Increasing duration overwrites subsequent notes
From the Handbook (https://musescore.org/en/handbook/note-input):
"Increasing the duration will overwrite the notes or rests that follow it".
Why must it be so? Is it due to a difficulty in programming? Surely it's not intentional, is it? If so, can it be optional?
I read that the work-around is to cut-and-paste all the subsequent notes, which might be okay if you just want to change a couple notes, but if you want to increase a dozen notes (spread over a few dozen bars), that seems less than ideal.
Thanks
Comments
This is intentional because in many cases it is the most logical thing to do. Obviously, sometimes you might want some unspecified number of subsequent notes to be shifted later in time. but MuseScore has no possible way of guessing how many notes you might want moved. That is why cut and paste is the way to do this - it allow you to be specific about how many notes you want to move.
Sounds like maybe you are wanting to o something else - not just move a bunch of notes, but actually change the duration of every note in a selection? For this, there are plugins that can help - see the Plugins page under the Downloads menu at top of this page. Eventually it's something we'd like to support more directly.
In reply to This is intentional because… by Marc Sabatella
For working on melodies, the approach I would prefer to take would be to enter all the notes of the melody without worrying about timings at all (e.g. entirely in quarter notes). Once I had selected the correct notes, then I would start from the beginning of that section and start adjusting note lengths and inserting rests to get the rhythm right. With this approach, the timings of the notes after my "focus point" wouldn't matter, as they would be handled later.
Another approach that doesn't cause the same problems in MuseScore would be to notate the proper timings first using the same note for everything, and then adjust the pitches. However, it's much harder to keep track of exactly where one is in the melody in this method, which is why I prefer the first.
With that said, please put in my vote for a mode where notes can be inserted, deleted, or adjusted in length and the subsequent notes shifted to open or close the gap. Of course it might be helpful to be able to specify an endpoint for these adjustments too, but I could live without that. Thanks!
In reply to For working on melodies, the… by oddacorn
Yes, what you describe is sometimes described as a "scratch pad" approach to note input. It's a valid idea, one that might be implement someday, but it is rather different than the way MuseScore works, which is designed to fit the workflow used by most musicians.
In reply to This is intentional because… by Marc Sabatella
Hi Marc,
I don't think we want MuseScore to guess how many notes we want moved. If a user inserts a quarter note, then surely the remainder of the score should be shifted back by one quarter note. Perhaps the user will then, in the next measure or two, eliminate an eighth note here and an eighth note there, and the duration of the score will remain unchanged. Am I missing something?
Also, do you see replies to other comments in a thread you commented upon? I think my reply to mike320 is also part of a good conversation about improving MuseScore on this issue.
It's possible that my understanding of music is not advanced enough to understand the problem, and if so, I apologize for taking your time, but in this case I think my semi-sophistication happens to be enough.
In reply to Hi Marc,… by jonathon.neville
If a user inserts a quarter note, then surely the remainder of the score should be shifted back by one quarter note.
And this is where your thinking goes wrong. It is equally likely that the user wants only that measure to change. Or only that system (up to next line break), or only that section (AKA movement), or any arbitrary selection of measures.
Any guess MuseScore could take here would be wrong most of the time.
And what happens when you insert an 8th note, what happes to the quarter at the end of the measure, should it turn into 2 tied 8th notes? What happens if you delete that 8th again, should those 2 tied 8th get joind into a quarter again? What happens to a triplet that now all of a sudden would need to cross a barline but can't?
In reply to _If a user inserts a quarter… by Jojo-Schmitz
Okay, so if the user wanted only that measure to change, how would they make that happen? They would insert a beat, and then eliminate a beat later so that other measures (or systems or sections) aren't affected. Right? So we don't need MuseScore to guess anything - just do what we specifically request - insert a beat here, eliminate a beat there.
Am I still missing something? If my thinking is wrong, I'm certainly happy to try to understand why... I'd like to think that level of understanding is not beyond me. :)
In reply to Okay, so if the user wanted… by jonathon.neville
And that is exactly what cut and paste does already, only that this is not even restricted to a single measure. You know what you want to move, so you select, cut and paste it.
