BELIEVEYOUME · Chapter Accuracy Verification

01

Asleep in the Trunk

residual risk: low

liveF minor, ♩=112.3 → verifiedF minor (verified:false) — F-minor home with a major-third vocal opening (sung A over the instruments' A♭) that darkens to the flat side in the back half; instruments-only chord field is minor

KEY CONFIRMED2 high2 medium3 low

MANUSCRIPT.md Chord Chart header (the '(composite chord: guitar + bass + vocal ... vocal supplies the 3rd to complete each triad)' card)

manuscriptChord = guitar + bass + vocal composite; the vocal supplies the 3rd to complete each triad (so the home chord reads as a completed major/minor triad including the sung note)

verifiedBook Chord Rule 0: chord = rhythm+lead+bass ONLY, vocal read separately as an interval. Under that rule the instrument chord field is minor (A♭ 14.9% vs A 0.6% in rhythm+lead+bass); the major third A exists ONLY in the Vocals track (A 11.5%/32 hits vs A♭ 6.5%/18 hits, 1.78x). The manuscript card uses the v2-website composite policy, opposite to the book.

MANUSCRIPT.md card line 24 + Figure 1 + prose line 53 (tempo)

manuscript♩=112.3 BPM

verifiedasleep-cdn.mid set_tempo metadata = 120.00 BPM (500000 us/qn) on all tracks; Songsterr id s5085156 displays ♩=173. Three different tempi; 112.3 matches the My Mirror Hates Me keys_resolved entry, suggesting a borrowed/erroneous DAW value. The draft flags 112.3 vs 173 but omits the MIDI's 120.

Draft D17 + asleep_stem_keycheck — 'the keys_resolved.yaml Asleep entry'

manuscript(draft) the keys_resolved.yaml Asleep entry stays verified:false

verifiedkeys_resolved.yaml (711 lines) has NO Asleep entry. The 7 entries are Alligator, So Far So, Broken Satellites, The Mutiny, My Mirror Hates Me, Lazarus, Endless Summer. The keycheck itself says it is 'evidence for one.' The entry was never appended.

Draft (3x) + keycheck — attribution of an 'A over F' note to the manuscript

manuscript(draft claims) the manuscript already gestured at this with its 'A over F' note

verifiedgrep of MANUSCRIPT.md and MANUSCRIPT_v3.md: 'A over F' does not appear anywhere. The manuscript's only third-language is the rule-violating composite caption 'vocal supplies the 3rd.' The draft credits the manuscript with an observation it does not contain.

Brandon ear-check

Confirm the voice reaches the major third A over the F in the opening 0-90 s (MIDI verse bars 41-54)
Confirm the back half darkens to the flat side (A♭, D♭, E♭) 120-270 s (MIDI home-return bars 136-146: C + A♭ in guitar, voice gone)
Count the bar: 3/4 vs 4/4 whole song (Songsterr transcriber chose 3/4; MIDI metadata is 4/4)
Confirm the felt pulse against 112.3 / 120 / 173 BPM whole song
Does the chorus resolve to C (the V) or to Fm chorus turns at MIDI bars 69-72, 133-135, 237-240

02

The Alligator

residual risk: low

liveG minor (prose + all 6 figures say G minor, banner "G minor · ♩=80.7"); BUT the Chord Chart header inside the same section says "Key: G major" — the manuscript is internally self-contradictory → verifiedG major (verified:false — major third lives in the VOICE (B 46:1 in CDN vocal) and LEAD (F#=M7), not the rhythm guitar, which leans minor on its own; keys_resolved.yaml confidence high, pending user ear-check of the 90-120s chorus)

KEY CORRECTED3 high3 medium2 low

MANUSCRIPT.md ~line 268 banner + ~296 prose vs the Chord Chart header

manuscriptprose/banner = G minor (♩=80.7) repeated throughout

verifiedG major; and the manuscript's OWN Chord Chart header already says 'Key: G major' — one section, two opposite keys

