{"slug": "revive-us-again", "name": "REVIVE US AGAIN", "meter_label": "11.11 with Refrain", "meter": {"pattern": [11, 11], "doubled": false, "refrain": true, "alleluia": false, "irregular": false}, "composer": "john-j-husband", "composer_date": "c. 1815", "canonical_key": {"tonic": "g", "mode": "major"}, "preferred_harmonization": "revive-us-again-1887", "provenance": {"layer": "tune", "source": "hymns-new-and-old-1887", "source_kind": "hand_encoded", "pd_basis": "pub_pre_1931", "pd_evidence": "Read from the dated engraving in Hymns New and Old, for Use in Gospel Meetings and Other Religious Services, comp. D. B. Towner (Chicago: Fleming H. Revell, 1887), no. 168 \u2014 archive.org item woldforus00town, leaf n176 \u2014 and re-read on the independent 1887 scan of the same plate, archive.org hymnsnewoldforus00town, leaf n174. G major, 3/4, one-quarter anacrusis, 16 bars, 50 soprano noteheads. US public domain by publication before 1929. THE ATTRIBUTION IS NOT THE 1887 PAGE'S. That page credits the music only as \"ENGLISH MELODY.\" and names no composer; it is the later dated printings that name John Jenkins Husband, and the one opened for this import is Tabernacle Hymns No. 2 (Chicago: Tabernacle Publishing, 1921), no. 330, archive.org tabernaclehymnsn02chic leaf n317, which heads the hymn \"Revive Us Again.\" and credits \"Wm. P. Mackay.\" at the left and \"J. J. Husband.\" at the right. That 1921 page is ALSO in G major and carries the same melody with a DIFFERENT lower-staff harmonization; none of its notes are encoded here; it is cited for the composer credit only. Husband died in 1825, so the melody itself long predates any printing consulted; the composer date \"c. 1815\" is the conventional one and is NOT printed on either page. NEITHER PAGE PRINTS A TUNE NAME on the staff \u2014 Towner 1887 heads it only with the hymn title \"Revive us Again.\" \u2014 so REVIVE US AGAIN is the name the hymnals and Hymnary carry for it, not a name read off the plate; no printed alternate name was seen, so no alt_name is recorded. tools/tune_fingerprint.py finds no corpus melody closer than 0.697 (WERE YOU THERE, a different meter), well under the 0.80 duplicate threshold, so this is not a duplicate of a tune already in the corpus under another name.", "pd_jurisdictions": {"US": true}}, "harmonization": {"slug": "revive-us-again-1887", "tune": "revive-us-again", "music_file": "revive-us-again-1887.ly", "harmonizer": null, "arrangement_type": "harmonization", "canonical_key": {"tonic": "g", "mode": "major"}, "time": "3/4", "pickup": "4", "system_breaks": [11, 23, 37], "breaks_manual": true, "markings": [{"soprano_note": 24, "kind": "refrain"}], "final_barline": "|.", "amen": false, "verse_variable_melisma": true, "refrain": {"soprano_note_start": 24}, "source_line": "Setting: Hymns New and Old, 1887, where the tune is credited only \"English Melody.\" Public domain.", "provenance": {"layer": "harmonization", "source": "hymns-new-and-old-1887", "source_kind": "hand_encoded", "pd_basis": "pub_pre_1931", "pd_evidence": "The four-part setting printed in Hymns New and Old, for Use in Gospel Meetings and Other Religious Services, comp. D. B. Towner (Chicago: Fleming H. Revell, 1887), no. 168, headed 'Revive us Again.' under the motto '\"O Lord, revive Thy work.