Better way to get correct slur shapes?

[Knute Snortum]

I am hoping to get some feedback from the power LilyPond users on this list.

I have a slurring problem that's not terribly difficult to solve, but I'm wondering if there's a better way to do it.  Here's my example:

%%%

\version "2.19"

\language "english"

slurShapeA = \shape #'((0.75 . 2) (0.5 . 1.5) (0.33 . 1) (0.25 . 0)) \etc

\relative {

  \time 6/8

  \key g \minor

  

  % Could something like this work?

  r8 << { <df'' f>4 } \\ { af8 f } \\ { \voiceTwo \hideNotes f'8 ^( f, ) } >> 

  

  % This works, but it's tedious and brittle

  r8 << { <ef' g>4 } \\ { bf8 -\slurShapeA ^( g ) } >> |

}

%%%

In the piece I'm engraving there are dozens of this kind of slurs.  In the second half of the measure, I've used the \shape function to modify the slur between the two eighth notes to look like it starts at the quarter note in the other voice.  I can do it this way, but it will take dozens of shapes, all controlled manually.  This is a tedious and brittle solution, because any change in the page layout that affects how the measures are spaced may change all the slur shapes (like switching from letter size to A4).

So I was hoping I could use some hidden notes in a third voice to make the slur, but as you can see from the first half of the measure, it's not going very well.  Can this or some other solution work?

---

Knute Snortum

(via Gmail)