27 rules

Beaming

How should notes be beamed together in this time signature?

Beams exist to make the beat visible. Almost every beaming rule follows from that one purpose: groups show the metric pulse, they stop at barlines and system breaks, and where a meter is genuinely ambiguous — 6/4, 5/8, additive meters — no default is correct and the grouping must be stated rather than guessed.

Breaks the readingCheckable from the score alone
A beam must never cross a notated barline
Beamed groups are always contained within a single measure; a beam that visually spans a barline is a fundamental engraving error in common-practice notation.
Breaks the readingCheckable from the score alone
A multi-note grace-note run renders as separate flagged notes instead of one beamed group
A run of two or more grace notes attached to a single main note (an appoggiatura pair, a slide, a turn-like figure) is conventionally beamed together as one small figure, not drawn as separate individually-flagged grace notes.
Breaks the readingCheckable from the score alone
A tuplet (and its beam) must not straddle an authored barline
Like an ordinary beam, a tuplet bracket and any beam under it must be fully contained within one notated measure.
Breaks the readingCheckable from the score alone
An anacrusis/pickup measure shifts where the true beat grid starts, but the beamer doesn't know that
Beat-group cycling for beaming purposes must start counting from the beginning of the first FULL measure, with the pickup notes grouped according to their own (usually shorter) partial-measure beat positions — not as if the pickup were beat 1 of a full bar.
Breaks the readingCheckable from the score alone
Odd/additive meters (5/8, 7/8, 5/4, 7/4) have no correct single default and no way to author one
Additive meters must show their internal subdivision via beam grouping: 5/8 conventionally as 3+2 or 2+3 eighths; 7/8 as 2+2+3, 2+3+2, or 3+2+2 eighths; 5/4 and 7/4 analogously as 3+2/2+3 or 2+2+3/2+3+2/3+2+2 quarters. None of these is a universal default — the grouping is a compositional decision that must be visible in the beam pattern, sometimes further cued by a dashed internal barline between sub-groups.
Breaks the readingCheckable from the score alone
Simple-meter beat grouping (2/4, 3/4, 4/4, and their multiples)
In simple time, beams group notes to make each metric beat visible: one beam per quarter-note beat in 4/4 and 3/4 (two eighths, or four sixteenths, per group), and correspondingly per beat in 2/4. The beam boundary must fall exactly on each beat, never spanning two beats unless a higher convention (tuplet, explicit compound feel) overrides it.
Breaks the readingCheckable from the score alone
Symbolic time signatures ("C", "C|", "¢") are mishandled by the default beam-grouping fallback
Cut time (alla breve, symbol ¢ or C|) must beam by the half-note beat; common time (C) must beam by the quarter-note beat. Whatever glyph is drawn on the stave, the beam grouping underneath it must match the meter it represents.
Breaks the readingCheckable from the score alone
Umbrella invariant — every generated beam group's boundaries land exactly on the meter's true nominal beat grid
This is the single check that subsumes simple/compound/odd-meter grouping, barline-crossing, system-break, and anacrusis correctness: at all times, a beam group's start and end tick must coincide with a genuine beat (or sub-beat, for secondary breaks) boundary as measured from the true start of its containing bar, never from an arbitrary note-index-zero origin.
Misleads the readerCheckable from the score alone
6/4, 9/4, 12/4 are metrically ambiguous and the naive fallback silently guesses
A numerator divisible by 3 does not by itself guarantee a compound feel. 6/4 carries two legitimate, mutually exclusive readings: (a) compound duple — 2 beats, each a dotted half note, grouped 3+3 quarters (VexFlow's actual default via its `tripleMeter` branch); or (b) simple triple with a doubled beat unit — 3 beats, each a half note, grouped 2+2+2 quarters — used when 6/4 functions as a slower relative of 3/2, including the classic hemiola against a surrounding 3/2 bar. 9/4 and 12/4 carry the analogous ambiguity (3 or 4 dotted-half beats vs. a triple/quadruple half-note reading). The composer's intended grouping must be explicit, not inferred from the numerator alone.
Misleads the readerCheckable from the score alone
Compound-meter beat grouping (6/8, 9/8, 12/8, 6/16, 12/16)
In compound time, the beat unit is a dotted note; beams group in threes of the written unit (three eighths per beam in 6/8, 9/8, 12/8 — three sixteenths per beam in 6/16, 12/16), yielding 2, 3, or 4 groups per bar respectively.
Misleads the readerCheckable from the score alone
Half-note meters (2/2, 3/2, 4/2) group by the half-note beat
In alla breve / cut time, the beat unit is the half note, so beams group eighths and sixteenths in fours/eights per half-note beat, not per quarter note. Quarter notes themselves are unbeamed in this meter (they take stems only).
Misleads the readerCheckable from the score alone
Secondary (16th+/32nd) beams must break to expose the primary beat — Gradus currently does NOT get this for free
When a beamed group contains sixteenth notes or shorter and spans more than one primary beat, the secondary (lower) beam levels should break at each beat subdivision so the primary pulse remains visible inside the dense run, rather than running one continuous double/triple beam across the whole group.
Misleads the readerCheckable from the score alone
Vocal (lyric-bearing) beaming: one-beam-per-syllable vs. modern instrumental-style beaming — currently unauditable AND unauthorable, a real gap for Gradus's chorale curriculum
Traditional vocal engraving beams notes only within a single syllable — a melisma of several notes on one syllable is beamed together, but two notes each carrying their own new syllable are given separate flags even if they would instrumentally beam together — with a hyphen/extender line showing the syllable spans multiple notes. Modern engraving practice increasingly uses full instrumental-style beaming by meter regardless of syllable count, relying on hyphens/extender lines under the lyric alone to show melisma, since it is simpler and more consistent with the rest of the score's beaming.
