re: the .slb leading region is 1392+n*2048 — and my fix for it is withdrawn

The structure is exact. In all five resupply banks the first RIFF sits at
HEADERLESS_DATA_OFFSET + n*2048, where 1392 is a constant this crate already
had and 2048 is the XMA1 packet size: n = 8, 1, 7, 22, 29. No free parameter
to tune, and the raw bytes agree -- high entropy from offset 0, then a zero
run immediately before the RIFF. VOICE_D_451 is the control, its single
packet being all zeros.

So I made the obvious fix, emitting that region as a sub-wave, and then
withdrew it on two measurements:

* It does not recover audio. Coverage went 5.4% -> 89.9% for VOICE_D_453, but
  the emitted stream decodes through FFmpeg to 1792 PCM bytes -- silence --
  while the RIFF sub-waves from the same banks decode to 150-270 KB. Byte
  coverage was the wrong success metric and it looked like progress.
* It is not narrow. The rule matches 1524 of the 8021 RIFF-bearing entries in
  sound.pak, including RT* movie banks that decode correctly today. Landing
  it would have risked a wide regression in order to not-fix five banks.

to_xma_riffs is back to its previous behaviour, verified by re-measuring:
coverage is 5.4% / 9.7% again. The refuted attempt is recorded in the code
beside the branch it would have changed, so the next person does not
re-derive the arithmetic and re-make the change.

XMA1_PACKET is kept as a named constant because the blast-radius scan uses
it. Artifacts: examples/voice_bank_shape.rs (structure), voice_bank_dump.rs
(sub-waves for decoding), slb_hybrid_scan.rs (the 1524 count).

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

crates/sylpheed-formats/examples/slb_hybrid_scan.rs

@@ -0,0 +1,31 @@
+//! How many `.slb` banks would a "leading headerless stream" rule affect?
+//!
+//! The rule fires when the first `RIFF` sits at exactly
+//! `HEADERLESS_DATA_OFFSET + n*XMA1_PACKET` with a non-zero leading region.
+//! Before trusting it, count how many banks it would change — including the
+//! `RT*` movie banks that already decode correctly today.
+use sylpheed_formats::{slb, PakArchive};
+
+fn main() {
+    let disc = std::env::var("SYLPHEED_DISC").expect("set SYLPHEED_DISC");
+    let snd = PakArchive::open(format!("{disc}/dat/sound.pak")).expect("sound.pak");
+    let (mut total, mut has_riff, mut hybrid, mut hybrid_nonzero) = (0, 0, 0, 0);
+    for e in snd.entries() {
+        let Ok(b) = snd.read(e) else { continue };
+        total += 1;
+        let Some(ri) = b.windows(4).position(|w| w == b"RIFF") else { continue };
+        has_riff += 1;
+        if ri > slb::HEADERLESS_DATA_OFFSET
+            && (ri - slb::HEADERLESS_DATA_OFFSET) % slb::XMA1_PACKET == 0
+        {
+            hybrid += 1;
+            if b[slb::HEADERLESS_DATA_OFFSET..ri].iter().any(|x| *x != 0) {
+                hybrid_nonzero += 1;
+            }
+        }
+    }
+    println!(
+        "sound.pak entries {total}; with a RIFF {has_riff}; \
+         first RIFF at 1392+n*2048 {hybrid}; of those with a NON-ZERO leading region {hybrid_nonzero}"
+    );
+}

crates/sylpheed-formats/examples/voice_bank_dump.rs

@@ -0,0 +1,19 @@
+//! Dump each resupply voice bank's sub-waves to `.xma` RIFFs for decoding.
+use sylpheed_formats::{slb, PakArchive};
+
+fn main() {
+    let disc = std::env::var("SYLPHEED_DISC").expect("set SYLPHEED_DISC");
+    let out = std::env::var("OUT_DIR_X").unwrap_or_else(|_| "/tmp/voicebanks".into());
+    std::fs::create_dir_all(&out).unwrap();
+    let snd = PakArchive::open(format!("{disc}/dat/sound.pak")).expect("sound.pak");
+    for n in 450..=454 {
+        let path = format!("eng\\etc\\VOICE_D_{n}.slb");
+        let Some(entry) = snd.find_by_name(&path) else { continue };
+        let bytes = snd.read(entry).expect("read");
+        for (i, r) in slb::to_xma_riffs(&bytes).iter().enumerate() {
+            let f = format!("{out}/VOICE_D_{n}_{i}.xma");
+            std::fs::write(&f, r).unwrap();
+            println!("{f} {} bytes", r.len());
+        }
+    }
+}