MANUSCRIPT.md banner + body (~268, ~296)

manuscript♩=80.7 BPM

verified120 BPM — audio-truth.mid single tempo event = 120.00, GP + T7 stem folder both '120bpm', Moises beat-track 117.5; 80.7 unreproducible from any file

MANUSCRIPT.md loop line + chord-anatomy/voice-leading figures

manuscripta chromatic B major chord (III, B/D#/F#) = 'the song's single most striking harmonic event'

verifiedno B-major triad exists in any MIDI; the only B event is the bar-12 G/B FIRST-INVERSION tonic (bass B1, upper G+D, no D# anywhere)

MANUSCRIPT.md predecessor paragraph (~end of section)

manuscript♭VII♯5 (F# augmented in G minor) = 'the defining borrowed chord of the Shiner idiolect', a refrain-substitute

verifiedno augmented chord in any file; F# is the lead's sustained MAJOR SEVENTH over G major (audio-truth lead F#=54.5%), diatonic to major

MANUSCRIPT.md loop + chord-anatomy figure + chord chart body

manuscriptloop includes Cm (iv), E♭ (♭VI) and F (♭VII) as chords

verified12-bar audio-truth model is I–♭III only (G↔B♭) closing on G/B; no Cm or F triad present; E♭ exists only as a bass color (CDN bass 13%), not a voiced ♭VI chord

Brandon ear-check

Does the voice land a clear B♮ (major third of G) over the verse/chorus G chord, or ever a B♭? This is THE pitch that flips the key. T7 stem '1-02 The Alligator-vocals-G major-120bpm-441hz.wav', ~90–120s chorus window
Does the lead's F# over G read as a bright major seventh (clean Gmaj9) or as something darker/passing? Same 90–120s window, lead/'other' stem
Does the bar-12 cadence land as a G/B (first-inversion home) or as a separate bright B-major 'shaft of light' shock? The seam into each chorus/cadence, end of each 12-bar cycle
Are E♭/♭VI and F/♭VII ever heard as CHORDS (not just bass notes) in the later/CDN sections the 12-bar model omits? Full-mix outro section

03

The Mutiny

residual risk: low

liveF# major → verifiedBb major with heavy bVI (Gb) / bVII (Ab) modal mixture (Bb-major-third over a b6/b7 field, Mixolydian-with-b6 leaning). Tonal center Bb is unambiguous from the MIDI; the MAJOR quality rests on the sung third (D 1.91x Db) and stays verified:false pending the user's ear-check at 180-240s.

KEY CORRECTED5 high3 medium

MANUSCRIPT.md L488/L509 + card banner

manuscriptF# major

verifiedBb major (tonal center unambiguous: bass home A#1 @bar39, A#102 in body window 96-119, top sung pitch Bb 27%, opening Bb5). 'F#' is the enharmonic Gb = b6 of Bb, the loudest rhythm-gtr PC (F# 554) rung by Drop-C rootless dyads.

MANUSCRIPT.md Figure 1 + Chord Anatomy table

manuscriptloop F#, G#m, A#m, E#dim (I ii iii vii)

verifiedbass arrivals are Bb(I), Gb(bVI), Ab(bVII), F(V); the named 'A#m' IS the Bb tonic re-spelled, 'G#m' is the Ab(bVII) pedal, 'F#' is the Gb(bVI) pedal. None of F#/G#m/A#m/E#dim appear as bass arrivals.

MANUSCRIPT.md prose ('no flat-side borrowing, no parallel-minor intrusion, brightness is the point')

manuscriptone clearly major-keyed track, no borrowed colors

verifiedthe flat side IS the verses + bridge: Gb(bVI) pedals verses (bass 40-47 top F#3), Ab(bVII) pedals bridge (bass 64-71 G#30); vocal sings Gb(b6) 21x vs G(nat6) 2x. This is the album's heaviest-mixture track, not its borrow-free one.

charts/the_mutiny_chord_lyric_FROMMIDI.md + charts/the_mutiny_chord_lyric_FIXED_HYBRID.md

manuscriptKey F#, loop F#/A#m/G#m/Fdim (FROMMIDI even claims 'root from bass MIDI')

verifiedthe bass-MIDI home root is Bb (A#); both live chart files carry the same enharmonic mis-key as the manuscript and will contradict a corrected manuscript.

