Another tuplet request/query

[judddanby]

@kgdrums, you can distinguish Steinberg/Dorico team members by the red Steinberg volume knob logo that either surrounds or appears in the lower-R part of their photos.

As a fellow forum member, I am disappointed by the “shouty” tone that has emerged in this thread. Good people, it’s just a request for alternate ways of notating tuplets countered by questions and comments about common practice, after all — not an assault on all that’s right and good in this world! (And while I recognize the ironic way in which many invoke the term “notation police” to capture something real about resistance to admitting newer notational practices, I think we should be more mindful about taking this metaphor into the realm of actual and horrible police brutality (!).)

This forum functions on the good will of fellow members volunteering their time and expertise to help each other, as well as the Dorico team helping but also explaining and listening to requests. If we descend into ad hominum attacks and assumptions about one another’s imagined intentions, bedlam will ensue.

There are so many generous and thoughtful members here from whom I have learned and benefited in my Dorico use, and I look forward to this staying a “good space.”

[stephenw56]

Hi All, and in particular @dspreadbury

I would like to go back to the origin of this thread and voice my support for this feature in Dorico. As a long time Finale user, the implementation of this feature would be much appreciated for anyone who understands and uses the mathematical relationship between notes. The below example outlines these relationships and how Finale allows users to choose the note value that the user is creating a tuplet relationship to. This is a necessary feature due to the current evolution of rhythm in both the contemporary jazz and classical communities.

With the ability to choose the note relationship, a composer is able to better communicate a relationship to a performer. The polyrhythm 5:2 feels and sounds different than 5:4, therefore if a composer is to communicate a particular relationship they must have the necessary tools to do so.

I would argue that this notated relationship is that of a ‘courtesy tuplet’, meaning that it is a marking used not to display a strict mathematical accuracy, although all of the listed tuplet examples are mathematically accurate, but instead to communicate more effectively their relationship to the preceding music.
The second attached example is from Totentanz (Adès), where the irrational 2/6 bars have courtesy tuplet markings above the quarter notes. This communicates to the performer the relationship between the rational and non-irrational bars. The value of a quarter note in a bar of 2/6 is a quarter note triplet. The triplet marking above the quarter note is to show its relationship to the preceding quarter notes and is not to create a secondary quarter note triplet over the quarter note value of the bar.

It is beyond me why this request thread has devolved to non-dorico-employed forum members questioning the validity of a request. Since MattMitch and kgdrums are not going to flaunt their own accolades as composers and performers, I am happy to weigh in that they are on the forefront of rhythmically complex music and their requests should be considered with due diligence.

[dspreadbury]

It looks like @MattMitch has quickly organised a write-in campaign from his friends and collaborators to ensure that we take this request seriously

This is impressive, but unnecessary – if a requirement is genuine, it only needs one request to make it on to our backlog, though I can’t make any guarantees about when we might be able to actually tackle it.

The way Dorico represents tuplets internally is using a ratio expressed in a single value. The number on each side of the ratio has to be an integral value, and the unit of the tuplet is what we call a “notehead duration”, which means a length that can be notated by a single notehead and up to four rhythm dots. (So that would include, for example, a dotted half note, but would not allow a quarter note tied to a sixteenth note.)

Of course it’s possible for us to change this implementation, but it’s a decidedly non-trivial thing for us to do: tuplets are a pretty fundamental type of item in Dorico, and any change to a fundamental item requires significant refactoring and extensive regression testing.

This is why I was asking in my initial replies in the thread about how durations that cannot be expressed as what we call a “notehead duration” would work. If it’s critical that the duration of either side of the tuplet must be completely arbitrary, and allow a different (potentially fractional) value of any kind of unit on either side, then there would be little advantage to us trying to implement a kind of halfway house solution.

If, on the other hand, it’s normally the case (leaving aside examples like the Adès) that the overall played duration of the tuplet can be expressed as a “notehead duration”, then we could potentially make it possible to maintain the same internal representation of the tuplet (i.e. an integral ratio expressed in a single “notehead duration”) while adding options to display the quantities on either side of the ratio in other values.

