Accents plays on acciaccatura when only specified on primary note?

• Jul 11, 2021 - 23:51

If I add an acciaccatura before a note then add an accent to that primary note (but not to the acciaccatura), it still plays the acciaccatura with an accent - I can't imagine any human performer thinking that was natural - surely that's just a bug?


Further, having now played around a bit further and been in some discussions with other Musescore users, it seems playback for acciaccaturas/grace notes is totally broken, and it plays them at the start of the beat rather than before the beat. I assume that's a known issue, hopefully will be fixed for MU4.
As a classical pianist the idea that notes notated to be played simultaneously might be somehow be offset because of grace notes/acciaccaturas is just bizarre. Appoggiaturas are totally different of course.

In reply to by Dylan Nicholson1

It's not a bug that grace notes are played on the beat - it's a deliberate and musically valid choice, in keeping with historical practice (coming from the pianistic practice of playing them together). But of course, it's not the only valid choice, and there has long been a request to also support the alternative interpretation where the grace note comes before the beat. I don't think anyone has ever really made an effort to implement it. but the new playback architecture in MuseScore 4 may well make this simpler.

As for the accent on the main note affecting the grace note, indeed that seems a bug.

In reply to by Marc Sabatella

I'd accept as a pianist you can sometimes make a choice to "crush" a single grace note right onto the primary note, i.e. play them simultaneously, but I can't think of any case where I'd ever delay the primary note until after the beat (again, appoggiaturas aside), and I'd certainly always play all primary notes on that beat at the same time, regardless of whether they had grace notes or not.
It's not uncommon to have 3 or 4 grace notes as a scale before a primary beat, and the way MuseScore plays them back currently makes no musical sense at all.

In reply to by Dylan Nicholson1

If you do a web search, you'll see that there is considerable precedent for the approach of "grace note on the beat, main note after" - and yes, even when there are multiple grace notes. It may not be the most common interpretation by modern musicians playing modern music, but it definitely has a long and perfectly respected history to it.

In reply to by Marc Sabatella

If you search the web you can find almost any opinion! But musically it just makes no sense 99% of the time. As a composer it just wouldn't occur to me a performer might think to delay a primary note because of a grace note.
Do you know of a recording by any respected artist clearly demonstrating that style of playing grace notes?

In reply to by Dylan Nicholson1

It might not make sense to you, but different people have different views on this. Again, just do the search - and if you don't trust Wikipedia (, try any respected published source like the New Harvard Dictionary. You'll find it's not some bizarre outlier interpretation but is actually a very strong contingent. Historically, the shift from before to on the beat is said to have become more common around the 18th or 19th century, but even Chopin was well-known for insisting on grace notes on the beat, as per his handwritten notations in his students' editions of his works.

This is again, not some sort random 1% opinion - it's been established practice for centuries.

In reply to by Marc Sabatella

Ok, for Mozart (and composers of that very specific era) I'd accept that might be musically justifiable for single note acciaccaturas (at tempo the difference would be pretty subtle).
But even for Beethoven at his most Mozartian (think Sonata #19 - measure 14) I'd never play grace notes on the beat - and certainly I can't imagine modern composers expecting that, and it's arguably for them that playback is the most useful feature of MuseScore.

In reply to by Dylan Nicholson1

BTW I did read up on the discussion re Chopin, it seems he was a bit unusual in insisting on grace notes be played on the beat, but then somebody posted some links of performances supposedly done that way, and maybe I'm just missing something but this still clearly sounds like it's before beat - the primary note in the RH and LH are played simultaneously and the B grace note just before:

However in this recording the grace note is played actually slightly after the beat (effectively sounding like an arpeggio from the bass E to the grace B to the primary treble G). Which as a solo performer playing music with level of rubato is fine, but that would definitely be a 1% case (IMO!).

In reply to by Dylan Nicholson1

Something's wrong with the piano roll editor though, it's showing the primary note being delayed by far more than you actually hear it being done:


(BTW I originally copied this from a snare drum part, where the two-note drag rudiment would always be played before the beat, because I couldn't even see them shown as a separate notes in the piano roll editor for a drum part).

