Dorico presents a 7:8 tuplet as one of the given tuplets:
E. Gould gives 7:4 is the standard tuplet following the rule that tuplets replace fewer regular notes of the same value with the exception of duplets, which are exceptional.
One does see 7:8 tuplets in the literature, but one would think that Dorico would present what is considered to be the more correct form.
I agree with you, but Nadia Boulanger probably would not have. She insisted that a septuplet in the space of a quarter note be written as seven 32nds. So in that case a 7:8 ratio makes sense.
Thanks rkrentzman, but whoâs boss around here? Elaine Gould or Nadia Boulanger?
âSeptuplet
A group of seven notes of equal value to be played in the time normally occupied by four or six notes of the same value.â The Harvard Dictionary of Music. Fourth Edition.
If one follows a âclosest toâ rule and uses, say, a 15:16 tuplet, is it then a 14:16 or a 14:8, a 12:16 or a 12:8 etc.? This system would seem to lead to chaos.
Please donât take us down another semantic rabbit hole.
The description 7:8 is unambiguous. That you might prefer Dorico to show 7:4 or even 7:6 is irrelevant.
As an advanced user, I doubt you ever use that panel, preferring to use the popover instead.
Iâm not sure what semantics have to do with it, but I was rephrasing what Gould says on page 202 of Behind Bars. Her option 1 is the standard 7:4 tuplet. Then, she deals with option 2, the 7:8 tuplet. She recommends the 7:4 for the reasons given.
Exactly - youâve gone down the hole.
My point is that Dorico provides a few tuplet possibilities in the left panel. It is not a complete/exhaustive list.
It includes one that you do not like. But that one is not illegal or wrong. (Personally I may be really, really upset they didnât include 11:8 to facilitate my chromatic runs, but 7:8 is fine for diatonic scalesâŚ)
You see how ridiculous your pearl-clutching is?
I may be wrong, but I thought that Dorico attempted to present best notational practices. I donât think that it is living up to that goal in this case.
I agree that itâs generally better to be consistent and use contracting tuplets rather than occasionally mixing in expanding tuplets, even if closer in value. Gardner Read agrees too and uses this example from Charles Ivesâs Violin Sonata No. 4 as an example of a publisher doing this wrong.
I almost never have the left panel open in Write mode though, and certainly never use it for tuplets, so the default tuplets are sort of a non-issue for me personally.
I wouldnât consider the Ives example to be wrong. The most basic rule for me is that the tuplet must be derived from either the next faster or slower regular value, skipping in either direction is off the mark (unless you want to confuse performers deliberately, in this case ratios are mandatory).
Further rules - or better: conventions - apply of course if readability comes into play, and thatâs when traditions start to matter, but also the particular context of the musicâs structure.
In the given example, the septuplet directly succeeds regular 32nds. Writing the septuplet with 16th beams would be possible, right, but it would also disrupt the musicâs graphical representation to a degree that is not equivalent to the flow of the played music itself. In this case, choosing an extending tuplet is more appropriate for this context in my opinion - the only thing that may have been wise to add would have been a ratio to indicate the extending tuplet.
Looking at the preceding and following bars of the Ives sonata, my personal verdict might turn out different though, as there are chains of extending septuplets following as well as lots of contracting tuplets having been played before, which would support switching to 16th beaming right away.
I think the most important factors are readability and a consistent representation of the musical flow. In cases like this, there are pros and cons for both options (and thus, no definitive solution), and to me, it is perfectly readable the way it is. After all, in chamber music it is common for the pianist to have the other playersâ parts in cue note scale in their own score, so the former always has a reference frame that leaves no questions.
