Re: Verse-chorus-verse with repeat, multiple ending, pick-up, and varying rhythm

[Christopher R. Maden]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 03/22/2011 11:25 PM, Carl Sorensen wrote:
> On 3/22/11 6:59 PM, "Christopher R. Maden" <address@hidden> wrote:
> Yes, but that's precisely why I don't have a vision of it.  The semantics of
> repeat means that the repeat *is* exactly the same all four times.  I know
> of no way in general music to indicate that the body of the repeat varies
> from repetition to repetition.  In vocal music of my experience, it's
> generally done with notes at the finest time resolution that are connected
> with slurs and then melismata are used on the lyrics that span multiple
> notes.

Sure, when the exact same notes are used in slightly different rhythms,
that works.  But in this song, there are rests in some verses where
there pickups in others.  No amount of slurs or melismata is going to
handle that. (-:

In pop stuff I’ve seen (I’m thinking mostly Hal Leonard here), the first
verse usually has prominence, with the subsequent variants given cue
notes.  If I really cared, I could make the cues do the work, but it
would be nice to tell LilyPond to actually consider the music a slight
variation.

One way to do this would be to write out each verse discretely, and ask
LilyPond to fold it up, factoring in the commonalities.  (This could
work for the lyrics, too.)

Another way would be to have a different version of the \alternative
syntax, like \variant.  This could be used the same way, but set without
all the bracketing:

\variant {
  { r4 }
  { c8 c8 }
  { r8 c8 }
  { c8 c8 }
}

This would not only work for typesetting, but it would let the lyrics be
assigned correctly to notes in each verse without having to muck about
with \skip and faux-melisata.

(I tried chord <c8 r8> and multi-voice << {c8 c8} \\ {r4} >> variations
with rests and notes, but LilyPond rejected those.  That still wouldn’t
address the lyric-placement issue, anyway.)

> Interestingly enough, if you look at the last layout in my example, there is
> nothing semantically in the notes to indicate how they are to be played.
> All of the semantics of the repeat are in the structure variable, and the
> notes just go along in increasing time.  The structure variable puts the
> repeat symbols and the volta brackets in on the staff.

Right.  I learned from your example, but didn’t use it blindly. (-:  I
am a semanticist to the bone, and the closer I can get to a formatting
of a description of the music, rather than a picture of the music, the
happier I am.

Ideally (not that I actually need this, but it’s how my brain works),
the single score should be able to generate the score, MIDI, a karaoke
or instructional presentation, and appropriate descriptions of the song
for blind and/or deaf people.  I am asking for a lot (but “all progress
depends on the unreasonable man”). (-:

> There has been some discussion on improving the semantics of repeats.  See
> issue 791 for a link to some discussion on it.  But nobody is currently
> working on it as far as I can see.

Someday when I have free time I’ll take a closer look...  (I only had
the time to work on this score because of a bad case of flu... which may
also have slowed my comprehension.)

~crism
- -- 
Chris Maden, text nerd  <URL: http://crism.maden.org/ >
“Those in power write the history, while those who suffer
 write the songs.” — Frank Harte
GnuPG Fingerprint: C6E4 E2A9 C9F8 71AC 9724 CAA3 19F8 6677 0077 C319
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAk2JfJEACgkQGfhmdwB3wxkmhgCgxe4CjH4dgSwjQuzVIFl4JImd
JcIAoMB+DXBn7qAt4kaWxI+e49SqIKx3
=5iIs
-----END PGP SIGNATURE-----