re: the voice decoder discards up to 87% of a bank — "multi-subwave" refuted

The record table gives a DIRECT binding hokyu_DS_s13A -> VOICE_D_452, where
the corpus records the movie as unbound and movie_manifest_disc.rs asserts
None, citing an in-game verdict that this exact value was "the wrong
recording". That is the only place on the disc where a runtime observation
disagrees with the record table, so it was worth settling.

First, shape: these banks are SHARED. Five slots bind VOICE_D_452, five bind
451, four 450, four 453, three 454 -- 21 hokyu slots over five banks, and the
movies repeat too. Generic resupply cutscenes, not per-stage recordings.

The recorded explanation for 453 decoding to 0.14 s and 454 to 0.43 s was
that the banks are "likely multi-subwave / not cleanly sliced". Refuted: the
count of RIFF magics EQUALS the number of sub-waves recovered in all five
banks, and the last data chunk ends exactly at EOF in four of them. Nothing
between or after sub-waves is being missed.

The real defect: slb::to_xma_riffs finds audio by searching for the RIFF
magic, and a large region PRECEDES it. 87% of VOICE_D_453 and 85% of
VOICE_D_454 sit in front of the first RIFF -- 21-27% zero over 256 distinct
byte values, i.e. content, not padding. VOICE_D_451 is the control: its
leading region is 100% zero, 1 distinct value, real padding.

So the in-game verdict listened to a decode that had discarded most of the
bank, for exactly this bank class. It is evidence about the decoder, not
about the mapping. Note also that what was rejected was a value INFERRED
from a shared demo id; the record table supplies the same value as a stored
field, and only the inference was ever tested.

This does NOT establish the binding is right -- it removes the only recorded
evidence against it. What the leading region actually holds is undecoded, and
confirming the binding needs a human listening.

Artifact: examples/voice_bank_shape.rs.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PMRJjbxLqZtsb5Vb7KunPE

docs/re/structures/movie-subtitles.md

@@ -98,9 +98,13 @@ the **correct join key is not yet known**:
   was tried and is **WRONG** — it plays the wrong recording in-game. Do not use.
 - Only `VOICE_D_450..454` exist (no 44x/45x neighbours). Decoded durations are
   suspicious — `450`=2.8s, `451`=1.6s, `452`=2.2s, but `453`=**0.14s**,
-  `454`=**0.43s** — far too short for the spoken line, so these `.slb` banks are
-  likely **multi-subwave / not cleanly sliced** by the current extractor (same
-  class as the deferred B/C banks).
+  `454`=**0.43s** — far too short for the spoken line. ❌ The guess that follows
+  was **REFUTED 2026-08-25**: these banks are *not* multi-subwave-and-missed. The
+  RIFF-magic count equals the number of sub-waves recovered in all five banks, so
+  nothing between or after them is lost. The audio is missing because a large
+  region **precedes the first RIFF** and the decoder searches for that magic —
+  87 % of `453` and 85 % of `454` sit in front of it, high-entropy and not
+  padding. See [voice-bank-leading-region](../voice-bank-leading-region.md).
 
 ⇒ The unbound-hokyu voice mapping is **open** (needs either the real join key
 from mission data, or a proper multi-subwave `.slb` decode + audio verification).