Re: Constructive Criticism and a Question

[Brett Duncan]

Rick Hansen (aka RickH) wrote:
> David Rogers wrote:
>   
> > Orm Finnendahl wrote:
> >
> >     
> > > Am 28. Dezember 2006, 11:30 Uhr (-0800) schrieb David Rogers:
> > >       
> > > > > bf16[d, f ef] \tuplet 4 { { { d16 ef f } { g a } } { bf32a c bf d c 
> > > > >           
> > > bf a 
> > >       
> > > > > g f
> > > > > g ef } }
> > > > >
> > > > > The above would generate a parent tuplet with the number "5" and two
> > > > > sub-tuplets with "3" and "2", followed horizontally by the "12" 
> > > > >           
> > > tuplet.
> > >
> > > If you intend to think of tuplets as collections of notes of the same
> > > duration, the syntax is fine. But what happens, if the elements of the
> > > tuplet contain things of different durations (for example, the first
> > > part of your suggested tuplet is of an quarter duration and the second
> > > half containing the 12 32nds should last a dotted quarter within the
> > > parent "5" tuplet)?
> > >       
> > In the Beethoven example, Op.31 Nr.3 at bar 53, the 5-let is a
> > quarter-note duration, and the 12-let is another quarter-note duration.
> >
> > But this was Rick Hansen's proposal, for how to solve a problem pointed
> > out by David Fedoruk, and I was only admiring his solution.
> >
> > David
> >
> >     
> When the notes in a bracket are of mixed durations then by default no digit
> would be generated only the bracket.  If the user wanted a digit on a
> mixed-note tuplet they can add the "\tupletNbr x" indicator within that
> bracket of notes.
>
> As for summing the durations to match the parent stated "\tuplet x"
> duration...  This would not happen, instead durations stated within a
> \tuplet structure would be there solely for the purpose of determining the
> STYLE of the noteheads being printed.  IOW durations in a known-duration
> tuplet contruct would not participate in any further evaluation math-wise of
> expired time, just as a convenient way of setting the noteheads.  Because
> the true duration of the whole construct has already been stated on the
> "\tuplet x" there is no need to validate further, the tuplets total duration
> is still whatever they coded it to be at the outset.
>   

Hmm, I think I agree with Rick about the nested braces - that seems to 
be more "Lilypond-ish" than my original suggestion.
OTOH, I don't think the situation with mixed durations is really a problem.

For example,

   \tuplet 4 {{ a4 b8 } { c16 b a b8. } { a16 b c b a }}

(using nested brackets, as Rick suggested) should produce a 3 over the 
first and second bracket groups, and a 5 over the third, all based on 
the ratio of the cumulative duration of the notes within the brackets to 
the duration given after \tuplet. (For the first two groups, the ratio 
is 3:2, for the third group of notes, 5:4)

I think the digit on the tuplet bracket SHOULD be generated, regardless 
of what the notes and durations are within the tuplet. I don't like the 
idea that the user could put a number on the bracket that doesn't make 
sense musically - if you have a group of five notes in the time of four, 
then either a "5" should appear, or the ratio "5:4". The only exception 
I can see to this is to be able to suppress the number altogether for 
those situations where the pattern of tuplets is repeated through the 
piece (e.g. Schubert's Impromptu No. 3, Op. 90). So I don't think a 
\tupletNbr function is really necessary. But others may it differently 
to me.

My $0.02 (Australian, so worth even less!)

Brett