[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: modern-straight and flat- flags too thick and too spaced apart
From: |
Thomas Morley |
Subject: |
Re: modern-straight and flat- flags too thick and too spaced apart |
Date: |
Sun, 17 Feb 2019 21:41:20 +0100 |
Am So., 17. Feb. 2019 um 20:57 Uhr schrieb Carl Sorensen <address@hidden>:
>
>
>
> On 2/16/19, 7:22 PM, "edes" <address@hidden> wrote:
>
>
> Hello, list.
>
> Some time ago I reported a bug with modern-straight and flat flags, and it
> was accepted as issue 5412:
>
> https://sourceforge.net/p/testlilyissues/issues/5412/
>
> I only have a very basic knowledge of lilypond, but I assume that this
> problem happens at the font level, and that there is no way to improve the
> output by tweaking the lilypond code?
>
> It turns out that the straight family of flags is actually implemented in
> scheme.
>
> The file is scm/flag-styles.scm
>
> The thicknesses are hard-coded at lines 111 (modern-straight-flag), 117
> (old-straight-flag), and 122 (flat-flag).
>
> The names of the variables (which show up only as numbers in these calls) are
> found at line 60.
>
> Unfortunately, at this point there does not appear to be any grob property
> for the thickness, so there is not a straightforward override. You'll need
> to edit the scheme file. Or, you could define your own straight flag style
> by copying (and altering) one of the flag definitions.
>
> If you get some better numbers, and especially if you have some evidence from
> nicely-engraved scores, it's likely that we would replace the defaults.
>
> Thanks,
>
> Carl
Hi Carl,
I'm currently working on it.
Better default values are easy to obtain:
#(define-public (flat-flag grob)
"Flat flag style. The angles of the flags are both 0 degrees"
((straight-flag 0.48 0.81 0 1.0 0 1.0) grob))
Though, there is a design-decision we need to do.
Currently the straight-flags are modeled like default Flag-glyphs.
This means if there are more than 3 flags the spacing is not adjusted
(like for beams). Furthermore the end of the stem and flag match
exactly, as opposed to beams (at least they should, there's a small
discrepancy, already reported by the OP).
If we keep this thinking, than (flat) straight-flags stay different
compared to Beams.
The OP claims straight-flags, especially flat ones, should behave like
Beams. If we follow, than straight-flags need to adjust according to
the amount of flags, like Beams. Furthermore they should be placed so
that the stem ends in the _middle_ of the first flag, as for beams.
Disadvantage would be that a stem with default flag would often have a
different visible length compared to a stem with straight flags.
Speaking only for myself I think the OP is correct saying flat-flags
should behave like Beams, though I don't agree for
modern/old-straight-flags.
So the question is, which route to follow?
I already have code for the mentioned possibilties, but currently I'm
not able to do one thing for flat-flag and a different for other
straight-flags.
In any case the length of beamed and unbeamed Stems is different in
many cases. One would need to tackle Stem not Flag, so a different
issue, imho.
Cheers,
Harm