> From: "Simons, Don" <[EMAIL PROTECTED]>
> Date: Sun, 2 Jul 2000 10:59:21 -0700
Hello Don,
> Since there is so much interest in dynamic markings among the post-baroque
> PMX users, I'm working on implementing them. Here are some issues I'm
> facing. I'll gladly consider any feedback, the sooner the better.
Does that mean that it will be implemented already in 2xxx? :-)
> 1. For letter groups, should default horizontal position be
> centered (\ccharnote) or left-justified (\zcharnote)? Should it be tweaked
> depending on stem direction, so that for up-stems it is indexed to the
> middle of the notehead while for down-stems it is indexed to the stem
> itself?
It depends :-(
f for forte is so much slanted that it looks moved to the right
compared e.g. against p for piano. Combined texts like ffff or mf or
fp or sfz look so different so that it doesn't matter whether to
use \ccahrnote or \zcharnote. I would prefer \zcharnote and following
proposal: Dxxx produces a MusiXTeX input with \zcharnote with "correct"
pitch and some tweaking which refers to a TeX-Macro \xxx. The user may
then provide something like \def\f{\kern-0.5pt\ppff f\kern 0.5pt} or
\def\p{\ppff p} or to meet some baroque usage e.g. by Bach:
\\def\pia{\ppff pia.}\
...
... a8 Dpia b ...
If I look at some old editions I have the impression that there was
no fixed rule for positionig the dynamics. Instead it was highly
dependent on the surrounding.
For a specific edition I asked some players. Some of them voted for
something like \hloff{\zcharnote....} to see the dynamic a little bit
earlier than the note.
> 2. All the same questions apply for hairpin starts and hairpin endings.
\zcharnote + tweaking
> 3. If there is both a letter-group and a hairpin start or stop on the same
> note, should the position of one and/or the other be automagically tweaked
> to avoid a crash? If so, which, the hairpin, or the letter-group? Should the
> tweak be horizontal or vertical?
Yes, please. Horizontal.
But there is conflict with my answer to 1., because PMX cannot know the
width of \xxx. The conflict could be partially solved by the assumption
that \xyz has the width of \ppff xyz or maybe simply 3*\qn@width. Fine
tuning is possible by tweaking
> 4. If there is both a letter-group and a hairpin start or stop on the same
> note, should the vertical positions be made the same by default? Without
> any special checks, they could come out differently if the hairpin's
> vertical position is controlled by the note at the other end of the hairpin.
Like with figures there could be a common height for all dynamic indications
of one staff.
The base-line for letters and hairpins should be the same. AFAIK this
requires that letters are typeset one pitch lower than hairpins.
There should be an option that dynamics for some voices should be typeset
above the staff e.g. for a singer's staff. A tweaking option like n-tuplet's
f could be used to flip the dynamics.
> 5. Syntax for hairpins. I am now using "D<" and "D>". They act as toggles:
> the first one starts it, and the next one in the same line of music ends it.
> There is a slight incompatibility with this use of ">" and "<". They have
> already been used to trigger horizontal tweaks of accidentals. (But ">" is
> also used as an ornament.) It would be possible to use "c" and "d" instead
> of or in addition to "<" and ">"
Like with beams I would prefer something which switches on and off. This
allows early error reporting.
> 6. I'm certain to use the default position as the reference point for
> user-defined vertical and horizontal tweaks, since that's the way it is with
> ornaments. But I could allow the user to give an absolute position. For
> example "Dp-2-1" will mean to move the "p" down 2\internote and left one
> notehead width from the default position. But "Dp=-2-1" could mean to
> position the letter 2\internote below the bottom staff line and one notehead
> width to the left of the left edge of the note. Then "Dp=" would be
> equivalent to "Dp=+0+0". Would that option be useful enough to put it in?
Yes. But think about a common height for all dynamics of one staff.
(And then it's usefull, too[*]).
> 7. Some may want symbols vertically centered between staves using \zmidstaff
> etc. I could make it an option to switch the vertical default to this. Is
> this worth it? Possible syntax "Dp*" and "Dp*-2-1". Should it be possible
> to make this the global default, e.g. with "AD*" or maybe just "A*"?
If you do not choose a common height for all dynamics of one staff then
the default should be just below the staff or just above the staff.
For certain instruments it would be nice to change that to something
like \zmidstaff.
> Comments?
Following my syntax proposal there would be
Dxxx resulting in ...\xxx...
D>, D< resulting in ...\crescendo... or ...\decrescendo...
May I add
D"blabla" resulting in ...\medtype\it blabla... for 20pt music and
...\normtype\it blabla... for 16pt music.
That's useful e.g. for D"cresc." or D"dim." or D"moriendo".
The same could be done with D'...' omitting "\it", e.g. D'pizz.' or
D'arco' or D'rit.'
Because MusiXTeX does not provide hairpins longer than 68mm and because
there is no line breaking for hairpins, PMX could do the job partially.
hairpins longer that 68mm are used as is, i.e. generating max. length
hairpins. Line breaking could be supervised by PMX breaking hairpins
into multiple (two?) hairpins, one that ends with the line and one
which starts with the new line.
Happy programming!
-- Werner
[*] "Dp=..." reminds me that I always wanted to have something like
that for slurs and beams. If I generate beams or slurs with different
clefs in score and parts then it may happen that there is e.g.
an up-slur in the score and a down-slur in the part. If such
a slur ends at a note with e.g. a sharp then for an up-slur I use
"(+1 .... )+1" whilst the down-slur needs "(-1 ... )-1". So if
PMX would be re-invented I would propose that a lot of tweakings
should depend on the direction of the object. But as PMX is already
invented I would like to have a syntactical specification for
relative vertical tweakings. Same applies to x-tuplets and maybe
other objects.
