Am 12.01.2016 um 17:28 schrieb Mark Knoop:
> At 17:15 on 12 Jan 2016, Werner LEMBERG wrote:
>>> b) have *no* beamlets at all and let the subdivision be calculated
>>>    as usual (fourth attachment)  
>> This is what I prefer.
>>
> +1. Gould seems to always use stemlets with beamlets when the beam
> count > 2.
>

I'll see how far I can get in this direction. I like the clarity of not
disturbing the subdivision through beamlets. However, it's pretty
complicated because there are so many different constellations (with
regard to placement of the rest(s) in relation to the subdivision and
with regard to the relation of previous and following beam count).

Urs

_______________________________________________
lilypond-user mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/lilypond-user

Reply via email to