4cce44a64d
Applying my own METHOD entry one iteration after writing it found four refuted
statements still asserted unmarked where a reader lands:
* envelope correlation 'has no resolving power' -- in three places including
HANDOFF. The port controlled the same estimator on a single track and got
r=1.0000 at zero offset; the saturation needs CONCURRENT streams sharing
timing. I agreed to this in a message and never landed it.
* '8 of 10 three-chunk regions' -- still asserted in HANDOFF in a different
section from its own correction.
* 'r9 is a wild pointer, never a guest address' -- still asserted inside the
kept-for-the-record section.
* the ALSA channel permutation, stated without scope, when a later capture
measured the identity and labelling from it put the silent channel on the
wrong name.
All four marked in place, striking the sentence and pointing forward.
Two lessons added: a 'kept for the record' section still asserts, so labelling
the heading is not enough; and naming a refuted claim keeps it greppable, so the
audit returns its own corrections as hits and every hit needs reading.
The first item is the one worth admitting: I acknowledged that correction in a
message, wrote the entry about corrections that never land, and then did not land
my own for a full iteration.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Wuu56cE8vJGTBtn1ppsk8v
51 lines
2.4 KiB
Plaintext
51 lines
2.4 KiB
Plaintext
# Which ADV voice stream sits where -- the assignment, and how it was settled.
|
|
#
|
|
# 2026-08-30. Chunks dumped by examples/adv_voice_dump.rs from the resolved
|
|
# movie voice region (dat/sound 433930240..437044592), decoded with ffmpeg's
|
|
# xma decoder to f32le 48 kHz stereo.
|
|
#
|
|
# chunk bytes byte_size probe ctx dur L rms R rms L/R r R silent
|
|
# 0 806972 806912 ctx0 TAIL 84.55s -24.79 -24.81 +0.932 53.1%
|
|
# 1 1118268 1118208 ctx1 137.32s -20.33 -inf +0.000 100.0%
|
|
# 2 1171516 1171456 ctx2 137.32s -30.67 -30.68 +0.962 53.6%
|
|
#
|
|
# ctx0's full byte_size is 1294336; the region resolver starts at the
|
|
# predecessor cue's trailer, so chunk 0 is its clipped tail (62 %).
|
|
#
|
|
# 🔴 WHAT DID NOT WORK: envelope correlation cannot discriminate.
|
|
# Every residual channel shares the dialogue's activity timing, so a
|
|
# ⚠️ IN THIS REGIME ONLY (corrected 2026-08-30): the port controlled the same
|
|
# estimator on a single dialogue track and got r=1.0000 at zero offset with
|
|
# -0.08..+0.08 elsewhere -- it localises sharply. The saturation below needs
|
|
# CONCURRENT streams sharing timing at zero lag. Not a general limit.
|
|
# per-pair lag search returns 0.86-0.95 for EVERY chunk against EVERY
|
|
# channel. Recorded because it looks like a strong result and is not.
|
|
#
|
|
# 🔴 Sample-level correlation also fails: the chunks do not start with the
|
|
# movie and the XMA decode's framing offset is unknown, so r ~ 0.
|
|
#
|
|
# ✅ WHAT SETTLES IT: level, under the SAME 0.600 gain the bed uses.
|
|
#
|
|
# chunk level x0.600 nearest residuals (|error| dB)
|
|
# 0L -24.79 -29.23 FL 0.05 FR 0.06 FC 3.96
|
|
# 0R -24.81 -29.25 FL 0.03 FR 0.04 FC 3.98
|
|
# 1L -20.33 -24.77 FC 0.50 FL 4.51 FR 4.52
|
|
# 2L -30.67 -35.11 BL 0.35 BR 0.38 FR 5.82
|
|
# 2R -30.68 -35.12 BL 0.34 BR 0.37 FR 5.83
|
|
#
|
|
# Ratio test, immune to any worry about chunk 0 being clipped:
|
|
# chunk0L - chunk2L = +5.88 dB ; FL - BL = +6.18 dB -> agree to 0.30 dB
|
|
# swapped, the ratio would be wrong by 11.76 dB
|
|
#
|
|
# Structural confirmation: chunk 1 is the ONLY chunk with a digitally silent
|
|
# channel (R, 100 %), and LFE is the ONLY output channel with an empty
|
|
# residual (-115.73 dBFS). One-to-one.
|
|
#
|
|
# Internal L/R correlation also tracks:
|
|
# chunk0 +0.932 <-> FL/FR residual +0.918
|
|
# chunk2 +0.962 <-> BL/BR residual +0.929
|
|
#
|
|
# ==> ctx0 (1294336) -> FL, FR
|
|
# ==> ctx1 (1118208) -> FC, LFE silent
|
|
# ==> ctx2 (1171456) -> BL, BR
|