MusicXML: Export/Import of Mordents and Inverted Mordent
It seems that import/export of Mordents and Inverted Mordent swaps around these two ornaments.
See attached example produced with MuseScore 3.2.3.7635.
Am I correct that this is an unreported bug?
| Attachment | Size |
|---|---|
| testOrnaments.mscz | 4.03 KB |
| testOrnaments.xml | 8.57 KB |
Comments
Hmm, ISTR a code change for that.
Ah, none that yet made it into the code base, see https://github.com/musescore/MuseScore/pull/4883
And yes, if it is a bug, then a yet unreported one.
In reply to Hmm, ISTR a code change for… by Jojo-Schmitz
Ah OK so if I understand it correctly it is something already in the pipeline as part of a bigger change right?
In reply to Ah OK so if I understand it… by joejoe_
Not sure it is related, it just rang a bell to me...
Seems there's a change in the meaning of these glyphs in SMuFL, probably to fix a bug.
In reply to Not sure it is related, it… by Jojo-Schmitz
Shall I see if I can reproduce the issue after the change you quoted merged into master?
In reply to Shall I see if I can… by joejoe_
Yes, please. If you're on Windows I can get you a build to try with, just let me know...
In reply to Yes, please. If you're on… by Jojo-Schmitz
Yes I'm on Windows and I'll try a build if you have.
In reply to Yes I'm on Windows and I'll… by joejoe_
Try https://ci.appveyor.com/api/buildjobs/nep6k5cv25rdw2sw/artifacts/MuseSc…
In reply to Try https://ci.appveyor.com… by Jojo-Schmitz
This version has actually more issues with the mordents.
If I open the mscz file (as attached earlier), the mordent becomes (a) an inverted mordent as symbol but as text still mordent. And the inverted mordent dissappears altogether.
If I open the xml file (as attached earlier), both become as (a).
In reply to This version has actually… by joejoe_
As written in the PR, https://github.com/musescore/MuseScore/pull/4883#issuecomment-482094234, I'm not really sure ornamentMordentInverted vs. ornamentMordent and ornamentShortTrill.
And even if correct, whether a compatibility layer is needed to convert older scores into newer.
But here's a text build: https://ci.appveyor.com/api/buildjobs/5eie44aago8f2pry/artifacts/MuseSc…
In reply to As written in the PR, https:… by Jojo-Schmitz
I did some more tests with your new build and some more reading.
I now better realise there is a lot of confusion around mordents, where in different historical style periods and languages different names have been used.
Now I prefer the names "Upper Mordent" and "Lower Mordent" as they are hardly open to misinterpretation.
However, now I see that:
https://en.wikipedia.org/wiki/Mordent
Says Inverted Mordent is one of the names for Lower Mordent, having a vertical line.
While
https://usermanuals.musicxml.com/MusicXML/Content/CT-MusicXML-ornaments…
Contradicts this and says Inverted Mordent does not have the vertical line.
Intuitively I think wikipedia is right here and the MusicXML page wrong.
If I'm correct, I would propose for within MuseScore to resolve this confusion by using "Upper Mordent" and "Lower Mordent" as names. And in standardisation contexts like MusicXML, SMuFL I would argue similarly if possible.
For conversion from and to MusicXML I'm not so sure now. If current MusicXML is indeed wrong, should the wrong convention be used as per above link pending perhaps a MusicXML upgrade, or should the right convention be used, so assuming that link is just incorrect doc not reflecting the actual standard?