crates/sylpheed-formats/examples/voice_bank_shape.rs

@@ -55,6 +55,15 @@ fn main() {
             };
             let tail = &bytes[last..];
             let zeros = tail.iter().filter(|b| **b == 0).count();
+            // Dump the first bytes of the leading region so its structure is
+            // visible rather than guessed at.
+            let hex: String = head
+                .iter()
+                .take(48)
+                .map(|b| format!("{b:02x}"))
+                .collect::<Vec<_>>()
+                .join(" ");
+            println!("  head[0..48] {hex}");
             println!(
                 "{:<14} {:>7} {:>6} {:>6} {:>6.1}%  1st RIFF @{:>6}  data chunks {}  head {:>5.1}% zero/{:>3} distinct  tail {:>5} B ({:>5.1}% zero)  {:?}",
                 format!("VOICE_D_{n}"),

crates/sylpheed-formats/src/slb.rs

@@ -23,6 +23,10 @@
 /// Fixed offset of the raw XMA1 stream in a headerless `.slb` (no `RIFF`).
 pub const HEADERLESS_DATA_OFFSET: usize = 1392;
 
+/// XMA1 packet size. A headerless stream is always a whole number of these, which
+/// is how a leading stream is told apart from arbitrary bytes before a `RIFF`.
+pub const XMA1_PACKET: usize = 2048;
+
 /// Voice language for cutscene audio. Only English and Japanese voice exist on
 /// the disc (subtitles cover more languages, voice does not).
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]
@@ -138,6 +142,22 @@ pub fn to_xma_riffs(slb: &[u8]) -> Vec<Vec<u8>> {
         }
         return out;
     }
+    // ❌ A "leading headerless stream" rule was tried here and WITHDRAWN.
+    //
+    // The structure is real: in all five resupply banks the first `RIFF` sits at
+    // exactly `HEADERLESS_DATA_OFFSET + n*XMA1_PACKET` (n = 8, 1, 7, 22, 29), and
+    // 87 % of `VOICE_D_453` lies in front of it. Emitting that region as a
+    // sub-wave raised byte coverage from 5.4 % to 89.9 %.
+    //
+    // But byte coverage was the wrong success metric. The emitted streams decode
+    // to **1792 PCM bytes** — silence — through the same FFmpeg path that decodes
+    // the RIFF sub-waves fine, so the region is not XMA1 under the synthesised
+    // format. And the rule is not narrow: it matches **1524 of the 8021**
+    // RIFF-bearing entries in `sound.pak`, including `RT*` banks that decode
+    // correctly today. Landing it would have risked a large regression to fix
+    // five banks it does not actually fix.
+    //
+    // See docs/re/voice-bank-leading-region.md.
     let mut pos = 0usize;
     while let Some(ri) = find(slb, b"RIFF", pos) {
         // Parse this sub-wave's fmt + data (declared size is honest per sub-wave).

