Re: wrong beamlet direction in 6/8 and 3/4 measure for dotted quaver and semiquavers

[Carl Sorensen]

On 1/15/12 3:06 PM, "Janek Warchoł" <address@hidden> wrote:

>2012/1/15 Martin Straeten <address@hidden>:
>>> I'm not top posting.
>> in 6/8 an 3/4 measure the beamlet of the last/first semiwaver
>> points to the dotted quaver.
>>
>> semiquavers should be beamed together
>
>I may be wrong, but i don't agree.  Beaming them together would be
>against the rhythmical subdivision in 6/8 and 3/4 meters.
>

I agree with Martin. One of the important things to recognize (which
LilyPond doesn't do yet in the beaming pattern) is that complex meters
(like 3/4 and 6/8) have a fundamental beat unit of 3, and so the beaming
that is chosen is wrong.

The fundamental unit of 6/8 is not 1/8, but 3/8.  6/8 is two groups of
3/8.  The division at half of the 3/8 group is a more important division
than the division at 1/3 of the 3/8 group, so the beaming should reflect
that division.

In the old code, this was handled by trying to avoid "sticking out" flags.
 That was tossed in order to recognize beat divisions, and the beat
divisions work properly in simple meter, but not it complex meter.

Please make this a regression, with a title that says "Beamlet orientation
is wrong in complex meters".

Thanks,

Carl