558,2026-05-16T06:33:25,K558LL9Z.JP0W,Long,0.4060,0.4102,0.990,0.2,0
+5ca23f47-2a38-4a54-99dd-041287aa9510,BE9558,2026-05-16T06:36:17,K558LL9Z.OH0W,Long,0.4054,0.4102,0.988,0.2,0
+03f371cf-81bc-4681-80f0-ef6fa437da85,BE9558,2026-05-16T06:39:09,K558LL9Z.T90W,Long,0.4052,0.4102,0.988,0.2,0
+ed386f54-140d-45e8-ad51-6585a58b4375,BE9558,2026-05-16T06:44:54,K558LLA0.2U0W,Long,0.3993,0.4004,0.997,0.2,0
+579f1c3f-4c1e-42a3-aa89-69aa93514033,BE9558,2026-05-16T06:47:47,K558LLA0.7N0W,Long,0.4001,0.4053,0.987,0.2,0
+57024de6-07ea-41f4-b916-c6adc3f51636,BE9558,2026-05-16T06:50:40,K558LLA0.CG0W,Long,0.4103,0.4150,0.989,0.2,0
+36bf5adf-afb5-4a5e-a106-1afbd1f56c7b,BE9558,2026-05-16T06:54:49,K558LLA0.JD0W,Long,0.4069,0.4150,0.980,0.2,0
+3e2f831b-5228-46ff-9acb-75d8c86f8bc1,BE9558,2026-05-16T06:57:43,K558LLA0.O70W,Long,0.4108,0.4150,0.990,0.2,0
+1775fa20-568b-4294-b9ca-561201a66c5b,BE9558,2026-05-16T07:00:38,K558LLA0.T20W,Long,0.3967,0.4004,0.991,0.2,0
+fd391469-03f9-4d8d-bccd-3af50c91e5b7,BE9558,2026-05-16T07:03:33,K558LLA0.XX0W,Long,0.3927,0.3955,0.993,0.2,0
+02b52ef6-b707-4dcb-b58f-43da02698832,BE9558,2026-05-16T07:06:27,K558LLA1.2R0W,Long,0.3824,0.3857,0.991,0.2,0
+8d45333f-e101-4249-9e27-913c583a0a8d,BE9558,2026-05-16T07:09:21,K558LLA1.7L0W,Long,0.4085,0.4150,0.984,0.2,0
+361d11ce-f42a-4d04-8270-ee3771fe4f09,BE9558,2026-05-16T07:12:14,K558LLA1.CE0W,Long,0.4017,0.4053,0.991,0.2,0
+59218125-ed5d-4457-aa8b-35ba03441363,BE9558,2026-05-16T07:15:07,K558LLA1.H70W,Long,0.3934,0.3955,0.995,0.2,0
+a603e55a-2ad6-4fc9-8c8c-0dc7be95c3e7,BE9558,2026-05-16T07:23:43,K558LLA1.VJ0W,Long,0.3806,0.3857,0.987,0.2,0
+c8d07902-93ea-4b03-9e69-9ad6ed5bffb1,BE9558,2026-05-16T07:26:36,K558LLA2.0C0W,Long,0.3807,0.3857,0.987,0.2,0
+4c46c0c7-42f0-4648-827e-1995f733b18f,BE9558,2026-05-16T07:30:44,K558LLA2.780W,Long,0.3667,0.3711,0.988,0.2,0
+04759a00-ba3e-4959-a4cb-be8fc1f3004e,BE9558,2026-05-16T07:36:28,K558LLA2.GS0W,Long,0.3514,0.3516,1.000,0.2,0
+0b6b0856-bd87-4a07-a5b4-f13d99e04a12,BE9558,2026-05-16T07:39:20,K558LLA2.LK0W,Long,0.3512,0.3516,0.999,0.2,0
+3d81a8bd-464d-4050-8a76-25e8373d1ddb,BE9558,2026-05-16T07:42:12,K558LLA2.QC0W,Long,0.3350,0.3369,0.994,0.2,0
+b8b9c0c4-c388-4acb-955a-be4ac356afce,BE9558,2026-05-16T07:45:04,K558LLA2.V40W,Long,0.3563,0.3564,1.000,0.2,0
+cb3d142a-fb84-4a21-a298-f61bb869b036,BE9558,2026-05-16T07:49:13,K558LLA3.210W,Long,0.3663,0.3711,0.987,0.2,0
+7401b19b-bc9d-43f0-b395-f4323aca5c72,BE9558,2026-05-16T07:52:06,K558LLA3.6U0W,Long,0.3612,0.3662,0.986,0.2,0
+2236df76-a9e6-401f-9176-ad565242f2cf,BE9558,2026-05-16T07:57:52,K558LLA3.GG0W,Long,0.3366,0.3369,0.999,0.2,0
+a645846c-cae4-4aa5-a2e6-85166bd4dd62,BE9558,2026-05-16T08:00:44,K558LLA3.L80W,Long,0.3265,0.3271,0.998,0.2,0
+7212bd33-99a9-4659-9f0b-c98cc8d1e8bc,BE9558,2026-05-16T08:04:51,K558LLA3.S30W,Long,0.3418,0.3467,0.986,0.2,0
+2cc4bf33-c63a-4c65-b0e2-5e61a9146a5d,BE9558,2026-05-16T08:07:43,K558LLA3.WV0W,Long,0.3405,0.3418,0.996,0.2,0
+4fedb486-5f1f-4623-a01f-083e829f0565,BE9558,2026-05-16T08:10:35,K558LLA4.1N0W,Long,0.3515,0.3564,0.986,0.2,0
+58aff0ba-0f9b-45e1-9bbe-a990d16192ae,BE9558,2026-05-16T08:13:27,K558LLA4.6F0W,Long,0.3513,0.3516,0.999,0.2,0
+e56d87e5-e1d4-4947-b6a5-3bbd2773c54a,BE9558,2026-05-16T08:20:03,K558LLA4.HF0W,Long,0.3564,0.3613,0.986,0.2,0
+0e96e61e-6a21-4e07-8e09-72d09dbfa1e6,BE9558,2026-05-16T08:25:21,K558LLA4.Q90W,Long,0.3565,0.3613,0.987,0.2,0
+19611f28-2da8-4362-a58e-dbed2e061379,BE9558,2026-05-16T08:34:19,K558LLA5.570W,Long,0.3647,0.3662,0.996,0.2,0
+bd580c1d-44d4-41fd-9f83-f462cda4099a,BE9558,2026-05-16T08:37:15,K558LLA5.A30W,Long,0.3536,0.3564,0.992,0.2,0
+2782df3b-3cb4-471a-a6f5-bbc4903a8ab9,BE9558,2026-05-16T08:44:00,K558LLA5.LC0W,Long,0.3612,0.3662,0.986,0.2,0
+d15a08f3-1c7c-4f80-8d9b-f5008c58c3d9,BE9558,2026-05-16T08:49:18,K558LLA5.U60W,Long,0.3610,0.3613,0.999,0.2,0
+03a10a6a-193a-476c-9543-f63818f69417,BE9558,2026-05-16T08:54:36,K558LLA6.300W,Long,0.3515,0.3516,1.000,0.2,0
+1c271856-8100-4c0d-bc8c-9357b47d7626,BE9558,2026-05-16T08:59:54,K558LLA6.BU0W,Long,0.3617,0.3662,0.988,0.2,0
+a85f125e-bfa2-42be-b877-2604b8d068c8,BE9558,2026-05-16T09:05:17,K558LLA6.KT0W,Long,0.3510,0.3516,0.998,0.2,0
+dde9aa71-d2e3-41ac-8388-f2bbddb6dae6,BE9558,2026-05-16T09:08:11,K558LLA6.PN0W,Long,0.3447,0.3467,0.994,0.2,0
+2f3f2be2-0ee3-422e-b3ad-83fb87512a2b,BE9558,2026-05-16T09:14:39,K558LLA7.0F0W,Long,0.3580,0.3613,0.991,0.2,0
+357a3034-8a9c-485e-8cb6-f526dd2baf6f,BE9558,2026-05-16T09:29:11,K558LLA7.ON0W,Long,0.3708,0.3760,0.986,0.2,0
+f8ad0274-8e6f-4714-82ff-d229e60f1442,BE9558,2026-05-16T09:32:06,K558LLA7.TI0W,Long,0.3567,0.3613,0.987,0.2,0
+42bb007a-2726-401a-b085-c25151b0dc24,BE9558,2026-05-16T09:35:02,K558LLA7.YE0W,Long,0.3575,0.3613,0.989,0.2,0
+c99bfc4a-db02-4bea-bea4-943c67485dba,BE9558,2026-05-16T09:37:58,K558LLA8.3A0W,Long,0.3527,0.3564,0.989,0.2,0
+8fc17de0-73f7-48bd-8d93-5451dce47e8e,BE9558,2026-05-16T09:40:52,K558LLA8.840W,Long,0.3613,0.3662,0.987,0.2,0
+bbee5276-03dd-4ca4-a7b9-aec02baac7e8,BE9558,2026-05-16T09:48:13,K558LLA8.KD0W,Long,0.3590,0.3613,0.993,0.2,0
+8dcdee10-db01-4224-9506-4181c3105d6f,BE9558,2026-05-16T09:53:31,K558LLA8.T70W,Long,0.3588,0.3613,0.993,0.2,0
+48c8be72-64d9-4da3-95b5-bbec2c5e1dbd,BE9558,2026-05-16T09:58:49,K558LLA9.210W,Long,0.3582,0.3662,0.978,0.2,0
+0da20919-7ce9-4feb-8ae9-12d9bb40c2b5,BE9558,2026-05-16T10:04:07,K558LLA9.AV0W,Long,0.3506,0.3516,0.997,0.2,0
+c2f7e9bc-8b93-42c3-96d3-e3af217dde35,BE9558,2026-05-16T10:09:25,K558LLA9.JP0W,Long,0.3506,0.3516,0.997,0.2,0
+522a0e42-d443-4447-bc71-af1832c6604e,BE9558,2026-05-16T10:14:43,K558LLA9.SJ0W,Long,0.3624,0.3662,0.990,0.2,0
+67f7e732-3c0b-439f-87c5-3c478a381418,BE9558,2026-05-16T10:20:01,K558LLAA.1D0W,Long,0.3494,0.3516,0.994,0.2,0
+251767e8-56f7-4091-a001-364d8e75c15c,BE9558,2026-05-16T10:25:19,K558LLAA.A70W,Long,0.3463,0.3516,0.985,0.2,0
+9457239d-a4ca-4476-b385-5e01fc8f6420,BE9558,2026-05-16T10:30:37,K558LLAA.J10W,Long,0.3436,0.3467,0.991,0.2,0
+43718d7c-9353-49e1-89b6-0841aeb1b276,BE9558,2026-05-16T10:35:55,K558LLAA.RV0W,Long,0.3452,0.3516,0.982,0.2,0
+f54ba557-ddfe-41d3-8f4c-8bbec9a3783c,BE9558,2026-05-16T10:41:13,K558LLAB.0P0W,Long,0.3419,0.3467,0.986,0.2,0
+dc26c758-ad99-49c1-b518-931368cdf8cf,BE9558,2026-05-16T10:46:31,K558LLAB.9J0W,Long,0.3463,0.3516,0.985,0.2,0
+10917eb2-d151-461c-a8cb-e9660da5a5b0,BE9558,2026-05-16T10:51:49,K558LLAB.ID0W,Long,0.3520,0.3564,0.987,0.2,0
+4938a5e6-d53b-40ee-be2b-06ec0e60398c,BE9558,2026-05-16T10:57:07,K558LLAB.R70W,Long,0.3541,0.3564,0.993,0.2,0
+0711b51d-55bd-48be-b51e-2aad67107a60,BE9558,2026-05-16T11:13:01,K558LLAC.HP0W,Long,0.3661,0.3711,0.986,0.2,0
+065564b6-aa49-40e7-9ea6-c75defdc60fe,BE9558,2026-05-16T11:18:19,K558LLAC.QJ0W,Long,0.3622,0.3662,0.989,0.2,0
+ffb6b230-1d70-4e0e-ba30-e945339a122a,BE9558,2026-05-16T11:23:37,K558LLAC.ZD0W,Long,0.3711,0.3760,0.987,0.2,0
+52b4da82-c1e0-4a8a-8365-51b090787005,BE9558,2026-05-16T11:28:55,K558LLAD.870W,Long,0.3644,0.3662,0.995,0.2,0
+cdbff3d8-fbfe-4bea-8bd6-c898800e81ba,BE9558,2026-05-16T11:39:31,K558LLAD.PV0W,Long,0.3609,0.3662,0.986,0.2,0
+3fb0bc30-1463-4cfc-9e7f-386c0eb2a34a,BE9558,2026-05-16T11:44:49,K558LLAD.YP0W,Long,0.3553,0.3564,0.997,0.2,0
+ba1f5ec8-aef7-4a9e-ac1f-276cf016e285,BE9558,2026-05-16T11:50:07,K558LLAE.7J0W,Long,0.3479,0.3516,0.990,0.2,0
+b1dd4df7-34a9-4512-a903-f8242b236fd2,BE9558,2026-05-16T11:55:25,K558LLAE.GD0W,Long,0.3427,0.3467,0.988,0.2,0
+1f23aa28-c026-4bc7-b3d1-7c78fd76a673,BE9558,2026-05-16T12:00:43,K558LLAE.P70W,Long,0.3454,0.3467,0.996,0.2,0
+0d91f54d-b20d-4a50-b7dd-d071c39a0690,BE9558,2026-05-16T12:06:01,K558LLAE.Y10W,Long,0.3431,0.3467,0.990,0.2,0
+12ccdecd-5fd9-4465-9813-3d6006ec32f3,BE9558,2026-05-16T12:11:19,K558LLAF.6V0W,Long,0.3433,0.3467,0.990,0.2,0
+a7024f1a-f071-48ab-820f-99bdf87366b5,BE9558,2026-05-16T12:16:37,K558LLAF.FP0W,Long,0.3361,0.3369,0.998,0.2,0
+de0fa439-30b8-4607-a3c3-7e17191e4624,BE9558,2026-05-16T12:21:55,K558LLAF.OJ0W,Long,0.3274,0.3320,0.986,0.2,0
+f9374ec3-05e5-477f-8e1a-0c862fed2a50,BE9558,2026-05-16T12:27:13,K558LLAF.XD0W,Long,0.3449,0.3467,0.995,0.2,0