In reply to Okay, so if the user wanted… by jonathon.neville
Two problems with this:
1) it still takes two steps, so it's really no more efficient in general and less efficient than the current method in cases where in fact you actually don't want anything else moved (which is the majority of edits for many people)
2) in practice, actually shifting everything in the score forward then back would often be disastrous, as the beaming, ties, alignment with other staves, and much else would be altered. So if you had carefully customized how you had notes beamed or tied or how you aligned things with other staves, this work would be lost.
In reply to Two problems with this:… by Marc Sabatella
Point 2 --- ah, I see. .... and those problems don't exist when you cut and paste.... (?) ... I think I see that you're right: If you cut from beat 1 of a measure and paste into beat 3 of a measure, these same problems could exist, but people would not do that (?), they would paste such that beat 1 after the paste matches beat 1 before the cut. (I expect I could say that better, but I think I understand what you're saying.)
Point 1 --- "it still takes two steps" --- not exactly, cut and paste takes more than two steps, or if we want to call the process a two-step process, surely we need to acknowledge that these steps are comprised of sub-steps. The process is: select start note, navigate to end, select end note, cut, insert note (or change duration), paste. Doing that once is not a problem. Doing it several times, spread over several measures/systems/sections - that's a pain.
Point 1b --- "[the suggestion would be] less efficient than the current method"
If MuseScore offered options like "shift to next measure / system / section / end / bar # __", then I could see how inserting here and deleting there (a two-step process) could be less efficient than the current method, but the current method doesn't include any such one-step options, does it? Could it? Would that eliminate the need for guessing? It seems not - it would still need to guess what notes/rests to eliminate/contract. So perhaps include, as another option, "just insert"....
Oh - new thought - perhaps a bit messy - perhaps ignore this: allow, as an option, that a measure changes time signature, e.g. 4/4 becomes 5/4, then the next measure returns to 4/4. Feel free to ignore that.
In reply to Point 2 --- ah, I see. … by jonathon.neville
That indeed is the only sensible option, adjust the measure's actual duration, in some special 'scratch pad' mode
In reply to Point 2 --- ah, I see. … by jonathon.neville
Yes, cut and paste is a process consisting of sub-steps. But so would insert and then delete something else, in some cases, depending on how things flowed across barlines etc. Very possibly many steps if you have to redo all the work you did in specifying beaming, ties, etc. And as mentioned, it's actually extremely common to not need to do the cut and paste step right now - adding a note or changing duration is possible with only a single step in the very common case where you don't want anything else moved at all.
Basically, you are talking about a very different scenario than the common one I am describing. I am talking about entering music that is mostly correct but you want to change some small thing somewhere and want the least amount of harm done to your carefully entered, mostly correct music. You are talking about a case where you have deliberately entered notes of wrong durations and then want to change all of them one by one to the correct ones. That in itself is a much less efficient way of working than simply entering the correct durations in the first place - it takes two steps to do what by rights should have taken only one. But I can see why sometimes, if you don't already know the rhythm but are figuring it out as you go, it can make sense. I do this when, for example, I am attempting to transcribe an improvised jazz solo. However, I use paper and pencil for this type of work because it is a far more efficient tool for the purpose than any software aid I have ever seen.
As I have said before, a scratch pad mode put an entire passage of music in one large bucket to play with could conceivably work well for this type of working, but such a mode would be useless for ther nmore normal type of editing MuseScore currently is optimized to support well. So, rather than destroy what is actually an extremely efficient and useful way of editing music that is already mostly correct, it would be better to implement a totally new mode for working out rhythms that you don't already know to be correct.
In reply to Yes, cut and paste is a… by Marc Sabatella
Thank you Marc for your clear examples. I imagined and described several scenarios but I should have stated the original cause for my writing. You say "adding a note or changing duration is possible with only a single step in the very common case where you don't want anything else moved at all."
That's the exact opposite of my experience, whether working with a 10-page nearly-perfected score or just wanting a scratchpad. In the 10-page score I was fixing, I changed an eighth note to a quarter note, and that eliminated the subsequent note. So I tried to insert the now-missing note, but inserting was not possible. So I proceeded to change all the subsequent notes one by one until I got to a rest I could shorten. I now know I could have cut-and-paste, but again, when handling over a hundred measures, it's a bit awkward navigating to select it all, especially when the process needs to be repeated several times in several different areas. Besides, cut-and-paste is far less intuitive.
