Re: flags, beams and stem length in forced directions - output improvement

[Janek Warchoł]

Hi,

after fixing my problems with virtualBox (again...) i'm trying to
implement the changes i suggested long ago and test them.
Unfortunately i have 2 problems:
- it seems that no matter what i do inside the metafont files in mf
directory, the changes to the font are not made (i.e. after running
make all from /build, and compiling a file with binary found in
build/out/bin/, the pdf output remains the same)
- i was unable to find where the point in which flag is attached to
the stem is determined.

May i ask for some pointers?

cheers,
Janek







2010/12/27 Janek Warchoł <address@hidden>:
> Hi all,
>
> There is a problem with LilyPond default output that i want to
> discuss. Look at the attachment1:
> - in the first line: the stems of the b's are too short,
> - in the second and third line: the beams of the 'a' eight notes are too high,
> - in the fourth line: the 16th flags look bad.
> http://tinypic.com/r/9rhr2v/7 shows how things could be done better.
> All this may sound nitpicky, but at least the first one is quite
> serious - and it affects virtually all scores!
> These problems are connected and a proper solution should be a
> comprehensive one. I've analyzed the situation and divided it into a
> list of consecutive questions; i'll post them one by one in this
> thread.
> Afterwards i'll gladly help with making a patch reflecting our decisions.
>
> The first question is: we all know that stems in forced directions
> (like this one: { \clef treble \stemUp a'' } ) should be shortened,
> but what shall LilyPond do when there is a flag attached to such short
> stem? (see attachment2):
> (A) don't do anything (current behaviour),
> (B) make the flag shorter. Adding shorter versions of the flags to the
> font would be great, but very laborous. Simply scaling the flags
> vertically, as i did manually to create the attachment, seems to work
> quite fine (i hope it's possible with LilyPond).
>
> I strongly prefer changing the default behaviour to B because:
> - the flags look much better,
> - it allows for shorter flagged stems*.
>
> *regular stem length is 3.5 staff space, while 2.5 is the length for
> shortened stems. Unfortunately unchanged downward 16th flags start
> looking bad at 3.1 stem length, while shortened flags look acceptable
> even at 2.7 stem length (and if we'd add shorter versions of the flags
> to the font, i suppose they would look great even at 2.5 and below!).
>
> I'll post the next question as soon as we decide on this.
>
> I'm looking forward to reading your answers,
> Janek
>
> PS What about 32nd, 64th and 128th flags? I think it'll be obvious
> what to do with them after we decide about 8th and 16th notes.
>