MANUSCRIPT.md key vs keys_resolved.yaml (The Mutiny entry)

manuscriptF# major

verifiedkeys_resolved.yaml resolves Bb major with heavy bVI/bVII mixture, verified:false

MANUSCRIPT.md banner + body (x4 occurrences)

manuscripttempo 86.1 BPM

verified129-130 BPM (GP-MIDI 130, Moises beat-track + yaml 129.2, CDN transcription clock 120). 86.1 reproduces from no source.

MANUSCRIPT.md mixture sentence

manuscriptthe b6 is D natural (b6 of F#)

verifiedthe b6 is Gb (sung 21x); D natural is the MAJOR 3rd of Bb (sung 21x) - the opposite function from what the manuscript assigns.

Brandon ear-check

Tonic chord quality: does the home chord sound MAJOR (Bb) or MINOR (Bb minor)? This is the one call that sets verified true/false. Body 180-240 s (bars 96-119), where the sung D sounds against the Bb root
Outro: is there a real B-natural bass drift, or is it the maj7 (A-natural) / a lead-line color? 240-270 s (outro)
Is the sung major third heard as the chord's defining 3rd, or as a passing/suspension tone? The F-D-F-D 5-3-5-3 hook phrase in the body (~bars 96-119)
Once re-keyed to Bb, does the refrain still FEEL like it ends open on a dominant? Refrain landings per the lyric chart (F(V)->Bb region)

04

So Far So

residual risk: low

liveD major → verifiedD minor (D natural minor) — D-centered with a B♭ (♭VI) pedal that takes over the choruses. verified: false (only the user's ear-check flips it; I did not self-flip).

KEY CORRECTED3 high3 medium1 low

MANUSCRIPT.md lines 656/660/671 (header banner + body) — Key

manuscriptD major

verifiedD minor — voice sings min3 F over maj3 F# at 14.4% vs 2.1% (6.8:1); B♭ beats B-natural in every stem; the chord-detector itself labels more Dm(64) than D(60) segments

MANUSCRIPT.md ~line 810 — chord-anatomy table, the 'A' chord

manuscriptA = A–C♯–E, function 'dominant' (V)

verifiedA minor (v). C# is only 0.2% of sounded harmony (bass+rhythm+lead) and 0.3% of the real bass; C-natural (10.9%) is the third. There is no audible C# in the recording, so the chord is Am, not A major. The reset bars (bridge bars 66/70) read A3,C4,E4 = A minor

MANUSCRIPT.md lines 694/712/730 — the ending / 'closes on V (A)' / 'outro lands on A' / 'refrain-on-V cluster with The Mutiny and Endless Summer'

manuscriptthe song closes on V, an open unresolved dominant A; half-cadence ending

verifiedthe song ends on a D pedal — last sounding pitch is D1/D2 in bar 139 (bass+drums only, rhythm/lead/vocal out after bar 134). No A arrival at the end; no half-cadence on V

MANUSCRIPT.md lines 690 & 783–790 (verse loop) and 694 (refrain loop)

manuscriptverse D→G→B♭→A (I–IV–♭VI–V) and refrain D→B♭→F→A (I–♭VI–♭III–V) with functional G, A, F bass arrivals

verifiedD-pedal verses (bass D 91–100%) and B♭-pedal choruses (B♭ 68%) with a one-bar Am (v) reset. Real bass roots: G 1%, A 4%, F 2% — none arrive as structural roots. G is the 4 carried by inner voices, not a G-major bass chord

