Re: hymns: chords vs. voices

[Carl Sorensen]

On 1/18/10 2:24 AM, "mike99" <address@hidden> wrote:


> Carl's method of using partcombine allows for keeping the voice inputs
> separate. This makes the code clearer, and is encouraging since it's still
> undecided which notation to adopt; by changing a few declarations I could
> use either method (minus tweaks).
> 
> Carl, I went about implementing my own instance of a song and quickly ran
> into three snags, which, although not covered by your example, are common.
> The snags are described below, and I've extracted four measures and attached
> them as an example for illustration. Would you mind sharing your experience
> on how you deal with them?
> 
> http://old.nabble.com/file/p27207893/sample2.png
> 
> 1.  Notes of common pitch and duration result in the text "a2" being set
> above the staff (to notate a due), as in measures 1, 2, and 4. The hymnal I
> have that uses chord notation ("Songs of Faith and Praise") sets displays
> this as two separate voices (i.e., an exception to the chord notation) with
> one note head and a stem in each direction, as this snippet would produce:
>    \relative c'' { << b \\ b >> }
> The manual says setting printPartCombineTexts to "false" suppresses the a
> due marking. Do you modify the behavior to add a second stem, or just accept
> the default with one stem?

I set printPartCombineTexts to  #f.  I have just lived with it.  But given
your experience with chords, I decided to try putting a chord with a
transparent note in the voice that needed the extra stem:

soprano =  \relative c'' {
  a f e d | 
  e g c b 
}

alto = \relative c' {
  c d c <d \tweak #'transparent ##t b> |
  c <g' \tweak #'transparent ##t e> g f
}


\score {
  \partcombine \soprano \alto
}


It worked (although in the first case (the d) there is an unwanted ledger
line; we could also make the ledger lines transparent to avoid that
problem).

I'm looking at the possibility of writing a music function that will
automatically do this, but I haven't got it to work yet.


> A related issue is how collision avoidance is handled. I changed the alto
> voice in measure 2 to have a b-flat in order to show that partcombine uses
> one stem with two note heads, whereas it's preferable to use standard
> polyphony (creating an exception from the chord notation), as discussed in
> my  previous post.

I suspect this can also be fixed by adding a chord with a second note having
a transparent head.

> 
> 2.  As in your example, the lyrics are set to an extra instance of the
> soprano voice, which is set an octave above and hidden. The slur in measure
> 3, however, is visible.

In the hidden voice, add

\override Slur #'transparent = ##t

This will hide the slurs.
> 
> 3.  Chords within a voice cause the voices to be typeset separately, as in
> the bass clef in measure four. I intuitively expected partcombine to use one
> stem.

I think that you want them to have different stems, so you can see which
part is split.

> 
> 
> I need to learn about using \context vs. \new declarations, and study up on
> the effects of \shiftOnn, which in this example caused some misalignment of
> lyrics, but this is looks promising.

\new creates a new context

\context switches to an existing context, and creates a new one if the
context doesn't exist.


> 
> Thanks for your help,

You're very welcome.

Carl