Propagating properties for slurs

Just came across what appears to be a pretty serious bug in Dorico Pro 4.0.31. In the full score below, the highlighted slurs are clearly over the upper voice:

In the piano solo part, they get moved to the lower voice, in effect turning them into ties:

Note that this is after “propagate properties” was applied.

This is probably a fairly rare occurrence, but as a small suggestion, maybe the program could check for two identical notes slurred together, and automatically place such slurs in another voice if possible, when creating parts?

If properties have been propagated (or if Global was applied before the slur was positioned) then the Direction property should be ON in the Part. Is it?

I can’t replicate the problem here: Propagating Properties when the slur is selected applies the direction to the Part.

Have you tried propagating again, making sure that the Slurs are selected?

Thanks for your reply. I literally just learned about “propagating properties” this morning, so I’m still just getting to know this feature.

I edited the piano part manually to force the slurs to curve upwards. There weren’t any visible changes that I could notice after propagating properties again just now.

Does there need to be a selection in order for “propagate properties” to work? Or does it always just work on the entire score?

No, it works on the present selection, which makes it possible to selectively propagate, e.g. only slur directions, but not, say, enharmonic spelling etc.

Do you want to attach a reduced file for someone to have a look at?

I’m working on a copyrighted, unpublished score which I’m afraid I don’t have permission to post. (And given my limited knowledge of Dorico, my effort to reduce this fairly complex score with several flows fell flat just now.)

Propagate Properties only works on a selection.

You can avoid having to use it by setting the Set local properties locally/globally switch (far right end of the properties panel) to Global, before altering properties (or flipping things). Again, this will only work for subsequent changes to properties.

It sounds like this was likely the issue then. Probably the process didn’t properly propagate the properties.

Here is the page in the manual about propagating properties. Note that the mode you’re in affects which properties set on the items selected at the moment you do the propagation, ultimately get propagated: i.e. if you’re in Write mode, any graphical offset properties don’t get propagated (because they’re only available in Engrave mode).

Since there’s no bug here where Dorico changes slurs to ties, I’ve taken the liberty of changing the title of this thread.

Happy to learn of the “propagate properties” feature of Dorico, so thank you to all who kindly filled me in.

@dspreadbury: Is this really not a bug? This example may well be just an outlier, though slurs were made indistinguishable from ties in a way that a human engraver would no doubt avoid. If it doesn’t already, maybe the program could check if the notes in any voice are identical when slurs are added, and place the slur above or below non-identical notes if they exist? (Or maybe this doesn’t occur frequently enough to be worth the trouble?)

It’s not a bug because the software isn’t designed to do anything special with this particular scenario (chords where the lower note is repeated). You could call it a deficiency, I guess, but it’s only a bug if it doesn’t work the way the designer expects it to work.

If it doesn’t already, maybe the program could check if the notes in any voice are identical when slurs are added, and place the slur above or below non-identical notes if they exist?

I’m not sure this would always be the right thing to do. Imagine a situation involving multiple voices, for example. Dorico’s default placement of the slur is correct as far as the algorithms that we have created are concerned, which is not to say that said algorithms could not be further extended in future.

You’re far more knowledgeable about music notation than I am, so I’ll happily defer to your expertise. That said, is it ever the case that slurs would preferably connect two identical notes, when there is a clear alternative in another voice? In other words, why would it ever be considered correct to write this:
… instead of this?

The first one is indistinguishable from a tie and only creates confusion or a misinterpretation, whereas the second version is crystal clear.

As a bit of an aside, you ask, “Imagine a situation involving multiple voices”—but these are multiple voices. Just to make sure we’re on the same page, I think we’re running into the issue where what music theory defines as a “voice” and what music notation software defines as a “voice” are not always one and the same. I guess you mean that this example would count as one “voice” as far as the software is concerned, even though these are actually two separate voices stemmed together? (Two notes can’t be sung simultaneously by one singer, so these are by definition two voices.)