+424cb275-3743-4174-95f0-0ca1205b374c,BE9558,2026-05-16T12:32:31,K558LLAG.670W,Long,0.3370,0.3418,0.986,0.2,0
+9accde13-cefa-41b8-bd0a-cef2ffed59b0,BE9558,2026-05-16T12:37:49,K558LLAG.F10W,Long,0.3465,0.3467,0.999,0.2,0
+4665be38-1956-4b14-b931-068ba9af5c5e,BE9558,2026-05-16T12:48:25,K558LLAG.WP0W,Long,0.3453,0.3516,0.982,0.2,0
+1aedf624-1a99-43e1-a8f2-f9d358b97f61,BE9558,2026-05-16T12:53:43,K558LLAH.5J0W,Long,0.3366,0.3418,0.985,0.2,0
+c15e1799-91d7-425e-ad17-2ba3d170d81e,BE9558,2026-05-16T13:04:19,K558LLAH.N70W,Long,0.3348,0.3369,0.994,0.2,0
+72d4c4ea-19ca-4cc6-9b04-7a2b70122c67,BE9558,2026-05-16T13:09:37,K558LLAH.W10W,Long,0.3272,0.3320,0.986,0.2,0
+b4ebf2c9-1f6d-4594-9810-40479e87c0a5,BE9558,2026-05-16T13:14:55,K558LLAI.4V0W,Long,0.3297,0.3320,0.993,0.2,0
+bb297e9e-548f-4dae-8b1a-0f161af2e766,BE9558,2026-05-16T13:20:13,K558LLAI.DP0W,Long,0.3272,0.3320,0.986,0.2,0
+87562c24-63db-48bd-9962-72a20a40f2ff,BE9558,2026-05-16T13:36:07,K558LLAJ.470W,Long,0.3174,0.3223,0.985,0.2,0
+01e9cc58-3bea-4f38-8499-a2c20e5f8550,BE9558,2026-05-16T13:46:52,K558LLAJ.M40W,Long,0.2880,0.2930,0.983,0.2,0
+5ae7503d-ae17-4063-9208-6a420699e1ab,BE9558,2026-05-16T14:00:35,K558LLAK.8Z0W,Long,0.2879,0.2930,0.983,0.2,0
+e009ea3c-1534-4154-8623-e180ed4db9fb,BE9558,2026-05-16T14:06:19,K558LLAK.IJ0W,Long,0.2882,0.2930,0.984,0.2,0
+cd4e21f5-512b-4b2b-8054-bec01aa06400,BE9558,2026-05-16T14:12:03,K558LLAK.S30W,Long,0.2857,0.2930,0.975,0.2,0
+ae5b780f-2038-4ccf-8bfc-b15b87be139e,BE9558,2026-05-16T14:17:22,K558LLAL.0Y0W,Long,0.2899,0.2930,0.989,0.2,0
+d8c3fb81-8c66-4ebc-b33a-0559bc03334b,BE9558,2026-05-16T14:22:40,K558LLAL.9S0W,Long,0.3009,0.3027,0.994,0.2,0
+990a4c7a-3c51-4d4a-8ab7-416ccb97d29c,BE9558,2026-05-16T14:27:58,K558LLAL.IM0W,Long,0.3057,0.3076,0.994,0.2,0
+68d7a2c3-d899-4e75-96fb-3bf16ae6ea6f,BE9558,2026-05-16T14:33:16,K558LLAL.RG0W,Long,0.3110,0.3125,0.995,0.2,0
+73505b8c-f912-4116-b903-a8a4c75524d3,BE9558,2026-05-16T14:38:34,K558LLAM.0A0W,Long,0.3172,0.3223,0.984,0.2,0
+8d3c9602-7efa-488a-801b-3228bd85ab14,BE9558,2026-05-16T14:43:51,K558LLAM.930W,Long,0.3136,0.3174,0.988,0.2,0
+243e16ee-860a-4960-bc15-b797ad7e9735,BE9558,2026-05-16T14:49:08,K558LLAM.HW0W,Long,0.3135,0.3174,0.988,0.2,0
+e90c8d90-6d41-4788-9ee6-a29e4bfcc16b,BE9558,2026-05-16T15:19:35,K558LLAN.WN0W,Long,0.3132,0.3174,0.987,0.2,0
+4182b0e7-7811-46b8-87d1-a4dd0c152f3b,BE9558,2026-05-16T15:28:45,K558LLAO.BX0W,Long,0.3023,0.3076,0.983,0.2,0
+28a790da-0e48-4fe4-b8ba-2c82cc6bfa79,BE9558,2026-05-16T15:34:02,K558LLAO.KQ0W,Long,0.3176,0.3223,0.985,0.2,0
+093d2fa3-7a4b-4f0f-abda-71155ebe2dfb,BE9558,2026-05-16T15:39:19,K558LLAO.TJ0W,Long,0.3223,0.3271,0.985,0.2,0
+68dc9560-4082-49b6-af04-9244773ffbc1,BE9558,2026-05-16T15:44:36,K558LLAP.2C0W,Long,0.3227,0.3271,0.986,0.2,0
+ad6b268a-165d-4b1d-b0b3-f58db7c9b0e4,BE9558,2026-05-16T15:49:53,K558LLAP.B50W,Long,0.3177,0.3223,0.986,0.2,0
+046c5cb4-56b3-48c0-83c2-bf1f30837940,BE9558,2026-05-16T15:55:10,K558LLAP.JY0W,Long,0.3174,0.3223,0.985,0.2,0
+1b76edf4-3b1c-4bcb-a73c-27334981c350,BE9558,2026-05-16T16:00:27,K558LLAP.SR0W,Long,0.3127,0.3174,0.985,0.2,0
+a0264171-a682-4396-ad8a-cda3b2964533,BE9558,2026-05-16T16:06:55,K558LLAQ.3J0W,Long,0.2885,0.2930,0.985,0.2,0
+707da1f6-ee81-4e57-abd5-98295c14651b,BE9558,2026-05-16T16:12:30,K558LLAQ.CU0W,Long,0.2880,0.2930,0.983,0.2,0
+b8115b64-8739-42c8-b541-713ab6b68bcd,BE9558,2026-05-16T16:17:53,K558LLAQ.LT0W,Long,0.2881,0.2930,0.983,0.2,0
+fb179c36-fef1-4df6-b0a0-2bd7792ea910,BE9558,2026-05-16T16:23:11,K558LLAQ.UN0W,Long,0.2933,0.2979,0.985,0.2,0
+0cec8793-c3c1-486b-a33a-4dbf9cb069b0,BE9558,2026-05-16T16:33:46,K558LLAR.CA0W,Long,0.3017,0.3027,0.997,0.2,0
+c531b052-ee04-4c66-888b-21e97445a615,BE9558,2026-05-16T16:49:37,K558LLAS.2P0W,Long,0.2922,0.2979,0.981,0.2,0
+0dc402f9-cab6-4c4a-8317-b8fc3c4b1ced,BE9558,2026-05-16T16:55:08,K558LLAS.BW0W,Long,0.2898,0.2930,0.989,0.2,0
+fd73733e-b42d-4a51-81d0-4acad6edfbd4,BE9558,2026-05-16T17:16:17,K558LLAT.B50W,Long,0.3115,0.3125,0.997,0.2,0
+6a254daa-ae27-4ac1-b334-58f776a6b276,BE9558,2026-05-16T17:26:51,K558LLAT.SR0W,Long,0.3174,0.3223,0.985,0.2,0
+0cc8e82f-5270-4508-beae-0d3af56095ca,BE9558,2026-05-16T17:37:27,K558LLAU.AF0W,Long,0.2899,0.2930,0.989,0.2,0
+00557a73-8dc3-4cd2-b880-c7820dc8b171,BE9558,2026-05-16T17:43:17,K558LLAU.K50W,Long,0.2881,0.2930,0.983,0.2,0
+8e5518d9-ff0a-4c09-9cf9-a681c84c0b6e,BE9558,2026-05-16T17:53:53,K558LLAV.1T0W,Long,0.2898,0.2930,0.989,0.2,0
+91dbd7e9-7344-400b-82b1-379f2ac4b49a,BE9558,2026-05-16T17:59:11,K558LLAV.AN0W,Long,0.2925,0.2930,0.998,0.2,0
+fa256408-d480-4376-95b5-d1a549c483ea,BE9558,2026-05-16T18:04:29,K558LLAV.JH0W,Long,0.2977,0.3027,0.983,0.2,0
+4ab6e66a-8364-49b6-82f1-86e86199e676,BE9558,2026-05-16T18:09:46,K558LLAV.SA0W,Long,0.2944,0.2979,0.988,0.2,0
+0a65fb7c-96e1-4ba0-85de-4620d88658c1,BE9558,2026-05-16T18:15:03,K558LLAW.130W,Long,0.2965,0.2979,0.995,0.2,0
+92b1ef3e-1696-4123-8662-163947737c83,BE9558,2026-05-16T18:20:20,K558LLAW.9W0W,Long,0.2976,0.3027,0.983,0.2,0
+9e3cb24c-4426-4220-9a73-389d0a96c94f,BE9558,2026-05-16T18:25:37,K558LLAW.IP0W,Long,0.2983,0.3027,0.985,0.2,0
+a3473215-4c46-45ab-9aa7-1720486ec4a8,BE9558,2026-05-16T18:30:54,K558LLAW.RI0W,Long,0.2956,0.2979,0.993,0.2,0
+449665b1-47c8-4bd2-984c-8f45b627e6d8,BE9558,2026-05-16T18:36:11,K558LLAX.0B0W,Long,0.2989,0.3027,0.987,0.2,0
+600cfabc-cffd-42e1-800c-70e75f312316,BE9558,2026-05-16T20:00:12,K558LLB0.WC0W,Long,0.3077,0.3125,0.985,0.2,0
+12bf5944-0194-4f5b-b486-a7acd7141c7f,BE9558,2026-05-16T20:04:28,K558LLB1.3G0W,Long,0.2888,0.2930,0.986,0.2,0
+f1fc913a-91bc-4240-a595-7dfe3091f9a2,BE9558,2026-05-16T20:07:29,K558LLB1.8H0W,Long,0.2977,0.3027,0.983,0.2,0
+e74f7ec3-20bb-4177-beb2-e802e5f801fd,BE9558,2026-05-16T20:10:26,K558LLB1.DE0W,Long,0.2978,0.3027,0.984,0.2,0
+27491c3a-b631-4643-818a-714523911178,BE9558,2026-05-16T20:16:21,K558LLB1.N90W,Long,0.2921,0.2930,0.997,0.2,0
+9d9196ce-11e2-4f28-b01c-3dcc60e60d42,BE9558,2026-05-16T20:22:28,K558LLB1.XG0W,Long,0.2881,0.2930,0.983,0.2,0
+a592d7c8-b643-4b25-9186-3d611ae22709,BE9558,2026-05-16T20:25:37,K558LLB2.2P0W,Long,0.2893,0.2930,0.987,0.2,0
+f709a3d5-27cd-4bb1-b91c-669e8ea11d77,BE9558,2026-05-16T20:28:57,K558LLB2.890W,Long,0.2880,0.2930,0.983,0.2,0
+cc9cad87-b87f-4b67-854a-ffb21e876a92,BE9558,2026-05-16T20:32:45,K558LLB2.EL0W,Long,0.2873,0.2930,0.980,0.2,0
+10732d26-6b49-4b14-a474-6c75e259d278,BE9558,2026-05-16T20:49:32,K558LLB3.6K0W,Long,0.2881,0.2930,0.983,0.2,0
+aa63957a-fdb2-4f17-af52-a63bb015119d,BE9558,2026-05-16T21:21:45,K558LLB4.O90W,Long,0.2880,0.2930,0.983,0.2,0
+8350440f-019d-45a4-b9f9-23058628e6f0,BE9558,2026-05-16T21:32:43,K558LLB5.6J0W,Long,0.2925,0.2930,0.999,0.2,0
+edfc10b2-dbe1-4527-a061-b0d670b5e5d2,BE9558,2026-05-16T21:35:37,K558LLB5.BD0W,Long,0.2891,0.2930,0.987,0.2,0
+f98001a3-88b2-4830-8112-b7c5164e99c3,BE9558,2026-05-16T21:38:40,K558LLB5.GG0W,Long,0.2877,0.2930,0.982,0.2,0
+f1e53e7b-b124-4fc5-82c2-504fb0c8b80d,BE9558,2026-05-16T21:41:46,K558LLB5.LM0W,Long,0.2881,0.2930,0.983,0.2,0
+7890637f-142e-4775-b48c-fd4f3b8ada77,BE9558,2026-05-16T21:45:14,K558LLB5.RE0W,Long,0.2880,0.2930,0.983,0.2,0
+7c97c10f-b413-467c-9b56-c18e423cdcfa,BE9558,2026-05-16T21:48:26,K558LLB5.WQ0W,Long,0.2886,0.2930,0.985,0.2,0
+8178cf22-8428-4f85-8199-0a43f7334a8f,BE9558,2026-05-16T21:51:47,K558LLB6.2B0W,Long,0.2881,0.2930,0.983,0.2,0
+eceb2997-deb4-403a-802f-50e93d23d423,BE9558,2026-05-16T21:57:18,K558LLB6.BI0W,Long,0.2881,0.2930,0.984,0.2,0
+9c26690c-1983-4b21-b8af-7c7967c014c3,BE9558,2026-05-16T22:12:21,K558LLB7.0L0W,Long,0.2876,0.2930,0.982,0.2,0
+78f67fc3-08b7-4317-96ba-ee5544e0b95b,BE9558,2026-05-16T22:19:34,K558LLB7.CM0W,Long,0.2880,0.2930,0.983,0.2,0
+7f9dad14-4f1b-492d-aa26-845a908ea894,BE9558,2026-05-16T22:27:40,K558LLB7.Q40W,Long,0.2881,0.2930,0.983,0.2,0
+f75c52b4-21b9-4065-baeb-5cfbd27d130a,BE9558,2026-05-16T22:33:50,K558LLB8.0E0W,Long,0.2881,0.2930,0.984,0.2,0
+10f7496d-fa76-40a7-8bef-dc14e8a41869,BE9558,2026-05-16T22:40:48,K558LLB8.C00W,Long,0.2878,0.2930,0.982,0.2,0