Falls short of best practiceCheckable from the score alone
A beam group must never span a system (line) break
When music wraps onto a new engraved system, every beamed group must be complete within one system; a beam cannot visually continue across the line break.
Falls short of best practiceNeeds the rendered page
A beam over same-pitch or near-unison repeated notes should be flat, not wedge-shaped
When all notes under a beam are the same pitch (or very close), the beam should be drawn flat/level rather than tilted, since there is no melodic contour to justify a slope, and a spuriously sloped beam over identical pitches reads as a visual error ('wedge').
Falls short of best practiceNeeds the rendered page
A beam's vertical placement relative to the staff should read cleanly — clear of staff lines, not ambiguously resting on one
Gould's guidance (and general engraving practice) is that a beam should visually clear the staff lines with enough margin that it doesn't appear to be sitting on, straddling, or merging with a staff line — the beam should read as clearly above or below the affected noteheads.
Falls short of best practiceCheckable from the score alone
A tuplet beam may legitimately span what would otherwise be two simple-meter beat groups
A tuplet of note values that would not normally take a beam on their own (e.g., a quarter-note triplet, where quarter notes are unbeamed) is joined by a plain beam so the figure reads as one rhythmic unit — this beam is permitted to cross the notional beat division the tuplet itself creates, because the tuplet bracket/beam IS the higher-level grouping signal.
Falls short of best practiceCheckable from the score alone
Beam groups must reveal syncopation against the beat grid, never regroup to obscure it
Beam boundaries always mirror the underlying metric beat grid, even across syncopated rhythms; syncopation is shown through note spelling and ties, never by re-grouping the beam to visually smooth over an off-beat entrance. Keeping group boundaries fixed on the beat is precisely what makes an off-beat note's entrance visible — it starts mid-group or crosses into the next beat's ties rather than sitting at a group's left edge.
Falls short of best practiceNeeds the rendered page
Beam slope follows Ross's half-angle heuristic via fixed constants, with no per-passage override
A beam's slope should reflect the overall rise/fall of the group's pitches but be moderated — roughly half the angle implied by a straight line between the first and last stem tips — to avoid an exaggeratedly steep beam.
Falls short of best practiceCheckable from the score alone
Beam stroke count must match each note's duration (single beam = eighth, double = sixteenth, triple = 32nd)
The number of beam strokes connecting a group of notes must correspond exactly to each note's written duration — a mismatch (e.g., a sixteenth note drawn with only one beam stroke) misrepresents the rhythm.
Falls short of best practiceCheckable from the score alone
Beams break at interior rests by default
A rest interrupts a beamed group unless the engraver deliberately chooses to beam through it (with a stemlet) to preserve a beat's visual grouping — the conservative, safer default is to end the beam at the rest.
Falls short of best practiceCheckable from the score alone
Cross-staff beaming (a beam whose notes visually span two staves) is not representable in Gradus's schema at all
In keyboard writing where a passage crosses between the two hands' staves, a beam is sometimes drawn spanning both staves to show the notes are one continuous rhythmic figure even though they're drawn on different staves.
Falls short of best practiceCheckable from the score alone
Fractional beams (beamlets) point toward the note(s) they group with
A partial beam on an isolated shorter note within a group points toward whichever neighboring note it forms a sub-group with (e.g., in 6/8's eighth-sixteenth-eighth-sixteenth, the beamlet direction implies whether the rhythm reads as three eighth-groups or two three-sixteenth-groups).
Falls short of best practiceNeeds the rendered page
Stems must maintain a minimum length where they meet the beam
A note's stem, when extended to reach a sloped or multi-level beam, must not be shortened below a legible minimum length, and multi-level (16th+) beams require progressively longer stems to accommodate each additional beam stroke.
Falls short of best practiceCheckable from the score alone
Tremolo/repeat beams are not representable in Gradus's schema
Measured tremolo notation uses short unattached beam strokes across a stem (or between two stems) to indicate rapid repetition rather than a rhythmic beam group.
Falls short of best practiceCheckable from the score alone
VexFlow determines a beamed group's stem direction by summed average line position, not by the single note farthest from the middle line
The classic engraving rule (per Gould/Read) is that a beamed group's stem direction is set by whichever note is farthest from the middle staff line — an extreme-note rule. VexFlow's actual implementation instead sums `(line - 3)` across every note (and every chord pitch) in the group and takes the sign of the total — an averaging rule. These two methods can disagree.
Falls short of best practiceCheckable from the score alone
Voices sharing one stave (SATB soprano+alto, tenor+bass) must keep separated, non-recomputed stem directions so their beams don't lattice
When two independent voices share one staff (the SATB grand-staff convention: S+A on treble, T+B on bass), each voice's stems — and therefore its beams — must consistently point one fixed direction (upper voice up, lower voice down) so the two voices' beamed groups don't visually merge or cross through each other, even with zero actual voice-crossing in pitch.

