#5691 French Beams are longer than expected

French beams are longer than
normal:

  \relative c'' {
    r16 f d f
    \override Stem.french-beaming = ##t
    r16 f d f
  }

Looking into scores from French publishers this behaviour seems to be
incorrect.

[Anonymous]

On 25/01/2020 08:27, Thomas Morley wrote:

Am Sa., 25. Jan. 2020 um 08:58 Uhr schrieb Kevin Barry barrykp@gmail.com:

The minimum length of the stem of the middle note is forcing the beams down.

Not here. It's more the relevant value of
beamed-extreme-minimum-free-lengths. See:

 \relative c'' {
    r16 f d f
    \override Stem.details.beamed-extreme-minimum-free-lengths =
      #'(2.0 0.8)  %% original:  '(2.0 1.25)
    \override Stem.french-beaming = ##t
    r16 f d f
  }

Cheers,
Harm

[Anonymous]

On 25/01/2020 09:43, Malte Meyn wrote:
Another problem that has to do with stem lengths in french beaming: articulations use the shorter beam length for positioning, resulting in collisions:

\version "2.19.83"

\relative c'' {
  \voiceTwo
  r16 f-> d-> f->
  \override Stem.french-beaming = ##t
  r16 f-> d-> f->
} 

Both beam and articulation positions should be the same for default and french beams. Probably the easiest solution would be to have shorter stem stencils without decreasing the stems’ Y extents.

[Malte Meyn]

tuplet-number.cc has a workaround for a similar issue but I think instead of workarounds for other grobs we need a solution coming from the stems/beams themselves.

[Torsten Hämmerle]

Done.
I've created a new ticket Issue 5788 - New French Beaming Approach

[Thomas Morley]

Must be a long standing issue, I could follow it back to the oldest running version on my machine, i.e. 2.12.3. Maybe even older

[Torsten Hämmerle]

In the quest for a better, much more straight-forward general solution, I've created the ticket

Issue 5788 - New French Beaming Approach.

As it is a complete re-design of LilyPond's French beaming implementation, this new approach should be able to solve all the various problems caused by current French beaming, because all grob positioning behaviour will exacly match the standard beaming case without the need for any distiction between French and stanard beaming cases.

Everything now is exactly identical, and only at the very end and only in the ly:stem::print stencil, the (printed) stem length will be shortened by the appropriate french-correction value.
For all the other objects and, of course, the beams, even French stems will be treated as standard-length stems (property "Length" will always keep the standard stem length value), so that all positioning calculations of anything will automatically be correct (in the sense of: identical to standard beaming).

Please have a look! --- Especially at the two old/new comparison PDFs attached there.

For that reason, I'll set this issue to "shelved" for the time being, waiting for the general solution that will also cover this issue.

Cheers,
Torsten

[Torsten Hämmerle]

Solved by more general
Issue 5788: New French Beamimg Approach

Set to Duplicate for that reason.