\"\u2014Hab. 3: 2', credited 'Rev. WM. PATON MACKAY, 1866.' at the left and 'ENGLISH MELODY.' at the right. G major, 3/4, one quarter anacrusis, 16 bars, no repeats, no accidentals, no amen. US public domain by publication before 1929, and no modern arranger stands anywhere in the chain: the plate names no harmonizer at all and the book itself is 1887. NO SCAFFOLD WAS USED. There is no MIDI, no NoteWorthy file and no modern edition behind this encoding; every chord was read off the page image. PRIMARY WITNESS: archive.org item woldforus00town, leaf n176 \u2014 2420x3765 px, staff space measured at 29.9 px, so one diatonic step = 14.95 px. SECOND, INDEPENDENT WITNESS: the other 1887 scan of the same plate, archive.org item hymnsnewoldforus00town, leaf n174 \u2014 1785x2934 px, staff space 23.6 px \u2014 read end to end by the same rig with its own binarisation and its own staff-line fits. (That second item's stored JP2 was fetched too and is the same 1785x2934 raster as its page render, so 23.6 px really is its native staff space; the higher-resolution copy was therefore made the primary reading and the lower-resolution copy the check. Both are dated 1887 printings of the same plates.) ALL 50 CHORD POSITIONS x 4 VOICES WERE READ ON BOTH SCANS AND AGREE \u2014 the matched filter found 45 of the 50 on the second scan and the remaining five, plus the four the filter merged at system edges, were probed by hand from their ink envelopes there. That set includes the pickup's alto b below its C4 ledger, the five hollow half-note chords, the ledgered d'-over-b chords of 'Je-sus who', and the single shared notehead at ordinal 46. HOW THE NOTEHEADS WERE MEASURED: for each head the five staff lines were re-found LOCALLY in a +-45 px window around that head's own x, with the five-line comb pinned to within 0.62 staff space of the globally located staff \u2014 an unpinned fit slipped a whole space near a system edge and mis-read the last chord of system 1 as b'/e' when the plate prints d''/g', which is exactly why the reading of every chord was also confirmed by eye at 3-8x. Paper level was taken per column window; heads were isolated by hole-filling and stem-masked centroids under an elliptical matched filter; each centroid became a diatonic step offset from the bottom line and was accepted only within ~0.15 step. Hollow half-note heads are invisible to a filled-head filter, so all five of them (ordinals 5, 11, 23, 37, 50) were read from their ring envelopes and looked at by eye at 6-8x on both scans. ACCIDENTALS WERE CHECKED SEPARATELY, BY EYE: a notehead probe reads staff positions and not sharps or flats, so the automated figure above covers no accidental at all. Every one of the six staves on BOTH leaves was examined at 2-8x, and an independent shape sweep for accidental-sized ink (1.45-3.2 staff spaces tall, 0.38-1.35 wide) was run over both scans; the only components it returned were clefs, time signatures and stems. THERE IS NOT ONE ACCIDENTAL IN THIS HYMN, and the only sharps on either leaf are the single F-sharp key signatures at the head of each staff, counted by eye on all six. THE FINGERPRINTS OF THIS READING, all visible at high magnification on both scans: the anacrusis is a four-note chord g,/b/b/d' with alto and tenor in unison on b, the alto's head sitting below a drawn C4 ledger; bar 2 sets 'God!' as a half note and 'for the' as two plain flagged eighths (no dots) while the lower staff jumps to d' over b with