{"slug": "higher-ground", "name": "HIGHER GROUND", "meter_label": "L.M. with Refrain", "meter": {"pattern": [8, 8, 8, 8], "doubled": false, "refrain": true, "alleluia": false, "irregular": false}, "composer": "charles-hutchinson-gabriel", "composer_date": "1898", "canonical_key": {"tonic": "aes", "mode": "major"}, "preferred_harmonization": "higher-ground-hcl-1904", "provenance": {"layer": "tune", "source": "hymns-of-the-christian-life-3-1904", "source_kind": "hand_encoded", "pd_basis": "pub_pre_1931", "pd_evidence": "Charles H. Gabriel's tune for Oatman's text, first copyrighted by J. Howard Entwisle in 1898 (so stated on the 1912 and 1921 printings; the 1904 printing read for this record says 1899) and published from 1898 onward. Read here from Hymns of the Christian Life No. 3, 1904, no. 200 (archive.org hymnsofchristi1904simp leaf 213), A-flat major, 3/4, dotted-quarter anacrusis; the same reading is in Great Songs of the Church, 1921, no. 127 (archive.org greatsonchur00jorg leaf 94) and Make Christ King, 1912, no. 336 (archive.org makechristkingse00exce leaf 246). PUBLIC DOMAIN ON THE PRINTING'S DATE, not on the composer's life: Gabriel died in 1932, so his death date clears nothing; what clears the tune is first publication before 1931, which three dated pre-1931 printings establish. Identity check on the melody, made only after the notes were read off the plate: the soprano's first fifteen notes measure as scale degrees 551332162165131 in A-flat, matching Hymnary's published incipit for tune authority higher_ground_gabriel digit for digit.", "pd_jurisdictions": {"US": true}, "hymnary_ids": {"tune_auth": "higher_ground_gabriel"}}, "harmonization": {"slug": "higher-ground-hcl-1904", "tune": "higher-ground", "music_file": "higher-ground-hcl-1904.ly", "canonical_key": {"tonic": "aes", "mode": "major"}, "time": "3/4", "pickup": "4.", "system_breaks": [16, 32, 48], "breaks_manual": true, "final_barline": "|.", "amen": false, "verse_variable_melisma": false, "refrain": {"soprano_note_start": 33}, "markings": [{"soprano_note": 33, "kind": "refrain"}], "provenance": {"layer": "harmonization", "source": "hymns-of-the-christian-life-3-1904", "source_kind": "hand_encoded", "pd_basis": "pub_pre_1931", "pd_evidence": "Gabriel's own four-part setting, read notehead by notehead off the engraving in Hymns of the Christian Life No. 3 (New York/Nyack: Christian Alliance Publishing Co., 1904), no. 200, printed page 200 \u2014 archive.org hymnsofchristi1904simp, scan leaf 213 \u2014 whose plate notice reads \"Copyright, 1899, by J. Howard Entwisle. Property of John J. Hood. Used by per.\" The licensing rests on that pre-1931 PUBLICATION, not on the composer's life: Charles H. Gabriel did not die until 1932, so his death date clears nothing. Every chord was measured, not eyeballed: the five staff lines were least-squares fitted per staff from 717-1016 ink-free columns (so scan skew is followed), staff lines and stems were stripped morphologically, and each notehead centroid was converted to a diatonic step offset from the fitted bottom line at its own x. This plate prints space-noteheads off the geometric centre of the space by a constant amount per staff (treble +0.227..+0.258 step, bass -0.246..