COMPARE VIDEO-EDITING:
Perhaps it's because I've done a lot of work in video editing software, but needing to cut everything following a point in order to make an edit, and then paste it back in - that is very counter-intuitive. When I insert or delete or change the duration of a video clip, the whole timeline shifts accordingly, including all other tracks and markings. I guess the presence of musical bars makes music different. ...?
P.S. In almost every other way, I am astounded with how great MuseScore is.
In reply to Thank you Marc for your… by jonathon.neville
Your example is not what I am talking about. You mention changing an eight into a quarter. That's fine, and if that was the only thing you wanted to happen, then you'd have been done. The problem is it is not the only thing you wanted to happen. You actually wanted the next note - and perhaps the next note after, and some unspecified of other notes, to also move to the right. This is a vastly different use case than what I am talking about. I am talking about the case where the other notes are all in their correct places already, and you just want to change one note. This is what MuseScore supports extremely well right now. What you are describing is, again, a totally different use case, one better served with a "scratch pad" mode.
As for cut and paste being "counter-intuitive", you have to realize this is subjective. What is intuitive to one might not be to another. But FWIW. cut and paste is how things are moved from one place to another tons of programs, starting with the text editor in the web browser I am using to type this response. There is nothing unusual about using cut and paste to move notes from one time position to another.
But yes, the presence of meter, barlines, etc makes music infinitely different from video editing, or text editing for that matter. And yet, both video editors and text editors do you cut and paste extensively. It's just a question of knowing when cut and paste is the right tool, and when something else is.
In reply to Hi Marc,… by jonathon.neville
No, it is not true that "surely the remainder of the score should be shifted back". That's just a guess based on what happens to be what you want in one particular situation. In fact, for most users, it would virtually never be the case that they would want the entire score shifted. Consider, I have just finished a 30-page symphony, and now i decide to lengthen a quarter note at the end of a phrase to a half note somewhere on page 7. All I really want is that one note to get longer, and the rest that follows it to get corresponding shorter, so the next phrase continues to start in the same place it did before. That is, I don't want the rest of my symphony affected in any way whatsoever.
For most musicians, these types of edits are far more common than any sort of edit in which you actually want something done in one measure to affect everything to the end of the score - indeed, that would be practically unheard of.
And yet in your case, at least in whatever specific instance you have in mind, it actually seems plausible that you might want all notes to the end of the piece moved. in another, maybe it's just the next three measures - like you entered your piece in sections and got the other sections right but this one is wrong and you are trying to fix it.
So again, any decision MuseScore makes here is going to be a guess that is right sometimes but s=wrong at other times. And from experience with programs that do attempt to make such a guess, I can say with some certainty that you often end up spending more time fixing things up after the fact than if you had simply used cut and paste in the first place.
In reply to No, it is not true that … by Marc Sabatella
We agree!
"Consider, I have just finished a 30-page symphony, and now i decide to lengthen a quarter note at the end of a phrase to a half note somewhere on page 7. All I really want is that one note to get longer, and the rest that follows it to get corresponding shorter, so the next phrase continues to start in the same place it did before. That is, I don't want the rest of my symphony affected in any way whatsoever."
This is a perfect example of my point. We don't want MuseScore (or Finale) to guess anything. We want "that one note to get longer, and the rest that follows it to get corresponding shorter" - but we want to be the ones who decide what gets correspondingly shorter - we don't want it decided by an algorithm. Therefore, if we want to insert a note, or increase the duration of a note, MuseScore should let us do what we ask, and trust that we will take care of making any corresponding adjustments - or at least give us that option.
Am I still missing the issue? I guess I'll let it drop if I'm still wrong, but it seems a heck of a lot easier to let us do insert, change duration and/or delete than to require that we cut and paste - especially if we want to change a dozen notes spread out across several measures.
(Thanks for leading me to discover the plugins - some very cool, but since the notes I want to change are not continuous, I don't think the doubletime plugin or any other plugin would help.)
In reply to We agree!… by jonathon.neville
No, you are missing the point. I don't want MuseScore moving anything in this *extremely common) scenario. I don't want it moving my entire score forward and thus screwing up my carefully considered ties, beaming, and adjustments. I don't want to then have to subsequently take the extra step of shortening something else. I am changing only one note out of thousands, and want everything else unchanged. Moving everything to the end of the score would be, as I said, disastrous.