In reply to by Dylan Nicholson1

Another very bizarre interpretation:

(I also noticed that the default slur isn't right for this, which is why I flipped it to a down slur. Admittedly it looks ever so slightly inelegant even in my Schnabel/Belwyn Mills edition but at least it doesn't draw itself through the grace notes. Interestingly most other editions don't seem to bother with the slur, but this one has quite a decent looking one. Seems to already be logged as

In reply to by Dylan Nicholson1

BTW, I hadn't noticed you can actually attach grace notes like this to the previous note, which does in fact produce more reasonable playback (it's not exactly right, but good enough).
But obviously the problem with doing that is a) you might not have a previous note to attach to - it won't let you attach them to rests for example, and b) if it's on the first beat of a measure, it will show them in the previous measure, which may well be logical, but it's not conventional, except it seems for trill terminations (the only example I could readily think of are the trills in 1st movement of Beethoven's Pathétique at the end of the development section leading back into the repeat of the main subject, and they're definitely notated in this fashion, with the grace note terminations shown just before the barline). I also note that the grace notes in the RH in the closing parts of the Moonlight sonata final movement need to be played on the beat as the pedal is depressed to stay in harmony (note these are notated after the barline, but the alignment is still between the principal notes). All of which leads me to accept that whether what are notated as grace notes should be played on or before the beat depends entirely on what their function is and how they musically fit into what's around them (this seems to be especially true for Beethoven!).

In reply to by Ziya Mete Demircan

It's not quite correct if the trill starts on the lower note though, it should also terminate on the lower note in that case, which obviously means you can't just assume 1/64th notes are suitable for trill playback (which is the other issue I found - it seems to always use 1/64th notes regardless of tempo meaning trills sound very odd at any tempo below about 80 and above about 120. Typically I'd expect trills to be typically played "as fast as musically possible", regardless of tempo.)

In reply to by Dylan Nicholson1

I've used a baroque trill here so it can properly approach the ending grace notes.

The real playing of trills is never like that of software.
In my school: It starts a little slower at first, then speeds up, and again slows down slightly on the two gracenotes at the end (if it ends with the trill we're used to).
But we don't expect software to do that. This is human interpretation, contains a bit of ad-lib, and can sometimes vary depending on the school.

In reply to by Ziya Mete Demircan

Sure, I wouldn't expect MuseScore to be able to play a trill that had the musicality of a human performer, but I'm rather surprised how little control you do have over it. In fact there's not even the option to turn off the trill playback, which might be what's wanted if you're using a soundfont/sample that has the trill built in.
The ability to specify the speed and turn off playback entirely would seem fairly trivial things to add and would significantly enhance the product - I'd jump in and implement them myself (along with options around grace note playback) except MuseScore 4's playback engine is undergoing a fairly radical rewrite and now's not a good time to muck around with it (also it would actually require some input from the UI/UX designers).

In reply to by Dylan Nicholson1

Again, not just Mozart or Chopin - this really was standard interpretation for centuries. Please do the research. Both styles are worth supporting, and I'd even be open to changing the default to the new interpretation. But it's not a bug that the current default is one that was common use during the common practice era.

In reply to by Marc Sabatella

Centuries? When were acciaccaturas first used? Maybe Scarlatti, but supposedly he used the same notation for appoggiaturas and acciaccaturas and left it up to the player, and probably intended them to be "crushed" onto the primary note. So perhaps most of the 18th century they were played that way, with a bit of leakage either side.
I still don't know of a case you'd ever play a run of grace notes (especially as part of a trill) on the beat though. It may have been a deliberate design choice, but it surely against 99% of any modern user's expectations.

In reply to by Dylan Nicholson1

Acciaccaturas have been used for centuries, yes. Although often they were simply called "short appoggiaturas", sometimes notated with slash and other times with, and scholarly articles will debate at great length about whether, for example, a grace note in some particular Beethoven sonata notated as an eighth note to a quarter main note is meant to be played long or short (see for example, if you have access to JSTOR - most major university music departments do).

But again, to claim that in 99% of cases grace notes belong before the beat is to ignore the overwhelming published evidence to contrary.

