Beam positions according to typographical rules.

• Oct 16, 2010 - 02:19
Type
Functional
Severity
S5 - Suggestion
Status
closed
Project

Attached is an example of beam typesetting recommendations for notes starting on C pitch classes (in treble clef) or E pitch classes (in bass clef). The recommendations are taken from the book:
The Art of Music Engraving and Processing by Ted Ross, Hansen Books 1970 (2nd edition c. 1975), page 104.
(There are several other pages in the book which display notes starting on other pitch classes in the book, and general guidelines for more complex beaming cases involving more than two notes in a beam group).

In the attached example, MuseScore's beam locations are shown on the top staff of the system, and Ross's recommended positions for the beams are given in the bottom staff of the system.

The codes underneath the notes on the Ross line:

  • 0 = the beam is centered on the staff line.
  • 1 = the beam sits on top of the staff line.
  • m1 = -1 = the beam hangs from a staff line.

These positions are intended to minimize "wedges" which are thin white areas in the angle between the beams and the staff lines. These are important to avoid in traditional printing so that ink does not get stuck on the paper in the points of the wedges. It is also useful so that there is less visual clutter between the notes and the staff, making the music appear "cleaner" which allows sight-reading of the music to be more efficient.

It would be useful if MuseScore adopted these recommendations, or had them as an alternate beam default position option alongside the current Lilypond style of beam positions (and perhaps people might want to define their own beaming defaults, so you can allow user beam definitions in the future as well :-).

In particular, in the MuseScore examples on the top staff of the systems:

m1: unnecessary wedges (no wedges present in the Ross version).
mm11 & m12: In these two measures the low notes should be attached
to the center line of the staff as shown in the Ross recommendation.
In the MuseScore default, these notes are attached to the second
staff line from the bottom. This is a fairly serious violation of music
typography. To quote the book "Essential Dictionary of Music Notation"
by Tom Gerou and Linda Lusk (Alfred Publishing1996), page 302:
"When a note extends beyond one ledger line, the stem
must touch the middle staff line." (emphasis is the authors').
mm 20 & 21: the same case as above but for notes above the staff.

mm29, 30: Ross recommends on page 103: "Under normal spacing
conditions, preference is given to shortening the stems instead
of lengthening them;" So notice that his lengths are shorter in these
cases in particular.

m2, 4, 6: In these measures, MuseScore starts the beam in the middle of the
space between two staff lines. This is not very commonly done in music
where strict music typesetting rules are followed, primarily because it
causes small angle wedges which are likely to fill in with ink when printed.
But also on the scree, this will cause the bit mapped images to form
distracting clumps of black in a similar process.

Attachment Size
ross-ce-beams.mscz 4.36 KB
ross-ce-beams.pdf 25.96 KB

Comments

craig wrote: "It would be useful if MuseScore adopted these recommendations, or had them as an alternate beam default position option alongside the current LilyPond style of beam positions"

Just to clarify, MuseScore's beaming in no way represents LilyPond's. LilyPond's music typography is more sophisticated than any other software I know. Attached is a copy of LilyPond's beaming of the passage. LilyPond avoids the "small angle wedges which are likely to fill in with ink when printed" although there are some different interpretations of the beaming than Ross's recommendation. Among the differences, LilyPond's rendering of measure 6 is more logical than Ross's recommendation (Ross uses the same beaming for m. 6 and m. 7, LilyPond stair-steps the beams from m. 5 through m. 7). In measure's 13-14 LilyPond uses down-stems. Ross uses up-stems (for the sake of reference only). LilyPond allows steeper beam angles m. 16-19. In measures 20-23 LilyPond centers the beam on the middle line, Ross only touches it.

What is needed is a comprehensive list of rules for beaming. It is not possible to hard code every conceivable combination of beamed notes. Here are the rules I'm able to discern:

  1. No stem should be shorter than an octave (for exceptions see rule number 3 below)
  2. When a note extends beyond one leger line, the stem must be lengthened to touch the middle staff line
  3. If the start or end of a beam causes "small angle wedges" preference is given to shortening the stems instead
    of lengthening them. Use the following four rules to avoid small angle wedges
    1. An upward sloping beam must not start immediately above a staff line (it should be centered or immediately below a staff line)
    2. An upward sloping beam must not end immediately below a staff line (it should be centered or immediately above a staff line)
    3. A downward sloping beam must not start immediately below a staff line (it should be centered or immediately above a staff line)
    4. A downward sloping beam must not end immediately above a staff line (it should be centered or immediately below a staff line)

What is missing from my list of rules is how to determine the angle of the beam. Also missing is the range of angles that are considered "small angle wedges" and when larger angles mean that you can disregard the third rule.

Attachment Size
lilypond-beams.pdf 49.09 KB
Status (old) needs info fixed

current release (r5546) implements most of Ross.
The regression tests in mscore/mtest/libmscore/beam show examples.
If you want to change something, please refer to this testfiles or create a new testfile for missing special cases.