The best solution is the one that impacts the score the least - the one that causes the last amount of harm. If you only change one note, then you'd like to think that means, only that note gets changed, but as mentioned, this harms your score by creating a measure with too many beats, requiring an extra operation to fix this - assuming you even realize this error now exists. So the the least harmful option is to keep your score musically correct by adjusting the next note/rest only. This has the least effect on the rest of the score. Yes, sometimes it isn't what you want, but it's the least harmful thing to have to recover from compared to the other alternatives, which are, again, disastrous in comparison.
The reason for this lies in the question with what to do if you shift everything after the note whose duration has changed. If you start a 4/4 measure with 2 8th notes followed by 3 1/4 notes and increase one of the 8th notes, what is to be done with the extra 1/2 beat on the last 1/4 note? Is the measure expanded to allow for extra beats? This is not likely to be your intention. Do you move it to the upbeat of the 4th beat and tie it to the beginning of the next measure as two 8th notes? If yes, do you continue possibly creating these ties to the end of the score? If yes, then if you make the other 8th note a 1/4 note, then the tied 8th notes will be moved to start on the first beat of the next measure rather than as a 1/4 note. I haven't even mentioned what will happen to the following measures. Even if there are rests sprinkled in among the notes, the rests will also need to move, because MuseScore has no way of determining which rests you do or do not want. So you can end up with a lot of tied notes you will need to fix after this move, which is far more time consuming than cutting and pasting a section of music. As you can see, this can become very messy and more frustrating than needing to move the notes yourself.
In reply to The reason for this lies in… by mike320
Thank you Mike. In your comment, I see you raise two problems that would occur if MuseScore let us insert notes:
1. inserting a note (or changing the duration of a note) would make the score no longer fit the time signature (I expect there's a better way to say that).
2. inserting a note could lead to a lot of tied notes (and rests). (You're saying that inserting one note could push a longer note across a bar, resulting in a tie, and inserting another note within the original measure could push the first part of that tied note also into the next measure, thus resulting in an unnecessary tie.
For #2, would the tied notes problem disappear if MuseScore automatically simplified durations, or if we could click a "simplify durations" button? MuseScore has a Simplify Durations option when importing from MIDI, so this should be possible, no?
For #1, I'm not clear why we can't let the user take care of any such problems.
In my case, I don't think any such problems exist because I'm wanting to insert a note to fix an incorrect score with an incorrect time structure.
If a score had a perfect time structure, and inserting notes would distort the time structure, based on the types of scenarios I can imagine (below), I think the user can take care of any problems resulting from shifting the time.
(a) Perhaps they want to add a full measure. If the composition is already 16-bar blues, that 17th bar might be a problem - or might not, depending how it's played. But many types of music can add an extra measure without a problem, so ... if the user wants to add a full measure, I imagine that's okay.
(b) Perhaps they want to add a beat in bars 77, 83, 89, and 95. If they are writing in 4/4 time, now there are no extraneous beats. I realize that pushing the rest of the score forward by a beat each insertion would cause the primary beat of a 4/4 rock song to go out of time, no longer at the start of each measure, but not all music requires that structure - not even all dance standards.
(c) If inserting a single quarter note to a 4/4 song meant that an extra measure was added to the end and that last measure ended with 3 beats of automatically-filled-in rests, I don't see a problem with that.
My understanding of music theory could be better, so perhaps I'm missing something, but I thought I should make a case for what still seems to fix a somewhat-serious flaw in MuseScore.
P.S. I completely agree with user 'oddacorn' 's reply to Marc's comment (would you be notified of his reply on Marc's comment?): "For working on melodies, the approach I would prefer to take would be to enter all the notes of the melody without worrying about timings at all (e.g. entirely in quarter notes). Once I had selected the correct notes, then I would start from the beginning of that section and start adjusting note lengths and inserting rests to get the rhythm right. With this approach, the timings of the notes after my "focus point" wouldn't matter, as they would be handled later." That's my story too [edit: when I don't have MIDI input]. If you have objections to this quote, and you haven't seen his full comment, know that he goes into more detail.
Edit: As I wrote to Marc, "I don't think we want MuseScore to guess how many notes we want moved. If a user inserts a quarter note, then surely the remainder of the score should be shifted back by one quarter note. Perhaps the user will then, in the next measure or two, eliminate an eighth note here and an eighth note there, and the duration of the score will remain unchanged. Am I missing something?"