docs/re/BACKLOG.md

@@ -1047,9 +1047,19 @@ premise was wrong.**
   its leading region being 100 % zero / 1 distinct value. 🟡 So the in-game
   verdict tested a decode that had thrown away most of the bank and is **not**
   evidence against the binding — though it does not confirm it either.
-  ▶️ First step: decode the leading region (it is not padding and not a RIFF —
-  directory? seek table? raw stream?). Then a human has to listen; audio
-  judgement cannot be done in this container. See
+  ✅ **(same day) The region's SIZE is now exact**: the first `RIFF` sits at
+  `1392 + n*2048` in all five banks (n = 8, 1, 7, 22, 29) — 1392 being the
+  crate's own `HEADERLESS_DATA_OFFSET` and 2048 the XMA1 packet size. No free
+  parameter.
+  ❌ **But my fix for it is WITHDRAWN.** Emitting that region as a sub-wave took
+  `VOICE_D_453` from 5.4 % to 89.9 % byte coverage — and the stream decodes to
+  **1792 PCM bytes**, silence, while the RIFF sub-waves decode to 150–270 KB.
+  Byte coverage was the wrong success metric. The rule also matches **1524 of
+  8021** RIFF-bearing `sound.pak` entries, including `RT*` banks that work today,
+  so it risked a wide regression to not-fix five banks.
+  ▶️ Next cheap probe: vary the synthesised `fmt` (channels / streams / sample
+  rate) instead of assuming the container. Then a human has to listen; audio
+  judgement cannot be done here. See
   [`voice-bank-leading-region.md`](voice-bank-leading-region.md).
 * ❌ **(2026-08-25) My own boot-nav diagnosis, MEASURED AND WITHDRAWN.** I said
   the run died because `skip_intro.sh` gates the title test at `rmse <= 1500`

docs/re/voice-bank-leading-region.md

@@ -84,11 +84,49 @@ tested.
 ⚠️ This does **not** establish that the binding is right. It removes the only
 recorded evidence against it.
 
+## ✅ The region's structure — and ❌ my own fix for it, withdrawn
+
+The leading region is not shapeless. In **all five** banks the first `RIFF` sits
+at exactly `HEADERLESS_DATA_OFFSET + n * 2048` — where `HEADERLESS_DATA_OFFSET`
+(1392) is a constant this crate already had, and 2048 is the XMA1 packet size:
+
+| bank | first RIFF | − 1392 | ÷ 2048 |
+|---|---|---|---|
+| `VOICE_D_450` | 17 776 | 16 384 | **8** |
+| `VOICE_D_451` | 3 440 | 2 048 | **1** |
+| `VOICE_D_452` | 15 728 | 14 336 | **7** |
+| `VOICE_D_453` | 46 448 | 45 056 | **22** |
+| `VOICE_D_454` | 60 784 | 59 392 | **29** |
+
+Exact on 5/5, with no free parameter to tune. The byte layout is a 1392-byte
+header, a whole number of 2048-byte packets, then the RIFF section — and the
+raw bytes agree: high-entropy from offset 0, then a zero run immediately before
+the `RIFF`. `VOICE_D_451` is again the control: its one packet is all zeros.
+
+**So I made the obvious fix — emit that region as a sub-wave — and then withdrew
+it.** Two measurements killed it:
+
+* **It does not recover audio.** Coverage rose from 5.4 % to 89.9 % for
+  `VOICE_D_453`, but the emitted stream decodes through FFmpeg to **1792 PCM
+  bytes** — silence — while the RIFF sub-waves from the same banks decode to
+  150–270 KB. Byte coverage was the wrong success metric, and it looked like
+  progress. The region is shaped like a packet stream but is not XMA1 under the
+  synthesised format.
+* **It is not narrow.** The rule matches **1524 of the 8021** RIFF-bearing
+  entries in `sound.pak` (1524 with a non-zero leading region), including `RT*`
+  movie banks that decode correctly today. Landing it would have risked a wide
+  regression in order to not-fix five banks.
+
+The refuted attempt is recorded in `slb.rs` beside the code, so the next person
+does not re-derive the arithmetic and re-make the same change.
+
 ## What this does not settle
 
-* ❔ **What the leading region is.** It is not padding and not a RIFF. Whether it
-  is a directory, a seek table, an alternate codec stream, or the audio itself is
-  unestablished — I did not decode it.
+* ❔ **What the leading region holds.** Its *size* is now exact (1392 + n·2048)
+  and it is neither padding nor a RIFF, but it does not decode as XMA1. A seek
+  table, a different codec, or a different channel/rate configuration all remain
+  open. The next cheap probe is to vary the synthesised `fmt` (channels, streams,
+  sample rate) rather than to assume the container.
 * ❔ Whether `hokyu_DS_s13A` really plays `VOICE_D_452`. That needs the leading
   region decoded *and* a human listening; audio judgement cannot be done here.
 * ❔ Whether the same leading region exists across the other ~9 500 `sound.pak`