Re: partcombine, but including rests in quiet voices?

[Reinhold Kainhofer]

Am Tuesday, 11. December 2007 schrieb Trevor Daniels:
> > In this case, temporary voices won't work,
> > because then I no longer have both
> > instruments separately.
>
> If you already have the full parts for both
> instruments separately then I think the approach
> using two voices is still the best, 

Actually, I'm trying to not only write a score, that can be read by the 
orchestra, but a professional-looking score, which adheres to the principles 
and standards that are employed by professionally published scores.
Comparing with professionally engraved scores, here are my observations as far 
as partcombine is concerned:

-) if (almost) all notes or at least a large part of the notes in a line use 
the same rhythm, those with the same duration are combined to chords
-) If the instruments cross over, the notes are never combined to make the 
inversion clear
-) If notes are combined to a chord, the stem follows the usual directions for 
a chord, if the two instruments are printed separately, the first instrument 
uses up, the second instrument down (the same if one instrument is quiet)
-) If the two instruments have largely a different rhythm, or different slurs 
or ties on one staff line, then those parts (or even whole staff lines) are 
printed in two voices.
-) Rests for both instruments are combined to one staff-centered rest, rests 
for only one instrument are displayed in their own voice (since those parts 
are printed like two voices)
-) If one instrument has only a one- or two-measure "solo", the rest of the 
other instrument is printed, for longer soli, the a1/a2 (or I/II) notation is 
used (in newer engravings, old engravings never use the a1/a2 notation, but 
always print the rests of the quiet instrument).


Lilypond's partcombine comes quite close tho these guidelines, but in a few 
(but important) cases, it does not, most notably in the last case (where 
lilypond detects a "solo" also if one instrument is quiet for only one 
quarter!)


> > 1) notes of the same length are not combined into
> > a chord,
>
> No, but doesn't this make it easier to tell when
> the two instruments cross over, for example?

Those parts should always be printed in two voices by partcombine (and they 
are if you use partcombine).

> > 2) stem directions have to be fixed with
> > \voiceOne / \voiceTwo,
>
> Defining a couple of variables makes this easier

Manually doing this in a few thousand places is not what I call "easier"...

> > 3) When both instruments have a rest, two rests
> > are printed in strange
> > locations...
>
> This can be fixed, as follows.  You have to tag
> sections where the horns are in unison and where
> they are separate in each part, made easier by
> defining a couple of variables, in order to get
> the rests right.

That's infeasible: I have a full score of 100 pages, where I want to use 
partcombine for the flutes, oboes, fagotti, horns, tromboni and violas (not 
all instruments appear on all 100 pages, but its still too much to do 
manually). The unison measure are so numerable that doing this manually this 
would be a really huge effort, for no apparent reason, as those things could 
be auto-detected.

Cheers,
Reinhold

-- 
------------------------------------------------------------------
Reinhold Kainhofer, Vienna University of Technology, Austria
email: address@hidden, http://reinhold.kainhofer.com/
 * Financial and Actuarial Mathematics, TU Wien, http://www.fam.tuwien.ac.at/
 * K Desktop Environment, http://www.kde.org, KOrganizer maintainer
 * Chorvereinigung "Jung-Wien", http://www.jung-wien.at/