In reply to Thank you Mike. In your… by jonathon.neville
Regarding 1), it's not really the case that inserting as note necessarily means measures would have the wrong number of beats. That's only the case if MuseScore does not do something to account for this - like delete the last note of the measure, or move it to the next measure and then keep shifting from there until some arbitrary point in the future (the next rest? the next double bar? the end of the piece?). As I've observed previously, since only you know how many notes you actually want moved, the chances that this sort of algorithm would happen to guess correctly how many notes you want moved and what you want done with the extras at the end is small, so in practice, when using programs that work this way, you end up spending more time fixing the problems than if you had simply used cut and paste in the first place.
In reply to Thank you Mike. In your… by jonathon.neville
The basic problem with what you are trying to do is that you know what you want it to do. Someone else may want notes to disappear at the end of the measure, or stop moving at some random point later in the score. Each person will have a different idea.
For your (c), moving every note and adding a measure at the end will ruin most scores. Most music has a rhythm with a strong first beat, moving all of the notes will put in on a different beat, perhaps even an up beat. Also, if you have multiple instruments you will have made the instrument you are changing play everything later than the rest of the ensemble. This is rarely the desired result.
As far as seeing other replies, I have read all of the rest of them.
Marc's scratch pad idea is one I would like to see implemented one day, because this will allow the user to create a melody and then apply a rhythm as desired. To fix a single note, cut and paste is always the easiest way to do it. To redo a section of music, the scratch pad would be ideal.
If I were concerned with notes rather than rhythm to keep my concentration on the subject at hand, I would make a temporary file and enter all 1/4 notes with the tune. I would then select the view menu and check Documents stacked so I can see both scores at the same time. I would then apply a rhythm to the notes. I have used this method for writing music, such as when I want to invert or reverse a series of notes. Actually I view different parts of the same score in each window so I don't have to create a temporary score.
Finally, under the Layout menu there is a Regroup Rhythm option that needs to be disabled. It will do what you suggest as far as simplifying rhythms and then the reverse of simplifying rhythms. The problem is that it causes several elements to be deleted from the score and undo will not restore them.
To all who are "against" the insert mode with reasons like "In my 30 pages symphony I don't want that MuseScore moves anything"...
All this is completely pointless: of course the default behavior would be exactly the current behavior of MuseScore, and insert mode would only be activated by explicit choice of users.
And of course the insert mode is mainly useful (for those who want to use it) in the beginning of the process: when notes are being introduced and not too much work has been done yet around the score such as slurs, line break, ....
And there is no question about "what must MuseScore do with the rest of the score" => just push and pull until the end, point. The problem to be solved is that notes over a measure line must not be split in insert mode, only when MuseScore comes back to "normal" mode the split can occur. Which is from an implementation point of view perhaps difficult to do due to the internal data structure of MuseScore.
In reply to To all who are "against" the… by frfancha
And there is no question about "what must MuseScore do with the rest of the score" => just push and pull until the end, point.
Once again: exactly this the wrong approach, a wrong assumption and only the desired effect in a fraction of the use cases.
In reply to _And there is no question… by Jojo-Schmitz
I disagree: this is the need of myself and numerous other people asking for it. Why would you say that this wrong? You think all people asking for that are stupid or bad musicians?
In reply to I disagree: this is the need… by frfancha
I'm not saying that you or anyone is stupid, nor that your workflow or use case is wrong.
Butyour apparent assumption this to be the one and only use case is wrong.
In reply to _And there is no question… by Jojo-Schmitz
Sorry, but is this statement based on any vaid user imput or just on your personal workflow or gut feeling?
Being able to insert at any given point is a feature I heavily used in all other notation programs I've used so far. Somehow it never caused any sort of discomfort so far. As a matter of fact MuseScore's only-overwrite "feature" is probably the only reason I have to revert to using Lilypond for any serious ediring work. I do transcriptions of music notated in mensural notation, often from manuscripts with many lacunae where the interpretation of rythmic values often is anything but clear. So I'm often forced to jump around in the score and change note durations (make a brevis imperfect, make a longa perfect). Dirt cheap with Lilypond (and Finale or Sibelius or Capella ...) but a pain in ... you get it.