a c' ledger between them; 'Je-sus who' is three quarters of d''/g'/d'/b, each lower-staff chord ledgered at c'; the verse does NOT cadence but ends on D major ('a-bove' = a'/fis'/d'/d), running straight into the chorus; the chorus's four 'Hal-le' figures are a dotted eighth plus a sixteenth (two flags on the second, dots on both heads of the first) and the bass holds a g pedal under the entire chorus except 'a-men' (ordinal 37, d) and the last three notes; and the lower staff prints ONE notehead, on g, at ordinal 46 ('re-'), which is encoded as a tenor/bass unison. THE CYBER HYMNAL WAS NOT CONSULTED AT ALL. Collated for TEXT only (not for notes) against the words-only printing in Gospel Hymns Consolidated, Nos. 1 to 4 (1883), no. 25 \u2014 archive.org hymnsconso00blis, which prints the hymn without music and refers the singer to the tune on its page 26 \u2014 and against Tabernacle Hymns No. 2 (Chicago, 1921), no. 330, archive.org tabernaclehymnsn02chic leaf n317, which prints a different lower-staff harmonization of the same G-major melody, and is used here only as the dated authority that names the composer 'J. J. Husband.'", "pd_jurisdictions": {"US": true}, "encoding_origin": "Hand-encoded from the 1887 Hymns New and Old engraving, archive.org woldforus00town leaf n176, against staff lines re-fitted locally at every notehead, and read a second time end to end from the independent 1887 scan archive.org hymnsnewoldforus00town leaf n174. No MIDI, no Cyber Hymnal asset and no modern edition was used at any stage.", "encoding_license": "PD-dedication"}, "music_ly": "%% REVIVE US AGAIN (11.11 with Refrain) \u2014 four SATB voice streams, key G major.\n%% Corpus harmonization source for the HymnBookGenerator close-score engine.\n%% RULES (CI-enforced): exactly the variables soprano/alto/tenor/bass/global;\n%% absolute octaves; each voice rhythmically complete; NO chords <>, NO <<>>,\n%% NO \\secondShift/\\unisonWholeShift/tightSpacing, NO \\break, NO \\bar, NO lyrics.\n%%\n%% SOURCE OF THE NOTES \u2014 ONE DATED PLATE, READ TWICE FROM TWO SEPARATE SCANS.\n%% Hymns New and Old, for Use in Gospel Meetings and Other Religious Services,\n%% comp. D. B. Towner (Chicago: Fleming H. Revell, 1887), no. 168, headed\n%% \"Revive us Again.\", with the motto \"O Lord, revive Thy work.\"\u2014Hab. 3: 2,\n%% credited \"Rev. WM. PATON MACKAY, 1866.\" at the left and \"ENGLISH MELODY.\" at\n%% the right. PRIMARY WITNESS: archive.org item woldforus00town, leaf n176\n%% (2420x3765 px; staff space 29.9 px, so one diatonic step = 14.95 px).\n%% SECOND WITNESS: the OTHER 1887 scan of the same plate, archive.org item\n%% hymnsnewoldforus00town, leaf n174 (1785x2934 px; staff space 23.6 px), read\n%% end to end by the same rig with its OWN binarisation and its OWN staff-line\n%% fits. No scaffold of any kind was used \u2014 no MIDI, no Cyber Hymnal asset, no\n%% modern edition. All 50 chord positions x 4 voices agree between the two scans.\n%%\n%% HOW IT WAS MEASURED. Per notehead: the five staff lines were re-found LOCALLY\n%% in a +-45 px window around that notehead's own x (the comb fit pinned to\n%% within 0.62 staff space of the globally located staff, after an unpinned fit\n%% slipped a whole space at a system edge and mis-read the last chord of system\n%% 1); the paper level was taken per column window; heads were isolated by\n%% hole-filling plus stem-masked centroids under an elliptical matched filter;\n%% each centroid was converted to a diatonic step offset from the bottom line and\n%% accepted only within ~0.15 step. Hollow (half-note) heads, which a filled-head\n%% filter cannot see, were read separately from their ring envelopes and every\n%% one of them was ALSO looked at by eye at 6-8x on both scans.