hymns: chords vs. voices

[mike99]

Greetings,

After a few days of reading and coding, I'm entering the steep part of the
learning curve as a new user. Years ago I used MusiXTeX with pmx and M-tx,
and the different is appreciable. I'm gearing up for a song-book project,
and it looks like LP will win out over MuseScore and any commercial packages
because of its freedom, power, and beautiful output.

This song book is to be filled with hymns, and after typesetting a few dozen
songs I expect to be somewhat experienced, but I'd like to get things as
straight as possible early on and avoid developing bad habits. (So if you
see me using poor practice, please mention it.) I already have several
questions—mostly related to lyrics—and I want to start out with a fairly
easy albeit long-winded one.

I am evaluating which of two notational systems to adopt. The hymns will be
set on a two-staff system, mostly with four voices (SATB), i.e., two voices
per staff. I'm in Europe, where it is common to consistently write the two
voices of a staff separately, as with \voiceOne and \voiceTwo. Some hymnals
use an alternative system that writes, for example, the soprano and alto as
a chord where possible, only splitting into two voices (in LP terminology)
where needed, such as with different simultaneous ryhthms, or as an
alternative to shifting note heads for collisions. These will be illustrated
below. This is largely preference in a congregation, although I am leaning
toward the former simply because it is common practice here. The advantage
of the second system seems to be legibility in many cases (often easier to
distinguish rhythms), and the ability to shift stems away from the lyrics,
thereby saving vertical space. This can be quite useful in making a hymnal,
as the final print size will inevitably be small (possibly A5).

I've prepared a small example to contrast the two systems and ask my first
question. As the v2.12 Learning Manual describes in section 3.2.1 "I’m
hearing Voices," you can split chords into separate voices. (I am trying to
be careful to use voice in the LP sense and avoid speaking of soprano, alto,
tenor, or bass as voices, unless they are in the LP sense.) In my example,
the soprano and alto use the chord system while the tenor and bass use
separate voices. To keep it simple, there is one set of lyrics, which is set
to the upper staff.

http://old.nabble.com/file/p26989867/lyric-alignment_chordvsvoice.png 

The fourth beat in the first measure shows the soprano note head shifted to
the right; it is LP's practice to shift the upper note. The sight of this is
already a point of potential confusion for congregational singers (most of
whom have little or no formal training) used to seeing parts written as in
the bass clef. In fact, the "chord system" does not do it this way. To make
things more immediately clear, it instead temporarily creates a second
voice, and then resumes the chord system. Measure two illustrates this,
wherein the alto note is written as a second voice. The source is below.

%=======================
\version "2.12.2"

global = {\key ees \major \time 4/4}

sopalt = \relative c' { 
  <bes f>2. <bes aes>4 |
  <bes f>2. <<{bes4}\\{aes4}>> |
  <ees' bes>2. <f bes,>4
}
tenor = \relative c { d2. f4 | d2. f4 | g2. g4 }
bass = \relative c { bes2. bes4 | bes2. bes4 | ees2. d4 }
groupWords = \lyricmode { Why does it skip a -- bout? }

\score {
  \new ChoirStaff <<
    \new Staff = upper
      \new Voice = "sopranos" {
        \oneVoice
        << \global \sopalt >>
      }
    \new Lyrics = "group" { s1 }
    \new Staff = lower <<
      \clef bass
      \new Voice = "tenors" {
        \voiceOne
        << \global \tenor >>
      }
      \new Voice = "basses" {
        \voiceTwo << \global \bass >>
      }
    >>
    \context Lyrics = group \lyricsto sopranos \groupWords
  >>  
  \layout { indent = 0 }
  \midi {  }
}
%=======================

Fair enough, but there are the lyrics, set here to the soprano voice, which,
unintended by myself, skips the fourth beat in the second measure. In the
documentation's first example on divisi lyrics (Notation Reference, 2.1.4,
version 2.12), it does not, because the authors have explicitly created a
new voice. Sure enough, changing the sopalt expression to:
%=======================
sopalt = \relative c' { 
  <bes f>2. <bes aes>4 |
  <bes f>2. << 
    { \voiceOne bes4 }
    \new Voice = "splitapart" { \voiceTwo aes4 }
  >>
  \oneVoice |
  <ees' bes>2. <f bes,>4
}
%=======================

yields this output:
http://old.nabble.com/file/p26989867/lyric-alignment_chordvsvoice2.png 

If such a technique is required for all exceptions to the chord structure,
it seems as if the "chord method" could become patchwork if many exceptions
are needed in a piece. Ten exceptions might be common on a one-page hymn,
requiring the creation of as many new voices. With consistent technique it
should be doable; for commonly occurring exceptions it should even be
possible to define some clever commands that reduce clutter. Extra work in
any case.

Enter my question: What is your opinion to the two methods, given the
direction of the project and in terms of readability and complications that
it would cause in the score? It would not be a simple thing to switch
between the two methods once several songs are written, and the songbook
should be as consistent as possible. That's why I'm looking for feedback on
the technical side before I discuss with others involved in the project.

Best regards,
Mike
-- 
View this message in context: 
http://old.nabble.com/hymns%3A-chords-vs.-voices-tp26989867p26989867.html
Sent from the Gnu - Lilypond - User mailing list archive at Nabble.com.