Looser enforcement of musical grammar

• Jan 21, 2021 - 00:32

Please don’t assume the only use case for scoring software is the musician with a completed composition doing nothing more than transcribing their work.
Scoring applications can be great writing tools allowing us to try ideas beyond our ability to audiate and play instruments.
But the creative process works better if you can be formally wrong - eg, five beats in a 4/4 bar. MS is too fussy to allow easy experimentation and discovery.

Guitar Pro is one application that has balanced creative freedom with tools that support correct composition very effectively. I think MS could better support “creative” scoring if if did not rigidly enforce musical grammar all the time.


In reply to by DanielR

Yes I have. That does rather open up a whole other discussion.

Creating UI modes is - IMHO - a sign of bad design. Yes sometimes they can help, but they always add complexity and effort (cognitive overhead) and should not be a default design paradigm for anything. The entire (interactive) software industry has settled on a basic select-do paradigm. Once I select an object the software should make every legal function, in the current context, available to. And a function should be available in any and every context where it makes intuitive sense - not just the ‘modes’ to which was arbitrarily assigned. Humans can learn to make difficult tasks second nature - like playing most musical instruments - but no matter how natural modes seem to long time power users of MS I am confident they are objectively awful and send many a beginner to commercial alternatives.

I'm all for experimentation and discovery. Consider that MuseScore allows you the opportunity to learn the accepted way to have 4 beats in one measure and 5 in the next. Just like any skill we try to learn.

In reply to by bobjp

Let me make an analogy using MS-word: let's say that you misspelled and wrote "hoelo". You want to correct it to "hello". We all write typos from time to time, right ?

Now you have two options:

A) you correct it as you do in MS-word, that is: with any combination of del, backspace, adding the right letters first and deleting the wrong letters later,or the other way round, in short: a thousand different ways of correcting it, and they all work.


B) you have to plan the correction as a "chess movement" because the word can never exceed 6 letters- not even temporarily- nor have 4 letters so you have to solve the puzzle of which sequence of keystrokes will correct the typo.

Now which word processor do you think the general public appreciate more?

This discussion went too philosophical about experimentation and things like that. The question is how many keystrokes and milliseconds it takes a beginner to correct a typo in MuseScore and in GuitarPro. and if there is - please note the "IF" - if there is a difference in favor of GuitaPro why do we then keep year after year defending a slower, less intuitive interface.

And this could finish being a matter of taste, because it could be measured. Just take 3-4 randomly selected beginners and film their behaviour in using the keys to correct a similar typo but in musical language and draw conclusions.

Do you get my point ?

I put here a video about my frustration with the key "del". MuseScore tells me by means of a context menu ( right click) that "del" deletes...but nothing happens.




In reply to by ArielAr

@Ariel, what you wrote sounds like music to my ears :-)
The true reason why it is so difficult to add that to Musescore is because the internal C++ data structures are organized around mesures/bars and all the code is written with that assumption.
That also explains lots of other Musescore limitations.
Getting rid of all these would require a major rewrite of the code using "string of notes" as internal data basis instead of measure (like lilypond)

In reply to by frfancha

@frfancha thank you for your kind words and thank you for the most interesting answer I got so far. Yes, I agree with you. Another probable example of the consequences of a design where measures and beats are the fundamental objects is the work around needed to write meterless music: you have a create an exotic e.g. "247/8" time signature and hide it. Finale also has need of such workaround.

In reply to by ArielAr

For the record, no, you don't need to resort to invisible time signatures. msueScore provides several methods of adding beats to measures without the need to create or hide fake time signatures. Inserting notes is the simplest for many case, or joining and splitting measures.

In reply to by Marc Sabatella

oh ! thanks for the info, Marc.

I am interested in meterless music ( specially arabic lute improvisations), and in this thread:


I was recommended creating "a custom time signature of, say, 120/4." This is the same solution that Finale offers for writing meterless music. That is why I mentioned it.

you write "adding beats to measures", please note that when I say "meterless" I mean several rows of staff without a single measure beat/measure/bar, a total lack of beats and measures for many staffs, often for the whole piece..