\n%%\n%% NO ACCIDENTALS ANYWHERE. A notehead probe reads staff positions, not sharps\n%% and flats, so accidentals were checked SEPARATELY and by eye, at 2-8x, over\n%% every one of the six staves on BOTH leaves, and by an independent shape sweep\n%% for accidental-sized ink (1.45-3.2 staff spaces tall, 0.38-1.35 wide) across\n%% both scans. There is not one accidental in this hymn. The only sharps on\n%% either leaf are the single F-sharp key signatures at the head of each staff \u2014\n%% one sharp, G major, confirmed by eye on all six staves of both scans. The alto\n%% therefore spells its four F's as `fis'` (ordinals 21, 22, 23 and 37, the D\n%% chords on \"gone a-bove\" and \"a-men\"): the plate prints no accidental there, so\n%% the key signature governs and they are F-sharps. They are the ONLY F's in any\n%% voice; nothing else in the hymn leaves the G major scale.\n%%\n%% STRUCTURE. 3/4, one-quarter anacrusis, 16 bars. 50 soprano noteheads: 23 in\n%% the verse (bars 1-8 beat 1) and 27 in the refrain (bar 8 beat 3 to the end).\n%% Each poetic line is a one-beat pickup + three bars + a half note, so every\n%% phrase end falls on beat 1 of a bar and the refrain begins on beat 3 of bar 8.\n%%\n%% THE ONE SLUR, soprano notes 12-13 (bar 4, g'8 to b'). The plate draws a curve\n%% under the treble chord and a mirroring curve over the lower-staff chord there.\n%% Stanza 1 sings ONE syllable across it (\"For\"); stanzas 2-5 print two syllables\n%% under it in small type (\"Who has\" / \"Who has\" / \"Who has\" / \"May each\"), which\n%% is why this harmonization is verse_variable_melisma. Encoded on the soprano\n%% (the voice that moves) and on the tenor (whose repeated b is what the printed\n%% lower-staff curve joins). DEPARTURE, STATED: the plate's curve in the treble\n%% is a single stroke under the whole chord; here only the soprano carries it, so\n%% the alto's repeated d' prints unslurred.\n%%\n%% THE ONE SHARED NOTEHEAD, lower staff, soprano ordinal 46 (\"re-\" of \"revive\",\n%% bar 14 beat 3). The plate engraves ONE notehead there, on g; both tenor and\n%% bass are encoded on g and LilyPond's default collision resolver is left to\n%% kiss them. Confirmed as a single head on BOTH scans.\n%%\n%% THE BASS PEDAL IS THE PAGE'S. Through the whole refrain the bass sits on g\n%% (G3) \u2014 every chord from ordinal 24 to 46 except \"a-men\" (ordinal 37, d) \u2014 and\n%% only the last three notes leave it (c, d, d) before the final g,. Re-read on\n%% both scans.\n%%\n%% HIGH TENOR AND BASS ARE THE PAGE'S TOO. \"Je-sus who\" (ordinals 14-16) prints\n%% the lower staff as d' over b, both above the bass staff with a c' ledger line\n%% drawn between them; \"For\" (ordinal 6, \"for the\") likewise prints d' over b.\n%% Every one of those ledgered chords was re-read at 7x on both scans.