lilypond-user
[Top][All Lists]
Advanced

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

Re: Gould &c. question


From: Simon Albrecht
Subject: Re: Gould &c. question
Date: Fri, 22 Jul 2016 00:08:48 +0200

On 12.07.2016 10:16, David Kastrup wrote:
Simon Albrecht <address@hidden> writes:

On 11.07.2016 22:33, Carl Sorensen wrote:
On 7/11/16 7:54 AM, "Simon Albrecht" <address@hidden> wrote:
Hello,

what do the authorities say on beaming something like this:

%%%%%%%%%%
\version "2.19.45"
{
    \time 2/2
    \repeat unfold 4 { \tuplet 6/4 { c16 e g c' g e } }
}
%%%%%%%%%%

Currently, it¹s beamed by half-measure, which I think impairs
legibility, so I introduced a patch changing this to beaming by quarter
note value (issue 4919).
The authorities say to beam by the beat, which is the half measure.  I
think that's why the default is what it is.
However, somewhere one has to draw the line – it’s certainly not an
option to have 16 32nd notes beamed together (unless subdivided).
In 2/2 splitting up the beams sub-beat is just wrong.  Now we have the
"unless subdivided" moniker here and LilyPond should likely do more in
_that_ respect.  In particular when beaming over several tuplet groups,
a subdivision corresponding to the number-carrying groups seems pretty
much mandatory to me irrespective of whether one uses binary subdivision
schemes (which I often find overdoing it) elsewhere.

And I’d draw it just below 16th notes.
For subdivision, yes.  But splitting beams completely inside of a beat
does not make sense automatically since we are then talking about a
judicious breach of rules, and that really should be done manually.

Now the patch is on countdown (again) and I have to return to this argument.
It’s probably a rare case that we’re talking about; 2/2 and 3/2 measures don’t often contain notes smaller than a 16th. But if they do, beaming all of them together without subdivision makes the rhythm very difficult to grasp. Plus, as Urs argued, the number of beams retained at a subdivision should correspond to the duration of the subdivided group, which in this case would mean zero beams. So after all I still consider this a sensible change, regardless of the general rule of beaming by beat („Ausnahmen bestätigen die Regel“ – ‘exceptions confirm the rule’ :-) ). And now, after writing these elaborations, I notice that the question is quite outside the scope of this issue: The previous code already beamed 32nd notes by quarter note instead of by beat. Tuplets may not even have been considered when that beaming exception was introduced, and in fact I just lowered the threshold a bit instead of creating it in the first place. So I would like to not take the patch out of countdown again.

Best, Simon



reply via email to

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