Re: Chapter "Collision resolution"

[Trevor Daniels]

David Kastrup wrote Wednesday, January 12, 2011 3:40 PM


> Polyphony is not required here, as a proper notehead merge will do 
> the
> trick, like the fourth example shows.  The fourth example is 
> correct
> after shifting an _unrelated_ notehead.

In LilyPond terms polyphony is required when two (or
more) notes of different duration start at the same
moment.  Two (or more) voices are then required to
code this.  Merging noteheads takes place only between
notes in different voices.

LilyPond currently has a documented restriction that
merging cannot take place if there are three or more
notes in the same note column.  Here there are three,
so the third note has to be moved out to permit the
first two notes to merge.  See the Real music example
in the Learning Manual where this is explained.

> But \mergeDifferentlyHeadedOn is used in the two examples before
> already.  Why does it only start working correctly in the fourth
> example?

See above.

> > > Can we do better?
> >
> > Yes.  A better example should be found.
>
> I think we should first figure out why \mergeDifferentlyHeadedOn 
> does
> not work properly without shifting a note head _unrelated_ with 
> the
> collision.

Well, we know why.  It would be good if LilyPond
could be improved to cope with this, but at present
it can't, so we have this documented work-around.

> After we get \mergeDifferentlyHeadedOn to work properly on 
> examples two
> and three, example four is a step backwards, and we should indeed 
> find
> something better.

I'd code that example this way:

<<
 {
   \mergeDifferentlyHeadedOn
   \mergeDifferentlyDottedOn
   c8 d e d c d c4
   \shiftOnn
   g'2 <fis d>
 } \\ {
   c2 c8. b16 c4
   e,2 r
 } \\ {
   s1
   e8 a b c s2
 }
> >

Trevor