\n\\version \"2.24.0\"\n\nglobal  = { \\key g \\major \\time 3/4 \\partial 4 }\n\nsoprano = {\n  d'4 |\n  g'4 g'4 g'4 |\n  g'2 g'8 a'8 |\n  b'4 b'4 b'4 |\n  b'2 g'8( b'8) |\n  d''4 d''4 d''4 |\n  e''4 d''4 b'4 |\n  b'4 a'4 a'4 |\n  a'2 d''8. d''16 |\n  e''4 d''4 b'8. a'16 |\n  g'4 g'4 d''8. d''16 |\n  e''4 d''4 b'4 |\n  a'2 d''8. d''16 |\n  e''4 d''4 b'8. a'16 |\n  g'4 g'4 b'4 |\n  c''4 b'4 a'4 |\n  g'2\n}\n\nalto = {\n  b4 |\n  d'4 d'4 d'4 |\n  d'2 d'8 d'8 |\n  d'4 d'4 d'4 |\n  d'2 d'8 d'8 |\n  g'4 g'4 g'4 |\n  g'4 g'4 d'4 |\n  g'4 fis'4 fis'4 |\n  fis'2 g'8. g'16 |\n  g'4 g'4 d'8. d'16 |\n  d'4 d'4 g'8. g'16 |\n  g'4 g'4 g'4 |\n  fis'2 g'8. g'16 |\n  g'4 g'4 d'8. d'16 |\n  d'4 d'4 d'4 |\n  e'4 d'4 d'4 |\n  d'2\n}\n\ntenor = {\n  b4 |\n  b4 b4 b4 |\n  b2 d'8 c'8 |\n  b4 g4 g4 |\n  g2 b8( b8) |\n  d'4 d'4 d'4 |\n  c'4 b4 b4 |\n  d'4 d'4 d'4 |\n  d'2 b8. b16 |\n  c'4 b4 d'8. c'16 |\n  b4 b4 b8. b16 |\n  c'4 b4 d'4 |\n  d'2 b8. b16 |\n  c'4 b4 d'8. c'16 |\n  b4 b4 g4 |\n  g4 g4 c'4 |\n  b2\n}\n\nbass = {\n  g,4 |\n  g,4 b,4 d4 |\n  g2 b8 a8 |\n  g4 d4 b,4 |\n  g,2 g8 g8 |\n  b4 b4 b4 |\n  g4 g4 g4 |\n  d4 d4 d4 |\n  d2 g8. g16 |\n  g4 g4 g8. g16 |\n  g4 g4 g8. g16 |\n  g4 g4 g4 |\n  d2 g8. g16 |\n  g4 g4 g8. g16 |\n  g4 g4 g4 |\n  c4 d4 d4 |\n  g,2\n}\n"}, "harmonizations": [{"slug": "revive-us-again-1887", "tune": "revive-us-again", "music_file": "revive-us-again-1887.ly", "harmonizer": null, "arrangement_type": "harmonization", "canonical_key": {"tonic": "g", "mode": "major"}, "time": "3/4", "pickup": "4", "system_breaks": [11, 23, 37], "breaks_manual": true, "markings": [{"soprano_note": 24, "kind": "refrain"}], "final_barline": "|.", "amen": false, "verse_variable_melisma": true, "refrain": {"soprano_note_start": 24}, "source_line": "Setting: Hymns New and Old, 1887, where the tune is credited only \"English Melody.\" Public domain.", "provenance": {"layer": "harmonization", "source": "hymns-new-and-old-1887", "source_kind": "hand_encoded", "pd_basis": "pub_pre_1931", "pd_evidence": "The four-part setting printed in Hymns New and Old, for Use in Gospel Meetings and Other Religious Services, comp. D. B. Towner (Chicago: Fleming H. Revell, 1887), no. 168, headed 'Revive us Again.' under the motto '\"O Lord, revive Thy work.\"\u2014Hab. 3: 2', credited 'Rev. WM. PATON MACKAY, 1866.' at the left and 'ENGLISH MELODY.' at the right. G major, 3/4, one quarter anacrusis, 16 bars, no repeats, no accidentals, no amen. US public domain by publication before 1929, and no modern arranger stands anywhere in the chain: the plate names no harmonizer at all and the book itself is 1887. NO SCAFFOLD WAS USED. There is no MIDI, no NoteWorthy file and no modern edition behind this encoding; every chord was read off the page image. PRIMARY WITNESS: archive.org item woldforus00town, leaf n176 \u2014 2420x3765 px, staff space measured at 29.9 px, so one diatonic step = 14.95 px. SECOND, INDEPENDENT WITNESS: the other 1887 scan of the same plate, archive.org item hymnsnewoldforus00town, leaf n174 \u2014 1785x2934 px, staff space 23.6 px \u2014 read end to end by the same rig with its own binarisation and its own staff-line fits. (That second item's stored JP2 was fetched too and is the same 1785x2934 raster as its page render, so 23.6 px really is its native staff space; the higher-resolution copy was therefore made the primary reading and the lower-resolution copy the check. Both are dated 1887 printings of the same plates.) ALL 50 CHORD POSITIONS x 4 VOICES WERE READ ON BOTH SCANS AND AGREE \u2014 the matched filter found 45 of