[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Enhancement: Non-power-of-2 note values

From: Simon Albrecht
Subject: Re: Enhancement: Non-power-of-2 note values
Date: Thu, 26 Nov 2015 00:22:50 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.3.0

On 25.11.2015 23:59, Simon Albrecht wrote:
On 25.11.2015 23:45, Noeck wrote:
Hi Simon,

do we have an issue in the tracker for notation like
\relative { c'8  d24 c d }
instead of
\relative { c'8  d16*2/3 c d }
How should it be defined what the shown duration is?

Well, the most common use case is a triplet without indication (like when you indicate the tuplets in the first bar only). And it’s pretty straightforward that a 24th note is 1/24 the duration of a whole note.

Always 2/3 of the duration* if the duration is of the form 3*2^n?
Or the next smaller power of 2?
Would d25 or d15 also be allowed?

The question is: why not? We can document it appropriately, and then it would be the user’s responsibility to choose sensible values, of course, but generally it seems to me to be well-defined. A 16th quintuplet could be written as a 20, as there are 20 of them to a whole note.

Sorry, too short thinking of mine. You mean, that if arbitrary values were possible, it would be difficult to draw a line and to compute a sensible "length" (as first argument to ly:make-duration). And, for an absurdly extreme example, what about a Chopin flourish, fitting 58 single-beamed notes into a 3/4 measure? So, I am getting your point. Normally, we have to code 8*58/6, which makes sense, whereas 48/58 to represent the same duration would be madness and impossible to correctly interpret (as having only one beam! – per proximity we’d use three or four of them).

But, to return in realistic realms, what about limiting it to triplets and (maybe) quintuplets, i.e. 3, 6, 12, 24, 48; 5, 10, 20, 40. In other words, 3*2^n to be interpreted as 2^(n+1)*2/3.

At least it’s worthwhile considering.

And we already have time signatures with non-power-of-2 denominators supported, such as occur in complex modern music.

This looks like a very mathematically and programming-wise inspired
feature request to me. Some syntactic sugar for a rare use case which is
much more understandable if written as 16*2/3.

I think it’s not so difficult to understand, and music theory, such as we have to know when writing LilyPond code, certainly features mathematical aspects. The use case of ‘implicit tuplets’ is not quite so rare, and it’s tedious to write alternations of 16*2/3 and 8, much more so than of 24 and 8.

Yours, Simon

bug-lilypond mailing list

reply via email to

[Prev in Thread] Current Thread [Next in Thread]