For the original example in @MattMitch’s first post, a tuplet of 5:4 in eighth notes could indeed show its right-hand value as one half note or two quarter notes. The left-hand side couldn’t be expressed another way, as five eighth notes is not expressible as a single notehead with up to four rhythm dots.

Would this be sufficient, or for the purposes of @MattMitch and friends, would it still be insufficiently flexible?

[FredGUnn]

This has come up quite a few times before with lots of published examples. Just searching my own posts turned up this,, this (7e=w), this, and this … I’m sure there are plenty more. It is trivial in Finale where you can specify any value in the space of any value.

[Janus]

Thank you for these.

I see the notation is used neither consistently (in any of the examples), nor (IMO) helpfully for the player.

[MattMitch]

Thanks, @dspreadbury. I swear I didn’t intentionally do so
And I can certainly imagine that such a change is decidedly non-trivial!

@dspreadbury

If it’s critical that the duration of either side of the tuplet must be completely arbitrary, and allow a different (potentially fractional) value of any kind of unit on either side, then there would be little advantage to us trying to implement a kind of halfway house solution.

If I’m correct in inferring that you’re referring to potential “partial tuplets” here - well I wasn’t even originally talking about those, but yes that would be definitely incredibly useful and welcome to have them integrated in such a way. Having both sides of the tuplet be arbitrary to that degree would indeed be the true holy grail of rhythmic flexibility, and this would be the ideal scenario, to be quite honest.

@dspreadbury

If, on the other hand, it’s normally the case (leaving aside examples like the Adès) that the overall played duration of the tuplet can be expressed as a “notehead duration”, then we could potentially make it possible to maintain the same internal representation of the tuplet (i.e. an integral ratio expressed in a single “notehead duration”) while adding options to display the quantities on either side of the ratio in other values.

For the original example in @MattMitch’s first post, a tuplet of 5:4 in eighth notes could indeed show its right-hand value as one half note or two quarter notes. The left-hand side couldn’t be expressed another way, as five eighth notes is not expressible as a single notehead with up to four rhythm dots.

This would indeed cover my original request and absolutely be better than nothing.

To be honest, there is an element of composing where I self-impose limitations as a result of the relative difficulty of implementing more ambitious ideas smoothly in notation software sans clunky workarounds. Granted, that is my problem, ha, but again - the “any unit or number on either side of the tuplet, including fractions” would be truly stellar.

Thanks very much, Daniel, for the consideration, and thanks all for the continued discussion.

-m

[MattMitch]

@FredGUnn

Just searching my own posts turned up this,, this (7e=w), this, and this …

thanks for citing these, I missed most of them in my previous searches!

@judddanby

placing the tuplet grouping within a durational span

This is a great way to put it and has real relevance to the experience of performing this type of rhythmic material. Thanks.

-m

[judddanby]

@Janus

Thank you for these.

I see the notation is used neither consistently (in any of the examples), nor (IMO) helpfully for the player.

Always glad to contribute as my library (and recall of it) allows, @Janus.

I imagine you’re talking more about the inconsistency within the second example, which I note as well.

As for the “helpful” aspect, I’ll leave that to performers more immersed in reading/performing this kind of notation to address.

[Janus]

@judddanby

I imagine you’re talking more about the inconsistency within the second example, which I note as well.

Also in the first…

Why is the boxed quintuplet not using traditional notation like the triplet?

(The reason I find the novel notation odd is because they often have tied notes at the input or exit, so the notation as a quaver or whatever is sonically meaningless)

[FredGUnn]

I’m out of town and don’t have access to my usual texts, but I would imagine Gardner Read’s book Modern Rhythmic Notation will have plenty of published examples of similar tuplets if anyone has access to it. (I can check when I’m home in a few days.)

[kgdrums]

Thank you so much for this response!! If I understand correctly, what you’re talking about would totally solve it for me, personally. This is REALLY exciting that it’s a possibility.

