Re: tie behavior (was: New slur/tie behaviour)

[Steve D]

Han-Wen Nienhuys <address@hidden> wrote,
regarding general default tie rules or behavior:

> Well, AFAIK engraving conventions are:
>
>   * outer ties point outward
>   * double directions for seconds
>   * direction otherwise determined by vertical position.
--- ---

Piano is my instrument, and as a polyphonic instrument it can 
have complex and sometimes crowded chord structures, multiple 
voices and chords per staff, etc. So ties are personally very 
important to me. A default behavior and rules for placing ties is 
very helpful, but there can also occur numerous specific 
situations in which default behavior is not optimal. So I 
personally hope that, in any future rewriting of the LilyPond tie 
code, users will remain able to override default behavior and 
specify at least the vertical placement and arc-direction for 
each individual tie, between single notes (that are not part of a 
chord) as well as for the notes of a chord, and perhaps also be 
able to stretch or compress the tie horizontally along the 
X-axis, either the tie as a whole or the individual beginning and 
ending points of a tie, if the person notating the score thinks 
the tie would look better a little shorter or longer in any 
particular circumstance. However, I have no idea whether this 
last feature (overriding the horizontal length of an individual 
tie) would be worth the time and effort to implement.

Regarding the engraving conventions listed above, they seem 
generally sound to me and applicable in most situations, although 
I don't claim to be an expert in such matters. All three 
conventions would need to be altered in some cases, which is why 
I hope that users will be able to override default tie behavior.

The only thing I would add, personally, to the conventions listed 
above is a corollary to the first one, "outer ties point 
outward."

To me (and I wonder if others agree), although outer notes of a 
chord should ideally arc outward from the chord, the vertical or 
Y-axis placement of the beginning and ending points for those 
outermost ties should be different for tied notes that fall 
within a staff or ledger *space*, than for ties between notes 
that fall on a staff or ledger *line*.

In the case of notes that fall within a staff space rather than 
on a line, and merely in my opinion, the ties would look better 
if the beginning and ending points of the ties were placed in the 
staff space *adjacent* to the space in which the notes exist, 
rather than in the same staff space as the notes themselves.

For example, in a chord within a treble clef staff, the uppermost 
note of which is c'', the tie connecting that note would begin 
and end in the staff space that would be occupied by the note 
e'', which is more or less exactly the same place that such a tie 
would be drawn if the uppermost note of the chord were a d'' 
(placed on a line) instead of a c'' (placed in a space).

Just my opinion. ;-)

Han-Wen recently mentioned that, in rewriting the tie code a week 
or 10 days ago, he thought of a possibly better conceptual 
approach to and foundation for LilyPond's tie code. I would be 
very willing, in a couple months (like in November for example) 
to help sponsor such a re-thinking and rewriting of the tie code.

Although the recent rewrite is a lot better than the older 
default behavior for ties, if Han-Wen thinks it can be improved, 
I'm personally enthusiastic about helping to sponsor an 
improvement in this (to me) very important aspect of music 
notation. I am thinking about helping to sponsor such a project 
by making monthly payments of perhaps 200 Euros or more each, for 
4 or more consecutive months, depending on how much time and 
effort Han-Wen spends on this functionality, should he wish to 
direct his thought and attention toward it. I would hope that 
others might be interested in helping to sponsor such a project 
as well.

Best wishes,

Steve D
New Mexico US