Re: GDP: NR 1.5 Simultaneous, second draft

[Karl Hammar]

Daniel:
> On Thu, Jul 17, 2008 at 10:53:03PM +0200, Karl Hammar wrote:
> > Daniel:
> > > Why not? I find myself wanting to go into two (or three) voices and back 
> > > again very frequently when typesetting percussion parts, and the 'right' 
> > > way is far too long-winded - often it would be longer than the music it 
> > > encloses. I always use the method given in the second example in NR 
> > > 2.5.1.3 Percussion Staves, i.e. explicitly instantiating the voices 
> > > beforehand and using \\, in combination with skip-of-length. Does this
> > That is strange, why do you need to do the \new DrumVoice-lines in 
...
> > > count as the right way, or is it still the wrong way? As a user, it 
> > > would be much easier for me to just be able to tell Lilypond once that 
> > > I'm doing drums, and then just put the music in, without using any kind 
> > > of "method" at all.
> > 
> > Ok, what happens if you replace the bd4 sn4 etc. with the snares from
> > last example of 2.5.1.2 ? By doing it this way, a tie is missing:
> 
> Isn't the missing tie just the usual "ties can't cross voices" problem?

Yes.

> I tend to work around that by the trick of 'hoisting' one end of the tie
> into the other voice, replacing it with a space in its original voice.
> So, in your example
> 
> >     \version "2.11.52"
> >     \new DrumStaff <<
> >       \new DrumVoice = "1" { s1 *2 }
> >       \new DrumVoice = "2" { s1 *2 }
> >       \drummode {
> >         sn16 sn8 sn16 sn8 sn8:32~ sn8 sn8 sn4:32~ |
> >         <<
> >           { \repeat unfold 16 hh16 }
> >           \\
> >           { sn4 sn8 sn16 sn16 sn4 r4  }
> >         >>
> >       }
> >     >>
> 
> the relevant lines become
> 
>     sn16 sn8 sn16 sn8 sn8:32~ sn8 sn8
>     <<
>       { s4 | \repeat unfold 16 hh16 }
>       \\
>       { sn4:32 ~ | sn4 sn8 sn16 sn16 sn4 r4  }
>     >>

Ok, this is not a problem that you cannot handle.

In vocal music, one wants to relate text with music using \lyricsto. 
That is harder if you create new voices, which makes that coupling 
harder to see and maintain.

My way of handling these problems is to always explicitly name any new 
voice I use and letting the "main thread of music" belong to the same 
voice throughout. Not making new a new voice makes my typesetting 
easier.

So, it becomes:
 Which way of creating the second voice and handling problems like
 the above is most easily explained and understood by users?
 Which one should go into NR 1.5?

Regards,
/Karl