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

[Janek Warchoł]

2010/12/28 Carl Sorensen <address@hidden>
>
> On 12/28/10 12:28 PM, "Janek Warchoł" <address@hidden>
> wrote:
>
> > Because i believe the situation is quite different.
> > Would it be feasible for an engraver to have different flags for many
> > different stem lengths? No, perhaps there would even be problems with
> > manufacturing such dies. Is it feasible for a computer program like
> > LilyPond? Of course!
> > Generally speaking, i don't like to prove things by providing
> > examples; i prefer to prove them by reasoning.
> > I'm that kind of guy
> > who doesn't take anything for granted, you know :) I'm sure there are
> > things that we can do *better* than the hand engravers because of the
> > technological advantage, and therefore i don't want to do things in
> > some manner only because engravers did so.
>
> I'm not opposed to reasoning, but I don't think that the goal of LilyPond is
> to invent new standards for engraving.

Maybe. But from what i've heard, there are few things that actually
can be called standards. I heards that engraving books don't agree
with each other quite often; i'd gladly check myself if this is true,
but i don't have those books. I've searched for engraving resources -
see "music engraving resources in the internet" thread on -user - but
i didn't found much. And there are different opinions even in these
resources that i found! Look here
http://mpa.org/music_notation/standard_practice.pdf , page 9: it says
to do things differently (key cancellation) than LilyPond does. Since
MPA is some kind of authority, how do we defend ourselves? If we cite
opinion of some engravers in our defense and they cite other engravers
in their defense, how can it be decided who is right after all? If
there are two accustomed practices, we have to choose one - or choose
a compromise - and there are many choices like this to do.  LilyPond
is about _automated_ engraving (we want to make a program that does as
many things by default as possibly), so when it (hopefully) will
become the most widespread notation program (don't we want this to
happen?) it's decisions *will* became the new standards!

> > Surely, they had their reasons, but looking at the scores won't
> > explain what the reasons were. Maybe they don't exist now? For
> > example, according to this article
> > http://icking-music-archive.org/lists/sottisier/sottieng.pdf the
> > reasons behind steep angles of the beams in the 19th century were the
> > physical limits of printing technology (page 4) - this sounds
> > reasonable. Now we have better printers and the standards changed.
>
> Personally, I much prefer the older standard, and LilyPond implements the
> older standard.  I'm glad it does.  Just because this article suggests it
> could be different doesn't necessarily make it right....

There are 3 "correct" examples of beaming of the same passage there:
one labeled "Peters", one "Henle" and one "recommended". Certainly i
don't like Henle, but doesn't LilyPond output look more like
"recommended" than "Peters"? LilyPond rarely puts the beam parallel to
the noteheads.

> > I hope this doesn't sound too arrogant...
>
> I don't think it does, but I also think that you will have a hard time
> convincing the LilyPond community that we should change the standard
> behavior just because you like it.

Perhaps... But i promise that i'll always have logical arguments to
back up any suggestions :)

> > 2010/12/27 Carl Sorensen <address@hidden>:
> >>
> >> I have now reviewed the engraving books.  I will give the answers that they
> >> have.
> >
> > Thanks for sharing this!
> >
> > I must admit that i'm surprised to see a shorter variation of flag in
> > the Stone scan. However there is (should be?) at least one
> > intermediate stem length between 3.5 and 3. What shall we do with it?
>
> Why is there an intermediate stem length between 3.5 and 3?  If there is,
> then we use the size 3 flag, I think.

If there is no intermediate stem length, the point of transition
betweeen 3.5 and 3 looks bad.
This leads me to another related question i was going to ask. Look at
the attachment and tell me which line looks best for you? (i'm asking
this question to anyone who is still following this thread - by the
way, is there anyone else? :) )

> >> The bottom line, for me?  I think there should be separate single and 
> >> double
> >> flags for 3-space stems.  I don't think they should be scaled from the
> >> existing flags.  I think they should be designed separately and added to 
> >> the
> >> font.  But in order to get this to happen, I think we need some examples
> >> from well-engraved scores.  And I don't have a good collection.
> >
> > Hopefully i'll learn Metafont and try to add it myself, but before
> > this happens maybe it would be good to have those scaled flags?
> > Wouldn't it be better than the current situation?
>
> I'll be happy to add 3-space flags to the font.  It won't take too long.
> But before I do it, I'd like to see some prototype code that would use them.
> Here's a suggestion.  How about if you write code that uses some other flag
> for short stems (take your pick; it doesn't matter what flag you choose) and
> then send me a copy of your patch. At that point, I'll make shorter versions
> of flags.d3, flags.d4, flags.u3, and flags.u4.

I'll do this as soon as we decide on the issues related to this one
(for example about the attachment).
And I hope it won't take too long - but, as i've said before, i'm
quite inexperienced.

cheers,
Janek

5.png
Description: PNG image