Loading…

A beam must never cross a notated barline

A multi-note grace-note run renders as separate flagged notes instead of one beamed group

A tuplet (and its beam) must not straddle an authored barline

An anacrusis/pickup measure shifts where the true beat grid starts, but the beamer doesn't know that

Odd/additive meters (5/8, 7/8, 5/4, 7/4) have no correct single default and no way to author one

Simple-meter beat grouping (2/4, 3/4, 4/4, and their multiples)

Symbolic time signatures ("C", "C|", "¢") are mishandled by the default beam-grouping fallback

Umbrella invariant — every generated beam group's boundaries land exactly on the meter's true nominal beat grid

6/4, 9/4, 12/4 are metrically ambiguous and the naive fallback silently guesses

Compound-meter beat grouping (6/8, 9/8, 12/8, 6/16, 12/16)

Half-note meters (2/2, 3/2, 4/2) group by the half-note beat

Secondary (16th+/32nd) beams must break to expose the primary beat — Gradus currently does NOT get this for free

Vocal (lyric-bearing) beaming: one-beam-per-syllable vs. modern instrumental-style beaming — currently unauditable AND unauthorable, a real gap for Gradus's chorale curriculum

A beam group must never span a system (line) break

A beam over same-pitch or near-unison repeated notes should be flat, not wedge-shaped

A beam's vertical placement relative to the staff should read cleanly — clear of staff lines, not ambiguously resting on one

A tuplet beam may legitimately span what would otherwise be two simple-meter beat groups

Beam groups must reveal syncopation against the beat grid, never regroup to obscure it

Beam slope follows Ross's half-angle heuristic via fixed constants, with no per-passage override

Beam stroke count must match each note's duration (single beam = eighth, double = sixteenth, triple = 32nd)

Beams break at interior rests by default

Cross-staff beaming (a beam whose notes visually span two staves) is not representable in Gradus's schema at all

Fractional beams (beamlets) point toward the note(s) they group with

Stems must maintain a minimum length where they meet the beam

Tremolo/repeat beams are not representable in Gradus's schema

VexFlow determines a beamed group's stem direction by summed average line position, not by the single note farthest from the middle line

Voices sharing one stave (SATB soprano+alto, tenor+bass) must keep separated, non-recomputed stem directions so their beams don't lattice