#4665 Create a balance in (optical) spacing of different note-values

[Simon Albrecht]

Urs Liska reported a case where the nice optical spacing for the semiquavers results in inacceptably irregular spacing for simultaneous quavers:

:::TeX
\version "2.18.2"

\score {
  <<
    \new Staff {
      \repeat unfold 4 { e'16 g' c'' g' }
    }
    \new Staff {
      \repeat unfold 8 c''8
    }
  >>
}

(png attached)

This means that the spacing engine should find some sort of balance, at any rate take into account the spacing of other note values when computing the optical spacing.

The hand-engraving at http://imslp.org/wiki/File:PMLP06099-Saint-Sa%C3%ABns,_Camille,_Le_Cygne,_Vc_and_Pno._Compl._score.Inside_sheet.pdf doesn’t seem to do much optical spacing here, but the spacing isn’t very tight either, so the result looks quite amiable.

[Simon Albrecht]

The issue may be worked around (as Trevor suggested) by using on of

\override SpacingSpanner.common-shortest-duration = #(ly:make-moment 1/16)
\override SpacingSpanner.uniform-stretching = ##t

in a Score context. However, both seem to disable optical spacing (or do my eyes deceive me?)

[Urs Liska]

I think this is not an enhancement (aka feature request) but a defect of type "ugly". If I'm not mistaken that type is exactly for this kind of issue: not really a bug but a general shortcoming in engraving quality.

[Urs Liska]

Four attachments of the original example:
LilyPond's rendering as

• default
• with common-shortest-duration set
• with uniform-stretching set
• Henle's edition from 2009

I think uniform-stretching gives the best result. While it definitely makes the semiquavers look worse this is much less severe than the original case.

[Simon Albrecht]

Type:Ugly is the right choice, of course.

Well, Henle 2009 doesn’t do any optical spacing either, but as a computer engraving that’s not much authoritative I think.

[Urs Liska]

No, but it shows that they also value the spacing of the eighth notes over those of the 16ths