-0.278 step, line-noteheads within 0.03), so a per-staff per-parity median offset was subtracted; after that, 249 of the 251 noteheads land within 0.16 step of an exact staff position (largest 0.158, only three over 0.12). The two exceptions are the two chords whose noteheads are a SECOND apart and whose blobs merge \u2014 the refrain's sixth chord in the lower staff and the second phrase's first anacrusis chord in the upper staff \u2014 both read by hand at 10x with the fitted grid drawn over them. The whole page was then measured a second time, by the same method, in an independent dated printing: Great Songs of the Church (Louisville: Word and Work, 1921), no. 127, archive.org greatsonchur00jorg leaf 94, headed \"Copyright, 1898, by J. Howard Entwisle. John J. Hood, owner. Chas. H. Gabriel.\" The two agree on every chord of all four systems, including both hand-read seconds and all five printed unisons; where the 1921 scan was too blurred for the detector (the tail chords of four staves) those chords were read off 5x-10x crops with the grid drawn, and they agree too. A third pre-1931 witness, Make Christ King (Chicago: Glad Tidings Co., 1912), no. 336, archive.org makechristkingse00exce leaf 246, prints the same key, metre and anacrusis but abbreviates the last phrase as a D.S., so it corroborates without being transcribed. No Cyber Hymnal file, MIDI, existing encoding or modern hymnal was used at any stage, as a scaffold or otherwise. Departures from the page, all presentational: the page's \"CHORUS.\" is printed as the house marking \"Refrain\" at the same soprano note; no amen is added; the brace and staff spacing are the engine's. Nothing in the pitches or the rhythm departs from the 1904 engraving.", "pd_jurisdictions": {"US": true}, "encoding_origin": "Hand-encoded from the 1904 page image (archive.org hymnsofchristi1904simp leaf 213) by measuring notehead centroids against a locally fitted staff grid; cross-measured against the 1921 printing (greatsonchur00jorg leaf 94)", "encoding_license": "PD-dedication", "hymnary_ids": {"tune_auth": "higher_ground_gabriel"}}, "music_ly": "%% HIGHER GROUND (L.M. with Refrain) \u2014 \"I'm pressing on the upward way\"\n%% Four SATB voice streams, canonical key A-flat major, 3/4, dotted-quarter\n%% anacrusis, 16 bars (4 four-bar phrases).\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 the page, and only the page. Every notehead here was\n%% read off the four-part engraving in HYMNS OF THE CHRISTIAN LIFE No. 3 (New\n%% York / Nyack: Christian Alliance Publishing Co., 1904), no. 200, printed page\n%% 200, archive.org identifier hymnsofchristi1904simp, scan leaf 213\n%% (hymnsofchristi1904simp_0213). That page carries the plate's own notice,\n%% \"Copyright, 1899, by J. Howard Entwisle. Property of John J. Hood. Used by\n%% per.\" No Cyber Hymnal asset, no MIDI, no existing encoding and no modern\n%% hymnal was used at any point, as a scaffold or otherwise.\n%%\n%% HOW THE NOTEHEADS WERE MEASURED (not eyeballed). For each of the eight staves\n%% the five staff lines were fitted LOCALLY: columns carrying nothing but staff-\n%% line ink were found by ink-count, the five line centres were read in each such\n%% column, and a quadratic y_i(x) was least-squares fitted per line (717 to 1016\n%% clean columns per staff), so the scan's skew and warp are followed rather than\n%% assumed away. Staff lines and stems were then removed morphologically\n%% (vertical opening at 2.2x line thickness, horizontal opening at 0.55x the line\n%% spacing) and the surviving notehead blobs kept by size and fill; each blob's\n%% centroid was converted to a diatonic step offset from the fitted BOTTOM line\n%% at that blob's own x. Line spacing on this scan is about 23.3 px, so one\n%% diatonic step is about 11.6 px and 0.15 step is 1.7 px.