Menu ▾ ▴

#2368 wrong position of a tie between notes in chord

Accepted
nobody
None
Ugly
2019-01-26
2012-03-01
Anonymous
No

Originally created by: *anonymous

Originally created by: colingh...@gmail.com

Reported by Karol Majewski here:

http://lists.gnu.org/archive/html/bug-lilypond/2012-02/msg01194.html

% Tie that should connect 'g' and 'g'
% is located very low, so that it looks like a tie between 'd' and 'es'

\version "2.15.30"

#(set-default-paper-size "a4")
#(set-global-staff-size 18)

LiniaJeden =  \relative c'' {
  \clef "treble" \key c \major \numericTimeSignature \time 3/4
  << { g'4 f2 } \\ { < d, g ~ c ~ >2. } >>
  << { g'4 c2 } \\ { < es,, g c >2. } >>

}

\score {
    \new Staff = "LiniaJeden" \LiniaJeden
}

1 Attachments
bug.ly

Discussion

  • Google Importer

    Google Importer - 2012-03-01

    Originally posted by: colingh...@gmail.com

    Sorry, Karol, I used the wrong version of lilypond to create that png.

    Result of running bug.ly with 2.15.30 is attached to this comment.

     
    Thumbnail
    bug.preview.png

  • Google Importer

    Google Importer - 2012-03-01

    Originally posted by: plros...@gmail.com

    This was rendered by Lilypond 2.14.2.  Not stellar,  but at least it's clear that the tie starts at g, not at d.  Lilypond may be better at avoiding tie collisions now, but it's worse at making unambiguous output, at least in this particular case.

     
    Thumbnail
    tie.png

  • Google Importer

    Google Importer - 2012-03-28

    Originally posted by: Karol.Ma...@gmail.com

    Is there any way to move this tie up? I tried

    once \override Tie #'Y-offset = #1

    before the chord, but that's not what I was looking for, because this command refers to both ties in chord. And unfortunately it's not possible to use this command inside the chord. Also

    \once \override Tie #'staff-position = #1

    doesn't work. Maybe there is any way to allow tie-accidental collision?

    Karol Majewski

     

  • Google Importer

    Google Importer - 2012-03-28

    Originally posted by: Karol.Ma...@gmail.com

    OK, so here's simpler version generated with 2.15.35

    //

    \version "2.15.35"

    #(set-default-paper-size "a4")
    #(set-global-staff-size 18)

    LiniaJeden =  \relative c'' {
      \clef "treble" \key c \major \numericTimeSignature \time 3/4
      < d, g ~ c ~ >2.
      < es g c >2.

    }

    \score {
        \new Staff = "LiniaJeden" \LiniaJeden
    }

    //

    And a tie between 'g' and 'g' is still misplaced.

     
    Thumbnail
    regression.png

  • Google Importer

    Google Importer - 2012-04-11

    Originally posted by: PhilEHol...@googlemail.com

    http://lsr.dsi.unimi.it/LSR/Item?id=407

      \clef "treble" \key c \major \numericTimeSignature \time 3/4
      \override TieColumn #'tie-configuration = #'((0.0 . 1) (-2.0 . -1))
      < d, g ~ c ~ >2.
      < es g c >2.

     
    Thumbnail
    tiepos2.png

  • Valentin Villenave

    Valentin Villenave - 2019-01-26
    • Description has changed:

    Diff:
    • Needs: -->
    • Patch: -->
     

  • Valentin Villenave

    Valentin Villenave - 2019-01-26

    I’ve been trying different values for some Tie grob-properties, and this is what I came up with. For every one of the -penalty-factor properties, there’s a threshold below which the bug doesn’t manifest (this values are linked; lowering one lessens the threshold for the others).

    Of course one can also unset one of these by commenting it in the alist (for example outer-tie-length-symmetry-penalty-factor in the example below).

    \version "2.21.0"
{
  \time 3/4
  \override Tie.details =
    #'(
        ;; directly copied from define-grobs.scm
        (ratio . 0.333)
        (center-staff-line-clearance . 0.6)
        (tip-staff-line-clearance . 0.45)
        (note-head-gap . 0.2)
        (stem-gap . 0.35)
        (height-limit . 1.0)
        ;; default is 10; any value higher than 2
        ;; triggers the bug:
        (horizontal-distance-penalty-factor . 10)
        (same-dir-as-stem-penalty . 8)
        (min-length-penalty-factor . 26)
        (tie-tie-collision-distance . 0.45)
        (tie-tie-collision-penalty . 25.0)
        (intra-space-threshold . 1.25)
        ;; default is 10; any value higher than 10
        ;; triggers the bug:
        (outer-tie-vertical-distance-symmetry-penalty-factor . 10)
        ;; default is 10; any value higher than 5
        ;; triggers the bug:
        (outer-tie-length-symmetry-penalty-factor . 10)
        ;; default is 7; any value lower than 7
        ;; triggers the bug:
        (vertical-distance-penalty-factor . 7)
        (outer-tie-vertical-gap . 0.25)
        (multi-tie-region-size . 3)
        (single-tie-region-size . 4)
        (between-length-limit . 1.0)
    )
  <d' g'~ c''~ >2.
  <es' g' c''>2.
}

    I’ve found that some other properties also trigger ugly-Tie bugs; for example the one mentioned by Colin above is fixed by commenting the center-staff-line-clearance property (or setting to a low enough value).

    Actually, most of the bugs I encountered appear to be fixed by getting rid of all these properties in bulk: simply set Tie.details to an empty list, and Bob’s your uncle.

    \override Tie.details = #'()

    If so, why bother at all?

     
MongoDB Logo MongoDB