If you want one really authoritative example to perhaps finally convince you of this - and the fact that yes, it even applies to multiple grace notes - consider this article from Grove's Dictionary (1904):…

In case the link doesn't work for you (I never understand how Google Books excerpts work), the relevant text says:

"the time required for its performance, whether long or short, being always taken from the value of the principal note"

and then in discussing double appoggiaturas:

"in all cases the time required for both notes is subtracted from the value of the principle note"

with the following shown as an illustration:

Screenshot 2021-07-13 4.05.01 PM.png

As I said, this is not some sort of contrarian viewpoint. It really is exactly how these notes were interpreted for centuries by the composers and performers of the era.

In reply to by Marc Sabatella

That Groves example is specifically for appoggiaturas though? Appoggiaturas by definition to me are played on the beat - I didn't think multiple-note appoggiaturas were all that common though.
Anyway, fine, maybe my initial 99% claim wasn't justified - I'd still think the vast majority of users would use MuseScore expecting grace notes (other than appoggiaturas) to be played before the beat, and there lots of cases where the current behaviour is rather nonsensical.

In reply to by Dylan Nicholson1

BTW after having dug out that Rach 2 example earlier, I realised later in the same movement at Rehearsal Number 15 there's a extended grace note run that is almost always played on the beat even though it significantly delays where the primary note is played:

The notation alignment and harmony makes it pretty obvious how it's intended to be performed (how can you get MuseScore to notate it like this where the grace notes are visibly shown starting on the beat?)
In fact there's not really any musical way of performing that passage without considerable rubato - the first YouTube example I found had the first half note lasting no less than 5 times as long as the second one!
But I wouldn't be surprised if computer software tried to play that by playing that run before the beat, whereas I do find the current MuseScore behaviour very surprising.

In reply to by Dylan Nicholson1

There is no problem in engraving the notes, it can be written in any way.

However, the playing of this arpeggio in the software will not be as expected. Because voice-1 contains a full note, the grace notes attached to it will be played in a longer time. Meanwhile, the notes in the second voice will already start playing. //A few workarounds can be applied for this, but that's out of our scope.

However, the term grace notes denotes more than one thing in the same or similar engraving. For example: what you show in this example is an arpeggio.
Some cadences or runs (played as "ad libitum") are also indicated by such grace-notes.
Many ornaments have a short spelling (symbol), but in a few special cases that are not included, they can be indicated with grace-notes. eg: When playing a special glissando consisting of only certain notes, this can also be written as grace-notes.
These cases aren't covered by the "playing before the beat" issue.

Our object, acciaccatura (as the very short form of appoggiatura), usually contain one or two notes // single and double(*1). Beyond that, it's either: arpeggio, cadence or other kinds of ornaments/embellishments.

(*1): There are theoretical reasons for this that might take more than a few paragraphs to explain, but this is not the right place to go into detail.

In reply to by Dylan Nicholson1

As I said, the term "acciaccatura" is not used by all, historically they were simply called "short appoggiaturas". So, that is what the article in question is about.

Once again, you will find that what I am saying - and have backed up with published examples, which I didn't cherry pick but are practically unviersal in their agreement - is not even the slightest bit controversial if you do the research.

So, no, not 99% of cases that need them before the beat, and not the "vast majority" either. The published record is quite clear - most common practice period composers wrote their music absolutely expecting the acciaccaturas / short appoggiaturas to be played on the beat, each and every time. It's really only music post 19th-century where this ceases to be true.

And yes, maybe that in itself is reason to change the default - maybe most people are playing 20th/21st century music, not baroque/classical/romantic era music. But let's at least be clear that this is the choice we are faced with - whether the default should honor the intent of Bach, Mozart, Beethoven, and Chopin, or whether it should honor the intent of, say, pop sheet music arrangers (the most obvious case where before-the-beat absolutely makes sense).

Of course, one could also argue the default should be on the beat for single or double grace notes, before the beat for groups of three or more. Probably this should all be configurable.

In reply to by Marc Sabatella

Definitely agree it should configurable and in fact be a property of the primary chord the grace notes precede, given in the same piece you may need both styles. The only implementation challenge may be if they occur at the very start of a piece (e.g. the first 3 timpani notes of the Star Wars main theme!)

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