scratch/offset_scan.py

@@ -0,0 +1,171 @@
+#!/usr/bin/env python3
+"""Scan series-3 waveform binaries for the 'offset' hardware fault.
+
+A healthy geophone trace is centred on zero.  An offset unit sits displaced,
+so the channel mean approaches its own peak.  Detector (unchanged from the
+2026-08-25 run, see memory note `offset-archive-analysis-backlog`):
+
+    dominant-axis  |mean| / peak > 0.7
+    AND            |mean| >= 0.9 * the unit's geo trigger level
+
+Trigger level is read from a paired _ASCII.TXT where one exists, otherwise
+from a per-serial median learned across that unit's ASCII files, otherwise
+--default-trigger.
+
+Serial is decoded from the BW filename: prefix letter encodes thousands
+(chr(ord('B') + n)), next 3 digits the remainder -- T193 -> BE18193.
+
+Usage:
+  python scratch/offset_scan.py --dir <path> [--jobs N] --out offsets.csv
+"""
+from __future__ import annotations
+
+import argparse, csv, json, re, sys
+from collections import defaultdict
+from concurrent.futures import ProcessPoolExecutor, as_completed
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+
+from minimateplus.event_file_io import read_blastware_file
+from minimateplus.bw_ascii_report import parse_report
+
+GEO = ("Tran", "Vert", "Long")
+_GEO_FS_COUNTS = 32000.0
+_WAVE_RE = re.compile(r"\.[A-Za-z0-9]{2}0[Ww]$")
+_STEM_RE = re.compile(r"^([B-Z])(\d{3})")
+
+MEAN_OVER_PEAK_MIN = 0.7
+TRIGGER_FRACTION   = 0.9
+
+
+def serial_from_name(name: str):
+    m = _STEM_RE.match(name)
+    if not m:
+        return None
+    letter, digits = m.group(1), m.group(2)
+    return f"BE{(ord(letter) - ord('B')) * 1000 + int(digits)}"
+
+
+def counts_to_ips(c, gr):
+    return c * (gr or 10.0) / _GEO_FS_COUNTS
+
+
+def scan_one(path_str: str, default_trigger: float) -> dict | None:
+    p = Path(path_str)
+    try:
+        gr, trig = 10.0, None
+        ap = p.with_name(p.name.replace(".", "_", 1) + "_ASCII.TXT") \
+            if False else p.parent / (p.stem + "_" + p.suffix.lstrip(".") + "_ASCII.TXT")
+        if ap.exists():
+            rep = parse_report(ap.read_text(errors="replace"))
+            gr = rep.geo_range_ips or 10.0
+            trig = rep.geo_trigger_level_ips
+        ev = read_blastware_file(p)
+        s = ev.raw_samples or {}
+        if not all(s.get(c) for c in GEO):
+            return None
+        best = None
+        for ch in GEO:
+            arr = s[ch]
+            n = len(arr)
+            if n == 0:
+                continue
+            mean = sum(arr) / n
+            peak = max(abs(v) for v in arr)
+            if peak == 0:
+                continue
+            ratio = abs(mean) / peak
+            if best is None or peak > best["peak_counts"]:
+                best = {"channel": ch, "mean_counts": mean,
+                        "peak_counts": peak, "ratio": ratio}
+        if best is None:
+            return None
+        ts = ev.timestamp
+        return {
+            "serial": serial_from_name(p.name) or "?",
+            "timestamp": (f"{ts.year:04d}-{ts.month:02d}-{ts.day:02d}T"
+                          f"{ts.hour:02d}:{ts.minute:02d}:{ts.second:02d}") if ts else "",
+            "filename": p.name,
+            "channel": best["channel"],
+            "offset_ips": round(counts_to_ips(best["mean_counts"], gr), 4),
+            "peak_ips": round(counts_to_ips(best["peak_counts"], gr), 4),
+            "mean_over_peak": round(best["ratio"], 3),
+            "trigger_level_ips": trig if trig is not None else "",
+            "geo_range_ips": gr,
+        }
+    except Exception:
+        return None
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--dir", required=True)
+    ap.add_argument("--jobs", type=int, default=4)
+    ap.add_argument("--limit", type=int, default=0)
+    ap.add_argument("--default-trigger", type=float, default=0.2)
+    ap.add_argument("--out", required=True)
+    a = ap.parse_args()
+
+    # The DL2 export keeps a byte-identical `Sent/` mirror of the root, so
+    # enumerate paths but keep only the first occurrence of each basename —
+    # otherwise every event is counted twice.
+    seen = set()
+    files = []
+    for q in sorted(Path(a.dir).rglob("*")):
+        if q.is_file() and _WAVE_RE.search(q.name) and q.name not in seen:
+            seen.add(q.name)
+            files.append(q)
+    if a.limit:
+        files = files[: a.limit]
+    print(f"waveform binaries to scan: {len(files)}", flush=True)
+
+    rows = []
+    with ProcessPoolExecutor(max_workers=a.jobs) as ex:
+        futs = [ex.submit(scan_one, str(p), a.default_trigger) for p in files]
+        for n, f in enumerate(as_completed(futs), 1):
+            r = f.result()
+            if r:
+                rows.append(r)
+            if n % 2000 == 0:
+                print(f"  {n}/{len(files)}", flush=True)
+
+    # learn per-serial trigger levels from the rows that had an ASCII
+    by_serial = defaultdict(list)
+    for r in rows:
+        if r["trigger_level_ips"] != "":
+            by_serial[r["serial"]].append(float(r["trigger_level_ips"]))
+    med = {}
+    for k, v in by_serial.items():
+        v.sort()
+        med[k] = v[len(v) // 2]
+
+    for r in rows:
+        if r["trigger_level_ips"] == "":
+            r["trigger_level_ips"] = med.get(r["serial"], a.default_trigger)
+        r["suspect"] = int(
+            r["mean_over_peak"] > MEAN_OVER_PEAK_MIN
+            and abs(r["offset_ips"]) >= TRIGGER_FRACTION * float(r["trigger_level_ips"])
+        )
+
+    cols = ["serial", "timestamp", "filename", "channel", "offset_ips", "peak_ips",
+            "mean_over_peak", "trigger_level_ips", "geo_range_ips", "suspect"]
+    with open(a.out, "w", newline="") as fh:
+        w = csv.DictWriter(fh, fieldnames=cols)
+        w.writeheader()
+        w.writerows(rows)
+
+    sus = [r for r in rows if r["suspect"]]
+    print(f"\nscanned {len(rows)} decodable waveforms")
+    print(f"suspect events: {len(sus)}")
+    per = defaultdict(int)
+    for r in sus:
+        per[r["serial"]] += 1
+    print(f"units with >=1 suspect event: {len(per)} of {len({r['serial'] for r in rows})}")
+    for s, n in sorted(per.items(), key=lambda x: -x[1])[:20]:
+        print(f"  {s:10} {n}")
+    print(f"\nwrote {a.out}")
+
+
+if __name__ == "__main__":
+    main()

scratch/offset_scan2.py

@@ -0,0 +1,108 @@
+#!/usr/bin/env python3
+"""Offset detector v2 — per-channel MEDIAN pedestal.
+
+Supersedes the dominant-axis / mean detector in offset_scan.py, which had two
+flaws that manufactured false "recoveries":
+
+  1. It scored only the axis with the largest peak, so a real event on one axis
+     hid a persistent pedestal on another.  BE12599 2026-08-21 read "clean"
+     because Long had a 1.065 in/s event, while Tran sat at +0.47 in/s.
+  2. It used the MEAN, which a real transient perturbs.  The median is the
+     resting baseline: most samples sit at it, so a blast does not move it.
+     Same event, Long: mean +0.0783 vs median -0.0050.
+
+Flags a CHANNEL when |median| >= --floor in/s (default 0.025 = 5 A/D counts,
+Instantel's own criterion; 1 A/D count = 0.005 in/s).
+
+Emits one row per (event, channel) so persistence can be tracked per channel.
+"""
+from __future__ import annotations
+import argparse, csv, re, statistics, sys
+from concurrent.futures import ProcessPoolExecutor, as_completed
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+from minimateplus.event_file_io import read_blastware_file
+
+GEO = ("Tran", "Vert", "Long")
+K = 10.0 / 32000.0                      # ADC counts -> in/s at the 10 in/s range
+_WAVE_RE = re.compile(r"\.[A-Za-z0-9]{2}0[Ww]$")
+_STEM_RE = re.compile(r"^([B-Z])(\d{3})")
+
+
+def serial_from_name(n):
+    m = _STEM_RE.match(n)
+    return f"BE{(ord(m.group(1))-ord('B'))*1000+int(m.group(2))}" if m else "?"
+
+
+def scan_one(ps):
+    p = Path(ps)
+    try:
+        ev = read_blastware_file(p)
+        s = ev.raw_samples or {}
+        if not all(s.get(c) for c in GEO):
+            return None
+        ts = ev.timestamp
+        stamp = (f"{ts.year:04d}-{ts.month:02d}-{ts.day:02d}T"
+                 f"{ts.hour:02d}:{ts.minute:02d}:{ts.second:02d}") if ts else ""
+        out = []
+        for ch in GEO:
+            a = s[ch]
+            out.append({
+                "serial": serial_from_name(p.name), "timestamp": stamp,
+                "filename": p.name, "channel": ch,
+                "median_ips": round(statistics.median(a) * K, 4),
+                "mean_ips":   round(statistics.fmean(a) * K, 4),
+                "peak_ips":   round(max(abs(v) for v in a) * K, 4),
+            })
+        return out
+    except Exception:
+        return None
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--dir", required=True)
+    ap.add_argument("--jobs", type=int, default=4)
+    ap.add_argument("--floor", type=float, default=0.025)
+    ap.add_argument("--out", required=True)
+    a = ap.parse_args()
+
+    seen, files = set(), []
+    for q in sorted(Path(a.dir).rglob("*")):
+        if q.is_file() and _WAVE_RE.search(q.name) and q.name not in seen:
+            seen.add(q.name); files.append(str(q))
+    print(f"unique waveform binaries: {len(files)}", flush=True)
+
+    rows = []
+    with ProcessPoolExecutor(max_workers=a.jobs) as ex:
+        for n, f in enumerate(as_completed([ex.submit(scan_one, p) for p in files]), 1):
+            r = f.result()
+            if r: rows.extend(r)
+            if n % 2000 == 0: print(f"  {n}/{len(files)}", flush=True)
+
+    for r in rows:
+        r["offset"] = int(abs(r["median_ips"]) >= a.floor)
+
+    cols = ["serial","timestamp","filename","channel","median_ips","mean_ips","peak_ips","offset"]
+    with open(a.out, "w", newline="") as fh:
+        w = csv.DictWriter(fh, fieldnames=cols); w.writeheader(); w.writerows(rows)
+
+    from collections import defaultdict
+    ev_flagged = {(r["serial"], r["filename"]) for r in rows if r["offset"]}
+    ev_all     = {(r["serial"], r["filename"]) for r in rows}
+    per = defaultdict(set)
+    for r in rows:
+        if r["offset"]: per[r["serial"]].add(r["filename"])
+    tot = defaultdict(set)
+    for r in rows: tot[r["serial"]].add(r["filename"])
+    print(f"\nfloor = {a.floor} in/s ({a.floor/0.005:.0f} A/D counts)")
+    print(f"events with >=1 offset channel: {len(ev_flagged)} of {len(ev_all)}")
+    print(f"units affected: {len(per)} of {len(tot)}")
+    for s in sorted(per, key=lambda s: -len(per[s])):
+        print(f"  {s:9} {len(per[s]):4} / {len(tot[s]):4} events")
+    print(f"\nwrote {a.out}")
+
+
+if __name__ == "__main__":
+    main()

scratch/offset_scan3.py

@@ -0,0 +1,95 @@
+#!/usr/bin/env python3
+"""Offset detector v3 — pre-trigger floor, with pre/mid/end consistency.
+
+Brian's method, and better than v2's whole-record median for one reason: the
+pre-trigger window is *definitionally* quiet (it is the buffer captured before
+the trigger fired), whereas a whole-record median is merely robust to the event.
+
+Per channel:
+    pre  = median of the first `pretrig_samples` samples (STRT record)
+    mid  = median of the middle third
+    end  = median of the final third
+    spread = max(pre,mid,end) - min(pre,mid,end)
+
+A DC offset is a *constant floor*: |pre| at or above the floor AND a small
+spread. A transient (settling, handling, a long-tailed event) moves one segment
+relative to the others and is rejected by the spread test.
+
+Floor default 0.025 in/s = 5 A/D counts (Instantel's own criterion; 1 count =
+0.005 in/s).  Quantisation is 0.005 in/s, so `spread` is measured in units of it.
+"""
+from __future__ import annotations
+import argparse, csv, re, statistics, sys
+from concurrent.futures import ProcessPoolExecutor, as_completed
+from pathlib import Path
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+from minimateplus.event_file_io import read_blastware_file
+
+GEO=("Tran","Vert","Long"); K=10.0/32000.0
+_WAVE=re.compile(r"\.[A-Za-z0-9]{2}0[Ww]$"); _STEM=re.compile(r"^([B-Z])(\d{3})")
+
+def serial_of(n):
+    m=_STEM.match(n)
+    return f"BE{(ord(m.group(1))-ord('B'))*1000+int(m.group(2))}" if m else "?"
+
+def scan(ps):
+    p=Path(ps)
+    try:
+        ev=read_blastware_file(p); s=ev.raw_samples or {}
+        if not all(s.get(c) for c in GEO): return None
+        pre_n=ev.pretrig_samples
+        ts=ev.timestamp
+        stamp=(f"{ts.year:04d}-{ts.month:02d}-{ts.day:02d}T"
+               f"{ts.hour:02d}:{ts.minute:02d}:{ts.second:02d}") if ts else ""
+        out=[]
+        for ch in GEO:
+            a=s[ch]; n=len(a); t=n//3
+            pre = a[:pre_n] if (pre_n and 0 < pre_n < n) else a[:t]
+            mid, end = a[t:2*t], a[2*t:]
+            if not pre or not mid or not end: continue
+            v=[statistics.median(x)*K for x in (pre,mid,end)]
+            out.append({"serial":serial_of(p.name),"timestamp":stamp,
+                        "filename":p.name,"channel":ch,
+                        "pretrig_n": pre_n or 0,
+                        "pre":round(v[0],4),"mid":round(v[1],4),"end":round(v[2],4),
+                        "spread":round(max(v)-min(v),4),
+                        "peak":round(max(abs(x) for x in a)*K,4)})
+        return out
+    except Exception:
+        return None
+
+def main():
+    ap=argparse.ArgumentParser()
+    ap.add_argument("--dir",required=True); ap.add_argument("--jobs",type=int,default=4)
+    ap.add_argument("--floor",type=float,default=0.025)
+    ap.add_argument("--max-spread",type=float,default=0.02)
+    ap.add_argument("--out",required=True)
+    a=ap.parse_args()
+    seen=set(); files=[]
+    for q in sorted(Path(a.dir).rglob("*")):
+        if q.is_file() and _WAVE.search(q.name) and q.name not in seen:
+            seen.add(q.name); files.append(str(q))
+    print(f"unique waveform binaries: {len(files)}",flush=True)
+    rows=[]
+    with ProcessPoolExecutor(max_workers=a.jobs) as ex:
+        for i,f in enumerate(as_completed([ex.submit(scan,p) for p in files]),1):
+            r=f.result()
+            if r: rows.extend(r)
+            if i%2000==0: print(f"  {i}/{len(files)}",flush=True)
+    for r in rows:
+        r["offset"]=int(abs(r["pre"])>=a.floor and r["spread"]<=a.max_spread)
+    cols=["serial","timestamp","filename","channel","pretrig_n","pre","mid","end","spread","peak","offset"]
+    with open(a.out,"w",newline="") as fh:
+        w=csv.DictWriter(fh,fieldnames=cols); w.writeheader(); w.writerows(rows)
+    from collections import defaultdict
+    per=defaultdict(set); tot=defaultdict(set)
+    for r in rows:
+        tot[r["serial"]].add(r["filename"])
+        if r["offset"]: per[r["serial"]].add(r["filename"])
+    print(f"\nfloor={a.floor} in/s ({a.floor/0.005:.0f} counts)  max spread={a.max_spread}")
+    print(f"units affected: {len(per)} of {len(tot)}")
+    for s in sorted(per,key=lambda s:-len(per[s])):
+        print(f"  {s:9} {len(per[s]):4} / {len(tot[s]):4} events")
+    print(f"\nwrote {a.out}")
+
+if __name__=="__main__": main()