the 50 on the second scan and the remaining five, plus the four the filter merged at system edges, were probed by hand from their ink envelopes there. That set includes the pickup's alto b below its C4 ledger, the five hollow half-note chords, the ledgered d'-over-b chords of 'Je-sus who', and the single shared notehead at ordinal 46. HOW THE NOTEHEADS WERE MEASURED: for each head the five staff lines were re-found LOCALLY in a +-45 px window around that head's own x, with the five-line comb pinned to within 0.62 staff space of the globally located staff \u2014 an unpinned fit slipped a whole space near a system edge and mis-read the last chord of system 1 as b'/e' when the plate prints d''/g', which is exactly why the reading of every chord was also confirmed by eye at 3-8x. Paper level was taken per column window; heads were isolated by hole-filling and stem-masked centroids under an elliptical matched filter; each centroid became a diatonic step offset from the bottom line and was accepted only within ~0.15 step. Hollow half-note heads are invisible to a filled-head filter, so all five of them (ordinals 5, 11, 23, 37, 50) were read from their ring envelopes and looked at by eye at 6-8x on both scans. ACCIDENTALS WERE CHECKED SEPARATELY, BY EYE: a notehead probe reads staff positions and not sharps or flats, so the automated figure above covers no accidental at all. Every one of the six staves on BOTH leaves was examined at 2-8x, and an independent shape sweep for accidental-sized ink (1.45-3.2 staff spaces tall, 0.38-1.35 wide) was run over both scans; the only components it returned were clefs, time signatures and stems. THERE IS NOT ONE ACCIDENTAL IN THIS HYMN, and the only sharps on either leaf are the single F-sharp key signatures at the head of each staff, counted by eye on all six. THE FINGERPRINTS OF THIS READING, all visible at high magnification on both scans: the anacrusis is a four-note chord g,/b/b/d' with alto and tenor in unison on b, the alto's head sitting below a drawn C4 ledger; bar 2 sets 'God!' as a half note and 'for the' as two plain flagged eighths (no dots) while the lower staff jumps to d' over b with a c' ledger between them; 'Je-sus who' is three quarters of d''/g'/d'/b, each lower-staff chord ledgered at c'; the verse does NOT cadence but ends on D major ('a-bove' = a'/fis'/d'/d), running straight into the chorus; the chorus's four 'Hal-le' figures are a dotted eighth plus a sixteenth (two flags on the second, dots on both heads of the first) and the bass holds a g pedal under the entire chorus except 'a-men' (ordinal 37, d) and the last three notes; and the lower staff prints ONE notehead, on g, at ordinal 46 ('re-'), which is encoded as a tenor/bass unison. THE CYBER HYMNAL WAS NOT CONSULTED AT ALL. Collated for TEXT only (not for notes) against the words-only printing in Gospel Hymns Consolidated, Nos. 1 to 4 (1883), no. 25 \u2014 archive.org hymnsconso00blis, which prints the hymn without music and refers the singer to the tune on its page 26 \u2014 and against Tabernacle Hymns No. 2 (Chicago, 1921), no. 330, archive.org tabernaclehymnsn02chic leaf n317, which prints a different lower-staff harmonization of the same G-major melody, and is used here only as the dated authority that names the composer 'J. J. Husband.'", "pd_jurisdictions": {"US": true}, "encoding_origin": "Hand-encoded from the 1887 Hymns New and Old engraving, archive.org woldforus00town leaf n176, against staff lines re-fitted locally at every notehead, and read a second time end to end from the independent 1887 scan archive.org hymnsnewoldforus00town leaf n174. No MIDI, no Cyber Hymnal asset and no modern edition was used at any stage.", "encoding_license": "PD-dedication"}, "music_ly": "%% REVIVE US AGAIN (11.11 with Refrain) \u2014 four SATB voice streams, key G major.