I do use irrational meters and ‘incomplete’ tuplets sometimes, but so far have always been able to take care of this via the time signature and haven’t needed fractional values in the actual tuplet label.

Also thanks for explaining about it only taking one request- that is of course great to know.

Re: the “write-in campaign”, that was probably my fault… There’ve been a fair number of people I know who have expressed hesitance to switch to Dorico primarily because of the tuplets thing (though I’ve always encouraged them that Dorico is awesome and worth learning, saying “I bet they’ll fix the tuplet thing eventually”) So when MattMitch posted, I of course shared the thread since I knew it is something lots of people I know care about. My intent was not for it to turn into the craziness here, but I hoped if it was clear that it wasn’t just a couple random people who cared about this, it might somehow positively affect things and communicate what a huge difference this would make.

It is really great to know Dorico might have this in the pipeline in some way, at least, and is taking it seriously- thank you so much!!

[ronn.pricer]

Hello there, I’m new to Dorico after using both Sibelius in Oregon public schools, and Finale as a full time music engraver for Christian Copyright Licensing International. I found your posted examples fascinating, and educational, esp, the B & H one. Personally, I had never encountered anything like what MattMitch led with in this thread (hope that’s right; I’m new to the whole forum thing as well), the whole “5 : 4”. I initially thought the 4 represented a “quarter note”. That’s my American background speaking. In my experience, notation of such things as measure 4 on the B & H example, rarely had more than just a single number with either a beam or a bracket, assuming the performer, and/or conductor would be able to discern how many beats the phrase took up.

For example in measure 4:
V1 is expected to play 7 notes in the space of one quarter note (crochet).
V2 is expected to play 9 notes in the same amount of time
Vla is expected to play 5 notes
Cello is expected to play 6
And none of the notations for these includes a ratio term like 5 : 4, but the convention in question may only be with respect to passages that cover more than one beat. Is that right?

In this second case, I do like the fact that your circled examples use the actual note as part of the ratio. Where a number could be interpreted as either representative of the number of beats or the number of notes, the actual note removes all doubt.

Well done. Thanks.

Ronn

[Janus]

@ronn.pricer

For example in measure 4:
V1 is expected to play 7 notes in the space of one quarter note (crochet).
V2 is expected to play 9 notes in the same amount of time
Vla is expected to play 5 notes
Cello is expected to play 6

I would probably notate these as…

Much easier to parse.

[rkrentzman]

I basically agree with you but I think the nine-tuplet (?) should be notated as a nested tuplet (three sixteenth tuplets under a eighth-note triplet).

I have a client who insists all septuplets be notated as 7:4, (seven 16ths in the space of 4 sixteenths) but without ratios on most other tuplets.

[SugarFree]

this doesn’t add up.

[klavierpunk]

To me this is readable as an eighths triplet with the last eighth divided in a sixteenths triplet. Triplets are so common that in many cases they do not need to be indicated throughout, in fact, indicating every single one of them would make a score look overly crowded, if they constitute a fundamental rhythmic feel.

[judddanby]

While I understand your point generally, @klavierpunk, in this case — given the rich mix of 5s and 7s in with those 9s — I think, like @SugarFree, I’d prefer the nested triplets to be explicit, at least much of the time. (The second violin probably doesn’t need them in the second measure of the second system.)

[klavierpunk]

Yes, it depends on the context. If this is not the first measure of the piece and the respective player has been confronted with this rhythmical feel multiple times before, that notation is sufficient, but if it is the first time showing up, the nested triplets should be indicated.

[John_Ruggero]

Janus’ notation was instantly clear to me and the first level of tuplets, obvious. I admire the composers of that past for whom simplicity was goal. Not a note or a marking too many or too few both played and written.

[SugarFree]

@John_Ruggero

I admire the composers of that past for whom simplicity was goal. Not a note or a marking too many or too little both played and written.

Yes, I agree with this. The original B&H notation was both clean and unambiguous.