scratch/verify_against_ascii.py

@@ -0,0 +1,221 @@
+#!/usr/bin/env python3
+"""Verify the series-3 decoder against preserved Blastware ASCII exports.
+
+Pairs each `<stem>_<ext>_ASCII.TXT` with its binary `<stem>.<ext>`, decodes the
+binary with the production codec, and compares against BW's own export:
+
+  waveform  — per-channel sample counts, then every sample value
+  histogram — interval count, then every per-interval channel peak
+
+ADC counts convert as  ips = counts * geo_range_ips / 32000  (1 decoder unit =
+16 counts = 0.005 in/s at the 10 in/s range; see CLAUDE.md).
+
+Usage:
+  python scratch/verify_against_ascii.py --dir <path> [--limit N] [--jobs N]
+                                         [--out results.json] [--kind w|h|all]
+"""
+from __future__ import annotations
+
+import argparse, json, re, sys, traceback
+from concurrent.futures import ProcessPoolExecutor, as_completed
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+
+from minimateplus.event_file_io import read_blastware_file
+from minimateplus.bw_ascii_report import parse_report_file, parse_report
+
+GEO = ("Tran", "Vert", "Long")
+_GEO_FS_COUNTS = 32000.0
+_ASCII_SUFFIX_RE = re.compile(r"_ASCII\.TXT$", re.IGNORECASE)
+
+
+def binary_for(ascii_path: Path) -> Path:
+    """H907KXOW_WC0H_ASCII.TXT -> H907KXOW.WC0H"""
+    stem = _ASCII_SUFFIX_RE.sub("", ascii_path.name)
+    if "_" not in stem:
+        return ascii_path.with_name(stem)
+    head, _, ext = stem.rpartition("_")
+    return ascii_path.with_name(f"{head}.{ext}")
+
+
+def counts_to_ips(counts, geo_range_ips):
+    r = geo_range_ips if geo_range_ips else 10.0
+    return counts * r / _GEO_FS_COUNTS
+
+
+def agrees(got, exp, geo_range_ips, tol=0.0006):
+    """True when decoded `got` matches BW's exported `exp`.
+
+    Saturation carve-out: when an event clips, BW clamps its export to the
+    channel's range maximum (and writes OORANGE for the summary PPV), while
+    the decoder faithfully reproduces raw counts that can sit a decoder unit
+    or two past nominal full scale (32016 counts observed = 10.005 in/s on
+    the 10 in/s range).  Same sign and both at/above the ceiling is agreement,
+    not a decode error.
+    """
+    if abs(got - exp) <= tol:
+        return True
+    r = geo_range_ips if geo_range_ips else 10.0
+    if abs(exp) >= r - tol and abs(got) >= r - tol and (got >= 0) == (exp >= 0):
+        return True
+    return False
+
+
+def parse_interval_table(text: str):
+    """Histogram interval rows: time, Tpk, Tfq, Vpk, Vfq, Lpk, Lfq, PVS, ..., micdB, micfq"""
+    rows = []
+    seen_header = False
+    for line in text.splitlines():
+        if "\t" not in line:
+            continue
+        cols = [c.strip().strip('"') for c in line.split("\t")]
+        cols = [c for c in cols if c != ""]
+        if not seen_header:
+            if any(c in ("Tran", "Vert", "Long") for c in cols):
+                seen_header = True
+            continue
+        if len(cols) < 7:
+            continue
+        if not re.match(r"^\d{1,2}:\d{2}:\d{2}$", cols[0]):
+            continue
+        def num(s):
+            try:
+                return float(s)
+            except ValueError:
+                return None
+        rows.append({"time": cols[0], "Tran": num(cols[1]),
+                     "Vert": num(cols[3]), "Long": num(cols[5])})
+    return rows
+
+
+def check_one(ascii_path_str: str) -> dict:
+    ap = Path(ascii_path_str)
+    bp = binary_for(ap)
+    res = {"ascii": ap.name, "binary": bp.name, "status": "?",
+           "kind": None, "detail": ""}
+    try:
+        if not bp.exists():
+            res["status"] = "no_binary"
+            return res
+        text = ap.read_text(errors="replace")
+        rep = parse_report(text, parse_samples=True)
+        ev = read_blastware_file(bp)
+        gr = rep.geo_range_ips
+        res["kind"] = kind = ("histogram"
+                              if (rep.event_type or "").lower().startswith(("full histogram", "histogram"))
+                              else "waveform")
+        samples = ev.raw_samples or {}
+        dec_n = {c: len(samples.get(c) or []) for c in GEO}
+
+        if kind == "histogram":
+            rows = parse_interval_table(text)
+            res["n_ascii"] = len(rows)
+            res["n_decoded"] = dec_n["Tran"]
+            if not rows:
+                res["status"] = "no_ascii_table"
+                return res
+            if dec_n["Tran"] == 0:
+                res["status"] = "decode_empty"
+                return res
+            if dec_n["Tran"] != len(rows):
+                res["status"] = "count_mismatch"
+                res["detail"] = f"decoded {dec_n['Tran']} vs ascii {len(rows)}"
+                return res
+            bad = 0
+            worst = 0.0
+            for i, row in enumerate(rows):
+                for ch in GEO:
+                    exp = row[ch]
+                    if exp is None:
+                        continue
+                    got = counts_to_ips(samples[ch][i], gr)
+                    if not agrees(got, exp, gr):
+                        bad += 1
+                        worst = max(worst, abs(got - exp))
+            res["worst_abs"] = round(worst, 6)
+            res["status"] = "exact" if bad == 0 else "value_mismatch"
+            if bad:
+                res["detail"] = f"{bad} interval-channel values off"
+            return res
+
+        # waveform
+        asc = rep.samples or []
+        res["n_ascii"] = len(asc)
+        res["n_decoded"] = dec_n["Tran"]
+        if not asc:
+            res["status"] = "no_ascii_table"
+            return res
+        if dec_n["Tran"] == 0:
+            res["status"] = "decode_empty"
+            return res
+        if len({dec_n[c] for c in GEO}) != 1:
+            res["status"] = "channel_len_mismatch"
+            res["detail"] = str(dec_n)
+            return res
+        if dec_n["Tran"] != len(asc):
+            res["status"] = "count_mismatch"
+            res["detail"] = f"decoded {dec_n['Tran']} vs ascii {len(asc)}"
+            return res
+        bad = 0
+        worst = 0.0
+        for i, quad in enumerate(asc):
+            for j, ch in enumerate(GEO):
+                exp = quad[j]
+                got = counts_to_ips(samples[ch][i], gr)
+                if not agrees(got, exp, gr):
+                    bad += 1
+                    worst = max(worst, abs(got - exp))
+        res["worst_abs"] = round(worst, 6)
+        res["status"] = "exact" if bad == 0 else "value_mismatch"
+        if bad:
+            res["detail"] = f"{bad} sample values off"
+        return res
+    except Exception as e:
+        res["status"] = "error"
+        res["detail"] = f"{type(e).__name__}: {e}"
+        return res
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--dir", required=True)
+    ap.add_argument("--limit", type=int, default=0)
+    ap.add_argument("--jobs", type=int, default=8)
+    ap.add_argument("--kind", choices=["w", "h", "all"], default="all")
+    ap.add_argument("--out", default=None)
+    a = ap.parse_args()
+
+    root = Path(a.dir)
+    files = sorted(p for p in root.rglob("*")
+                   if p.is_file() and p.name.upper().endswith("_ASCII.TXT"))
+    if a.kind != "all":
+        want = "0W" if a.kind == "w" else "0H"
+        files = [p for p in files
+                 if _ASCII_SUFFIX_RE.sub("", p.name).upper().endswith(want)]
+    if a.limit:
+        files = files[: a.limit]
+    print(f"pairs to check: {len(files)}", flush=True)
+
+    out = []
+    from collections import Counter
+    tally = Counter()
+    with ProcessPoolExecutor(max_workers=a.jobs) as ex:
+        futs = {ex.submit(check_one, str(p)): p for p in files}
+        for n, f in enumerate(as_completed(futs), 1):
+            r = f.result()
+            out.append(r)
+            tally[(r["kind"], r["status"])] += 1
+            if n % 500 == 0:
+                print(f"  {n}/{len(files)}", flush=True)
+
+    print("\n=== results ===")
+    for (kind, status), n in sorted(tally.items(), key=lambda x: -x[1]):
+        print(f"  {str(kind):10} {status:22} {n}")
+    if a.out:
+        Path(a.out).write_text(json.dumps(out, indent=1))
+        print(f"\nwrote {a.out}")
+
+
+if __name__ == "__main__":
+    main()

scripts/backfill_event_shape.py

@@ -11,7 +11,8 @@ from sfm.shape_metrics import shape_from_h5
 log = logging.getLogger("backfill_event_shape")
 
 def backfill_shape(db: SeismoDb, store: WaveformStore, *, dry_run: bool = False) -> dict:
-    counts = {"updated": 0, "skipped_no_h5": 0, "skipped_no_samples": 0}
+    counts = {"updated": 0, "skipped_no_h5": 0, "skipped_no_samples": 0,
+              "cleared_stale": 0}
     for row in db.query_events(limit=1_000_000):
         serial, filename = row.get("serial"), row.get("blastware_filename")
         if not serial or not filename:
@@ -21,6 +22,20 @@ def backfill_shape(db: SeismoDb, store: WaveformStore, *, dry_run: bool = False)
             counts["skipped_no_h5"] += 1; continue
         shape = shape_from_h5(h5_path)
         if shape is None:
+            # The .h5 can no longer yield a shape (fewer than 2 samples, or a
+            # flat trace).  Clear any previously stored value rather than
+            # leaving it behind — a stale shape outlives the decode it came
+            # from and silently feeds the false-trigger detector.  Seen after
+            # a decoder fix shrinks an event: 493 rows in the prod snapshot
+            # were carrying metrics from a superseded decode (2026-08-25).
+            if row.get("shape_crest_factor") is not None:
+                if not dry_run:
+                    with db._connect() as conn:
+                        conn.execute(
+                            "UPDATE events SET shape_crest_factor=NULL, "
+                            "shape_near_peak_count=NULL, shape_sample_count=NULL, "
+                            "shape_axis=NULL WHERE id=?", (row["id"],))
+                counts["cleared_stale"] += 1
             counts["skipped_no_samples"] += 1; continue
         if not dry_run:
             with db._connect() as conn:

scripts/backfill_sidecars.py

@@ -136,6 +136,7 @@ def main(argv=None) -> int:
     db    = SeismoDb(db_path)
 
     written = skipped = errors = 0
+    stale_h5_removed = 0
     for serial_dir in sorted(p for p in store_root.iterdir() if p.is_dir()):
         serial = serial_dir.name
         for path in sorted(serial_dir.iterdir()):
@@ -388,9 +389,16 @@ def main(argv=None) -> int:
                 # waveform_codec.decode_waveform_v2 or histogram_codec.
                 # decode_histogram_body.  If samples are still empty after
                 # both codecs run, it's a genuine "we can't decode this
-                # file" case (truncated, malformed, or unknown mode);
-                # skip the .h5 write so we don't replace whatever's
-                # there with an empty placeholder.
+                # file" case (truncated, malformed, or unknown mode).
+                #
+                # In that case we REMOVE any existing .h5 rather than leave
+                # it.  Leaving it was the old behaviour and it silently
+                # preserved output from a superseded decoder: after the
+                # 2026-08-25 record-chain fix, 415 histogram files stopped
+                # decoding (an unmapped block variant on BE18193/BE9440) but
+                # kept .h5 files whose peaks were up to 400x the device's own
+                # reported PPV — garbage that fed the charts and the
+                # false-trigger detector with nothing marking it.
                 has_samples = bool(
                     ev.raw_samples and any(
                         ev.raw_samples.get(ch) for ch in ("Tran", "Vert", "Long", "MicL")
@@ -398,6 +406,13 @@ def main(argv=None) -> int:
                 )
                 hdf5_path = store.hdf5_path_for(serial, path.name)
                 hdf5_filename = hdf5_path.name if hdf5_path.exists() else None
+                if not has_samples and hdf5_path.exists():
+                    log.warning("%s: no samples decode — removing stale %s",
+                                path.name, hdf5_path.name)
+                    if not args.dry_run:
+                        hdf5_path.unlink()
+                    hdf5_filename = None
+                    stale_h5_removed += 1
                 hdf5_action = "kept"
                 need_h5 = (
                     not args.skip_hdf5
@@ -458,7 +473,7 @@ def main(argv=None) -> int:
                 log.error("backfill failed for %s: %s", path, exc, exc_info=args.verbose)
                 errors += 1
 
-    print(f"\nDone.  written={written}  skipped(uptodate)={skipped}  errors={errors}")
+    print(f"\nDone.  written={written}  skipped(uptodate)={skipped}  errors={errors}  stale_h5_removed={stale_h5_removed}")
     return 0 if errors == 0 else 1
 
 

sfm/database.py

@@ -81,6 +81,7 @@ CREATE TABLE IF NOT EXISTS events (
     sample_rate             INTEGER,
     record_type             TEXT,                   -- "single_shot" | "continuous"
     false_trigger           INTEGER NOT NULL DEFAULT 0,  -- 0=no, 1=yes (manual flag)
+    reviewed_real           INTEGER NOT NULL DEFAULT 0,  -- 0=no, 1=operator-confirmed real (mutually exclusive with false_trigger)
     blastware_filename      TEXT,                   -- event file within waveform store; extension is per-event (AB0T encodes timestamp)
     blastware_filesize      INTEGER,                -- bytes; NULL if no event file saved
     a5_pickle_filename      TEXT,                   -- "<filename>.a5.pkl" sidecar
@@ -224,6 +225,7 @@ class SeismoDb:
             ("shape_near_peak_count", "INTEGER"),
             ("shape_sample_count",    "INTEGER"),
             ("shape_axis",            "TEXT"),
+            ("reviewed_real",         "INTEGER NOT NULL DEFAULT 0"),
         ):
             if col not in existing_cols:
                 log.info("_migrate: events ADD COLUMN %s %s", col, ddl)
@@ -601,13 +603,135 @@ class SeismoDb:
             ).fetchall()
         return [dict(r) for r in rows]
 
