Wrong(?) default placement of slurs

• Feb 4, 2021 - 19:05

It might be a matter of taste, but in some cases, I am dissatisfied with the default placement of slurs. See the examples. These slurs had been added with "block select" + "S", I haven't changed anything manually. I think the slur is placed too far away form the noteheads, it doesn't "stick" as it should. (More examples attached.)

2021-02-04 20_03_16-Window.png


It's much easier to assist given an actual score instead of just a picture. Slur positioning is very tricky and depends on many, many factors. In your example, it's almost certainly the natural sign on the C causing the problem. If the end point of the slur were closer to the note, it would then have to be a very steep curve to avoid the natural. So instead we choose to move the endpoint further away. it's a subjective call with no one right answer, but you can always manually adjust to whatever you prefer. You may sometimes need to disable autoplace first for the slur by pressing "=" or using the Inspector. In this case, you could also simply stretch the measure a bit so that natural sign isn't so much in the way. Or disable autoplace for the natural and allow the slur to go through it. Lots of possibilities you can select from.

In reply to by szekelyga

I am no professional engraver, but it should be something like this:

2021-02-05 08_40_47-Window.png

I took me more than 5 minutes to achieve this manually (by adjusting Bézier nodes), so that's a no-go. It is also painful that you have two - equally inconvenient - options to do this:
1) You keep "automatic placement" switched on - in that case, the program randomly changes the nodes' positions after you have worked on them (when you click away from the slur). This is especially painful if you have done a lot of adjustments, and everything is lost without warning when you are finally finished.
2) You switch off "automatic placement" - in that case, the program resets all positions of the slur: placing, size and shape to an equally random "default", and you have to start from scratch with the placement, and can't use the original "automatically placed" version as a starting point.

I think sometimes it is unavoidable for slurs to collide with accidentals. You will see that in professional scores, but you will never see something like this:

2021-02-05 11_25_02-Window.png

I would never pay money for a score that looks like that.

In reply to by szekelyga

And? Well:
1) What, if any, would be your "ideal" proposal in such cases ?
2) Above all, let's wait for Marc's comment, who knows this subject better than anyone else.

In reply to by szekelyga

This is not the answer I was expecting. You're talking about "a lot of other positioning defaults"? I'm talking about a specific case, for the moment, the one corresponding to one of your files, and my GIF. One thing at a time.

Naively, for my part, I would have said: why the slur suddenly changes direction between the A and the G (first note). It seems to me that maintaining the shape of the slur as with the A (and lower) would be a good thing.

But perhaps I will be told that it is much more complex than it seems at first glance, or that there must be a reason that I ignore, or that if one does that, it could break other more frequent cases. I don't know! As said, let's wait Marc's answer, it's preferable :)

In reply to by szekelyga

I quite honestly don't know what kind of "solution proposal" do you expect from me. I expect a better slur positioning, because this one is poor. I am no engraving expert. The team should consult the "master engravers" who advised them to do the changes they did for the new version, and ask THEM for a solution proposal, maybe.

In reply to by szekelyga

Yes, our engraving expert Simon is working on some proposals for better slur positioning in future versions.

But the complication is, there is no one right solution to thorny problems like this. No matter what algorithm you choose, there will be many cases where it looks good, and a few where it doesn't. It doesn't mean the algorithm is bad, it's just that there is no One True Algorithm or even One True Right Result for cases like this. Every single notation program ever written or that ever will be written has cases where you need to fiddle with things manually, that's just a fact of life.

So, hopefully we'll see some improvements to defaults in MuseScore 4, with fewer cases where you need to adjust manually, but the reality is, such cases will always exist, and for the time being, this is one of them. You are correct that professionally-edited scores won't have slurs like that, but that's not because the programs that produced them were magic - it's because professional engravers know they need to adjust slurs by hand, a lot. Many will say it's the one thing that takes the single biggest block of their time.

In reply to by szekelyga

To be clear, the slur layout algorithm hasn't changed in years. The only thing that changed about slurs for 3.6 is the default thickness of the slur. I actually did implement one small improvement but we elected not to include it because it would change behavior for existing scores in an unexpected way. If you're curious, see #314665: Slurs forced above staff when extending over more than one measure.

The issue you are seeing here only affects those very few slurs with large leaps from the first to second note and with an accidental on that second note. For MuseScore 3, we did make a lot of improvement having to do with how slurs responded to such large leaps as well as to accidentals, but this specific combination indeed is still not ideal - not that there is One True Ideal here.

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