Re: (no subject)

[Mats Bengtsson]

> address@hidden writes:
> > With
> > 
> > \time 6/8 [ c16 c c c8 c16 ]
> > 
> > I get
> > 
> > +--+--+--+---+
> > +--+  +- |  -+
> > |  |  |  |   |
> > |  |  |  |   |
> > 
> > 
> > instead of the desired
> > 
> > +--+--+--+---+
> > +--+--+  |  -+
> > |  |  |  |   |
> 
> > How do I correct this?
> 
> I assume that you know about the manual override. (\property
> stem{Left,Right}BeamSomethingICantRemember (see manual) )

stemLeftBeamCount and stemRightBeamCount.

> By default LilyPond tries to "split" beams at points that have lowest
> denomintor (in this case 1/4 gets precedence over 3/8).
> 
> A better solution is to have the initial split be tunable to allow
> ternary split points.  This should not be a hard hack to do.  Look at
> beaming.cc -- the code is not  complicated.

auto-beam-engraver.cc is more complicated but could still be useful
as inspiration for this extension. Finding the splitting points of
a given beam should be less complicated then finding both the start
and the end point of a beam. 

Even simpler but still very useful would be to add some intelligence
for the case you only specify one of the stem{Left,Right}BeamCount
properties. Now you almost always have to specify both, even though
it's obvious that if you set stemLeftBeamCount=2 in the example above,
you also want stemRightBeamCount=1. This wouldn't give less flexibility
since you could still specify both properties for the rare cases where
the heuristic doesn't work.

   /Mats