I appreciate our comment. I would like to use MuseScore for every use where GuitarPro falls short. Being the user of a few other notation packages ( Copyist 9.0, Finale, Forte v11) sometimes I end up confused and splitted mentally. After several months of not using a notation software I feel like I need to re-learn it to use it again.
Thanks for your messages and clarifications.

In reply to by ArielAr

Looks like that advice was very specifically about a particular need to get the actual duration of the measure correct in milliseconds, not about general meterless music concerns. But in case, that was just one random user suggesting one method. It doesn't mean it's the only or best way to solve the problem in general.

Anyhow, yes, I know what is meant by meterless music, and the methods I described are exactly how I recommend doing this. I've transcribed countless Renaissance pieces and composed modern compositions this way. I'm particularly fond of the join/split measure method personally. I join a bunch of measures, start entering notes, and when I reach the point where I think I want to end a system, I split (I have a shortcut for this since I do it so often) and then hit Enter to move to the next line. Two clicks and I'm on my way.

In reply to by ArielAr

I wouldn't use either of those word processors. Wouldn't you just right click the word and choose from the list that pops up?
The idea that a novice should be able to just open notation software and have at it, so to speak, is unrealistic. Suppose I want to write a poem in French. I know nothing about the French language. Don't I have to learn French first? And poetry in any language is different from the regular language.

In reply to by bobjp

Let me give you another example: let's say that you are transcribing a live recording of an ethnic music from caucasus writen in 12/8 and with a very fast tempo. And let's imagine that already the first 2-3 notes are giving you a headache, because you want to transcribe accurately and you feel unsure if the musician is syncopating the melody by just a 1/16th or not , and if he is playing exactly an 8th or not, and you just want to test small variations of those 2 initial notes and see what sounds like the most probably thing.

The question is: which tool will more easily allow you to tweak back and forth the durations of those two notes and hear them played...and this before you even know how the rest of the extensive measure sounds.

I've seen a lot of people that transcribe music this way, by trial and error (listen-> notate->check->correct). Not everyone is like Mozart who wrote in paper hardly without making any correction later.

Here is one example of a proffesional transcriber: https://www.youtube.com/watch?v=uBnJvcXuaPk

My dream is to see Musescore as the tool of choice even for those cases.

Attachment Size
finding.jpg 93.07 KB

In reply to by ArielAr

The problem with any ethnic performance, be it rom the Caucasus or African drummers, is that it was never "written" in any meter. It may have a certain feel to it. But Western notation standards don't, shouldn't, and can't apply to it. You may be able to come close. But that's about it. Players learn by ear, not by sheet music.
Mozart wouldn't be able to do it, either.

For the sake of performers, it's important to make it at least somewhat clear by default. You can change the number of beats in a bar by right-clicking on the bar and clicking on "measure properties," but the best way to do that imo, which is the most clear to the musicians playing your score, is to use a 5/4 bar and then going back to 4/4.

In reply to by L'Moose

You misunderstood me - I am not suggesting being able to create bad scores - which a 5 beat bar labelled 4/4 would be. And I know how to notate tempos.

What I am suggesting is that MS not enforce tempo on measures while editing - at least not in the heavy handed way it does now. An incorrect measure should be flagged - maybe by being rendered in red when the focus leaves the measure. But it would be much better if MS was not correcting me while is inside a measure doing my thing - and it shouldn’t prohibit incorrect measures at all. I may want to leave something half finished for a good reason and intend to come back. And I don’t want to have to set free time just so I can play around.

Playback could just add implied rests on under filled measures and skip extra notes on over filled measures - as is the case in other applications.

In reply to by Terracerulean

Are you aware that you can insert notes while in the default step time input mode by holding CTRL+SHIFT and pressing a note name (or zero for a rest)? See https://musescore.org/en/handbook/3/note-input#insert-notes. Similarly you can remove a note and its associated duration from the measure using CTRL+DELETE (scroll down a few more lines from that previous link). The insert mode is more helpful when many notes need to be inserted at one time (for example when writing a cadenza).

