Re: Ties on series of tied notes on center line flipping direction (randomly?)

[Richard Shann]

On Sun, 2022-01-09 at 17:43 +0100, Thomas Morley wrote:
> Am So., 9. Jan. 2022 um 17:16 Uhr schrieb Valentin Petzel <
> valentin@petzel.at>:
> > 
> > Hi Richard,
> > 
> > It is hard to try to understand the problem with just some
> > pictures. Do you
> > have some code you can send us?
> > 
> > Cheers,
> > Valentin
> > 
> > Am Sonntag, 9. Jänner 2022, 13:34:48 CET schrieb Richard Shann:
> > > Hi,
> > > 
> > > Has anyone experienced the direction of ties on a series of tied
> > > notes
> > > on the center line of the staff flipping direction, seemingly at
> > > random?
> > > 
> > > I attach an example, but I haven't been able to cut the file
> > > concerned
> > > down while keeping the effect - if I delete the lyrics for
> > > example the
> > > problem at that point goes away and if I delete bars likewise
> > > (*).
> > > 
> > > I guess this might be related to the question of how LilyPond
> > > decides
> > > the direction for ties on notes on the center line in any case? I
> > > am on
> > > version 2.22 of LilyPond.
> > > 
> > > Richard Shann
> > > 
> > > (*) I should add that as my lyrics are written separately from
> > > the
> > > notes, deleting bars without deleting the corresponding lyrics as
> > > I did
> > > does alter the underlay of the lyrics, so perhaps it is all
> > > lyrics
> > > related...
> 
> Below triggers it:
> 
> \relative b' { b2~ b1*3/8~ b1*1/8 }
> 
> Obviously it depends on spacing/tie-length, see also:
> 
> \relative b' {
>  \cadenzaOn
>   b1~
>   b1~
>   \override NoteColumn.X-offset = #-2
>   b
> }
> 
> No time to investigate further, though.

Thank you for the replies - I guess this means it is not a well-known
feature. Does anyone have an idea if there are guidelines which suggest
what choice should be made?

Richard