\n%%\n%% ONE CALIBRATION, STATED PLAINLY. As measured, a notehead standing in a SPACE\n%% does not centre on the geometric middle of the space: in the four treble\n%% staves it measures 0.227 to 0.258 step high, in the four bass staves 0.246 to\n%% 0.278 step low, while noteheads ON a line measure within 0.03 step of the line\n%% (median line offsets -0.030 to +0.026). The bias is constant per staff and per\n%% parity, so a per-staff, per-parity median offset was subtracted and the\n%% positions re-rounded. AFTER that calibration, 249 of the 251 noteheads the\n%% detector found on the page land within 0.16 step of an exact staff position;\n%% the largest of those 249 residuals is 0.158 step and only three exceed 0.12 \u2014\n%% far inside the half-step that would be needed to change a pitch. The two\n%% exceptions are the two second-interval chords described below; both were\n%% resolved by hand and against the 1921 printing.\n%%\n%% SECOND DATED WITNESS. The whole page was measured again, by the same method,\n%% in GREAT SONGS OF THE CHURCH (Louisville: Word and Work, 1921), no. 127,\n%% archive.org greatsonchur00jorg, scan leaf 94, whose header reads \"Copyright,\n%% 1898, by J. Howard Entwisle. John J. Hood, owner. Chas. H. Gabriel.\" The two\n%% printings are the same setting and they agree on every chord of all four\n%% systems: same key (A-flat), same 3/4, same dotted-quarter anacrusis, same\n%% 3+4+4+4+1 note layout per system, same voicing throughout. Where the 1921\n%% scan is too blurred for the blob detector (the tail chords of four staves, and\n%% one flag that the detector mistook for a low notehead) the chords were read\n%% instead off 5x-10x crops with the fitted grid drawn over them; they agree.\n%% A third pre-1931 printing, MAKE CHRIST KING (Chicago: Glad Tidings Co., 1912),\n%% no. 336, archive.org makechristkingse00exce leaf 246, prints the same key,\n%% metre and anacrusis but abbreviates the last phrase as a D.S., so it was used\n%% only as a corroborating witness and not transcribed.\n%%\n%% FOURTH, WHOLLY INDEPENDENT CHECK ON THE MELODY. Hymnary's published incipit\n%% for tune authority higher_ground_gabriel is 551332162165131. The soprano's\n%% first fifteen notes as measured here \u2014 ees' ees' aes' c'' c'' bes' aes' f'\n%% bes' aes' f' ees' aes' c'' aes' \u2014 are scale degrees 5 5 1 3 3 2 1 6 2 1 6 5\n%% 1 3 1 in A-flat. All fifteen digits match, and that string was never used to\n%% derive anything: it was compared only after the notes were read off the plate.\n%%\n%% TWO PLACES WHERE THE AUTOMATIC READING WAS OVERRIDDEN, both because two\n%% noteheads a SECOND apart share a stem and their blobs merge:\n%%   * refrain, lower staff, sixth chord (the word \"let\"): the merged blob\n%%     measured 9.38, i.e. residual 0.62 against position 10 and 0.38 against\n%%     position 9 \u2014 the automatic rounding could not be trusted either way. Read\n%%     at 10x with the fitted grid drawn over it, the upper notehead plainly\n%%     stands in the space above the top line and the lower sits ON the top line:\n%%     bes over aes. The 1921 printing's blobs merge in the same way (residual\n%%     0.45 there) but round to 9 over 8, agreeing. Encoded bes + aes.