Internal: sofarso-bridge/PROPOSED_BRIDGE_CARD.md & sofarso-outro/PROPOSED_OUTRO_CARD.md key-lines vs the deep-chapter draft

manuscriptbridge/outro cards label the song 'D Mixolydian, Drop C'

verifiedD natural minor. Mixolydian requires a major third (F#) and natural sixth (B); the recording has F#≈0 and B♭>B-natural, ruling out Mixolydian. The book's own sidecar cards contradict its own deep read

Brandon ear-check

The sung third over D — does the voice land F (minor 3rd) or F♯ (major 3rd)? This is THE major-vs-minor decider; MIDI says F beats F# ~7:1 but only the ear confirms the recording sings the flat third. a verse around 40–48 s (and the verse at ~150–185 s)
The chorus root — is the bass genuinely parked on B♭, felt as ♭VI under a still-ringing D-shape (not a key change to B♭ or G)? choruses 100–130 s and 200–248 s
The 'A' chord quality — minor or major? Listen for C-natural vs C♯ in/under the A-reset bar. the one-bar A reset inside chorus/bridge (e.g. bridge bars 66/70 and the verse-reset bars)
The ending — does the track end on a low D pedal with no resolving chord, or on an A? the fade-out, last ~10 s (bars 134–139)

05

My Mirror Hates Me

residual risk: low

liveE minor (Aeolian) in the banner (line 857) AND body (line 878) — but the same card's "Loop" line 859 carries contradictory OLD boilerplate "Borrowed-major loop (no tonic minor stated)" → verifiedE minor (Aeolian) — verified:false, dispute preserved (audit auto-detector called "E major" at conf 0.549, the lowest call in the audit; every independent re-read sides with E minor)

KEY CONFIRMED3 high3 medium

MANUSCRIPT.md line 859 (card header) vs line 878 (body)

manuscriptLoop: Borrowed-major loop (no tonic minor stated)

verifiedEm is the home chord; every chord but the flat-VII is minor; the tonic minor is stated constantly (E5 power chord + open-E drone + Em arpeggio; bass E 53.3%). There is no major-key frame to borrow into. The header is leftover OLD auto-detector boilerplate and is FALSE.

MANUSCRIPT.md line 888 (body)

manuscriptThe borrowed-major loop is a groove built from avoidance.

verifiedThe loop is a closed Aeolian i flat-VII v iv (Em D Bm Am); there is no borrowed-major content. Same false 'borrowed-major' label as the header.

MANUSCRIPT.md banner lines 857 + 859 (same box)

manuscriptE minor (Aeolian) ... Loop: Borrowed-major loop (no tonic minor stated)

verifiedCannot both be true in one banner; E minor / tonic-minor-stated is correct, 'Borrowed-major' is the error.

MANUSCRIPT.md tempo (112.3) vs source MIDI

manuscript112.3 BPM

verifiedCDN MIDI metaevent = 120 BPM, GP8 = 114 BPM. 112.3 is the audio-feel/published-tab tempo; the file clocks differ. Provenance gap, not a key error, but unannotated in the manuscript.

MANUSCRIPT.md Figure 2 grid vs its prose (snare count)

manuscriptsnare hits 13 of 16 (prose)

verifiedFigure 2 grid shows 14 filled cells (14/16). One-cell self-contradiction inside the manuscript.

Brandon ear-check

The SUNG third over the E tonic: minor G vs major G# ~30-60s (sung G loudest per yaml) and ~180-210s
The bridge sung material over A/D harmony ~120-150s and ~210-240s
The outro bass: does it ever land on F, or close on E1? ~270-300s
The Am7 cadence (A-C-E-G) on 'I'm losing hours alone' chorus close ~0:50-1:00
The 15s wall as the song's only peak / no chorus lifts 0:15 vs 0:30+

06

Endless Summer

residual risk: low

liveF Mixolydian, tonic F, ♩=143.6 (the old auto-detector card) → verifiedB♭ major (tonic B♭, true diatonic I-IV-V, with a recurring ♭VI/G♭ modal-mixture gesture and a blue-third D♭ vocal inflection). verified:false — never self-flipped; user ear-checks and flips.

