#332 Lilypond export: beaming triplets

Maybe
closed
nobody
lilypond (14)
1
2013-10-04
2007-03-18
Arnout Engelen
No

When exporting to Lilypond, with 'export beamings' turned on, a [rest-note-note]-triplet (as in: the pa-pa of humm-pa-pa :) ) is exported as:

\times 2/3 { r8 [ e' e' ] }

Instead of:

\times 2/3 { r8 e' [ e' ] }

I think lines 1488/1489 of LilypondExporter.cpp are suspicious, and indeed when I remove them the triplet is exported correctly, but I don't understand the code well enough to commit that with a clear conscience ;).

Discussion

  • Heikki Junes
    Heikki Junes
    2007-03-18

    Logged In: YES
    user_id=30776
    Originator: NO

    Confirmed.

    The logic of notation in LilyPond export should match with the logic of the NotationView.

    I have rewritten quite a lot LilyPondExporter.cpp, but I am not sure whether the logic here matches with the code of NotationView. Apparently not, because there are still differences.

     
  • Heikki Junes
    Heikki Junes
    2007-03-18

    Logged In: YES
    user_id=30776
    Originator: NO

    Committed fix in svn revision 7981.

    Before closing the tracker, I would like the triplet beaming would be tested in larger pieces.

     
  • Chris Cannam
    Chris Cannam
    2007-03-19

    Logged In: YES
    user_id=13489
    Originator: NO

    Perhaps the original code had something to do with the differently ordered grouping syntax (prefix/postfix) in older versions of Lilypond? I forget the details, and although I probably wrote this bit of LilypondExporter, I can't give a definitive explanation.

     
  • Heikki Junes
    Heikki Junes
    2007-08-26

    Logged In: YES
    user_id=30776
    Originator: NO

    This is not a bug, because currently two different groups, for example beamed and triplet groups, cannot exist at the same time.

    Converting to a feature request.

     
    • status: open --> closed
    • Group: --> Maybe
     
  • I'm not sure what feature is being requested. Triplet beaming is working whether beams are exported or not. The largest piece I tested is only some 80 measures long, but they're solid triplets from beginning to end. It's probably the best we can do, since getting that many triplets into existence is actually a gigantic pain in the ass.