Re: There no ligatures in Kievan notation (was Re: Kievan "hatchet" notes for lilypond?)

[Juergen Reuter]

On Thu, 30 Nov 2006, Monk Panteleimon wrote:

> On Thursday 30 November 2006 06:01, you wrote:
> > Hello, Panteleimon!
>
> Hello, thank you for your response
>
> > I fear the task of fully implementing kievan chant notation is more than
> > just a question of adding thirteen glyphs.
> > recognize at least pes ligatures (those pairwise stacked rhombic glyphs)
> > and clivis ligatures (e.g. the third glyph in the last line of the long
> > example).
>
> Actually, those are not ligatures, and there are no ligatures at all in Kievan
> notation. The two stacked diamond-shapes you refer to are simply a half note.
> They refer to the space inbetween the two diamonds, in this case "mi" or
> b-natural. If the diamonds where in spaces, they would refer to the line
> in-between. The "squiggly-with-stem" and "double-diamond-with-stem" notes
> that also seem to take up several spaces are not ligatures either, just 16th
> notes. (By the way, the durations I cite are doubled by some transcribers in
> order to have a whole-note as the longest duration, but this gives an
> incorrect impression of the flow of the chant).

Ok, I understand.  This, of course, simplifies things a lot.  Then the 
main task really should be to incorporate these glyphs into lily.  Though 
I unfortunately have to confess that I currently have no time at all to 
care for it (I am currently working hard on a thesis, and my remaining 
time is running out soon).  Maybe Han-Wen/Jan would do it as a sponsored 
feature?

I am not sure what is the latest state of incorporating new glyphs into 
lily; so far, we used to have to design glyphs by manually writing 
Metafont code, which may result in high-quality glyphs, but is also an 
extremely time-consuming process.  But nowadays, lily supports a couple of 
font formats.  Maybe there is a simpler way of getting font characters 
into lily, e.g. by scanning + tracing them?

Han-Wen, Werner, what do you think?

In any case, I guess you would at least need to have high-resolution scans 
of hand-writings or printings (which should be sufficiently old in order 
to not violate copyright laws).

> > To me, the notation looks structurally quite similar to gothic
> > (also called "hufnagel") chant notation, though the glyphs are somewhat
> > different.
>
> I would be interested to know what notation looked like in Poland in the
> 1600's, to see if it has  a parent for this style

Unfortunately, I am not familiar with Polish ancient notation.  Still, 
since (at least the northern and western part of) Poland is mostly 
Catholic rather than Orthodox, I strongly guess that they used gothic and 
Roman (i.e. square) neumes, as well as mensural notation.  I suspect, 
however, that it may be different for those parts of Poland near to the 
Ukrainian or Russsian border.

> As I understand it, the only things necessary *besides* adding the glyphs
> would be:
>
> 1. Spacing according text syllable. This is the only way notes are spaced in
> this chant, as you can see. There is a small &  even block of space between
> each textual syllable, all the notes whereof are equidistant and very close.
> %%%(This is part of what makes this notation better for Znamenny Chant, being
> closer in this regard to the neumatic origins and taking up less space for
> long, involved melismata. It looks ugly if you try to do it with round
> notes.)%%%
> My hope is that this can be accomplished simply making slurs invisible in the
> kievan-init.ly file |  Slur #'transparent = ##t  |
> and then convincing lilypond to put an appropriate block of space between
> syllables. At first I thought that this could be done with LyricSpace
> #'minimum-distance , but that would only work for short syllables.

Right.  There are a tricks that will force lily to typeset notes rather 
densely, thus resulting in almost what you probably would like to get. 
Still, I do not know how you could achieve strictly equidistantly and 
densely spaced noteheads within melismas.

> 2. Right now we can set shaped noteheads in LP like this:
> \set shapeNoteStyles = ##(fa # la fa # la mi)
> (or something like that).  We'd have to do something similar but different in
> the init.ly for kievan, since several symbols (the quarter and the eight)
> have slightly different forms for lines and spaces. These differences are
> very important to the overall look of the music. Note that they differ  not
> to according scale-position (like shapenotes) but according to whether they
> are on a line or space. That means that g,8 looks different from g8

There is currently support for line/space glyph distinction for custodes 
and mensural flags in the c++ backend.  It should not be too hard to also 
add it for noteheads, but I guess Han-Wen will not be too happy to add new 
program logic to the c++ backend.  Hence I guess, for the long term, this 
cries for migrating the whole logic to scheme.

> Furthermore, the eigth note (block with long tail) has not two but *four*
> slightly different forms: two for stem-up and two for stem-down.

Just like with the custos glyphs, I guess.

> I think that rest could be done just by thickening lines and such, except
> perhaps that the beams on the 16th notes (single beams) extend beyond the
> note-stems and half sort of blunt edges, but I'm sure that can be done
> without designing anything new, right?

I am not sure; but before implementing something completely new, one 
definitely should first try, if the targeted result can be achieved by 
tweaking existing functionality.

Greetings,
Juergen

> Thanks again for your response.
> I look forward to doing whatever I can to implement this feature.
>
> Sincerely,
> Monk Panteleimon
> Hermitage of the Holy Cross
> Wayne, WV USA
> address@hidden
> http://www.holycross-hermitage.com/
> http://www.holycrosskliros.org/
>
> PS. I have an adaption of a long chant like this example in which the sequence
> of notes is almost exactly as in the Slavonic original. I can post both if
> you like, so you how can see how simple it really is.