KEY CORRECTED3 high2 medium2 low

MANUSCRIPT.md l.1024/l.1042 card banner vs draft + keys_resolved + stem report

manuscriptF Mixolydian, tonic F

verifiedB♭ major, tonic B♭ — bass downbeat root B♭ 53% vs F 11%; duration-weighted Krumhansl B♭ major 0.852 vs F major 0.346; Moises stems confirm B♭ (bass yin B♭ 41.8% top in 9/9 chunks)

MANUSCRIPT.md self-contradiction: card banner (l.1024) vs body (l.~1050) and embedded FROMMIDI chart header

manuscriptcard says F Mixolydian, but body says 'the 2026-06-04 audio chroma audit favors the predecessor's [B♭] reading, bass plants on B♭' and the embedded chord chart header literally reads 'Key B♭'

verifiedB♭ is correct; the section already contains the right answer fighting its own banner

MANUSCRIPT.md l.1044 chord table + l.1140 Figure 1 Roman numerals

manuscriptF / E♭ / B♭ = I / ♭VII / IV in F

verifiedsame three roots re-rooted in B♭ = V / IV / I (F is the dominant, B♭ is tonic); the hook F→E♭ is V→IV not I→♭VII

MANUSCRIPT.md l.1046

manuscriptthe whole song lives inside the seven notes of F mixolydian, no chromatic borrowing at all

verifieda recurring ♭VI (G♭ = F#+A#+C#) at bars 37-43, 89-95, 165-171, plus a blue D♭ sung at 16.9%, are both chromatic to B♭ major — borrowing is present

MANUSCRIPT.md chord-anatomy figure (F = F,A,C)

manuscriptF chord is the major triad F-A-C

verifiedF is never voiced as F-A-C; F's maj3 (A) is 2.4% of the chord layer. The sounded F is F+C+E♭ (a power/dominant shape)

Brandon ear-check

Tonic home: does the song rest on B♭ (I) or on F (the V)? Chorus ≈0:55-1:05 of the record (MIDI bars 63-72), B♭ pedal with voice on D
Sung third: bright major D vs blue D♭ any verse; chorus ≈ bars 63-72
The ♭VI (G♭) darkening gesture refrain turnarounds, MIDI bars 37-43, 89-95, 165-171
Tempo feel: 142 vs 120 whole song

07

Lazarus

residual risk: low

liveC minor → verifiedC, sectional major-then-minor third (tonic C unambiguous; the third is mixed — bright C-major verses, flat-side refrain, third-less C7sus outro). Stays verified:false.

KEY CORRECTED3 high3 medium

MANUSCRIPT.md outro chord (chord-anatomy table + prose ~line 1154 region) vs 06-02 MIDI bars 153-206

manuscriptOutro terminus is B-flat m7 / B-flat m11, 'B-flat in the bass with A-flat and D-flat above', the flat-VII of C minor

verifiedBass floor is C2 (no B-flat in bass register at all); outro field is C 50.0% vs B-flat 15.4%; the A-flat that a B-flat m7/m11 REQUIRES is 0.0% (1 of 2168 notes). Terminus is C7sus, no third. B-flat is the flat-7 ringing ABOVE a C root.

MANUSCRIPT.md vocal-vocabulary paragraph vs 06-02 Vocals track (165 notes)

manuscriptVocal centers on scale degrees b7, b3, b6, 1, 6 = B-flat, E-flat, A-flat, C, A; phrase endings fall on b7, b3, b6, 1

verifiedB-flat (b7), the manuscript's #1 claimed vocal pitch, is sung ZERO times. Actual top sung pitches: D (9th) 25.5%, G (5) 23.0%, C (1) 15.2%, E (maj3) 10.9%, A (6) 8.5%. E-flat (b3) only 5 hits (3.0%). The described vocabulary cannot come from this sung line.

