Option for last bar to automatically adjust duration to account for pickup
Hi, that's an issue in response to:
#17563: Scores with pickup measure have last bar wih wrong beat number.
I just think that the now default behaviour shouldn't be default at all. That's for a very, very, very concrete type of music. Myself, who mostly works on the field of classical music and jazz had never, ever heard of having to remove beats from the last bar if there was a pickup measure.
See a mere example in a random piece of classical music:
http://erato.uvt.nl/files/imglnks/usimg/5/5b/IMSLP29439-PMLP02286-Chopi…
So, I'd suggest either reverting to the behaviour of 1.2. (last measure is always complete)
Or, creating the ability to set the duration of the last bar as with the pickup measure.
Thanks,
G.
Comments
However, the default behaviuor should be as in 1.2.
A Google search for "anacrusis music definition" yielded a large number of pages, the majority of the ones giving a definition (and not advertising music lessons or the band anacrusis) mention that the last measure should be incomplete.
I find it very strange that you have never come across this in your musical career, bearing in mind the fields you have been working in.
As an example every single piece of the Anna Magdalena Notebooks of J.S. Bach which uses an anacrusis culminates in an incomplete bar: http://imslp.org/wiki/Notebooks_for_Anna_Magdalena_Bach_%28Bach,_Johann…
Now I certainly played pieces from these when I was learning piano, and they are normally part of the standard didactic diet for piano students.
I have no problem with your suggestion for the default behaviour - after all it is easy enough to adjust the actual time signature of the final measure - I'm just surprised you have not met this before, however, my own preference would be for the last measure to conform to the norm as outlined in the Music Theory books - if we take the ABRSM theory exam definition as outlined here:-
http://www.mymusictheory.com/grade3/lessons/12-completing-a-rhythm.html
and
http://www.mymusictheory.com/grade5/lessons/13-composing-a-melody-for-i…
I think it is important for MuseScore to conform to the didactic standards of Western music, particularly if we wish it to be used in schools.
I'd vote for default being last bar compensates for the pickup (i.e. is "incomplete") but is over-ridable when you select pickup at the time of score creation. Random selection of easy folk-tunes attached. A quick perusal of 1000 or so pieces suggests that a) pick-up beats are common in folk music and b) about three-quarters of those that have pickups are also written as having an incomplete last measure to match.
I too am surprised that someone working in classical music hadn't encountered this - it's extremely common in music published 100 or more years ago. More the rule than the exception.
But, I still maintain, it's an archaic convention that started fradually dying off a long time ago and is almost never used any more in modern music publishing in my experience. While I have no problems with MuseScore supporting it as an option (a checkbox in the wizard next to where you specify the anacrusis), I do strongly feel it should not be the default. People familiar with this older convention and wishing to carry it forward into this century will be able to make the necessary adjustment to the last bar manually as they always have. But people unfamiliar with the convention will simply see it as a confusing and frustrating bug - they will have no idea why they can't the right number of beats in the last measure, because they will have no idea that an older convention they have never heard of says it is supposed to be that way.
Fwiw, neither Sibelius nor Finale even offer this as an option. Last bars are always complete unless you explicitly alter them, just as in MuseScore 1.X. Furthermore, they don't make it as easy as MuseScore's "actual time signature" to override it. And schools seem to have no problem using those programs.
I reverted the default behavior to the one in MuseScore 1.2. c20bda5856
This "archaic convention" is, however, part of the ABRSM theory syllabus, and this needs taking into account when deciding what the default should be.
I haven't checked the other examining board syllabi, but I would expect that where ABRSM has trodden the others will follow.
As long as it is *possible(* to create scores that fulfill the pedagogical guidelines of any one particular country's professional associations, I don't see how any one such set of guidelines should affect the *default*. Defaults should be about meeting the expectations of the majority, not the specific quirks of any one set of guidelines.
Musicians already familiar with that guideline will know they want the last measure to have a different number of beats than the rest, and their only question will be how to make that happenand since nothing about the behavior will seem like an outright bug. hopefully at least consider consulting the documentation to see if there is a way to get the job done. Just as they have been doing for as long as MuseScore has been out there.
But to musicians *not* familiar with that particular convention (and I believe this to be the majority), having anything other than the usual number of beats in that last measure will simply look like a bug in the program. They won't go looking for documentation on how to change the number of beats in a measure - they'll start posting bug reports because the behavior will make no sense to them. Probably on the order of several a week. Basically every single newcomer to MsueScore will encounter this sooner or later, and most will find it frustrating.
So I still say, we can solve this very neatly by putting one additional checkbox in the new score wizard. Those users not caring about the ABRSM convention can leave it unchecked. Those who care can, in a single click that is made obvious to them at score creation, get what they want.
In any case, my opinion stays that the last measure should not be compensated by default. In the majority of the case, users don't know the number of measures they want, and they will append measures anyway... The option will just be a one more checkbox, one more to maintain, one more to document, one more to confuse newcomers... So my opinion would be to even not have the option. It's easy enough to right click a measure and change the actual duration.
Still if there is a consensus to add this feature but I think there is not (?) it would be good to propose a title for this checkbox.
The point about adding / deleting measures after score creation is a good one. I retract my suggestion that a checkbox in the new score wizard would solve the problem.
*If* we need this solved, the I guess it would instead have to be a program or style option that caused the last bar to *automatically* be made incomplete as you add or delete bars, and in the case off adding bars, the former last bar would need to automatically be made complete again. Ugh.
The only way I can see this having any real chance of success is if it were somehow tied to the mechanism used to make the final bar line also track the insertion and deletion of measures. For instance, you could not bother to change the actual measure durations at all, but instead change the rendering algorithm to do something along the lines of "if barline style is final, then suppress display of last N beats". Or, you could piggy back off the code that actually adds and removes the final barline, and diddle with with actual durations at the same time.
Or do nothing, and let people discover how they can make the last bar shorter manually.
More discussion in #13457: [Trunk] Actual duration should be nominal duration for all bars, except anacrusis when selecting it during score creation
came up again in #123151: pickup measure doesn't add up with last measure of score
So, it's a feature request. Even aside from the questionable value of the feature (those people who want shrter last bars can do this already), it still not clear how this should work given the fact that measures can be added or deleted after score creation and it's pretty well impossible for MuseScore to know if the current last measure of a piece is intended to actually be the last measure or not, or what to do with the contents of the measure should more measures be added.