diff --git a/docs/port/HANDOFF.md b/docs/port/HANDOFF.md
index 0c414fe4..d8aa5983 100644
--- a/docs/port/HANDOFF.md
+++ b/docs/port/HANDOFF.md
@@ -6367,3 +6367,30 @@ observable changes.
 Next instrument, when it is worth it: a **write-watch on the UI group's clock
 field** in Canary. It would name every writer including the snap, and close the
 standing "which function advances the clock" question at the same time.
+
+## 🔴 F6 figures corrected — I measured against the wrong element
+
+`docs/re/f6-what-the-human-is-watching.md`. **No code change for you**; the
+conclusion is unchanged and stronger.
+
+I gave you the sweep-to-plate separation as **0.057 of a sweep loop**. That was
+measured against **`ptcopyright`**, which I had already established is *not* the
+plate — the plate is `ptbtn00f`, the pulsing element. Against the real plate:
+
+| no input | sweep leads `ptcopyright` | sweep leads **`ptbtn00f`** |
+|---|---|---|
+| `f6b` | 34 frames | **83** |
+| `f6` | 67 frames | **165** |
+
+Phase-free: **0.138** and **0.141** of a sweep loop, agreeing to 2.2 % — against
+the 0.057 I sent. **The lead is 2.4× larger than I told you**, and with Ⓐ it still
+collapses to ~0 (−1 frame). So "gate the sweep on the plate" is a worse fit than
+I made it sound, not a better one.
+
+Nothing you have implemented depends on this number — you gate on the declared
+`t=70`/`t=100`, which is unaffected. It matters only if you were carrying 0.057
+as a check value.
+
+⚠️ Third time a label has been the error rather than the measurement, all mine:
+`0x3003`/`0x3002`, ptcopyright-as-plate for the clock conflict, and now
+ptcopyright-as-plate again in a page written *after* I had corrected it.
diff --git a/docs/re/f6-what-the-human-is-watching.md b/docs/re/f6-what-the-human-is-watching.md
index a85b76de..567486a0 100644
--- a/docs/re/f6-what-the-human-is-watching.md
+++ b/docs/re/f6-what-the-human-is-watching.md
@@ -14,18 +14,24 @@ exit), and anything about the menu.
 
 ## The brief asked which of two declared events this is. It is neither.
 
-| | sweep pair enters | `ptcopyright` | separation |
-|---|---|---|---|
-| `f6b` no input | f577 | f611 | **34 frames** |
-| `f6` no input | f746 | f813 | **67 frames** |
-| `f5-early` Ⓐ during build-in | f438 | f436 | **2 frames** |
-| `f5` Ⓐ *after* the sweep already started | f427 | f456 | 29 frames |
+> 🔴 **Numbers corrected 2026-09-02.** This page first measured the separation
+> against **`ptcopyright`**, calling it "the plate". It is not — the plate is
+> `ptbtn00f`, the pulsing element, identified in
+> [`f6-plate-identity-and-clock-conflict-resolved.md`](f6-plate-identity-and-clock-conflict-resolved.md).
+> Measured against the real plate the lead is **2.4× larger**. The conclusion is
+> unchanged and stronger; the figures below are the corrected ones.
+
+| | sweep enters | **`ptbtn00f`** (plate) | separation | (vs `ptcopyright`) |
+|---|---|---|---|---|
+| `f6b` no input | f577 | f660 | **83 frames** | 34 |
+| `f6` no input | f746 | f911 | **165 frames** | 67 |
+| `f5-early` Ⓐ during build-in | f438 | f437 | **−1 frame** | −2 |
 
 Phase-free, as a fraction of the sweep's own 600-unit loop: the controls give
-**34/600 = 0.057** and **67/1168 = 0.057** — agreeing to **1.2 %** across runs
-that differ 2× in frames — against **~0.003** when Ⓐ lands early.
+**83/600 = 0.138** and **165/1168 = 0.141** — agreeing to **2.2 %** across runs
+that differ 2× in frames — against **~0.002** when Ⓐ lands early.
 
-**Ⓐ collapses the gap by a factor of ~18.** It reveals the plate and restarts
+**Ⓐ collapses the gap by a factor of ~80.** It reveals the plate and restarts
 both sweep leaves at their own `t=0` in the same frame
 ([`f5-verified-with-full-quad-reader.md`](f5-verified-with-full-quad-reader.md)),
 so the two events become simultaneous *because the player caused both*.
@@ -45,7 +51,7 @@ the press falls.
 ## 🔴 What the port must not do
 
 **Do not start the sweep when the plate appears.** That reproduces one boot and
-breaks the other: with no input the glow genuinely precedes the plate by 0.057 of
+breaks the other: with no input the glow genuinely precedes the plate by 0.14 of
 a sweep loop. The correct behaviour is both of the pieces already delivered —
 
 * the declared parent gate, α0 until `t=70`, full at `t=100`; **and**