MANUSCRIPT.md internal: over-lyrics Chord Chart vs its own Figure-1 + Chord-anatomy table

manuscriptLyric chart repeatedly uses Fm, Gb, Ddim (e.g. 'Cm Fm Gb DdimFm', 'Gm Eb E Cm Ddim Ab')

verifiedThe same section's Figure-1 loop (Cm-Ab-Gm-Cm) and chord-anatomy table (Cm, Ab, Gm, Eb, B-flat7, Fadd9) never list Fm/Gb/Ddim, and none of Fm/Gb/Ddim appear in the 06-02 grid. Two incompatible chord vocabularies in one section; the lyric chart is stray/contaminated.

MANUSCRIPT.md Lazarus bar-by-bar form map source label

manuscript'(bars from Shiner-MyMirrorHatesMe-FROMCDN.mid)'

verifiedThe Lazarus form map cites the My Mirror Hates Me MIDI — a different song. Copy-paste/source error.

MANUSCRIPT.md header key vs keys_resolved.yaml resolved_key

manuscriptC minor (flat)

verifiedkeys_resolved.yaml resolves 'C, major/minor-third mixture'. Verses are demonstrably C-major-bright (guitar C-E-G bars 1-2/9-10/33-36; voice E-natural 10.9% vs E-flat 3.0%). Flat 'C minor' erases the bright verses. (Counterpoint: recorded stem AGGREGATE leans minor 2.9x, so C minor is closer to the whole-song average than the draft's 'major-leaning' — only 'mixture' holds both ends.)

Brandon ear-check

Confirm the opening guitar C-E-G major and the voice's E-natural read bright/major (the case for 'mixture', against flat 'C minor') ~0-30s, 06-02 bars 1-10
Confirm the refrain flat-side bIII-bVI-bVII (E-flat/A-flat/B-flat) gravity and the C-major landing at bars 33-36 refrain, 06-02 bars 25-36
THE decisive call: confirm the heavy E-flat in the back half is the VOICE sustaining the flat third over the third-less drone, not vocal-stem bleed from the instruments ~180-240s
Confirm the outro drone sits on C (low C2), i.e. the B-flat is the b7 ringing ABOVE a C root, not a B-flat tonic outro, 06-02 bars 153-206

08

Not Too Much

residual risk: low

liveA♭ major → verifiedA♭ major (verified:false — stem + MIDI confirmed; ear-check is for the ♭VI middle, not the key)

KEY CONFIRMED2 high1 medium2 low

MANUSCRIPT.md line 1566 (predecessor block) vs line 1427/1413 + audit_not_too_much_bass_intervals_2026-06-04.json

manuscript100 percent stepwise motion in the bass, every move by half/whole-step, no skips (MIDI-verified)

verified31.7% stepwise, 68.3% skips (re-derived from the audit histogram, n=290). The manuscript itself states ~32% (line 1427) and 34% (line 1413) for the same bass line — it holds two mutually exclusive numbers.

MANUSCRIPT.md line 1566 vs MIDI re-read + line 1487 (Chord-anatomy figure)

manuscriptthe verse crosses through flat-6 via the iv chord (D-flat MINOR) carrying the flat sixth as a chord tone

verifiedEvery D-flat bar in the FROMCDN MIDI voices C#-F-G# = D-flat MAJOR (F is the major third). D-flat minor (needs F-flat=E) is never played. The manuscript's own Chord-anatomy figure (line 1487) labels D-flat as 1-3-5 major, contradicting its own line 1566.

MANUSCRIPT.md line 1427 (Figure 1 caption) vs verified verse motion

manuscriptVerse arc = I IV flat-6 flat-3 I

verifiedThe flat-6 (E/F-flat) and flat-3 (C) roots are bridge features (bars 57-99), not verse chords. Verified verse motion is ii<->I with IV and passing iii/V. The caption imports bridge material into the verse label.