\n%%   * verse second half, upper staff, first anacrusis chord (\"Still\"): the alto\n%%     blob measured residual 0.40 for the same reason. At 9x the soprano sits on\n%%     the bottom line and the alto in the space below it, displaced to the right\n%%     of the shared stem: ees' over des' \u2014 the seventh of the E-flat seventh\n%%     that opens the phrase. The 1921 printing shows the identical pair (its own\n%%     residual there 0.44, rounding the same way).\n%%\n%% UNISONS, PRINTED AS SUCH. Five chords print ONE notehead carrying two stems,\n%% i.e. the two voices of that staff in unison \u2014 which is why the detector finds\n%% 251 noteheads where 64 four-part chords would otherwise give 256. They are:\n%% soprano = alto on ees' at refrain note 9 (\"By\"), and tenor = bass on aes at\n%% refrain notes 1, 2, 3 and 7 (\"Lord, lift me ... me\"). Both printings agree.\n%% They are written out here as two voices on the same pitch, which is what they\n%% mean; LilyPond's default resolver prints the pair as one notehead again.\n%%\n%% NOTHING ELSE ON THE PAGE. There is not one accidental outside the key\n%% signature anywhere in the piece (checked staff by staff at 1.7x or closer on\n%% the 1904 page, and confirmed by the harmony, which is entirely diatonic in\n%% A-flat: A-flat, D-flat, B-flat minor seventh and E-flat seventh), not\n%% one slur or tie, and not one rest. All four voices strike together on all 64\n%% chords \u2014 this setting really is homorhythmic from first note to last, and the\n%% duration of every chord was read from the printed stems: three eighths for the\n%% anacrusis, then a dotted quarter and three eighths in every full bar, then a\n%% lone dotted quarter to close each phrase (which the next phrase's anacrusis\n%% completes). The final chord of the piece is a filled notehead with a dot, i.e.\n%% a dotted quarter, under a light-heavy final barline.\n%%\n%% DEPARTURES FROM THE PAGE (there are three, all presentational):\n%%   1. The 1904 page prints \"CHORUS.\" over the third system; the corpus prints\n%%      its house marking \"Refrain\" at the same soprano note (33).\n%%   2. The page has no amen and none is added (amen: false).\n%%   3. The engraved system breaks are reproduced exactly (after soprano notes\n%%      16, 32 and 48), but the page's brace/bracket and its two-staff spacing are\n%%      the engine's, not the plate's.\n%% Nothing in the pitches or the rhythm departs from the 1904 engraving.\n\\version \"2.24.0\"\n\nglobal  = { \\key aes \\major \\time 3/4 \\partial 4. }\n\nsoprano = {\n  ees'8 ees'8 aes'8 | c''4. c''8 bes'8 aes'8 | f'4. bes'8 aes'8 f'8 | ees'4. aes'8 c''8 aes'8 |\n  bes'4. ees'8 ees'8 aes'8 | c''4. c''8 bes'8 aes'8 | f'4. bes'8 aes'8 f'8 | ees'4. aes'8 c''8 bes'8 |\n  aes'4. ees'8 aes'8 c''8 | ees''4. ees''8 des''8 c''8 | des''4. ees'8 g'8 bes'8 | des''4. des''8 c''8 bes'8 |\n  c''4. ees''8 ees''8 c''8 | aes'4. c''8 bes'8 aes'8 | f'4. bes'8 aes'8 f'8 | ees'4. aes'8 c''8 bes'8 |\n  aes'4. |\n}\n\nalto = {\n  c'8 c'8 c'8 | ees'4. ees'8 ees'8 ees'8 | des'4. f'8 f'8 des'8 | c'4. c'8 ees'8 ees'8 |\n  ees'4. des'8 c'8 c'8 | ees'4. ees'8 ees'8 ees'8 | des'4. f'8 f'8 des'8 | c'4. c'8 ees'8 des'8 |\n  c'4. c'8 c'8 ees'8 | ees'4. ees'8 ees'8 ees'8 | ees'4. ees'8 ees'8 g'8 | bes'4. bes'8 aes'8 g'8 |\n  aes'4. ees'8 ees'8 ees'8 | ees'4. ees'8 ees'8 ees'8 | des'4. f'8 f'8 des'8 | c'4. c'8 ees'8 des'8 |\n  c'4. |\n}\n\ntenor = {\n  aes8 aes8 aes8 | aes4. aes8 g8 aes8 | aes4. aes8 aes8 aes8 | aes4. aes8 aes8 aes8 |\n  g4. g8 aes8 aes8 | aes4. aes8 g8 aes8 | aes4. aes8 aes8 aes8 | aes4. aes8 aes8 g8 |\n  aes4. aes8 aes8 aes8 | c'4. c'8 bes8 aes8 | bes4. g8 bes8 ees'8 | ees'4. ees'8 ees'8 ees'8 |\n  ees'4. c'8 c'8 ees'8 | c'4. aes8 g8 aes8 | aes4. aes8 aes8 aes8 | aes4. aes8 aes8 g8 |\n  aes4. |\n}\n\nbass = {\n  aes,8 aes,8 aes,8 | aes,4. aes,8 bes,8 c8 | des4. des8 des8 des8 | aes,4. aes,8 aes,8 c8 |\n  ees4. ees8 aes,8 aes,8 | aes,4. aes,8 bes,8 c8 | des4. des8 des8 des8 | ees4. ees8 ees8 ees8 |\n  aes,4. aes8 aes8 aes8 | aes4. aes8 aes8 aes8 | ees4. ees8 ees8 ees8 | ees4. ees8 ees8 ees8 |\n  aes4. aes8 aes8 aes8 | aes4. aes,8 bes,8 c8 | des4. des8 des8 des8 | ees4. ees8 ees8 ees8 |\n  aes,4. |\n}\n"}, "harmonizations": [{"slug": "higher-ground-hcl-1904", "tune": "higher-ground", "music_file": "higher-ground-hcl-1904.ly", "canonical_key": {"tonic": "aes", "mode": "major"}, "time": "3/4", "pickup": "4.", "system_breaks": [16, 32, 48], "breaks_manual": true, "final_barline": "|.", "amen": false, "verse_variable_melisma": false, "refrain": {"soprano_note_start": 33}, "markings": [{"soprano_note": 33, "kind": "refrain"}], "provenance": {"layer": "harmonization", "source": "hymns-of-the-christian-life-3-1904", "source_kind": "hand_encoded", "pd_basis": "pub_pre_1931", "pd_evidence": "Gabriel's own four-part setting, read notehead by notehead off the engraving in Hymns of the Christian Life No. 3 (New York/Nyack: Christian Alliance Publishing Co., 1904), no. 200, printed page 200 \u2014 archive.org hymnsofchristi1904simp, scan leaf 213 \u2014 whose plate notice reads \"Copyright, 1899, by J. Howard Entwisle. Property of John J. Hood. Used by per.\" The licensing rests on that pre-1931 PUBLICATION, not on the composer's life: Charles H. Gabriel did not die until 1932, so his death date clears nothing. Every chord was measured, not eyeballed: the five staff lines were least-squares fitted per staff from 717-1016 ink-free columns (so scan skew is followed), staff lines and stems were stripped morphologically, and each notehead centroid was converted to a diatonic step offset from the fitted bottom line at its own x. This plate prints space-noteheads off the geometric centre of the space by a constant amount per staff (treble +0.227..+0.258 step, bass -0.246..-0.278 step, line-noteheads within 0.03), so a per-staff per-parity median offset was subtracted; after that, 249 of the 251 noteheads land within 0.16 step of an exact staff position (largest 0.158, only three over 0.12). The two exceptions are the two chords whose noteheads are a SECOND apart and whose blobs merge \u2014 the refrain's sixth chord in the lower staff and the second phrase's first anacrusis chord in the upper staff \u2014 both read by hand at 10x with the fitted grid drawn over them. The whole page was then measured a second time, by the same method, in an independent dated printing: Great Songs of the Church (Louisville: Word and Work, 1921), no. 127, archive.org greatsonchur00jorg leaf 94, headed \"Copyright, 1898, by J. Howard Entwisle. John J. Hood, owner. Chas. H. Gabriel.