\n%% Corpus harmonization source for the HymnBookGenerator close-score engine.\n%% RULES (CI-enforced): exactly the variables soprano/alto/tenor/bass/global;\n%% absolute octaves; each voice rhythmically complete; NO chords <>, NO <<>>,\n%% NO \\secondShift/\\unisonWholeShift/tightSpacing, NO \\break, NO \\bar, NO lyrics.\n%%\n%% SOURCE OF THE NOTES \u2014 ONE DATED PLATE, READ TWICE FROM TWO SEPARATE SCANS.\n%% Hymns New and Old, for Use in Gospel Meetings and Other Religious Services,\n%% comp. D. B. Towner (Chicago: Fleming H. Revell, 1887), no. 168, headed\n%% \"Revive us Again.\", with the motto \"O Lord, revive Thy work.\"\u2014Hab. 3: 2,\n%% credited \"Rev. WM. PATON MACKAY, 1866.\" at the left and \"ENGLISH MELODY.\" at\n%% the right. PRIMARY WITNESS: archive.org item woldforus00town, leaf n176\n%% (2420x3765 px; staff space 29.9 px, so one diatonic step = 14.95 px).\n%% SECOND WITNESS: the OTHER 1887 scan of the same plate, archive.org item\n%% hymnsnewoldforus00town, leaf n174 (1785x2934 px; staff space 23.6 px), read\n%% end to end by the same rig with its OWN binarisation and its OWN staff-line\n%% fits. No scaffold of any kind was used \u2014 no MIDI, no Cyber Hymnal asset, no\n%% modern edition. All 50 chord positions x 4 voices agree between the two scans.\n%%\n%% HOW IT WAS MEASURED. Per notehead: the five staff lines were re-found LOCALLY\n%% in a +-45 px window around that notehead's own x (the comb fit pinned to\n%% within 0.62 staff space of the globally located staff, after an unpinned fit\n%% slipped a whole space at a system edge and mis-read the last chord of system\n%% 1); the paper level was taken per column window; heads were isolated by\n%% hole-filling plus stem-masked centroids under an elliptical matched filter;\n%% each centroid was converted to a diatonic step offset from the bottom line and\n%% accepted only within ~0.15 step. Hollow (half-note) heads, which a filled-head\n%% filter cannot see, were read separately from their ring envelopes and every\n%% one of them was ALSO looked at by eye at 6-8x on both scans.\n%%\n%% NO ACCIDENTALS ANYWHERE. A notehead probe reads staff positions, not sharps\n%% and flats, so accidentals were checked SEPARATELY and by eye, at 2-8x, over\n%% every one of the six staves on BOTH leaves, and by an independent shape sweep\n%% for accidental-sized ink (1.45-3.2 staff spaces tall, 0.38-1.35 wide) across\n%% both scans. There is not one accidental in this hymn. The only sharps on\n%% either leaf are the single F-sharp key signatures at the head of each staff \u2014\n%% one sharp, G major, confirmed by eye on all six staves of both scans. The alto\n%% therefore spells its four F's as `fis'` (ordinals 21, 22, 23 and 37, the D\n%% chords on \"gone a-bove\" and \"a-men\"): the plate prints no accidental there, so\n%% the key signature governs and they are F-sharps. They are the ONLY F's in any\n%% voice; nothing else in the hymn leaves the G major scale.