-    def set_false_trigger(self, event_id: str, value: bool) -> bool:
-        """Set or clear the false_trigger flag on an event. Returns True if found."""
+    def find_twins(self, event_id: str, *, window_seconds: int | None = None) -> list[dict]:
+        """
+        Find this event's histogram/waveform twin(s): the SAME physical event
+        recorded both as a scheduled histogram and as a triggered waveform.
+
+        A real trigger is captured twice — once as a triggered waveform (stamped
+        at the trigger instant) and once inside the scheduled histogram whose
+        interval contains it (stamped at the histogram's interval start, e.g. the
+        7am/7pm call-in). The two can be HOURS apart in time yet report the same
+        serial and identical peak_vector_sum. Twins are therefore matched by:
+
+          * same serial,
+          * identical peak_vector_sum,
+          * OPPOSITE record type (one histogram, one waveform), and
+          * the waveform's timestamp falls within the histogram's interval —
+            from a histogram's timestamp up to the next histogram (same serial).
+
+        This replaces the old ±``window_seconds`` heuristic, which silently
+        missed twins more than a few minutes apart (a histogram's interval-start
+        stamp and the trigger instant routinely differ by hours). ``window_seconds``
+        is still accepted for backward compatibility but is ignored.
+
+        Returns [] if the event or a required field (serial / peak_vector_sum /
+        timestamp) is missing.
+
+        Caveat: identical-PVS matching remains a proxy for "same physical event"
+        — if the device clamps/saturates PVS (to sqrt(3) * geo_range), two
+        distinct saturated events could share a PVS. The added opposite-type and
+        interval constraints make a false pairing far less likely than the old
+        time-window match, and false_trigger/reviewed_real stay re-derivable from
+        the sidecar source of truth.
+        """
+        def _parse(ts):
+            if not ts:
+                return None
+            try:
+                return datetime.datetime.fromisoformat(str(ts).replace(" ", "T"))
+            except ValueError:
+                return None
+
+        def _is_hist(rt):
+            return str(rt or "").lower().startswith("hist")
+
+        row = self.get_event(event_id)
+        if not row:
+            return []
+        serial = row.get("serial"); pvs = row.get("peak_vector_sum")
+        t_target = _parse(row.get("timestamp"))
+        if serial is None or pvs is None or t_target is None:
+            return []
+        target_hist = _is_hist(row.get("record_type"))
+
         with self._connect() as conn:
-            cur = conn.execute(
-                "UPDATE events SET false_trigger=? WHERE id=?",
-                (1 if value else 0, event_id),
-            )
+            cand_rows = [dict(r) for r in conn.execute(
+                "SELECT * FROM events WHERE serial=? AND id!=? AND peak_vector_sum=?",
+                (serial, event_id, pvs)).fetchall()]
+            hist_ts = [r["timestamp"] for r in conn.execute(
+                "SELECT timestamp FROM events WHERE serial=? AND lower(record_type) LIKE 'hist%'",
+                (serial,)).fetchall()]
+
+        # Histogram interval-start times for this serial, sorted, to bound intervals.
+        starts = sorted(x for x in (_parse(t) for t in hist_ts) if x is not None)
+
+        def _interval_end(h_start):
+            # The next histogram strictly after h_start bounds the interval; else open-ended.
+            for x in starts:
+                if x > h_start:
+                    return x
+            return None
+
+        def _covers(h_start, w_time):
+            end = _interval_end(h_start)
+            return h_start <= w_time and (end is None or w_time < end)
+
+        twins = []
+        for c in cand_rows:
+            if _is_hist(c.get("record_type")) == target_hist:
+                continue  # twins are strictly cross-type (one histogram, one waveform)
+            c_time = _parse(c.get("timestamp"))
+            if c_time is None:
+                continue
+            h_start, w_time = (t_target, c_time) if target_hist else (c_time, t_target)
+            if _covers(h_start, w_time):
+                twins.append(c)
+        return twins
+
+    def propagate_review_to_twins(self, event_id: str, *, window_seconds: int | None = None) -> list[str]:
+        """
+        Copy this event's `false_trigger`/`reviewed_real` columns onto each
+        of its histogram/waveform twins (see `find_twins`), so flagging one
+        twin flags both.  Returns the list of twin ids updated.
+
+        ``window_seconds`` is accepted for backward compatibility but ignored;
+        twin matching is now interval-based (see `find_twins`).
+        """
+        row = self.get_event(event_id)
+        if not row:
+            return []
+        ft = 1 if row.get("false_trigger") else 0
+        real = 1 if row.get("reviewed_real") else 0
+        twins = self.find_twins(event_id)
+        moved = []
+        with self._connect() as conn:
+            for tw in twins:
+                conn.execute("UPDATE events SET false_trigger=?, reviewed_real=? WHERE id=?",
+                             (ft, real, tw["id"]))
+                moved.append(tw["id"])
+        return moved
+
+    def set_false_trigger(self, event_id: str, value: bool) -> bool:
+        """
+        Set or clear the false_trigger flag on an event. Returns True if found.
+
+        Mirrors the 3-state exclusivity enforced by `update_event_review`:
+        setting false_trigger true also clears `reviewed_real` (a false
+        trigger can't also be a confirmed-real event). Clearing false_trigger
+        leaves `reviewed_real` untouched.
+        """
+        with self._connect() as conn:
+            if value:
+                cur = conn.execute(
+                    "UPDATE events SET false_trigger=1, reviewed_real=0 WHERE id=?",
+                    (event_id,),
+                )
+            else:
+                cur = conn.execute(
+                    "UPDATE events SET false_trigger=0 WHERE id=?",
+                    (event_id,),
+                )
         return cur.rowcount > 0
 
     def delete_event(self, event_id: str) -> Optional[dict]:
@@ -683,17 +807,23 @@ class SeismoDb:
         """
         Sync derived index columns from a sidecar's `review` block.
 
-        Currently the only derived index is `events.false_trigger` — kept
-        in sync so `/db/events?false_trigger=true` queries don't have to
-        scan every sidecar JSON on disk.  The sidecar JSON itself remains
-        the source of truth for the full review state.
+        The derived indexes are `events.false_trigger` and
+        `events.reviewed_real` — kept in sync so `/db/events` queries don't
+        have to scan every sidecar JSON on disk.  The sidecar JSON itself
+        remains the source of truth for the full review state.
+
+        The two columns are mutually exclusive: setting either one true
+        forces the other's column to 0.
 
         Returns True when the row exists, False otherwise.  No-op fields
-        (review without `false_trigger`) leave the column untouched.
+        (review without `false_trigger` or `reviewed_real`) leave both
+        columns untouched.
         """
         if not isinstance(review, dict):
             return False
-        if "false_trigger" not in review:
+        has_ft = "false_trigger" in review
+        has_real = "reviewed_real" in review
+        if not has_ft and not has_real:
             # Nothing derived to update; just confirm the row exists.
             with self._connect() as conn:
                 row = conn.execute(
@@ -701,12 +831,20 @@ class SeismoDb:
                 ).fetchone()
             return row is not None
 
-        flag = 1 if review.get("false_trigger") else 0
+        sets = {}
+        if has_ft:
+            sets["false_trigger"] = 1 if review.get("false_trigger") else 0
+        if has_real:
+            sets["reviewed_real"] = 1 if review.get("reviewed_real") else 0
+        # mutual exclusivity: a true in one forces the other column to 0
+        if sets.get("false_trigger") == 1:
+            sets["reviewed_real"] = 0
+        if sets.get("reviewed_real") == 1:
+            sets["false_trigger"] = 0
+        assign = ", ".join(f"{k}=?" for k in sets)
+        params = list(sets.values()) + [event_id]
         with self._connect() as conn:
-            cur = conn.execute(
-                "UPDATE events SET false_trigger=? WHERE id=?",
-                (flag, event_id),
-            )
+            cur = conn.execute(f"UPDATE events SET {assign} WHERE id=?", params)
         return cur.rowcount > 0
 
     # ── Monitor log ───────────────────────────────────────────────────────────

sfm/event_hdf5.py

@@ -77,6 +77,20 @@ _GEO_FS_BY_RANGE = {
 }
 _INT16_FS = 32768.0
 
+# Geophone full-scale count.  NOT 32768: the verified body codec emits geo
+# samples in 16-count units whose documented LSB is exactly 0.005 in/s, and
+# ``waveform_codec.decoded_to_adc_counts`` multiplies by 16 — so one ADC count
+# is 0.005/16 in/s and Normal range (10.000 in/s) is 10.0 / (0.005/16) = 32000
+# counts.  Using 32768 here made every geophone reading 2.3% low
+# (1 - 32000/32768 = 0.0234).
+#
+# Confirmed 2026-08-25 against 216 per-channel comparisons with the preserved
+# Blastware ASCII exports: 32000 gives 216/216 exact within 1 LSB (worst error
+# 0.005 in/s); 32768 gave 151/216 with a worst error of 0.238 in/s on a
+# 10 in/s event.  The mic path is unaffected — it back-solves its own scale
+# from the device-reported peak (see _mic_scale_factor).
+_GEO_INT16_FS = 32000.0
+
 # Default mic conversion: ADC count → psi.  Approximate; exact factor
 # depends on firmware reference voltage and mic sensitivity, neither of
 # which is independently confirmed.  We try to refine it from the device-
@@ -125,15 +139,14 @@ def _samples_to_float(
 ) -> np.ndarray:
     """Convert int16 ADC counts → float32 physical units.
 
-    Uses _INT16_FS=32768 (not 32767) so that a count of -32768 maps to
-    exactly -full_scale and +32767 maps to ~+full_scale * 32767/32768.
-    Matches the device firmware's documented mapping (see CLAUDE.md
-    geo_hardware_constant rationale).
+    Uses _GEO_INT16_FS=32000 (see the constant's rationale): one decoder
+    unit (16 ADC counts) is exactly 0.005 in/s, so full scale is 32000
+    counts, not 32768.
     """
     if not samples_int16:
         return np.array([], dtype=np.float32)
     arr = np.asarray(samples_int16, dtype=np.int32)  # int32 to avoid overflow during scale
-    return (arr.astype(np.float32) * (full_scale / _INT16_FS)).astype(np.float32)
+    return (arr.astype(np.float32) * (full_scale / _GEO_INT16_FS)).astype(np.float32)
 
 
 def _mic_scale_factor(

sfm/server.py

@@ -67,6 +67,7 @@ from minimateplus.blastware_file import write_blastware_file, blastware_filename
 from minimateplus.client import _decode_a5_metadata_into, _decode_a5_waveform, _decode_event_count
 from minimateplus.framing import build_bw_write_frame, SESSION_RESET, POLL_PROBE, POLL_DATA
 from minimateplus.protocol import SUB_STOP_MONITORING
+from minimateplus.event_file_io import TOOL_VERSION as SFM_VERSION  # single source for the service version (release-bumped)
 from sfm import event_hdf5
 from sfm.cache import SFMCache, get_cache
 from sfm.database import SeismoDb
@@ -90,7 +91,7 @@ app = FastAPI(
         "Implements the minimateplus RS-232 protocol library.\n"
         "Proxied by terra-view at /api/sfm/*."
     ),
-    version="0.24.0",
+    version=SFM_VERSION,
 )
 
 # Allow requests from the waveform viewer opened as a local file (file://)
@@ -371,7 +372,7 @@ def _backfill_events(events: list, info: "DeviceInfo") -> None:
 @app.get("/health")
 def health() -> dict:
     """Service heartbeat.  No device I/O."""
-    return {"status": "ok", "service": "sfm", "version": "0.1.0"}
+    return {"status": "ok", "service": "sfm", "version": SFM_VERSION}
 
 
 @app.get("/", response_class=FileResponse)
@@ -1957,7 +1958,10 @@ def db_set_false_trigger(
     value:    bool = Query(..., description="True to flag as false trigger, False to clear"),
 ) -> dict:
     """
-    Set or clear the false_trigger flag on a single event.
+    Set or clear the false_trigger flag on a single event.  Enforces the
+    same 3-state exclusivity as the sidecar review PATCH (flagging false_trigger
+    clears reviewed_real) and propagates the resulting flags to the event's
+    histogram/waveform twin(s), same as the sidecar path.
 
     Used by the terra-view event review UI.
     Returns 404 if the event_id is not found.
@@ -1965,6 +1969,12 @@ def db_set_false_trigger(
     found = _get_db().set_false_trigger(event_id, value)
     if not found:
         raise HTTPException(status_code=404, detail=f"Event {event_id} not found")
+
+    try:
+        _get_db().propagate_review_to_twins(event_id)
+    except Exception as exc:
+        log.warning("twin review-propagation failed for %s: %s", event_id, exc)
+
     return {"status": "ok", "event_id": event_id, "false_trigger": value}
 
 
@@ -2481,6 +2491,13 @@ def db_event_sidecar_patch(event_id: str, body: SidecarPatchBody) -> dict:
     if body.review is not None:
         _get_db().update_event_review(event_id, new_sidecar.get("review", {}))
 
+        # Propagate the review to the event's histogram/waveform twin(s) so
+        # flagging one flags both (column-level; twins share serial+PVS+near time).
+        try:
+            _get_db().propagate_review_to_twins(event_id)
+        except Exception as exc:
+            log.warning("twin review-propagation failed for %s: %s", event_id, exc)
+
     return new_sidecar
 
 

tests/test_event_file_io.py

@@ -626,3 +626,12 @@ if __name__ == "__main__":
                 failed += 1
         print(f"\n{passed} passed, {failed} failed")
         sys.exit(0 if failed == 0 else 1)
+
+
+def test_peaks_from_samples_uses_32000_full_scale():
+    """`_peaks_from_samples` must use the same 32000-count geo full scale as
+    the .h5 writer, or sidecar peaks disagree with the plotted waveform by
+    2.3%.  See test_event_hdf5.test_geo_full_scale_count_is_32000."""
+    from minimateplus.event_file_io import _peaks_from_samples
+    pv = _peaks_from_samples({"Tran": [32000], "Vert": [0], "Long": [0], "MicL": []})
+    assert abs(pv.tran - 10.0) < 1e-4

tests/test_event_hdf5.py

@@ -99,8 +99,14 @@ def test_hdf5_round_trip_preserves_metadata(tmp_path: Path):
 
 
 def test_hdf5_samples_in_physical_units_normal_range(tmp_path: Path):
-    """Vert hits ADC full-scale (32767) → with Normal range FS=10 in/s,
-    the HDF5 sample value should be ≈ 10 * 32767/32768 in/s."""
+    """Vert hits 32767 ADC counts → with Normal range FS=10 in/s that is
+    ``10 * 32767/32000`` in/s.
+
+    Geo full scale is 32000 counts, not 32768 (see
+    test_geo_full_scale_count_is_32000), so 32767 counts sits slightly
+    ABOVE nominal full scale -- the ADC has headroom past 10.000 in/s,
+    which is why Blastware reports peaks like 10.14 in/s.  This test
+    previously asserted the 32768 scale and was wrong by 2.3%."""
     ev = _make_event_with_samples()
     h5 = tmp_path / "n.h5"
     event_hdf5.write_event_hdf5(h5, ev, serial="BE11529", geo_range="normal")
@@ -110,7 +116,7 @@ def test_hdf5_samples_in_physical_units_normal_range(tmp_path: Path):
     assert vert.dtype.name == "float32"
     assert max(abs(v) for v in vert) > 9.99   # full-scale ≈ 10.0
     # The dirac was at n//2 → 32767 ADC counts.
-    expected_peak = 10.0 * 32767 / 32768
+    expected_peak = 10.0 * 32767 / 32000
     assert abs(max(vert) - expected_peak) < 1e-3
 
 
@@ -122,7 +128,7 @@ def test_hdf5_samples_in_physical_units_sensitive_range(tmp_path: Path):
     data = event_hdf5.read_event_hdf5(h5)
 
     vert = data["samples"]["Vert"]
-    expected_peak = 1.250 * 32767 / 32768
+    expected_peak = 1.250 * 32767 / 32000
     assert abs(max(vert) - expected_peak) < 1e-4
 
