lilypond-user
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Discussing typographical standards (was: Tuplet notehead shared...)


From: Richard Shann
Subject: Re: Discussing typographical standards (was: Tuplet notehead shared...)
Date: Tue, 25 Mar 2014 09:08:09 +0000

On Tue, 2014-03-25 at 01:10 +0100, Simon Albrecht wrote:
> 
> Am 24.03.2014 13:33, schrieb Richard Shann:
> 
> > An example of this, typeset using LilyPond is posted here:
> > 
> > http://imslp.org/wiki/Special:ImagefromIndex/278632
> > 
> > To do this I set tuplet timing around the entire bass part and used
> > doubled time signatures (one hidden IIRC)
> > 
> > Richard
> The following is completely off-topic, but I’d like to share some
> observations I often make and thoughts I have and ask for your
> opinion:
> Looking at this score confirms me in my opinion that LilyPond default
> output alone is no guarantee for a good-looking result in accordance
> with typographical good use.

This is a matter of degree: most people coming to LilyPond will have
experienced drawing-based gui programs which leave much more to your
skill as a music engraver.

>  This may be partly due to an older lilypond version used,
yes, I think that was 2.12. Nowadays Denemo is built with the latest
stable LilyPond release.
>  but there are some basic issues I see with this:
> 
> – For what I know of best practice in typography, it is normally
> unnecessary to use slurs for indicating melismata. Beaming
> (\autoBeamOff, melismata with []), placement of syllables and
> hyphenation/extender lines make the lyrics assignment unambiguous and
> easy to read in all but the most complex cases (that is, when the
> rhythmic complexity requires that the beaming corresponds to beat
> groups and legibility would suffer in the opposite case—which will
> rarely occur before 1900).
> Certainly I know that the Lily authors knew what they were doing, when
> they recommended using slurs for this purpose. This is used in
> excellent hand-engraved editions as well, I think especially later in
> the 20th century. Nevertheless I vote for the supposedly older use, as
> described before.
> 
> – The default Denemo output

hmm, I think you misunderstand Denemo here. It is a music *input*
program - I think in this case it is a question on garbage-in,
garbage-out. I was asked to generate that score for a concert, and hoped
for feedback from the consumers (I know *nothing* about vocal scores);
unfortunately, the consumers are so used to, and tolerant of, bad
computer music typesetting that they simply said it was "lovely" and
that was that. 

>  reflects the now common, but faulty practice of writing syl- la- ble
> instead of syl - la - ble (with the hyphens centered between
> syllables). The corresponding Lilypond code would be { syl -- la --
> ble }, see Learning Manual, Aligning lyrics to a melody.

So, you would type -- in Denemo to get that style, if you knew what you
were doing. As I say, I was too ignorant of vocal music to do that or to
devise a way that Denemo would hint to the user that they might want to
do it. But Denemo is just a LilyPond front end, you can (I hope) put
anything in the lyrics as needed.

> 
> – The beginning of the first recitative is a good example where
> inserting a line break at half-measure would significantly improve the
> visual impression by a more even horizontal spacing.

Denemo has the command to do this built in, I have been using it myself
recently to break a page at the second time (half-)bar.

>  I found that it was common in traditional hand-engraved scores to do
> such mid-measure breaks (if measures aren’t rather short), and thus I
> am often using \bar "" at half-measure. Sometimes I even use an extra
> voice for something like \repeat unfold 35 { s2 \bar "" s2 } and thus
> create more flexibility in line-breaking. The disadvantage is that
> there is no possibility to differ in likeliness between mid-measure
> and full-measure breaks, which would then be desirable.
> 
> – As always, the default margins are too small.

I have once or twice created Denemo defaults that emit the LilyPond to
alter the LilyPond defaults, but have then had occasion to regret it
later when LilyPond improves/changes. I don't think this is one of them
- though I place an enormous premium on lack of page turns for my
personal use.

>  This is already being discussed as issue 3808 and will hopefully be
> changed soon. I once read a comprehensive article (in German) on this
> topic from the German Tex user group’s magazine, and the author
> pointed out that in medieval manuscripts and renaissance prints an
> outstandingly pleasing appearance is achieved by page margins which
> cover up half of the page’s space! This is luxury, of course, and
> usually unaffordable, but I find it evident that having "unusually"
> large margins (and simple ratios between the measurements of the page
> and margins, and the top-margin smaller than the bottom-margin and so
> on…) much improves the look of the page. It might necessitate to
> decrease staff size, though, but anyway 16 pt are no way too small.
> 
> – In order to increase legibility and clarity it’s also much advisable
> to use at least one StaffGroup, e.g.
> \new StaffGroup {
>   \new Staff = "fl" {}
>   \new StaffGroup {
>     \new Staff = "vl1" {}
>     \new Staff = "vl2" {}
>   }
>   ...
> }
> 
> Using LilyPond unfortunately doesn’t in itself guarantee "flawless"
> typography (as Denemo advertises itself).

Yes, I agree I was over-egging things claiming that, but as I say it is
matter of degree. While not wanting to incense perfectionists I did want
to alert a very large potential user-base to the fact that you can churn
out eminently readable scores at a phenomenal pace using LilyPond's
default settings. The gulf between the expectations of average users of
music notation programs and the possibilities is enormous when it comes
to the time needed to create a pleasantly usable score.

It is the case that since writing that eulogy to LilyPond I have
experienced rather more of music typesetting outside of my own personal
interests, and was surprised to find LilyPond still has some way to go
in some areas. Choosing ideal tie shapes in some circumstances, merging
multi-measure rests in voices ... So I have been adding stuff to Denemo
to support those things as they are drawn to my attention. But the vast
majority of people are just grateful they don't get an unreadable score.

>  You need to use it correctly also, following the instructions in the
> manuals…
> 
> I hope I haven’t been too "moralist" there, nor too extensive… sorry
> if I have.

Not at all, it is only with feedback that anything improves.

Richard


> Best regards,
> 
> Simon





reply via email to

[Prev in Thread] Current Thread [Next in Thread]