\" The two agree on every chord of all four systems, including both hand-read seconds and all five printed unisons; where the 1921 scan was too blurred for the detector (the tail chords of four staves) those chords were read off 5x-10x crops with the grid drawn, and they agree too. A third pre-1931 witness, Make Christ King (Chicago: Glad Tidings Co., 1912), no. 336, archive.org makechristkingse00exce leaf 246, prints the same key, metre and anacrusis but abbreviates the last phrase as a D.S., so it corroborates without being transcribed. No Cyber Hymnal file, MIDI, existing encoding or modern hymnal was used at any stage, as a scaffold or otherwise. Departures from the page, all presentational: the page's \"CHORUS.\" is printed as the house marking \"Refrain\" at the same soprano note; no amen is added; the brace and staff spacing are the engine's. Nothing in the pitches or the rhythm departs from the 1904 engraving.", "pd_jurisdictions": {"US": true}, "encoding_origin": "Hand-encoded from the 1904 page image (archive.org hymnsofchristi1904simp leaf 213) by measuring notehead centroids against a locally fitted staff grid; cross-measured against the 1921 printing (greatsonchur00jorg leaf 94)", "encoding_license": "PD-dedication", "hymnary_ids": {"tune_auth": "higher_ground_gabriel"}}, "music_ly": "%% HIGHER GROUND (L.M. with Refrain) \u2014 \"I'm pressing on the upward way\"\n%% Four SATB voice streams, canonical key A-flat major, 3/4, dotted-quarter\n%% anacrusis, 16 bars (4 four-bar phrases).\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 the page, and only the page. Every notehead here was\n%% read off the four-part engraving in HYMNS OF THE CHRISTIAN LIFE No. 3 (New\n%% York / Nyack: Christian Alliance Publishing Co., 1904), no. 200, printed page\n%% 200, archive.org identifier hymnsofchristi1904simp, scan leaf 213\n%% (hymnsofchristi1904simp_0213). That page carries the plate's own notice,\n%% \"Copyright, 1899, by J. Howard Entwisle. Property of John J. Hood. Used by\n%% per.\" No Cyber Hymnal asset, no MIDI, no existing encoding and no modern\n%% hymnal was used at any point, as a scaffold or otherwise.\n%%\n%% HOW THE NOTEHEADS WERE MEASURED (not eyeballed). For each of the eight staves\n%% the five staff lines were fitted LOCALLY: columns carrying nothing but staff-\n%% line ink were found by ink-count, the five line centres were read in each such\n%% column, and a quadratic y_i(x) was least-squares fitted per line (717 to 1016\n%% clean columns per staff), so the scan's skew and warp are followed rather than\n%% assumed away. Staff lines and stems were then removed morphologically\n%% (vertical opening at 2.2x line thickness, horizontal opening at 0.55x the line\n%% spacing) and the surviving notehead blobs kept by size and fill; each blob's\n%% centroid was converted to a diatonic step offset from the fitted BOTTOM line\n%% at that blob's own x. Line spacing on this scan is about 23.3 px, so one\n%% diatonic step is about 11.6 px and 0.15 step is 1.7 px.\n%%\n%% ONE CALIBRATION, STATED PLAINLY. As measured, a notehead standing in a SPACE\n%% does not centre on the geometric middle of the space: in the four treble\n%% staves it measures 0.227 to 0.258 step high, in the four bass staves 0.246 to\n%% 0.278 step low, while noteheads ON a line measure within 0.03 step of the line\n%% (median line offsets -0.030 to +0.026). The bias is constant per staff and per\n%% parity, so a per-staff, per-parity median offset was subtracted and the\n%% positions re-rounded. AFTER that calibration, 249 of the 251 noteheads the\n%% detector found on the page land within 0.16 step of an exact staff position;\n%% the largest of those 249 residuals is 0.158 step and only three exceed 0.12 \u2014\n%% far inside the half-step that would be needed to change a pitch. The two\n%% exceptions are the two second-interval chords described below; both were\n%% resolved by hand and against the 1921 printing.