 
@@ -294,3 +300,35 @@ if __name__ == "__main__":
                 failed += 1
         print(f"\n{passed} passed, {failed} failed")
         sys.exit(0 if failed == 0 else 1)
+
+
+# ── Geophone full-scale count ───────────────────────────────────────────────
+
+def test_geo_full_scale_count_is_32000():
+    """Geo full scale is 32000 ADC counts, not 32768.
+
+    The verified body codec emits geo samples in 16-count units with a
+    documented LSB of exactly 0.005 in/s, and ``decoded_to_adc_counts``
+    multiplies by 16 — so one ADC count is 0.005/16 in/s and Normal range
+    (10.000 in/s) is 10.0 / (0.005/16) = 32000 counts.
+
+    Using 32768 made every geophone reading 2.3% low (1 - 32000/32768).
+    Confirmed 2026-08-25 against 216 channel comparisons with preserved
+    Blastware ASCII exports: 32000 → 216/216 exact within 1 LSB;
+    32768 → 151/216, worst error 0.238 in/s on a 10 in/s event.
+    """
+    from sfm.event_hdf5 import _GEO_INT16_FS
+    assert _GEO_INT16_FS == 32000.0
+
+
+def test_samples_to_float_lsb_is_exactly_5_milli_ips():
+    """One decoder unit (= 16 ADC counts) must be exactly 0.005 in/s."""
+    from sfm.event_hdf5 import _samples_to_float
+    out = _samples_to_float([16], 10.0)
+    assert abs(float(out[0]) - 0.005) < 1e-9
+
+
+def test_samples_to_float_full_scale_count_maps_to_full_scale():
+    from sfm.event_hdf5 import _samples_to_float
+    assert abs(float(_samples_to_float([32000], 10.0)[0]) - 10.0) < 1e-4
+    assert abs(float(_samples_to_float([32000], 1.25)[0]) - 1.25) < 1e-5

tests/test_find_twins.py

@@ -0,0 +1,71 @@
+import sqlite3
+from sfm.database import SeismoDb
+from minimateplus.models import Event, Timestamp
+
+
+def _ins(db, key, serial, pvs, ts, record_type="Waveform"):
+    ev = Event(index=0)
+    ev._waveform_key = bytes.fromhex(key)
+    ev.timestamp = ts
+    db.insert_events([ev], serial=serial)
+    row = [r for r in db.query_events(serial=serial) if r["waveform_key"] == key][0]
+    with sqlite3.connect(db.db_path) as c:
+        c.execute("UPDATE events SET peak_vector_sum=?, record_type=? WHERE id=?",
+                  (pvs, record_type, row["id"]))
+    return row["id"]
+
+
+def _ts(hour, minute, second=0, day=25):
+    return Timestamp(raw=b"", flag=0x10, year=2026, unknown_byte=0, month=2, day=day,
+                     hour=hour, minute=minute, second=second)
+
+
+def test_histogram_and_waveform_twin_across_hours(tmp_path):
+    # The real UM12947 case: histogram stamped at its 7pm interval start, the
+    # triggered waveform 75 min later — same serial + identical PVS. The old
+    # ±5-min window missed this; interval matching catches it, both directions.
+    db = SeismoDb(tmp_path / "s.db")
+    hist_pm = _ins(db, "01110001", "BE1", 0.4763, _ts(19, 31, 17), "Histogram")
+    wave    = _ins(db, "01110002", "BE1", 0.4763, _ts(20, 46, 44), "Waveform")
+    _ins(db, "01110003", "BE1", 0.0100, _ts(7, 0, 0, day=26), "Histogram")  # bounds the interval
+    assert {r["id"] for r in db.find_twins(hist_pm)} == {wave}
+    assert {r["id"] for r in db.find_twins(wave)} == {hist_pm}
+
+
+def test_same_type_not_twinned(tmp_path):
+    # Two waveforms, same serial + PVS, seconds apart → NOT twins (cross-type only).
+    db = SeismoDb(tmp_path / "s.db")
+    a = _ins(db, "01110001", "BE1", 0.4763, _ts(20, 19, 5), "Waveform")
+    _ins(db, "01110002", "BE1", 0.4763, _ts(20, 19, 45), "Waveform")
+    assert db.find_twins(a) == []
+
+
+def test_waveform_matches_only_the_containing_interval(tmp_path):
+    # Two overnight intervals with the same PVS; a waveform in the SECOND interval
+    # must twin with that histogram, never the first — even though PVS matches both.
+    db = SeismoDb(tmp_path / "s.db")
+    h1 = _ins(db, "01110001", "BE1", 0.4763, _ts(19, 0, 0, day=25), "Histogram")
+    h2 = _ins(db, "01110002", "BE1", 0.4763, _ts(7, 0, 0, day=26), "Histogram")
+    w = _ins(db, "01110003", "BE1", 0.4763, _ts(8, 0, 0, day=26), "Waveform")
+    assert {r["id"] for r in db.find_twins(w)} == {h2}
+    assert w not in {r["id"] for r in db.find_twins(h1)}
+
+
+def test_different_pvs_not_twinned(tmp_path):
+    db = SeismoDb(tmp_path / "s.db")
+    h = _ins(db, "01110001", "BE1", 0.4763, _ts(19, 0, 0), "Histogram")
+    _ins(db, "01110002", "BE1", 0.9999, _ts(20, 0, 0), "Waveform")  # different PVS
+    assert db.find_twins(h) == []
+
+
+def test_open_ended_latest_interval(tmp_path):
+    # A waveform after the latest histogram (nothing bounds the interval) still twins.
+    db = SeismoDb(tmp_path / "s.db")
+    h = _ins(db, "01110001", "BE1", 0.4763, _ts(19, 0, 0), "Histogram")
+    w = _ins(db, "01110002", "BE1", 0.4763, _ts(23, 30, 0), "Waveform")
+    assert {r["id"] for r in db.find_twins(h)} == {w}
+
+
+def test_missing_fields_returns_empty(tmp_path):
+    db = SeismoDb(tmp_path / "s.db")
+    assert db.find_twins("nonexistent-id") == []

tests/test_health_version.py

@@ -0,0 +1,17 @@
+"""The /health version must track the release, not a stale literal.
+
+terra-view's SFM Admin page displays whatever `/health` reports. It was
+hardcoded to "0.1.0" and never bumped, so the page showed 0.1.0 while the
+service was actually 0.26.0. These guard against that regression — and run
+without httpx (they call the endpoint function directly, no TestClient).
+"""
+from minimateplus.event_file_io import TOOL_VERSION
+from sfm.server import app, health
+
+
+def test_health_reports_current_tool_version():
+    assert health()["version"] == TOOL_VERSION
+
+
+def test_openapi_version_matches_tool_version():
+    assert app.version == TOOL_VERSION

tests/test_histogram_codec.py

@@ -354,23 +354,29 @@ _K558_INTERVAL_12_BLOCK = bytes.fromhex(
 
 
 def test_extension_byte_does_not_inflate_peak():
-    """The annotation byte at [7]/[11]/[15]/[19] must NOT contribute to
-    the peak count.  Decoded T_peak must be 3 (uint8 byte[6]), NOT
-    53763 (uint16 LE byte[6:8])."""
+    """The byte after each peak must NOT contribute to the peak count.
+
+    Still true, but for a different reason than originally recorded: the
+    block is uniformly **big-endian**, so T_peak is uint16 BE at [5:7]
+    (= 3 here) and the 0xd2 at [7] is the HIGH BYTE of the big-endian
+    half-period at [7:9], not an "annotation" field.  Reading the peak as
+    uint16 LE at [6:8] gave 53763 → 268 in/s, which is what this test was
+    written to prevent.
+    """
     body = _K558_INTERVAL_12_BLOCK
     records = decode_histogram_body_full(body)
     assert records is not None
     assert len(records) == 1
     r = records[0]
-    assert r["t_peak"] == 3,    f"T_peak should be 3 (uint8), got {r['t_peak']}"
+    assert r["t_peak"] == 3,    f"T_peak should be 3, got {r['t_peak']}"
     assert r["v_peak"] == 2
     assert r["l_peak"] == 2
     assert r["m_peak"] == 16
-    # Half-periods unchanged — still uint16 LE.
-    assert r["t_halfp"] == 0x0045  # 69 → 7.4 Hz
+    # Half-period is uint16 BE — 0xd245 = 53829 samples → 0.0095 Hz, which
+    # is exactly the sub-Hz drift BW rendered as "<1.0" for this interval.
+    assert r["t_halfp"] == 0xd245
+    assert half_period_to_hz(r["t_halfp"]) < 1.0
     assert r["m_halfp"] == 6       # → 85.3 Hz
-    # Annotation byte is preserved (for future RE) but does not affect peak.
-    assert r["annotations"] == (0xd2, 0x00, 0x00, 0x00)
 
 
 def test_extension_byte_decoded_to_correct_in_s():
@@ -383,3 +389,292 @@ def test_extension_byte_decoded_to_correct_in_s():
     assert channels["Vert"] == [2]
     assert channels["Long"] == [2]
     assert channels["MicL"] == [16]
+
+
+# ── Big-endian block layout + terminal block (2026-08-25) ───────────────────
+#
+# Verified against 1211 production histograms paired with their Blastware
+# ASCII exports: 1211/1211 decode exactly (interval count + every
+# per-interval peak), and 842,442 per-interval frequency comparisons match
+# with zero mismatches.
+
+def _mk_block(t_peak=3, t_halfp=69, v_peak=2, v_halfp=69, l_peak=2, l_halfp=69,
+              m_peak=16, m_halfp=6, seg=0, ctr=256, tail=b"\x1e\x0a\x00\x00",
+              b22=0x00):
+    """Build one synthetic 32-byte histogram block, big-endian throughout."""
+    b = bytearray(32)
+    b[0] = 0x00
+    b[1] = seg
+    b[2], b[3] = ctr & 0xFF, (ctr >> 8) & 0xFF        # block_ctr is uint16 LE
+    b[4] = 0x0A                                        # marker, uint8
+    for off, val in ((5, t_peak), (7, t_halfp), (9, v_peak), (11, v_halfp),
+                     (13, l_peak), (15, l_halfp), (17, m_peak), (19, m_halfp)):
+        b[off], b[off + 1] = (val >> 8) & 0xFF, val & 0xFF   # uint16 BE
+    b[22] = b22
+    b[28:32] = tail
+    return bytes(b)
+
+
+def test_geo_peak_is_uint16_be_not_uint8():
+    """A peak above the uint8 ceiling (255 counts = 1.275 in/s) must decode.
+
+    Real example: BE18193/T193LQ9K.OE0H's final interval reads 8.270 in/s
+    in the BW export = 1654 counts = 0x0676, which needs both bytes.
+    Reading only byte[6] gave 0x76 = 118 = 0.590 in/s.
+    """
+    r = decode_histogram_body_full(_mk_block(t_peak=1654))
+    assert r is not None and len(r) == 1
+    assert r[0]["t_peak"] == 1654
+    assert geo_count_to_ins(r[0]["t_peak"]) == pytest.approx(8.27)
+
+
+def test_half_period_is_uint16_be():
+    """Half-period spans two bytes big-endian; a large value is sub-Hz."""
+    r = decode_histogram_body_full(_mk_block(t_halfp=53829))
+    assert r[0]["t_halfp"] == 53829
+    assert half_period_to_hz(53829) == pytest.approx(512 / 53829)
+
+
+def test_terminal_block_tail_is_accepted():
+    """The LAST block of a histogram stream carries tail `9c 06 00 42`
+    instead of `1e 0a 00 00`.  Rejecting it dropped the final interval —
+    which is where the event peak often lives.  Observed in 1206 of 1211
+    production histograms, always after every standard-tail block."""
+    body = _mk_block(ctr=256) + _mk_block(ctr=257, t_peak=1654,
+                                          tail=b"\x9c\x06\x00\x42")
+    ch = decode_histogram_body(body)
+    assert ch is not None
+    assert ch["Tran"] == [3, 1654], "terminal block must not be dropped"
+
+
+def test_terminal_block_exempt_from_byte22_constraint():
+    """Terminal blocks carry arbitrary bytes at [22:24]; only standard
+    blocks hold 0x00 there.  Requiring it dropped the final interval on
+    files such as BE18438/T438LO30.DC0H."""
+    body = _mk_block(ctr=256) + _mk_block(ctr=257, tail=b"\x9c\x06\x00\x42",
+                                          b22=0x01)
+    ch = decode_histogram_body(body)
+    assert ch is not None and len(ch["Tran"]) == 2
+
+
+def test_standard_block_accepts_nonzero_byte22():
+    """block[22] is NOT a constant and must not be tested.
+
+    It was documented as always 0x00, but it carries data on loud blocks.
+    Rejecting those threw away the interval holding the event peak:
+    BE18350/T350L7HR.NL0H block 92 has block[22]=0x26 and a Tran peak of
+    0x0563 = 1379 counts = 6.895 in/s — exactly the device-reported PPV —
+    while the file as a whole decoded to 0.015 in/s.
+
+    block[0]==0x00, block[4]==0x0A and the 4-byte tail are six bytes of
+    constraint, which is what keeps trailer content out.
+    """
+    ch = decode_histogram_body(_mk_block(t_peak=1379, b22=0x26))
+    assert ch is not None
+    assert ch["Tran"] == [1379]
+    assert geo_count_to_ins(ch["Tran"][0]) == pytest.approx(6.895)
+
+
+def test_marker_is_single_byte_not_uint16():
+    """block[4] alone is the 0x0A marker.  Treating [4:6] as a uint16 LE
+    marker forced block[5] to zero, which capped every geo peak at 255
+    counts — block[5] is the peak's high byte."""
+    r = decode_histogram_body_full(_mk_block(t_peak=0x0676))
+    assert r is not None, "block[5] != 0 must not disqualify the block"
+    assert r[0]["t_peak"] == 0x0676
+
+
+# ── Multi-interval block variant (2026-08-26) ───────────────────────────────
+#
+# When the histogram interval is SHORTER than one minute the device packs
+# several intervals into one block, so that every block still covers exactly
+# one minute of data:
+#
+#     interval  intervals/block  stride
+#     1 min     1                32   (the standard big-endian block)
+#     15 s      4                92
+#     2 s       30               612
+#
+#     stride = 12 + n_intervals * 20
+#
+# Block layout:
+#     [0]      0x00
+#     [1]      segment_id (256 blocks per segment)
+#     [2:4]    block_ctr  uint16 LE
+#     [4]      0x0a marker
+#     [5]      0x00
+#     [6 ...]  n x 20-byte interval records, each 8 x uint16 LITTLE-endian:
+#                  T_peak, T_halfp, V_peak, V_halfp,
+#                  L_peak, L_halfp, M_peak, M_halfp
+#              followed by 2 words (first is 0x0000)
+#     [-6:]    6-byte block trailer
+#
+# NOTE the endianness flip: the standard 32-byte block is big-endian, this
+# variant is little-endian.
+#
+# Ground truth: BE9440/K440L3AQ.T70H (15 s intervals, 5,710 of them) decodes
+# with 17,130/17,130 geo peak counts, 22,840/22,840 frequencies and
+# 5,710/5,710 mic dB(L) values matching its Blastware ASCII export exactly.
+
+from minimateplus.histogram_codec import (        # noqa: E402
+    detect_multi_interval_stride,
+    walk_multi_interval_blocks,
+)
+
+
+def _mk_multi_block(intervals, seg=0, ctr=256):
+    """Build one multi-interval block from a list of 8-tuples."""
+    b = bytearray()
+    b += bytes([0x00, seg])
+    b += int(ctr).to_bytes(2, "little")
+    b += bytes([0x0A, 0x00])
+    for iv in intervals:
+        for v in iv:
+            b += int(v).to_bytes(2, "little")
+        b += (0).to_bytes(2, "little")
+        b += (5).to_bytes(2, "little")
+    b += bytes(6)
+    assert len(b) == 12 + 20 * len(intervals)
+    return bytes(b)
+
+
+def test_stride_is_twelve_plus_twenty_per_interval():
+    for n in (4, 30):
+        ivs = [(1, 1, 2, 2, 3, 3, 4, 4)] * n
+        body = _mk_multi_block(ivs, ctr=256) + _mk_multi_block(ivs, ctr=257)
+        assert detect_multi_interval_stride(body) == 12 + 20 * n
+
+
+def test_multi_interval_block_decodes_all_four_channels():
+    ivs = [(1, 1, 2, 2, 3, 3, 4, 4), (5, 6, 7, 8, 9, 10, 11, 12)]
+    body = _mk_multi_block(ivs) + _mk_multi_block(ivs, ctr=257)
+    recs = walk_multi_interval_blocks(body)
+    assert len(recs) == 4
+    r = recs[0]
+    assert (r["t_peak"], r["v_peak"], r["l_peak"], r["m_peak"]) == (1, 2, 3, 4)
+    assert (r["t_halfp"], r["v_halfp"], r["l_halfp"], r["m_halfp"]) == (1, 2, 3, 4)
+    assert recs[1]["t_peak"] == 5 and recs[1]["m_halfp"] == 12
+
+
+def test_multi_interval_values_are_little_endian():
+    """The standard 32-byte block is big-endian; this variant is not.
+
+    A peak of 0x0100 must decode as 256, not 1.
+    """
+    ivs = [(0x0100, 1, 1, 1, 1, 1, 1, 1), (1, 1, 1, 1, 1, 1, 1, 1)]
+    body = _mk_multi_block(ivs) + _mk_multi_block(ivs, ctr=257)
+    assert walk_multi_interval_blocks(body)[0]["t_peak"] == 0x0100
+
+
+def test_decode_histogram_body_falls_back_to_the_variant():
+    ivs = [(1, 1, 2, 2, 3, 3, 4, 4)] * 4
+    body = _mk_multi_block(ivs) + _mk_multi_block(ivs, ctr=257)
+    ch = decode_histogram_body(body)
+    assert ch is not None
+    assert len(ch["Tran"]) == 8
+    assert ch["Long"][0] == 3
+
+
+def test_standard_blocks_still_take_precedence():
+    """A body of standard 32-byte blocks must not be re-read as the variant."""
+    std = _mk_block(t_peak=7) + _mk_block(t_peak=9, ctr=257,
+                                          tail=b"\x9c\x06\x00\x42")
+    ch = decode_histogram_body(std)
+    assert ch is not None and ch["Tran"] == [7, 9]
+
+
+# ── Ground truth for the multi-interval variant ─────────────────────────────
+# Fixture is gitignored (like the rest of tests/fixtures); skips when absent.
+
+_MULTI_FIXTURE = os.path.join(
+    os.path.dirname(__file__), "fixtures", "histogram-multi-interval",
+    "K440L3AQ.T70H",
+)
+
+
+@pytest.mark.skipif(not os.path.exists(_MULTI_FIXTURE),
+                    reason="multi-interval fixture not present")
+def test_multi_interval_matches_blastware_ascii_exactly():
+    """BE9440/K440L3AQ.T70H — 5,710 intervals at 15 s, 4 per 92-byte block.
+
+    Every geo peak, every frequency and every mic dB(L) in the file matches
+    the Blastware ASCII export: 17,130 / 22,840 / 5,710 values, zero
+    mismatches.  Before this decoder the file produced nothing at all.
+    """
+    import math
+    import re as _re
+    from minimateplus import blastware_file as _bwf
+
+    raw = open(_MULTI_FIXTURE, "rb").read()
+    bs = _bwf._WAVEFORM_HEADER_SIZE + 21
+    pos, fp = bs, -1
+    while True:
+        pos = raw.find(b"\x0e\x08", pos)
+        if pos < 0 or pos + 26 > len(raw):
+            break
+        if 2015 <= ((raw[pos + 4] << 8) | raw[pos + 5]) <= 2050:
+            fp = pos
+            break
+        pos += 1
+    recs = walk_multi_interval_blocks(raw[bs:fp])
+
+    rows = []
+    for line in open(_MULTI_FIXTURE + "_ASCII.TXT", errors="replace"):
+        p = [x.strip() for x in line.split("\t")]
+        if len(p) >= 11 and _re.match(r"^\d{2}:\d{2}:\d{2}$", p[0]):
+            rows.append(p)
+
+    assert len(recs) == len(rows) == 5710
+
+    def want_count(x):
+        return round(float(x) / 0.005)
+
+    for rec, row in zip(recs, rows):
+        assert rec["t_peak"] == want_count(row[1])
+        assert rec["v_peak"] == want_count(row[3])
+        assert rec["l_peak"] == want_count(row[5])
+        # mic dB(L)
+        assert abs((81.94 + 20 * math.log10(rec["m_peak"])) - float(row[9])) <= 0.06
+        # frequency: half-period <= 5 is BW's ">100 Hz" sentinel
+        for hp, cell in ((rec["t_halfp"], row[2]), (rec["v_halfp"], row[4]),
+                         (rec["l_halfp"], row[6]), (rec["m_halfp"], row[10])):
+            hz = None if hp <= 5 else 512.0 / hp
+            if cell.startswith(">"):
+                assert hz is None
+            elif not cell.startswith("<"):
+                assert hz is not None and abs(hz - float(cell)) <= max(0.55, float(cell) * 0.02)
+
+
+def test_partial_final_block_is_not_disqualified_by_missing_third_header():
+    """A body can exceed two strides yet hold only two real blocks.
+
+    Regression for BE18193 `T193L0XM.CI0H` — 51 intervals at 2 s = one full
+    30-interval block plus a 21-interval remainder, in a body long enough to
+    demand a third block header at ``2 * stride`` that does not exist.  The
+    third-block confirmation used to be mandatory whenever the body was long
+    enough, so the correct stride was discarded and the file decoded to
+    nothing.  A missing third header means end-of-stream, not disqualification;
+    the block-counter check is the decisive anti-false-positive test.
+    """
+    full = [(1, 1, 2, 2, 3, 3, 4, 4)] * 30
+    partial = [(5, 5, 6, 6, 7, 7, 8, 8)] * 21
+    body = (_mk_multi_block(full, ctr=256)
+            + _mk_multi_block(partial, ctr=257)
+            + b"\xff" * 700)      # trailing padding past 2 * stride
+    stride = 12 + 20 * 30
+    assert 2 * stride + 6 <= len(body), "padding must reach past two strides"
+    # the whole point: a third header is absent, and that must not disqualify
+    assert detect_multi_interval_stride(body) == stride
+    recs = walk_multi_interval_blocks(body)
+    assert len(recs) == 51
+    assert recs[0]["t_peak"] == 1
+    assert recs[-1]["t_peak"] == 5
+
+
+def test_third_block_still_rejects_a_mismatched_counter():
+    """The corroboration must still bite when a third block IS present."""
+    ivs = [(1, 1, 2, 2, 3, 3, 4, 4)] * 4
+    body = (_mk_multi_block(ivs, ctr=256)
+            + _mk_multi_block(ivs, ctr=257)
+            + _mk_multi_block(ivs, ctr=999))   # counter jumps — not consecutive
+    assert detect_multi_interval_stride(body) != 12 + 20 * 4

