Re: repeat

[David Wright]

On Sat 13 Feb 2016 at 10:21:51 (+0100), Johan Vromans wrote:
> On Fri, 12 Feb 2016 23:14:18 -0600
> David Wright <address@hidden> wrote:
> 
> > So, if you write/print the sequence of notes once, why would you want
> > to copy and paste them all in again?

That comment of mine, as I explained in my opening paragraph, was
directed at the OP's problem, not yours.

> The basic questron is: If I have a sequence of measures that goes like
> this:
> 
>    AAA BBB CCC1 DDD EEE 
>    AAA BBB CCC2 DDD EEE 
> 
> can I take advantage of some repeat method to avoid duplicating all the AAA
> BBB DDD and EEE's, in the input and in the output.

With respect, your introduction of this structure with the phrase
"I've once wished for something like..." in
http://lists.gnu.org/archive/html/lilypond-user/2016-02/msg00266.html
was not the basic question, but a different problem. Some might say
your posting was hijacking the thread.

> The answer is: No, you can't. There is no officially accepted music
> notation that supports this. We can brew something in LilyPond using
> functions, variables etc

We sort of agreed on the problems presented by this in October.

> but there would still be no shortening of the
> sheet output, just the LP input.

That's not true. Back in October when this was being discussed here,
http://lists.gnu.org/archive/html/lilypond-user/2015-10/msg00064.html
I wrote "Well, it's funny how a piece immediately pops up that could
benefit from such a construction." Just for this private study† I did
typeset the piece, and I've attached the relevant section. It reduces
the number of measures from 18 to 10. The structure variable is

globalm = {
  \allowVoltaHook "|"
  \key ef \major
  \time 6/4
  \partial 2.
  %% \bar ".|:" % if desired
  s2.
  s1.
  s1.
  s1. \break % added to isolate this quotation
  s1.
  \time 4/4
  \set Score.repeatCommands = #'((volta "1."))
  s1 % alternative 1
  \set Score.repeatCommands = #'((volta #f) (volta "2."))
  s1 % alternative 2
  \set Score.repeatCommands = #'((volta #f))
  \time 6/4
  s1. \break % added to isolate this quotation
  \time 4/4
  s1
  s1
  \time 6/4
  s2.
  \bar ":|."
}

> > > For the sheet, I would resort to just writing the first note and add a
> > > markup silimar to "2nd time, play XXX".
> 
> > This will unnecessarily prevent LilyPond using the \repeat volta
> > construction and so, once again, no MIDI rendition among other
> > concerns,
> 
> LilyPond doesn't deal with segno, coda, d.s. etc either in the MIDI. The LP
> generated MIDI only matches the sheet music in most but definitely not all
> cases.

That's right, hence my inclusion of:

"and no correct MIDI rendition without juggling
the chunks in variables, ie
\score { { \AAA \BB1 \CCC \AAA \BB2 \CCC } \midi { } }"

which is the way I deal with those problems (unless I'm using the MIDI
just for aural proof-reading, which is usually the case).

> > I can't help wondering what you and the OP have against a conventional
> > method like those depicted in
> > http://lists.gnu.org/archive/html/lilypond-user/2016-02/msg00260.html
> 
> "Have against" is a big word... Technically, these methods are fine. But
> you lose the visual aspect of a repeated piece of music with just a small
> difference between the first and the second repetitions.

That sounds as if you want a musical analysis point of view, which is
not the impression I got from the OP's OP which was about finding a
LilyPond construction that would shorten the sheet music. This is the
standard case in countless songs, as attached (where the difference is
actually in the accompaniment, not the tune).

Your ossia method clutters up the beginning of the piece. The
conventional approach keeps the beginning uncluttered, which makes it
easier for the performer to make a good clean start. It can also
discourage amateur singers from slowing down at the end of the first
repeat because it makes it clear that the piece continues.

Cheers,
David.

† it's still in copyright.

rooks.png
Description: PNG image

song.pdf
Description: Adobe PDF document