\n%%\n%% SECOND DATED WITNESS. The whole page was measured again, by the same method,\n%% in GREAT SONGS OF THE CHURCH (Louisville: Word and Work, 1921), no. 127,\n%% archive.org greatsonchur00jorg, scan leaf 94, whose header reads \"Copyright,\n%% 1898, by J. Howard Entwisle. John J. Hood, owner. Chas. H. Gabriel.\" The two\n%% printings are the same setting and they agree on every chord of all four\n%% systems: same key (A-flat), same 3/4, same dotted-quarter anacrusis, same\n%% 3+4+4+4+1 note layout per system, same voicing throughout. Where the 1921\n%% scan is too blurred for the blob detector (the tail chords of four staves, and\n%% one flag that the detector mistook for a low notehead) the chords were read\n%% instead off 5x-10x crops with the fitted grid drawn over them; they agree.\n%% A third pre-1931 printing, MAKE CHRIST KING (Chicago: Glad Tidings Co., 1912),\n%% no. 336, archive.org makechristkingse00exce leaf 246, prints the same key,\n%% metre and anacrusis but abbreviates the last phrase as a D.S., so it was used\n%% only as a corroborating witness and not transcribed.\n%%\n%% FOURTH, WHOLLY INDEPENDENT CHECK ON THE MELODY. Hymnary's published incipit\n%% for tune authority higher_ground_gabriel is 551332162165131. The soprano's\n%% first fifteen notes as measured here \u2014 ees' ees' aes' c'' c'' bes' aes' f'\n%% bes' aes' f' ees' aes' c'' aes' \u2014 are scale degrees 5 5 1 3 3 2 1 6 2 1 6 5\n%% 1 3 1 in A-flat. All fifteen digits match, and that string was never used to\n%% derive anything: it was compared only after the notes were read off the plate.\n%%\n%% TWO PLACES WHERE THE AUTOMATIC READING WAS OVERRIDDEN, both because two\n%% noteheads a SECOND apart share a stem and their blobs merge:\n%%   * refrain, lower staff, sixth chord (the word \"let\"): the merged blob\n%%     measured 9.38, i.e. residual 0.62 against position 10 and 0.38 against\n%%     position 9 \u2014 the automatic rounding could not be trusted either way. Read\n%%     at 10x with the fitted grid drawn over it, the upper notehead plainly\n%%     stands in the space above the top line and the lower sits ON the top line:\n%%     bes over aes. The 1921 printing's blobs merge in the same way (residual\n%%     0.45 there) but round to 9 over 8, agreeing. Encoded bes + aes.\n%%   * verse second half, upper staff, first anacrusis chord (\"Still\"): the alto\n%%     blob measured residual 0.40 for the same reason. At 9x the soprano sits on\n%%     the bottom line and the alto in the space below it, displaced to the right\n%%     of the shared stem: ees' over des' \u2014 the seventh of the E-flat seventh\n%%     that opens the phrase. The 1921 printing shows the identical pair (its own\n%%     residual there 0.44, rounding the same way).\n%%\n%% UNISONS, PRINTED AS SUCH. Five chords print ONE notehead carrying two stems,\n%% i.e. the two voices of that staff in unison \u2014 which is why the detector finds\n%% 251 noteheads where 64 four-part chords would otherwise give 256. They are:\n%% soprano = alto on ees' at refrain note 9 (\"By\"), and tenor = bass on aes at\n%% refrain notes 1, 2, 3 and 7 (\"Lord, lift me ... me\"). Both printings agree.