tests/test_reviewed_real_column.py

@@ -0,0 +1,58 @@
+import sqlite3
+from sfm.database import SeismoDb
+def _cols(db):
+    with sqlite3.connect(db.db_path) as c:
+        return {r[1] for r in c.execute("PRAGMA table_info(events)")}
+def test_fresh_db_has_reviewed_real(tmp_path):
+    assert "reviewed_real" in _cols(SeismoDb(tmp_path/"s.db"))
+def test_existing_db_migrates_reviewed_real(tmp_path):
+    p = tmp_path/"s.db"; db = SeismoDb(p)
+    with sqlite3.connect(p) as c:
+        c.execute("ALTER TABLE events DROP COLUMN reviewed_real")
+    assert "reviewed_real" not in _cols(db)      # dropped (read via existing handle/connection)
+    SeismoDb(p)                                   # re-open migrates
+    assert "reviewed_real" in _cols(SeismoDb(p))
+
+def test_legacy_pre_migration1_db_rebuild_does_not_crash(tmp_path):
+    # Genuinely legacy (pre-Migration-1) events table: no UNIQUE(serial, timestamp)
+    # (still on the old UNIQUE(serial, waveform_key)) and columns only through
+    # false_trigger/created_at — i.e. none of the columns added later by the
+    # Migration-1b ADD COLUMN loop (blastware_filename, device_family,
+    # tran_zc_freq, shape_*, reviewed_real, ...). Opening this DB triggers the
+    # Migration-1 rebuild (`events_old` -> `events` via positional
+    # `INSERT OR IGNORE INTO events SELECT * FROM events_old`), which requires
+    # the rebuild's CREATE TABLE to have exactly the legacy column count.
+    p = tmp_path / "legacy.db"
+    with sqlite3.connect(p) as c:
+        c.execute("""
+            CREATE TABLE events (
+                id                      TEXT PRIMARY KEY,
+                serial                  TEXT NOT NULL,
+                waveform_key            TEXT NOT NULL,
+                session_id              TEXT,
+                timestamp               TEXT,
+                tran_ppv                REAL,
+                vert_ppv                REAL,
+                long_ppv                REAL,
+                peak_vector_sum         REAL,
+                mic_ppv                 REAL,
+                project                 TEXT,
+                client                  TEXT,
+                operator                TEXT,
+                sensor_location         TEXT,
+                sample_rate             INTEGER,
+                record_type             TEXT,
+                false_trigger           INTEGER NOT NULL DEFAULT 0,
+                created_at              TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%SZ', 'now')),
+                UNIQUE(serial, waveform_key)
+            )
+        """)
+        c.execute(
+            "INSERT INTO events (id, serial, waveform_key, timestamp) VALUES (?, ?, ?, ?)",
+            ("evt-1", "BE11529", "01110000", "2026-01-01T00:00:00"),
+        )
+    # Pre-fix this raised: OperationalError: table events has 19 columns but
+    # 18 values were supplied (reviewed_real had leaked into the rebuild's
+    # CREATE TABLE, but events_old — and the positional SELECT * — only had 18).
+    db = SeismoDb(p)
+    assert "reviewed_real" in _cols(db)

tests/test_reviewed_real_in_feed.py

@@ -0,0 +1,15 @@
+from __future__ import annotations
+import os, sys
+from pathlib import Path
+
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+from sfm.database import SeismoDb
+from minimateplus.models import Event
+
+
+def test_query_events_includes_reviewed_real(tmp_path: Path):
+    db = SeismoDb(tmp_path / "s.db")
+    ev = Event(index=0)
+    ev._waveform_key = bytes.fromhex("01110000")
+    db.insert_events([ev], serial="BE1")
+    assert "reviewed_real" in db.query_events(serial="BE1")[0]

tests/test_set_false_trigger_exclusivity.py

@@ -0,0 +1,29 @@
+from sfm.database import SeismoDb
+from minimateplus.models import Event
+
+
+def _ins(db, eid_key="01110000", serial="BE1"):
+    ev = Event(index=0); ev._waveform_key = bytes.fromhex(eid_key)
+    db.insert_events([ev], serial=serial)
+    return db.query_events(serial=serial)[0]["id"]
+
+
+def test_set_false_trigger_clears_reviewed_real(tmp_path):
+    db = SeismoDb(tmp_path / "s.db"); eid = _ins(db)
+    db.update_event_review(eid, {"reviewed_real": True})
+    assert db.get_event(eid)["reviewed_real"] == 1
+
+    db.set_false_trigger(eid, True)
+    row = db.get_event(eid)
+    assert row["false_trigger"] == 1 and row["reviewed_real"] == 0
+
+
+def test_clearing_false_trigger_leaves_reviewed_real_untouched(tmp_path):
+    db = SeismoDb(tmp_path / "s.db"); eid = _ins(db)
+    db.update_event_review(eid, {"false_trigger": True})
+    db.set_false_trigger(eid, False)
+    row = db.get_event(eid)
+    assert row["false_trigger"] == 0
+    # reviewed_real was never set true, so it stays 0 either way; the point
+    # is set_false_trigger(False) does not touch the column at all.
+    assert row["reviewed_real"] == 0

tests/test_twin_propagation.py

@@ -0,0 +1,36 @@
+import sqlite3
+from sfm.database import SeismoDb
+from minimateplus.models import Event, Timestamp
+
+
+def _ins(db, key, serial, pvs, ts, record_type="Waveform"):
+    ev = Event(index=0)
+    ev._waveform_key = bytes.fromhex(key)
+    ev.timestamp = ts
+    db.insert_events([ev], serial=serial)
+    row = [r for r in db.query_events(serial=serial) if r["waveform_key"] == key][0]
+    with sqlite3.connect(db.db_path) as c:
+        c.execute("UPDATE events SET peak_vector_sum=?, record_type=? WHERE id=?",
+                  (pvs, record_type, row["id"]))
+    return row["id"]
+
+
+def _ts(hour, minute, second=0, day=25):
+    return Timestamp(raw=b"", flag=0x10, year=2026, unknown_byte=0, month=2, day=day,
+                     hour=hour, minute=minute, second=second)
+
+
+def test_propagate_copies_flags_across_hours_apart_twins(tmp_path):
+    # Flagging the waveform FT propagates to its histogram twin 75 min earlier
+    # (the interval matcher pairs them; the old ±5-min window would have missed it).
+    db = SeismoDb(tmp_path / "s.db")
+    hist  = _ins(db, "01110001", "BE1", 0.4763, _ts(19, 31, 17), "Histogram")
+    wave  = _ins(db, "01110002", "BE1", 0.4763, _ts(20, 46, 44), "Waveform")   # twin, 75 min later
+    other = _ins(db, "01110003", "BE1", 0.9999, _ts(20, 20, 0),  "Waveform")   # different pvs
+
+    db.update_event_review(wave, {"false_trigger": True})
+    moved = db.propagate_review_to_twins(wave)
+
+    assert hist in moved
+    assert db.get_event(hist)["false_trigger"] == 1
+    assert db.get_event(other)["false_trigger"] == 0

tests/test_update_event_review_reviewed_real.py

@@ -0,0 +1,25 @@
+from sfm.database import SeismoDb
+from minimateplus.models import Event
+
+
+def _ins(db, eid_key="01110000", serial="BE1"):
+    ev = Event(index=0); ev._waveform_key = bytes.fromhex(eid_key)
+    db.insert_events([ev], serial=serial)
+    return db.query_events(serial=serial)[0]["id"]
+
+
+def test_confirm_real_sets_and_clears_ft(tmp_path):
+    db = SeismoDb(tmp_path/"s.db"); eid = _ins(db)
+    db.update_event_review(eid, {"false_trigger": True})
+    assert db.get_event(eid)["false_trigger"] == 1
+    db.update_event_review(eid, {"reviewed_real": True})   # confirming real clears FT
+    row = db.get_event(eid)
+    assert row["reviewed_real"] == 1 and row["false_trigger"] == 0
+
+
+def test_flag_ft_clears_reviewed_real(tmp_path):
+    db = SeismoDb(tmp_path/"s.db"); eid = _ins(db)
+    db.update_event_review(eid, {"reviewed_real": True})
+    db.update_event_review(eid, {"false_trigger": True})
+    row = db.get_event(eid)
+    assert row["false_trigger"] == 1 and row["reviewed_real"] == 0