Looking at the source code to me it seems as if the internal representation chosen by MuseScore makes inserts rather hard and changing that would involve a major code change (wouldn't the 3.0 switch be a perfect place to do this?). But if this is (one if the reasons) for MS's behavior then please say so. We all understand. Repeatedly telling Forum users how "wrong" their expectations are starts to get kafkaesk.
Just my 0.2$
In reply to Sorry, but is this statement… by rmattes
It is really easy in Capella to create utterly corrupted scores, where every measure contains a different amout of beats, completly disregarding the current set time sig. MuseScore keeps them strict and straight.
And yes, the current design makes it pretty difficult to implement an insert mode.
But before even thinkimg about how to implement, it needs to be though thru to the end what behavoir exactly to implement.
In reply to Sorry, but is this statement… by rmattes
rmattes: who are you talking to, and what statement do you mean when you ask what it is based on? I don't recall a single person ever making a statement that could possibly be construed as implying inserting at any given point could never be a useful feature. Nor does the internal representation of a score have any real bearing on this - it does not pose any significant obstacles. the complication is more about figuring out the msot desirable use model, which is tough when the people asking for the feature most strongly also disagree so strongly over how it should work. Which is why I am trying to propose solutions I think should satisfy all use cases.
In reply to To all who are "against" the… by frfancha
Also note that those you mention to be "against" this method of entry are exactly the same people saying that this kind of entry should exist; but in a scratchpad, non-measureless (hey, no borders to cross now) way. Once your done there, whatever you have now (correctly) notated can be then easily transferred into measures of the real score.
In reply to To all who are "against" the… by frfancha
Obviously the insert mode would need to be turned on and off. A method like scratch pad would be great, it would fix the what do I do with the rest of the notes dilemma.
In reply to Obviously the insert mode… by mike320
<< A method like scratch pad would be great, it would fix the what do I do with the rest of the notes dilemma. >>
There is no such dilemma. The scratch pad would mainly solve the problem that due to the internal data structure of MuseScore which is measure based insert mode is difficult if not impossible to program without major changes.
In reply to << A method like scratch pad… by frfancha
Those who are telling you that you are not suggesting the only fix realize that there are several options as to what inserting a note should do. There most definitely is more than one desired affect from inserting or lengthening a note. A fix that ONLY moves all of the notes to the end is not the way to go about implementing this feature.
If I were to implement it, the scratch pad method would allow for selecting a measure, several measures or even to the end of the score. In this mode ignoring barlines would be possible. I don't program, but if I did, I would allow the user to preview the affect the editing has on barlines in relation to notes so the user could then decide what to do about them. The user would be able to accept the barlines that may force ties, shorten notes (i.e. continue editing), or even change time signatures if needed. It would also indicate if a tuplet would cross a barline (which is not acceptable in most music and impossible in MuseScore) and indicate the error or possibly even split the tuplet in two and connect them with a tie if needed. I think it would be possible to implement this feature with the current measure structure as long as the scratchpad went into a mode that created a dynamic measure that would grow or shrink as needed. Following xml specs while in this mode would be of no concern. Once the edit is accepted, the notes would be inserted into legal measures according to MuseScore's internal structure.
In reply to To all who are "against" the… by frfancha
This is a red herring. Not a single person in this entire thread - or any similar thread - has ever said they are "against" adding a new mode for this special use case. My example was only meant to illustrate that the existing mode really does work very well for its intended use, contrary to the suggestions made elsewhere. A new mode for a new use is a fine idea, one that virtually everyone who has ever responded to one of these threads has expressed support for.
it is, however, far from the case that there no question about what MsueScore should actually do in this insert mode. Sure, there are a few people who believe it would always make sense to push and pull to the end of the entire score. There are at least as many who believe just as strongly it should only be a measure at a a time - that is, it should work more or less like pencil and paper, where inserting a note makes that measure too long but leaves the rest intact. I believe both use cases can be handled well by, again, not asking MuseScore to guess, but instead putting the responsibility on the user to specify for himself what region he wants affected. Select a region, give the command, and the contents of the selection become your scratch pad, with nothing outside the scratch pad affected. If you only want to affect one measure, then start by selecting a measure. If you want everything to the end of the score, then select all. If you want everything to the start of the next phrase, select that. This should satisfy all use cases.