Re: lilypond-user Digest, Vol 186, Issue 108

[Aaron Hill]

On 2018-05-28 06:00, Wols Lists wrote:
> On 22/05/18 21:11, Arle Lommel wrote:
> lmost right, but doesn’t function quite as Hindemith’s
> > notation does:
> >
> >   * The Hindemith editions use time signatures /both/ above and in 
> > line,
> >     but never at the same time. They serve different purposes. The
> >     snipped removes the ability to do the in-line time signatures
> >     because the time signature engraver is removed from the main 
> > staffs.
> >     Easy enough to put in, but then I need to brush up enough to
> >     selectively control where a time signature appears
> >   * Hindemith also limits the effect of a superior time signature to a
> >     single measure. The following measure reverts (with nothing
> >     displayed) to the original meter. So replicating the Schott 
> > practice
> >     would mean also hiding the time signature following such a 
> > measure.
> >
> >
> > But this is very useful as a suggestion of an approach. Maybe I can 
> > find
> > a way to selectively hide/show time signature changes in each context 
> > to
> > get what I want.
> Another possible approach, for the temporary time changes, is can you
> display the new time signature without actually setting it? (Ie if the
> piece is in 2/4, you display 3/4 but lilypond still thinks its a 2/4 
> bar
> as before with no change.
>
> Then just use \times 2/3 {} for that bar so the notation displays 
> correctly.
>
> Okay, that won't work for midi, but for scores it might be a good 
> solution.

You could do a hidden tempo change for the MIDI output that would 
correct for the scaled durations.

But I would think the best option is not to mess with scaled durations 
and simply have an actual time signature change.  All you need to do is 
visually offset the normal time signature, or you can hide it and simply 
replace it with suitable markup.  There was a recent thread on this:

http://lists.gnu.org/archive/html/lilypond-user/2018-05/msg00403.html

What could be handy, perhaps, is to have the ability to do a time 
signature change that is limited to a defined scope.  Consider the 
following:

%%%%
  \version "future"
  {
    \time 3/4 % Normal time signature change
    a4 b c' | e'2 d'
    \localtime 5/4 % Scoped time signature change
    {
       e'4 g'8 f' e' d' e'2
    }
    % Time signature implicitly reverts to 3/4 here
    c'4 b g | a2.
  }
%%%%

Providing these so-called local time signature changes would be 
adjustable/overridable independently of normal ones, folks could easily 
accommodate something like Hindemith.

-- Aaron Hill