Brandon ear-check

The 150-210s middle: is it a flat-VI tonicization, a C-major (III) chromatic-mediant patch, or a brief new E/C center? Bass pedals E (flat-VI scale degree) but the chord above is C-E-G (C/E), not F-flat major 150-210s / 05-05 MIDI bars ~57-99 (stem chunks 6-8)
What the VOICE sings over the E middle — does it hold the bright C-against-A-flat (color, draft's thesis) or follow E-relative material (F, b6, D#)? The stem sung-pitch (C>B all 12 chunks) and the FROMCDN MIDI vocal (F/C-as-b6/D# over E bass, bars 74-88) DISAGREE 150-210s
A-flat major third in the opening verses (sung C as the bright third at cadences, e.g. bar-31 'C(3)' phrase) 0-90s
Outro final D — confirm the song ends on a D natural that drifts outside A-flat ('resolution forgotten') 270-348s (use 05-05 MIDI/stems; FROMCDN stops at 236s)

09

Broken Satellites

residual risk: low

liveE-flat major (card-level call, correct) → verifiedE-flat major (verified:false — the disputed B-flat/A# is the structural DOMINANT (V), not the tonic; same E-flat-home/B-flat-dominant relationship as Jackie. Pending user ear-check.)

KEY CONFIRMED3 high2 medium1 low

Card header + body, ~line 1572/1585: "At 78.3 BPM, this is the slowest song on the album"

manuscripttempo = 78.3 BPM

verified~117.5 BPM. CDN file=120, session export=~115-116, audio-truth/README/stems=117.5. 78.3 is a 2/3 triplet-half mis-read of 117.5 (117.5 x 2/3 = 78.3). The 'slowest song' framing rests on a wrong number.

Body prose ~line 1585: "a bass-only chromatic descent, B-flat-A-A-flat-G"

manuscripta four-note falling bass line B-flat-A-A-flat-G opens the song

verifiedNo such line exists in any MIDI on disk. CDN bass opens D1 (8s), F1 (12s), then E-flat. Literal Bb-A-Ab-G (pc 10,9,8,7) descent = 0 occurrences anywhere in the bass track.

Body prose + Figure, ~line 1585: "The outro similarly avoids root-position tonic landing, the section closing on a non-root chord"

manuscriptthe tonic I is avoided at the close

verifiedCDN bass pedals E-flat hard from 270-320s (D# 83 notes / C 23), the most concentrated root-position E-flat in the whole song. The genuine non-resolution is the separate late D/Dm outro tail (sections 23-24), a chromatic chord, not an avoided tonic.

Body prose "Fadd9, a suspended arrival on the IV degree" vs Figure 1 / Chord-anatomy figure labeling the same chord "IIadd9" / "II add9"

manuscriptthe F chord is the IV degree (prose) AND simultaneously labeled II (figures)

verifiedIn E-flat major, F is scale-degree 2 = II (the IV is A-flat). The prose 'IV degree' is theory-wrong and contradicts the manuscript's own figure labels. Internal contradiction + theory error.

Chord-anatomy figure "Fadd9 = F-A-C-G" (F major) vs charts/broken_satellites_chord_lyric_FROMMIDI.md "Fm" (F minor)

manuscriptmanuscript figure spells F major; the FROMMIDI chart/draft call it F minor

verifiedSounded as an F5 power chord; where a third rings on the F root it is A-natural (F major), never A-flat (across 17 F-root bars: A-nat=6, A-flat=0). The two book artifacts disagree on this chord's quality; F5 / F-major (II / V-of-V) is the audio-honest call.

Brandon ear-check

The tonic — does the song feel like it comes home to E-flat (not B-flat)? E-flat-home verse ~30-90s; outro E-flat pedal ~270-320s
The third — does the sung melody land on a MAJOR third (G) and never the minor third (G-flat/F#)? hook/chorus ~180-210s; verse ~30-90s
The Cm (vi) pedal — does the loud middle sit on C-minor color without becoming a new key? ~120-150s and ~240-270s
The late chromatic tail — does the very end drift to a D-major/Dm color that does not resolve? ~330-363s (sections 23-24)
The E-flat-minor mixture moment — is there a brief parallel-minor darkening late? ~315-330s (section 22)