There’s a real-life example that springs to mind: Chopin clearly differentiated between slurs and ties in his manuscripts, but the distinction can’t always be shown in standard printed notation. This is how he wrote ties, carefully connecting two note heads:

Here are two examples of slurs from his E minor Prelude; the bottom one is the one in question:

In the 1982 Henle edition, the editor interpreted this slur thus:

Later, in the 2007 Henle edition, the editor printed this, which is technically correct but looks like a tie:

Eigeldinger, in his edition for Peters, copied Chopin’s slurs extending well beyond the identical note heads, thus clearly marking them as slurs rather than ties:

This may not quite conform to modern printed notation standards, but it at least avoids the confusion between slurs and ties. (Too bad the upper voice in the left hand also consists of identical notes, so the workaround of flipping the slurs doesn’t work in this case!)

Anyway, just hoping to learn a bit more about the often confusing distinction between slurs and ties. (Let’s not even get started on Beethoven’s pairs of tied notes, which make for heated debate!)

I don’t think the 2007 Henle edition looks like a tie at all: you would never (for some value of “never”!) position a tie between two stems. Ties always (for some value of “always”!) join noteheads. But of course it’s a different situation to your original post because there are multiple “voices” (i.e. notes with both stems up and stems down) active on the same staff at the same time, in which case the convention is obviously to put the slurs between the stems (which Dorico would also do by default, of course).

In your original situation, if there had been a down-stem note at the same position, then the slur would naturally have been positioned between the stems of the up-stem voice by Dorico, which would give you the result you want. But Dorico doesn’t do this when a single “voice” (i.e. no notes with opposing stems sounding at the same time) is active. Perhaps it could be made to do so in the future, but the absence of it doing so now isn’t a bug, because it’s not something that we designed the software to do.

I’ll certainly think about this as something we might add to Dorico’s automatic slur placement algorithms in future.


Thank you, Daniel! This all makes sense now. :slight_smile:

I’ve been going down rabbit holes trying to disentangle Beethoven’s, Chopin’s, and Debussy’s slurs and ties. Here’s an example from “Golliwog’s Cakewalk”:

Clearly this is a slur, based on the musical context. I suppose if the dot were outside the curved line, then it wouldn’t be so clear. (Normally I just need to read what publishers print, so there’s no need to split hairs and remember all the rules and corollaries and exceptions.) Anyway, these are the kinds of vexing questions that students (and myself, forever a student) understandably pose, and the answers are not always self-evident.

Returning briefly to the original question, does this look like a slur to anyone?

To me it unmistakably looks like a tie, even though it was input as a slur. That’s why it would seem that flipping the slur by default would be preferable:

Again, this may just be an edge case, so it might not be worth tweaking the algorithm, though I’d imagine a human engraver would place the slur above to avoid any ambiguity.

There are subtle difference between slurs and ties. Slurs sit further away, and the ends point to the middle of the notehead. Ties sit closer to the notes, and the ends are drawn in to the sides of the notehead.

The context of the notation (including how slurred notes might be played on the instrument) is going to help determine what’s going on.

It’s standard for down-stem chords to be slurred across the noteheads, even if the bottom notes are the same.

Dorico could automatically flip all slurs on pairs of chords whose note furthest from the stem were the same; though I’d bet there’d be someone asking for the opposite. As things stands, you can press F to get what you want.

I agree with Ben that with study, I’d perhaps guess a slur, but at a glance I’d assume a tie. The bigger issue, at least in this particular instance, and as I see it, is that in your original screen grabs further up the thread, there ARE notes tied across the bar, so I would presume that NOT tying these notes would be an editing mistake. Mm22 into 23 is a prime example. Why the g’s wouldn’t be tied but the A’s are. I would question.

1 Like

That’s literally the very reason why I took such care to notate this piece so precisely! It’s just a strange decision on the part of the composer. My instinct would also be to tie both the G’s and the A’s, but apparently that’s not what the composer had in mind. Too bad he’s no longer around to ask.