Tuplets misaligned
Please see attached score. The triplets on pages 15 to 17 have all gone for a sixer.
I don't know how this happened.
While writing the score it was correct. I tried to re-align the bar numbers by shifting them a little to the right. That was when I observed that the number '3' in the tuplet bracket also got shifted.
How can I rectify this? Would I have to re-write the tuplets all over again?
| Attachment | Size |
|---|---|
| Great_And_Marvellous_Are_Thy_Works_The_Holy_City,_Part_II_-_Chorus_No_16.mscz | 57.58 KB |
Comments
This is very interesting. Could you elaborate how exactly you re-aligned the bar numbers? Did you also have the triplets selected somehow, or what?
Did you somehow just select the "3" part of the triplets when you adjusted the measure numbers? That would clearly be a bug. But I can't seem to reproduce that action.
Clearly seems to be a bug in musescore here. Could you clarify, are you on MuseScore 2.1? And what OS?
So on inspection of the mscx file, I notice that for these Tuplets, the Number element "3" for each of them has an xpos and and ypos:
and deleting those lines for each tuplet seems to fix it.
So I opened the .mscx (uncompressed) in a text editor and manually removed all lines for elements which had "xpos" or "offset". Try the attached file, and let me know if that is corrected.
I am very curious about your exact steps (keyboard and mouse) you took which got these "3" numbers out of position... If you can remember exactly...ideally would like reproduction steps starting from a blank score.
In reply to So I opened the .mscx … by ericfontainejazz
Dear Sir,
Thank you. Yes your manually edited score is correct.
To answer your queries:
I'm using 2.1 version on Windows 10.
As requested I'm reproducing below the sequence :
I’m reproducing below the sequence of steps that led to the mis-alignment of the tuplet brackets.
• Create a new score sheet
• Go to Style --- General --- Numbers --- Interval --- 5 --- Apply --- ok
• Select a Bar number --- select All Similar elements
• Open Inspector ---- increase Horizontal Offset to bring bar number to middle of bar.
• Select a Bar number ---- select Text Style --- remove Frame
• Start entering the score for all voices
•
I have also attached a score sheet where this has been reproduced.
To my knowledge I think the mistake I made was selecting bar numbers universally and shifting their horizontal position.
I did not notice that when I selected bar numbers even the tuplet number (3) also got selected and got shifted.
This is my assumption.
In reply to Dear Sir,… by johnbhasme0315
"I did not notice that when I selected bar numbers even the tuplet number (3) also got selected and got shifted. This is my assumption."
True.
And so, don't use the Inspector for changing the horizontal/vertical offsets of the bar numbers, but do it in menu "Style" -> Text -> Bar numbers.
In reply to "I did not notice that when… by cadiz1
Thank you
In reply to "I did not notice that when… by cadiz1
This is interesting. If you use right click and Select >All similar elements on:
Tuplets - you will only select tuplets
Measure numbers - you will select measure numbers, tuplets and the Title
The Title - you will select measure numbers, tuplets and the Title
...and the inspector doesn't realize you have different items selected.
If you select the Title and a measure number (using ctrl+click) the inspector thinks you have similar items selected, if you add a tuplet number the inspector realized there are different items selected.
Very strange behavior.
In reply to This is interesting. If you… by mike320
So it seems that selecting
"Measure numbers - you will select measure numbers, tuplets and the Title
The Title - you will select measure numbers, tuplets and the Title"
Is the source of the bug. That sounds like an easy fix.
In reply to This is interesting. If you… by mike320
Internally, all of these elements are the same type: generic "Text" (as opposed to Staff Text, or Dynamics, or Lyrics, etc, which are different types). In theory you can use "subtype" to differentiate based on text style even for these generic text elements.
In reply to Internally, all of these… by Marc Sabatella
Interesting.
If I select a measure number, the inspector says it is "Frame Text" element with "Substyle" combo-box set to empty.
If I select a title text, the inspector says it is "Frame Text" element with "Substyle" set to "title".
If I select a coposer text, the inspector says it is "Frame Text" element with "Substyle" set to "Composer"
So I'm gonna try changing the select all logic to select Frame Text only of matching substyle. That will differentiate between Title text vs Composer text vs any other text.
I'm a bit curious why Measure Numbers are treated as "Frame Text" to begin with, and not some other new type of text. Because to me they aren't associated with a Frame.
In reply to Interesting… by ericfontainejazz
Interestingly for measure numbers, if instead of "Selecting All", if I instead select the "More..." option, I see the type listed as "Text", but the Substyle is greyed out. However when doing the same with Title Text, I can check the "Same subtype" checkbox, in which case it only selects title text.
So basically i think I will redefine "Select All" for these text elements to match the substyle.
In reply to Interestingly for measure… by ericfontainejazz
I'm modifying both functions:
void Score::selectSimilar(Element* e, bool sameStaff)
void Score::selectSimilarInRange(Element* e)
with additional case:
else if (type == ElementType::TEXT) {
pattern.subtype = e->subtype();
pattern.subtypeValid = true;
}
I'm noting that for measure number, the subtype is -1. Hard to say exactly what -1 means (maybe "unspecified"?), but that would explain why the inspector's SubStyle combo-box is empty, and would explain why in "More..." popup, the substyle is greyed out.
But turns out selecting by subtype isn't enough. This change does cause Title & Composer texts to not be selected when selecting Meausure Number text. However there is still the problem that the text of the number of tuplets still get selected as well.
So I'm thinking maybe there needs to be an additional subtype "Measure Number". That way selecting all for measure numbers won't select tuplet text as well. And wondering if should maybe also have a separate subtype for tuplets as well, instead of -1.
In reply to I'm modifying both functions… by ericfontainejazz
Actually I've discovered enum SubStyle already has MEASURE_NUMBER. So I guess what I need to do is make sure generated measure numbers are assigned to be this type of substle, cause right now their substyle appears to be left unspecified as -1.
In reply to Actually I've discovered… by ericfontainejazz
Turns out that Measure Number text elements do get assigned the proper subtype when they are created. However the function:
int TextBase::subtype() const
was not properly returning the subtype for case MEASURE_NUMBER. So I will simply add MEASURE_NUMBER to that select case.
In reply to Turns out that Measure… by ericfontainejazz
That fixed the selection problem, so now selecting all measure numbers only select measure numbers.
I'm figured out why the Inspector doesn't have anything listed in SubStyle when Measure Numbers were selected. That is because InspectorText::InspectorText() didn't add substyle for MEASURE_NUMBER. So I've added measure number to that combo box, and now when I select a measure number, I see the Substyle in inspector correctly indicate it is a measure number.
In reply to That fixed the selection… by ericfontainejazz
Made an issue: https://musescore.org/en/node/270532
and corresponding pull request ready for review: https://github.com/musescore/MuseScore/pull/3566
In reply to Made an issue: https:/… by ericfontainejazz
sadly a few hours after making that PR, werner pushed some new code which made it not work.
In reply to Interesting… by ericfontainejazz
I guess you are looking at master, not 2.2? In 2.X, these are (and are reported by the Inspector as) plain Text, not Frame Text.
In reply to I guess you are looking at… by Marc Sabatella
Yes, sorry when i made that comment I was running 3.0 in qtcreator.
At this point I don't think lasconic will accept any 2.2 PRs from me, so I'm not going to bother fixing it for 2.2. I'm not sure if I should even bother fixing up my 3.0 PR cause I don't know if werner's code is in flux at the moment.
In reply to Yes, sorry when i made that… by ericfontainejazz
Maybe here we need an exception, as esp. Text styles are handled much different in master vs 2.x?