\n%%\n%% STRUCTURE. 3/4, one-quarter anacrusis, 16 bars. 50 soprano noteheads: 23 in\n%% the verse (bars 1-8 beat 1) and 27 in the refrain (bar 8 beat 3 to the end).\n%% Each poetic line is a one-beat pickup + three bars + a half note, so every\n%% phrase end falls on beat 1 of a bar and the refrain begins on beat 3 of bar 8.\n%%\n%% THE ONE SLUR, soprano notes 12-13 (bar 4, g'8 to b'). The plate draws a curve\n%% under the treble chord and a mirroring curve over the lower-staff chord there.\n%% Stanza 1 sings ONE syllable across it (\"For\"); stanzas 2-5 print two syllables\n%% under it in small type (\"Who has\" / \"Who has\" / \"Who has\" / \"May each\"), which\n%% is why this harmonization is verse_variable_melisma. Encoded on the soprano\n%% (the voice that moves) and on the tenor (whose repeated b is what the printed\n%% lower-staff curve joins). DEPARTURE, STATED: the plate's curve in the treble\n%% is a single stroke under the whole chord; here only the soprano carries it, so\n%% the alto's repeated d' prints unslurred.\n%%\n%% THE ONE SHARED NOTEHEAD, lower staff, soprano ordinal 46 (\"re-\" of \"revive\",\n%% bar 14 beat 3). The plate engraves ONE notehead there, on g; both tenor and\n%% bass are encoded on g and LilyPond's default collision resolver is left to\n%% kiss them. Confirmed as a single head on BOTH scans.\n%%\n%% THE BASS PEDAL IS THE PAGE'S. Through the whole refrain the bass sits on g\n%% (G3) \u2014 every chord from ordinal 24 to 46 except \"a-men\" (ordinal 37, d) \u2014 and\n%% only the last three notes leave it (c, d, d) before the final g,. Re-read on\n%% both scans.\n%%\n%% HIGH TENOR AND BASS ARE THE PAGE'S TOO. \"Je-sus who\" (ordinals 14-16) prints\n%% the lower staff as d' over b, both above the bass staff with a c' ledger line\n%% drawn between them; \"For\" (ordinal 6, \"for the\") likewise prints d' over b.\n%% Every one of those ledgered chords was re-read at 7x on both scans.\n\\version \"2.24.0\"\n\nglobal  = { \\key g \\major \\time 3/4 \\partial 4 }\n\nsoprano = {\n  d'4 |\n  g'4 g'4 g'4 |\n  g'2 g'8 a'8 |\n  b'4 b'4 b'4 |\n  b'2 g'8( b'8) |\n  d''4 d''4 d''4 |\n  e''4 d''4 b'4 |\n  b'4 a'4 a'4 |\n  a'2 d''8. d''16 |\n  e''4 d''4 b'8. a'16 |\n  g'4 g'4 d''8. d''16 |\n  e''4 d''4 b'4 |\n  a'2 d''8. d''16 |\n  e''4 d''4 b'8. a'16 |\n  g'4 g'4 b'4 |\n  c''4 b'4 a'4 |\n  g'2\n}\n\nalto = {\n  b4 |\n  d'4 d'4 d'4 |\n  d'2 d'8 d'8 |\n  d'4 d'4 d'4 |\n  d'2 d'8 d'8 |\n  g'4 g'4 g'4 |\n  g'4 g'4 d'4 |\n  g'4 fis'4 fis'4 |\n  fis'2 g'8. g'16 |\n  g'4 g'4 d'8. d'16 |\n  d'4 d'4 g'8. g'16 |\n  g'4 g'4 g'4 |\n  fis'2 g'8. g'16 |\n  g'4 g'4 d'8. d'16 |\n  d'4 d'4 d'4 |\n  e'4 d'4 d'4 |\n  d'2\n}\n\ntenor = {\n  b4 |\n  b4 b4 b4 |\n  b2 d'8 c'8 |\n  b4 g4 g4 |\n  g2 b8( b8) |\n  d'4 d'4 d'4 |\n  c'4 b4 b4 |\n  d'4 d'4 d'4 |\n  d'2 b8. b16 |\n  c'4 b4 d'8. c'16 |\n  b4 b4 b8. b16 |\n  c'4 b4 d'4 |\n  d'2 b8. b16 |\n  c'4 b4 d'8. c'16 |\n  b4 b4 g4 |\n  g4 g4 c'4 |\n  b2\n}\n\nbass = {\n  g,4 |\n  g,4 b,4 d4 |\n  g2 b8 a8 |\n  g4 d4 b,4 |\n  g,2 g8 g8 |\n  b4 b4 b4 |\n  g4 g4 g4 |\n  d4 d4 d4 |\n  d2 g8. g16 |\n  g4 g4 g8. g16 |\n  g4 g4 g8. g16 |\n  g4 g4 g4 |\n  d2 g8. g16 |\n  g4 g4 g8. g16 |\n  g4 g4 g4 |\n  c4 d4 d4 |\n  g,2\n}\n", "preferred": true}], "license": "The hymn texts and tunes are in the public domain; per-layer provenance is recorded on each record."}