10

Jackie

residual risk: low

liveB♭ Mixolydian, ♩=95.7 → verifiedE♭ major (B♭ = V), disputed against B♭ Mixolydian — medium confidence, verified:false (user ear-check pending)

KEY CORRECTED5 high4 medium

MANUSCRIPT.md card header + body (lines ~1758, 1772)

manuscriptB♭ Mixolydian (B♭ = I, tonic)

verifiedE♭ major, B♭ = V — bass home E♭ in both clean MIDIs (E♭ 27-30% of bass PCs; B♭ only 11%; B♭ is the downbeat root of just 4% of bars and is never a section home); keycheck combined Krumhansl E♭ major 0.938 vs B♭ major 0.431. verified:false, ear-check.

MANUSCRIPT.md header + Figures 1/2 (♩=95.7 appears 4x)

manuscript♩=95.7

verified♩=144 — GP MIDI set_tempo metaevent = 144.0 BPM; Songsterr id s5418285 displays ♩=144 (SONGSTERR_AUDIT). 95.7 matches no source (FROMCDN carries a generic 120).

MANUSCRIPT.md prose ~line 1772

manuscriptB♭ is the most frequent chord, A♭ the next, then E♭

verifiedPer-bar downbeat-root rank in BOTH MIDIs is C, A♭, E♭, G, B♭ — B♭ is the LEAST frequent real root (4% of downbeats), not the most frequent. The ordering is inverted.

MANUSCRIPT.md (manuscript self-admission M16) vs its own header

manuscriptearlier per-song key detection placed this track a fifth away, so scale-degree counts are omitted

verifiedA fifth below B♭ IS E♭. The manuscript concedes the detector was a fifth off (pointing straight at E♭) yet keeps B♭ as the headline key — an internal self-undercut.

SONGSTERR_AUDIT_2026-06-03.md Jackie row vs MANUSCRIPT tempo

manuscriptJackie tempo verdict = '✓ matches'

verifiedThe audit's 'matches' is itself wrong: Songsterr shows 144, the card shows 95.7 — they do NOT match. Row should read mismatch (card 95.7 vs Songsterr 144).

MANUSCRIPT.md prose ~1772 + figures

manuscriptthe ♭VII (A♭) is the modal signature, without a single chromatic borrowing

verifiedUnder E♭, A♭ is the diatonic IV, not a ♭VII borrowing; and the OUTRO is overtly chromatic (GP bass walks C→D♭(C#)→D, bars ~90-102, plus a G+A♭ tritone rub). Two wrong claims in one sentence.

MANUSCRIPT.md Figure (chord anatomy)

manuscriptB♭ = B♭–D–F, a clean I tonic triad

verifiedNo prominent D-natural over B♭ in the MIDI; the sounded body harmony (bars 13-46) is the A♭-rooted cluster C–E♭–F–G–A♭–B♭. The B♭ tonic triad is never played.

Brandon ear-check

Does the song rest/resolve on E♭, or sit on B♭? (the whole key call) throughout; cleanest in the chorus/bridge body ~bars 47-69 (bass roots E♭) vs the verse A♭ floor
The sung third over the home chord — G (M3 of E♭) vs B♭-as-root-with-no-third verse/chorus vocal entries (the cold-open chorus, 'Why'd you make a mess')
Outro: true modulation to C, or still-in-A♭ ♭III tonicization? final ~8 bars (card bars 95-102; C-territory actually begins ~bar 70/83)
Meter: is there a recurring 7/8, or straight 4/4? throughout (GP MIDI inserts 7/8 between 4/4 on a repeating cycle; FROMCDN is pure 4/4; Songsterr display says 4/4)
'Opens on Gm (vi)' — does the song audibly open on a G-minor chord? intro 0:00-0:15 (bars 1-12)

Kid mode 🧒

Asleep in the Trunk

The Alligator

The Mutiny

So Far So

My Mirror Hates Me

Endless Summer

Lazarus

Not Too Much

Broken Satellites

Jackie