"Incorrect" measures are flagged as you are requesting but with a subtle "+" or "-" indicating actual lengths longer or shorter than implied by the time signature as described here: https://musescore.org/en/handbook/3/measure-operations#duration. There is no prohibition or warning (either would be rather presumptuous), just a rather discrete indication that the length of the measure does not match the time signature (which is sometimes what one wants in the finished score in cadenzas for example).

Having playback skip notes that are present in the score using an arbitrary rule to decide which are "extra" seems less than helpful. Whether notes are played or not is currently completely under the user's control by unticking or leaving ticked the "play" tick box in the inspector.

In reply to by Terracerulean

As someone who uses notation software strictly for composition, I'm not sure I see a problem.

"Playback could just add implied rests on under filled measures and skip extra notes on over filled measures - as is the case in other applications." I have no idea what this sentence means. MuseScore is obsessed with rests. So much so that if I what to add a note to beat four of an empty (of notes) measure, I have to deal with rests on the other three beats first. Skip extra notes? I presume you put notes on the staff for a reason. Why would you want them skipped and how would play back have any idea which note to skip.
Personally, I don't believe that the creative process is a free for all on any level. Any art form has rules that should be followed if it is to be recognized as a particular art form. Put another way, part of the art is working within the medium. I guess for me, my process of audiation doesn't let me add extra beats or do things out of sync with the style I am writing.
I don't think we learn to write language properly by relaxing grammar.

In reply to by bobjp

Playback is simple. It defaults to the end of the bar - if there are not enough notes in the bar it adds the requisite silence. If there are too many it just doesn’t play any after the bar is used up - That’s the way guitar pro works and it never creates a problem.

As for the rest - no one seems to have considered that I am talking about ease of use during the editing process - being able to change things in the order I want rather than being forced to do things in one specific order. I never asked for bad musical ‘grammar’

In reply to by Marc Sabatella

Hi Marc ! May I ask what the status is of such "scratch pad" ? Different threads around it confuse me a bit when it comes to the status. I dream of the day Musescore has a note entering method EXACTLY like GuitarPro, or Microsoft Word, something sooooo incredibly intuitive that it does not need to be "learned", anyone just guess right out of intuition. Please let me know the current status. Thanks

In reply to by ArielAr

I would claim such a thing doesn’t exist. What’s intuitive for person is incomprehensible to another. Word has nothing even remotely like more input, and I’ve never been able to figure out Guitar Pro despite extensive experience with music notation software. But it’s certainly possible that someday GP-like input could be implemented someday. That’s nothing like the “scratch pad” idea that’s been discussed though.

In reply to by oMrSmith

I know I'm just a grumpy old man. That's why the idea of a "scratchpad" or "creative scoring" don't make much sense to me. The score I'm working on is my scratch pad. I'm not sure I'd call suddenly needing an extra beat, creative. Possibly just the opposite. But that'a just me. I know.

In reply to by bobjp

Really wasn’t advocating for bad scoring- just more flexible editing.
You might ‘need’ an extra beat switching from 4\4 to 5\4. What’s wrong with being able to add it before adjusting the time signature? Why should I be forced by the software to change the time signature first?

In reply to by Terracerulean

Well of course we all work differently and have different needs. Interesting that you look at it as being forced to do something. I look at it as the software helping me to a finished product faster.

Guitar Pro way. So I have a phrase in mind. I enter the notes I want. Hmm, too many notes in my idea. The software won't play all the notes I want. So I have to go back and re-work my idea.

I'm probably missing the point. But I see that as being more time consuming. I realize that probably most people write music far more complicated than mine. They need to be able to switch time signatures and what-not a lot. When I sit down to write, I almost never have anything in mind first. I might have a mood in mind. I just start entering notes and let the notes take me where they want to go. They are in control. But it isn't a free for all. The software helps keep both of us (me and the notes) in line. Sure, there are times when the notes ( being smarter than me ) try to pull a fast one. Or times when I ( being kind of slow ) forget what I'm doing. But the software is relentlessly organized. Right from the get go. Someone, or something has to be.

But then again, we all have different needs.

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