\n%% They are written out here as two voices on the same pitch, which is what they\n%% mean; LilyPond's default resolver prints the pair as one notehead again.\n%%\n%% NOTHING ELSE ON THE PAGE. There is not one accidental outside the key\n%% signature anywhere in the piece (checked staff by staff at 1.7x or closer on\n%% the 1904 page, and confirmed by the harmony, which is entirely diatonic in\n%% A-flat: A-flat, D-flat, B-flat minor seventh and E-flat seventh), not\n%% one slur or tie, and not one rest. All four voices strike together on all 64\n%% chords \u2014 this setting really is homorhythmic from first note to last, and the\n%% duration of every chord was read from the printed stems: three eighths for the\n%% anacrusis, then a dotted quarter and three eighths in every full bar, then a\n%% lone dotted quarter to close each phrase (which the next phrase's anacrusis\n%% completes). The final chord of the piece is a filled notehead with a dot, i.e.\n%% a dotted quarter, under a light-heavy final barline.\n%%\n%% DEPARTURES FROM THE PAGE (there are three, all presentational):\n%%   1. The 1904 page prints \"CHORUS.\" over the third system; the corpus prints\n%%      its house marking \"Refrain\" at the same soprano note (33).\n%%   2. The page has no amen and none is added (amen: false).\n%%   3. The engraved system breaks are reproduced exactly (after soprano notes\n%%      16, 32 and 48), but the page's brace/bracket and its two-staff spacing are\n%%      the engine's, not the plate's.\n%% Nothing in the pitches or the rhythm departs from the 1904 engraving.\n\\version \"2.24.0\"\n\nglobal  = { \\key aes \\major \\time 3/4 \\partial 4. }\n\nsoprano = {\n  ees'8 ees'8 aes'8 | c''4. c''8 bes'8 aes'8 | f'4. bes'8 aes'8 f'8 | ees'4. aes'8 c''8 aes'8 |\n  bes'4. ees'8 ees'8 aes'8 | c''4. c''8 bes'8 aes'8 | f'4. bes'8 aes'8 f'8 | ees'4. aes'8 c''8 bes'8 |\n  aes'4. ees'8 aes'8 c''8 | ees''4. ees''8 des''8 c''8 | des''4. ees'8 g'8 bes'8 | des''4. des''8 c''8 bes'8 |\n  c''4. ees''8 ees''8 c''8 | aes'4. c''8 bes'8 aes'8 | f'4. bes'8 aes'8 f'8 | ees'4. aes'8 c''8 bes'8 |\n  aes'4. |\n}\n\nalto = {\n  c'8 c'8 c'8 | ees'4. ees'8 ees'8 ees'8 | des'4. f'8 f'8 des'8 | c'4. c'8 ees'8 ees'8 |\n  ees'4. des'8 c'8 c'8 | ees'4. ees'8 ees'8 ees'8 | des'4. f'8 f'8 des'8 | c'4. c'8 ees'8 des'8 |\n  c'4. c'8 c'8 ees'8 | ees'4. ees'8 ees'8 ees'8 | ees'4. ees'8 ees'8 g'8 | bes'4. bes'8 aes'8 g'8 |\n  aes'4. ees'8 ees'8 ees'8 | ees'4. ees'8 ees'8 ees'8 | des'4. f'8 f'8 des'8 | c'4. c'8 ees'8 des'8 |\n  c'4. |\n}\n\ntenor = {\n  aes8 aes8 aes8 | aes4. aes8 g8 aes8 | aes4. aes8 aes8 aes8 | aes4. aes8 aes8 aes8 |\n  g4. g8 aes8 aes8 | aes4. aes8 g8 aes8 | aes4. aes8 aes8 aes8 | aes4. aes8 aes8 g8 |\n  aes4. aes8 aes8 aes8 | c'4. c'8 bes8 aes8 | bes4. g8 bes8 ees'8 | ees'4. ees'8 ees'8 ees'8 |\n  ees'4. c'8 c'8 ees'8 | c'4. aes8 g8 aes8 | aes4. aes8 aes8 aes8 | aes4. aes8 aes8 g8 |\n  aes4. |\n}\n\nbass = {\n  aes,8 aes,8 aes,8 | aes,4. aes,8 bes,8 c8 | des4. des8 des8 des8 | aes,4. aes,8 aes,8 c8 |\n  ees4. ees8 aes,8 aes,8 | aes,4. aes,8 bes,8 c8 | des4. des8 des8 des8 | ees4. ees8 ees8 ees8 |\n  aes,4. aes8 aes8 aes8 | aes4. aes8 aes8 aes8 | ees4. ees8 ees8 ees8 | ees4. ees8 ees8 ees8 |\n  aes4. aes8 aes8 aes8 | aes4. aes,8 bes,8 c8 | des4. des8 des8 des8 | ees4. ees8 ees8 ees8 |\n  aes,4. |\n}\n", "preferred": true}], "license": "The hymn texts and tunes are in the public domain; per-layer provenance is recorded on each record."}