The difference between slurs and ties

• Sep 3, 2014 - 02:31

The lines per se look the same in either case. My question concerns the precise location where the end points of the lines are anchoring to the tied or slurred notes.

In MuseScore 1.3, the end points of a slur anchor to positions centered over or under a notehead (except, obviously, when one of the anchor points is the note's stem).

In MuseScore 1.3, the end points of a tie anchor to positions near the inside edge of the notehead instead of being centered over or under it. It's as if there were an imaginary rectangular box surrounding the noteheads, and each end point anchors to a corner of its respective box. The result is a line that appears to be placed just barely inside the noteheads

In MuseScore 2.0 Beta 1, the end points of both slurs and ties anchor over or under the centers of the noteheads. Any distinction between slurs and ties has been eliminated, because both of them now anchor as slurs.

The original behavior of these elements in MuseScore 1.3 is what I have always understood to be a customary distinction between slurs and ties. It is generally confirmed in scores from the hand-engraved era, though I don't expect that all publishers followed this convention or rigidly adhered to it with consistency; the placement of slurs in particular is very flexible depending on the presence of dynamic accents, articulation symbols and fingering - and whatever priority is assigned to each element in a hierarchy of which ones should move out of the way for which others.

Anyway, I'm curious what led to this change from the status quo. I don't own 'Behind Bars' or 'The Norton Manual of Music Notation' or other reference books that are commonly cited, so I don't actually know what most experts say is preferred as a matter of style. My own sources for what I've described are personal experience and a trade paperback called 'Music Engraving Today' by Steven Powell.

Thanks in advance for any insight into this decision.


Comments

Yes, ties used to anchor inside notes in 1.3, but I don't consider that a good thing. In 1.3, it was very often the case that ties would end up being so short they were all but invisible (try tying two eighths together). So a conscious choice was made to make them more visible in 2.0, and that does indeed mean they now anchor over the note rather than between them. But they are still not the same as slurs. Ties between single notes anchor about 1/3 of the way into the notehead; slurs about 1/2 way. Ties also anchor closer to the notehead than slurs do, and the shape is also flatter. So ties will ordinarily fit inside slurs. For chords, ties remain anchored *inside* the notes, but there is also code to automatically move the notes farther away from each other to enforce a minimum length for the tie (a Style parameter, defaults to 1sp) so you don't have the "so-short-they-are-invisible" problem.

So the bottom line *should* be that ties look considerably better than they did in 1.3 in virtually all cases. And the settings chosen are in very well accordance with Elaine Gould's "Behind Bars". In particular, the fact that ties now overlap noteheads but not as much as slurs, the fact that ties anchor closer, the fact that ties are flatter than slurs - these are all as Gould recommends.

But as you say, different publishers definitely do things differently, and if you mostly have editions that attach ties inside the noteheads, I guess the Gould-style ties might look odd. But it's not a new thing. I'm looking at editions going back a century or more, and ties over the notehead are found in the majority of them.

In reply to by [DELETED] 448831

You're welcome. These sort of details are important to me, and I do want to make sure we do the best job we reasonably can, but I realzie that every situation is different.

BTW, if you didn't already, see the post I made a few months ago detailing some of the various layout improvements made for 2.0: http://musescore.org/en/node/25102. It shows a very nice example of the problem with too-short ties and how moving the anchor point over the note - which again, seems to really be the more common approach anyhow for single notes as opposed to chords - helps.

Thanks to Marc for his very valuable work on these fastidious details, in particular and among other, for the improvements in the placement of accidentals, which I have highly welcomed.

It is true that occasionally ties were too short in MS 1.3 and having made them easily visible is definitely an improvement.

However, after reading this thread and looking at some examples, I share the concern of the OP about ties and I am asking if it is too radical to 'widen' the ties even when there is plenty of room to have them both visible and distinctively anchored.

An example is in the post quoted above by Marc : in meas. 14 of both versions of Mystic Reverie posted there, the two tied D's look very much as if there were slurred (the difference between 1/2 and 1/3 of note head width is too small to be readily noticeable while playing); unnecessarily so, I would say, as the distance between them is more than enough to have the tie clearly visible, even if anchored 'inside'.

It is true that in most well-known books about notation, the distinction between the two shapes is not as clear as one would have hoped:

In The art of music engraving, T. Ross simply says: «A tie usually extends from notehead to notehead, but does not touch them». Many of his examples show 'inside' ties but other are also 'across'. G. Read's Music notation is mostly inconclusive. K. Stone's "Music notation in the 20th century" says nothing about the difference between ties and slurs, but the examples clearly show a preference for 'inside' ties. More explicit S. Powell in Music engraving today: «...the most common arrangement has ties attaching notes from just inside the notehead to just inside the notehead, while slurs attach from just above the notehead to just above the notehead».

All in all, as the distinction between 'inside' ties and 'across' slurs is documented -- even if not exclusive -- and making them easy to distinguish at the quick glances one can afford while reading music is important, I am asking if it would not be possible to go a step further, for instance:

extend ties into notehead width only if otherwise too short, while keeping long enough ties at the 'inside' anchoring.

'Long enough' and 'too short' are subjective qualifiers, I agree, but I think that a reasonable threshold can be found.

Thanks,

M.

In reply to by Miwarre

Fwiw, I should mention that Gould explicitly says ties should attach to the mid points except in various special cases. When I look at published music, I have to say, this *does* seem to be the most common arrangement, at least for single notes. The appearance of ties in 2.0 Beta 1 seems very consistent with how they look in most the major publishers I have looked. So I'm not seeing much yet in the way of reason to change at this point, but I'm certainly open to it.

[ EDIT: also, to me, the difference between ties and slurs is always obvious, as ties are always to the same pitch, and slurs never are. ]

One thing I can tell you - for a couple of reasons, picking different attachment points based on distance between notes is not going to be a good option for us. For one thing, there are two different places where this attachment is calculated and used, and keeping them in sync is tough enough as it is. For another, as you (Miwarre) well know, most layout is performed before any stretch has been applied, so we get an artificially pessimistic estimate of the distance between notes. So if the algorithm was to move the ties further over the notes only when necessary, we'd probably end up deciding it was necessary much more often than we should.

Now, it *would* be very easy to make the default attachment point a configurable style parameter. So that people who prefer ties to be "inner" by default can have hat. But given that Gould actually recommends 50% (we used 35% I believe) and this seems well supported by actual practice amongst the editions I consulted, I am not really inclined to change the default.

Do you still have an unanswered question? Please log in first to post your question.