There are other imaginable cases in which an extending tuplet may be preferred: Imagine a piece set in 12/8, 9/8 etc. (3n/8 to be comprehensive) in which one player needs to play straight rhythms for quite a few bars. Excluding the option to have that player switch to n/4 signature for this section, one can either write a) contracting tuplets, b) extending tuplets or c) dotted values throughout. In the case the straight rhythm contains duplets interchanging with quadruplets, matters become complicated: An extending duplet (which is the regular choice for intermittent duplets) would have the same beaming as a contracting quadruplet, which makes things seriously awkward when theyâre in close succession, so if the dotted notation is to be avoided and readability is to be maintained, one has to bend the rules and include either extending quadruplets (4:6) or contracting duplets (2:1½). Personally I would prefer the latter while avoiding indicating the ratios altogether, but extending tuplets might be easier to read as most players are more used to quarter note beats than to half note beats - unless they are predominantly playing Early Music.
Aww, but itâs so exciting down there! Please! I promise weâll be home for dinner!
FWIW, a specific composer I work for consistently uses expanding ratios (7:8, 3:4, etc.).
While the math mind may not agree, how the musical mind feels (more or less notes than they should normally be, or, feeling like an accelerated or decelerated run) has its importance.
Sometimes we forget how the music should feel instead of just how it should look.
And yes, @klavierpunk, greetings from layer 666 of the rabbit hole! 
Genuine question: Wouldnât it be clearer to use 16th beaming to indicate at first glance that the note value is slower than the preceding 32nds?
In this case, the
would actually help getting this right at first sight⌠
B.
I engrave for a modernist colleague and we have scores say 90 pages or more long strewn with dozens of 7:8 tuplets, and 5:6, (and 5:6 inside 7:8 etc) and so on and so forth. Itâs perfectly clear. I have to say that Gould is not a divine authority residing on Mount Olympus.
The Dorico sidebar is only an abbreviated convenience, not intended to be comprehensive. I find it quicker and easier to manually specify tuplets anyway.
I donât think they have ever stated that. Dorico aims to provide the tools and functions for good common notation, but they also provide enough flexibility for you to âbreakâ rules, for say, modernist works. But it is squarely aimed at Common Era music, and over time they are expanding to other periods and styles (e.g. cutaway scores).
They do lean on Gould heavily as a guideline, and this has been explicitly stated if I recall.
Thatâs what Read was saying in the Ives example. Using a contracting 7:4 16th note tuplet instead of an expanding 7:8 32nd tuplet would show that it is slower.
There was a notat.io thread on some tuplets recently (as well as a MET post filled with laughably awful advice) where there was some discussion of expanding vs contracting tuplets. 7:8 obviously is quite commonly found, but Iâm in favor of always using contracting rather than nearest value so Iâll go with 7:4 as Read advises.
The one that is probably the biggest issue for me, and likely most controversial, is 2:3 in a compound meter. The modern trend seems to be always use contracting. Hereâs Hindemithâs Elementary Training for Musicians:
By always using contracting you avoid this possibility where you have both expanding and contracting in the same bar, which just seems confusing to me, even if historically not uncommon:
Was he contractually obligated to expand that? 
This topic always gets me as I try to wrap my head around the logistics of a 7:8 versus a 15:16, etc. I mean a straight triplet, all well and good. But when I have to break out a calculator, slide rule and calculus textbook, I get on more âshakyâ ground. I am working on a piece and put a 5:3 into Dorico and it took me quite a while in order that Dorico would get what I wanted in there properly, but I finally managed to achieve that goal. After that I needed a nap! Haha.
All tuplets follow exactly the same rule: x notes in the space of y notes, where notes all have the same duration [crotchets, quavers etc]. The base duration can be specified explicitly [as q=crotchet, e=quaver, even e.=dotted quaver is possible!], or you can use the currently selected note duration.
Where Dorico is really clever (and to some, mystifying if they forget) is that the definition ignores barlines. So tuplets can quite happily span multiple bars!
Aargh, thereâs the skipping I mentioned, thatâs 9:4 instead of 9:8⌠With the reference frame below, itâs perfectly readable though, despite the mathematical-notational atrocities.