tests/test_waveform_codec.py

@@ -14,6 +14,7 @@ import pytest
 
 from minimateplus.waveform_codec import (
     WaveformBlock,
+    decode_waveform_legacy,
     decode_tran_initial,
     decode_waveform_v2,
     decoded_to_adc_counts,
@@ -210,9 +211,11 @@ def test_parse_segment_header_decodes_fields():
     )
     decoded = parse_segment_header(block)
     assert decoded is not None
+    assert decoded["n_prev_deltas"] == 2
+    assert decoded["prev_deltas"] == [0, 0]
     assert decoded["counter"] == 0x47       # uint32 LE
-    assert decoded["fixed_pattern"] == b"\x02\x00\x00\x01"
-    assert decoded["anchor_bytes"] == b"\x00\x00\x00\x00"
+    assert decoded["marker"] == b"\x02\x00"
+    assert decoded["anchors"] == [1, 1]
 
 
 def test_segment_counter_increments():
@@ -516,3 +519,279 @@ def test_decode_a5_frames_empty():
     from minimateplus.waveform_codec import decode_a5_frames
     assert decode_a5_frames([]) is None
     assert decode_a5_frames(None) is None
+
+
+# ── Wide-NN RLE, wide 30 NN, and variable-width segment headers ──────────────
+#
+# Three framing cases discovered 2026-08-25 by diffing 75 production events
+# against their preserved Blastware ASCII exports.  Each caused ``walk_body``
+# to hit its ``else: break`` mid-stream, truncating every channel decoded
+# after that point (see CHANGELOG v0.25.1).
+
+_PREAMBLE = b"\x00\x02\x00\x00\x00\x00\x00"   # magic + Tran[0]=0, Tran[1]=0
+_STOP = b"\xff\xff"                            # unrecognised tag → walker stops
+
+
+def _synth(*chunks: bytes) -> bytes:
+    return _PREAMBLE + b"".join(chunks) + _STOP
+
+
+def test_walk_body_wide_rle_block():
+    """``0X NN`` is a 12-bit-NN RLE run (NN = ((t0 & 0x0F) << 8) | t1).
+
+    Observed as ``01 0c`` (NN=268) in BE9558/K558LKOF.460W and five other
+    production events.  A narrow ``00 NN`` maxes out at NN=0xFC, so runs
+    longer than 252 samples must use the wide form.
+    """
+    blocks = walk_body(_synth(b"\x01\x0c"))
+    assert len(blocks) == 1
+    assert (blocks[0].tag_hi, blocks[0].tag_lo) == (0x01, 0x0C)
+    assert blocks[0].length == 2
+
+
+# NOTE (2026-08-25): the four tests below assert the SUPERSEDED tag-dispatch
+# model — `40 NN` as a variable-width segment header, tagless headers, channel
+# from rotation.  The body format is really a chain of self-delimiting records
+# (see the record-chain section of waveform_codec.py), so `decode_waveform_v2`
+# no longer behaves this way.  They are retargeted at `decode_waveform_legacy`,
+# which still implements the old model and is pinned by micromate/idf_file.py
+# for Thor IDFW bodies.
+
+
+def test_decode_wide_rle_repeats_full_run():
+    """A wide RLE run repeats the running value NN times, not NN & 0xFF."""
+    decoded = decode_waveform_legacy(_synth(b"\x01\x0c"))
+    # 2 preamble anchors + 268 repeats
+    assert len(decoded["Tran"]) == 2 + 268
+    assert set(decoded["Tran"]) == {0}
+
+
+def test_walk_body_30_block_nn_above_16():
+    """``30 NN`` data blocks are not capped at NN=0x10.
+
+    ``30 18`` (NN=24) appears in BE18193/T193LQ45.NN0W; the old
+    ``0 < t1 <= 0x10`` guard rejected it and stopped the walk 1033 bytes
+    into a 4877-byte body.  Length is still NN * 1.5 + 2.
+    """
+    payload = bytes(36)                       # 24 deltas × 1.5 bytes
+    blocks = walk_body(_synth(b"\x30\x18" + payload, b"\x00\x04"))
+    assert [b.length for b in blocks] == [38, 2]
+
+
+@pytest.mark.parametrize("nn,hdr_len", [(1, 18), (2, 20), (3, 22)])
+def test_walk_body_segment_header_width_follows_tag_lo(nn, hdr_len):
+    """``40 NN``: NN is the count of previous-channel continuation deltas.
+
+    Header length = 2 * NN + 16.  ``40 02`` (the only form previously
+    handled) is the NN=2 case at 20 bytes; ``40 01`` (18) and ``40 03``
+    (22) both occur in production files.
+    """
+    data = bytearray(hdr_len - 2)
+    data[2 * nn + 8 : 2 * nn + 10] = b"\x02\x00"   # constant marker
+    blocks = walk_body(_synth(bytes([0x40, nn]) + bytes(data), b"\x00\x04"))
+    assert [b.length for b in blocks] == [hdr_len, 2]
+
+
+@pytest.mark.parametrize("nn,hdr_len", [(1, 18), (2, 20), (3, 22)])
+def test_segment_header_anchors_track_header_width(nn, hdr_len):
+    """Anchor pair sits at data[2*NN+10 : 2*NN+14] regardless of width."""
+    data = bytearray(hdr_len - 2)
+    data[2 * nn + 8 : 2 * nn + 10] = b"\x02\x00"
+    data[2 * nn + 10 : 2 * nn + 12] = (7).to_bytes(2, "big")     # anchor 0
+    data[2 * nn + 12 : 2 * nn + 14] = (9).to_bytes(2, "big")     # anchor 1
+    decoded = decode_waveform_legacy(_synth(bytes([0x40, nn]) + bytes(data)))
+    assert decoded["Vert"][:2] == [7, 9]
+
+
+# ── Tagless segment headers ─────────────────────────────────────────────────
+#
+# A segment header can appear WITHOUT its ``40 NN`` tag: just the 14-byte tail
+# ``[field2:2][len:2][channel_id:4][marker:2][anchors:4]``.  This is the NN=0
+# case — no continuation deltas for the previous channel, so no tag and no
+# delta bytes.  Found 2026-08-25: it is where the walk stopped in 7 of the 8
+# remaining truncating production events.
+#
+# The channel_id field (previously mis-labelled a "monotonic counter") is
+# ``[channel][00][00][segment_index]`` with 0x46=Tran 0x47=Vert 0x48=Long
+# 0x49=MicL — verified on 1697 of 1697 segment headers across the ground-truth
+# corpus, zero disagreements.
+
+def _tagless(chan_id=0x47, seg=2, marker=b"\x02\x00", a0=0, a1=0):
+    return (b"\x5d\xee" + b"\x00\xd0" + bytes([chan_id, 0, 0, seg]) + marker
+            + a0.to_bytes(2, "big", signed=True) + a1.to_bytes(2, "big", signed=True))
+
+
+def test_walk_body_accepts_tagless_segment_header():
+    """A bare 14-byte header is walked as a segment block, not a stop."""
+    blocks = walk_body(_synth(b"\x10\x04\x00\x00", _tagless(), b"\x00\x04"))
+    kinds = [(b.tag_hi, b.tag_lo, b.length) for b in blocks]
+    assert kinds == [(0x10, 0x04, 4), (0x40, 0x00, 14), (0x00, 0x04, 2)]
+
+
+def test_tagless_header_carries_full_14_bytes_as_data():
+    """The synthetic block's data includes the leading bytes (there is no tag
+    to strip), so decode_waveform_v2's ``2*nd + k`` offsets line up at nd=0."""
+    blocks = walk_body(_synth(_tagless()))
+    hdr = next(b for b in blocks if b.tag_hi == 0x40)
+    assert len(hdr.data) == 14
+    assert hdr.data[8:10] == b"\x02\x00"          # marker at 2*0 + 8
+
+
+def test_tagless_header_anchors_and_channel_id():
+    """Anchors decode from data[10:14]; the channel comes from the id byte."""
+    decoded = decode_waveform_legacy(_synth(_tagless(chan_id=0x48, a0=11, a1=13)))
+    assert decoded["Long"][:2] == [11, 13]        # 0x48 → Long, not rotation
+    assert decoded["Vert"] == []
+
+
+@pytest.mark.parametrize("chan_id,name",
+                         [(0x46, "Tran"), (0x47, "Vert"), (0x48, "Long"), (0x49, "MicL")])
+def test_segment_channel_comes_from_id_not_rotation(chan_id, name):
+    """Channel is taken from the header's id byte.  Two headers in a row for
+    the SAME channel must both land on that channel — rotation-by-position
+    would put the second one on the next channel and corrupt both."""
+    body = _synth(_tagless(chan_id=chan_id, seg=1, a0=5, a1=6),
+                  _tagless(chan_id=chan_id, seg=2, a0=7, a1=8))
+    decoded = decode_waveform_legacy(body)
+    # Tran additionally carries the body preamble's 2 anchors (both 0 here).
+    expected = [0, 0, 5, 6, 7, 8] if name == "Tran" else [5, 6, 7, 8]
+    assert decoded[name] == expected
+    for other in ("Tran", "Vert", "Long", "MicL"):
+        if other != name:
+            assert decoded[other] == ([0, 0] if other == "Tran" else [])
+
+
+# ── Record-chain body model (2026-08-25) ────────────────────────────────────
+#
+# The body is a chain of self-delimiting per-channel records, not a flat
+# tag-dispatch stream.  Verified over 1,388 production series-3 waveform
+# binaries: the chain terminates on a 0x06 record in 1,387 of them and all
+# four channels come out at identical length in 1,388/1,388 (was 156/1,388).
+# Against the 75 events with a preserved Blastware ASCII export: sample-count
+# exact 72/75 -> 75/75, fully exact 70/75 -> 73/75.
+
+from minimateplus.waveform_codec import (          # noqa: E402
+    CHANNEL_IDS,
+    MODE_ABSOLUTE,
+    MODE_DELTA,
+    MODE_RAW12,
+    STREAM_END_ID,
+    data_block_len,
+    find_first_record,
+    is_record,
+    unpack12,
+    walk_records,
+)
+
+
+def _rec(chan_id, mode, payload, seg=0, field2=b"\x00\x00", anchors=None):
+    """Build one self-delimiting record."""
+    head = bytearray()
+    head += bytes([chan_id, 0x00, 0x00, seg])
+    head += bytes(mode)
+    if anchors is not None:
+        for a in anchors:
+            head += int(a).to_bytes(2, "big", signed=True)
+    body = bytes(head) + payload
+    return field2 + (len(body) + 2).to_bytes(2, "big") + body
+
+
+def _terminator():
+    return b"\x00\x00" + (8).to_bytes(2, "big") + bytes([STREAM_END_ID, 0, 0, 0, 0, 0])
+
+
+def _body(*records, preamble=b"\x00\x02\x00", seg0=b"\x00\x00\x00\x00"):
+    return preamble + seg0 + b"".join(records) + _terminator()
+
+
+def test_forty_nn_is_a_data_block_not_a_segment_header():
+    """`40 NN` is an int16 BE data block of length 2*NN + 2.
+
+    The superseded model read it as a segment header of length 2*NN + 16,
+    which is what made walks drift and channels come out unequal.
+    """
+    assert data_block_len(b"\x40\x02\x00\x01\x00\x02", 0) == (6, 2)
+    assert data_block_len(b"\x40\x08" + bytes(16), 0) == (18, 8)
+    # NN > 8 is not a data block
+    assert data_block_len(b"\x40\x0c" + bytes(24), 0) == (None, None)
+
+
+def test_record_chain_is_followed_by_length_not_by_tag_sniffing():
+    payload = b"\x00\x04"                      # RLE hold x4
+    body = _body(_rec(0x47, MODE_DELTA, payload, anchors=(3, 5)))
+    recs = walk_records(body)
+    assert len(recs) == 1
+    assert recs[0]["channel"] == "Vert"
+    assert recs[0]["mode"] == MODE_DELTA
+
+
+def test_chain_terminates_on_channel_id_06():
+    body = _body(_rec(0x47, MODE_DELTA, b"\x00\x04", anchors=(1, 1)),
+                 _rec(0x48, MODE_DELTA, b"\x00\x04", anchors=(2, 2)))
+    assert [r["channel"] for r in walk_records(body)] == ["Vert", "Long"]
+    assert not is_record(body, len(body) - 10)   # the terminator is not a record
+
+
+def test_mode_delta_emits_anchors_then_accumulates():
+    # two anchors, then an int8 block of +1,+1,+1,+1
+    body = _body(_rec(0x47, MODE_DELTA, b"\x20\x04\x01\x01\x01\x01",
+                      anchors=(10, 11)))
+    d = decode_waveform_v2(body)
+    assert d["Vert"] == [10, 11, 12, 13, 14, 15]
+
+
+def test_mode_absolute_replaces_rather_than_accumulates():
+    """mode `01 00`: no anchors, and block values are ABSOLUTE samples."""
+    body = _body(_rec(0x48, MODE_ABSOLUTE, b"\x20\x04\x0a\x0b\x0c\x0d"))
+    d = decode_waveform_v2(body)
+    assert d["Long"] == [10, 11, 12, 13], "01 00 blocks are absolute, not deltas"
+
+
+def test_mode_absolute_rle_holds_the_previous_value():
+    body = _body(_rec(0x48, MODE_ABSOLUTE, b"\x20\x04\x07\x07\x07\x07\x00\x04"))
+    d = decode_waveform_v2(body)
+    assert d["Long"] == [7, 7, 7, 7, 7, 7, 7, 7]
+
+
+def test_mode_raw12_has_no_tags_at_all():
+    """mode `00 03`: the whole data section is raw 12-bit absolute samples.
+
+    Decoding these matters for the time base — skipping the record would
+    displace every later sample on that channel (observed on
+    BE9558/K558LOF2.820W, MicL shifted by exactly 512).
+    """
+    packed = bytes([0x01, 0x23, 0x04, 0x05, 0x06, 0x07])   # 4 samples
+    body = _body(_rec(0x49, MODE_RAW12, packed))
+    d = decode_waveform_v2(body)
+    assert d["MicL"] == unpack12(packed)
+    assert len(d["MicL"]) == 4
+
+
+def test_unpack12_sign_extends():
+    assert unpack12(bytes([0x00, 0x00, 0x01, 0x02, 0x03, 0x04])) == [1, 2, 3, 4]
+    # high nibble 0x8 -> negative
+    assert unpack12(bytes([0x80, 0x00, 0x00, 0x00, 0x00, 0x00]))[0] == -2048
+
+
+def test_channel_comes_from_the_record_id():
+    for cid, name in CHANNEL_IDS.items():
+        body = _body(_rec(cid, MODE_ABSOLUTE, b"\x20\x04\x01\x02\x03\x04"))
+        d = decode_waveform_v2(body)
+        assert d[name][-4:] == [1, 2, 3, 4], f"{name} misrouted"
+
+
+def test_raw12_preamble_is_scanned_not_block_walked():
+    """A `00 00 03` preamble carries raw 12-bit data from body[3] with no tags,
+    so find_first_record must scan rather than block-walk.  One production file
+    has this (BE13121/O121L4L1.KF0W); block-walking returns None on it."""
+    packed = bytes([0x00, 0x00, 0x01, 0x02, 0x03, 0x04])
+    body = _body(_rec(0x47, MODE_DELTA, b"\x00\x04", anchors=(1, 1)),
+                 preamble=b"\x00\x00\x03", seg0=packed)
+    assert find_first_record(body) == 3 + len(packed)
+    d = decode_waveform_v2(body)
+    assert d["Tran"] == unpack12(packed)
+
+
+def test_returns_none_when_no_record_chain():
+    assert decode_waveform_v2(b"\x00\x02\x00" + bytes(40)) is None
+    assert decode_waveform_v2(b"") is None