Hello,
last weekend I started to try out lilypond-1.1.42.
It compiled rather painlessly on my linux system
(glibc2.0). After looking a bit through the examples in
mutopia and reading the tutorial, I thought, well, let's
give it a try with some real music.
I entered the first of Dvorak's Five Bagatelles op.47 for
two violins, cello, and harmonium. It took many hours, as I
needed to try many things out. The result is quite nice,
and definitely worth the effort (I was happy looking at the score!)
however, there are many points which I think can be improved.
One question of all this to me is: how much of the output should
(or can) be automatically positioned and how many 'hints' has
one to include into the mudela definition. From my naive viewpoint
at this time (not knowing much about lilypond), it is best to
enter only 'musically relevant' information into the mudela file
and as few typographical hints (or hacks) as possible.
Of course, I have not yet started to 'turn the knobs', i.e.
adjusting the apparently many parameters which influence the
typesetting.
Now, let's start with the remarks:
1. It seems to be difficult to enter dynamical indications in word
form (like 'dim.', 'dimin.', 'sempre pi\'u p'). If I use e.g.
a8-"dimin." g f
there is much space around a and g, and 'dimin.' gets written in
roman. Should these be written as a lyrics staff? Is there a
nice font for these wordings? The same problem applies to
wordings above the staff like 'poco ritard.' or 'in tempo',
or 'pizz.', 'arco'.
2. Indications of p, mf, f etc. are IMHO always a little bit
too far to the right, in most of my printed music they go
_befor_ the corresponding note.
3. Vertical placement of cresc (<) and decr (>) seems to be
difficult.
4. Can the note on the middle line be defaulted to have a stem
going down instead of up? This may influence the direction of the
stems of a group of beamed notes, too. They go up a little bit too
often for my taste (and comparing to published printed works).
5. In bar timings < 4/4, 2/2, as e.g. 2/4, a rest r1 should have
a length of the bar. Or is it possible to have whole rests
(hanging on the 4th line from the botton of the staff) in such bars?
6. Is it possible to have two large piano (in my case: harmonium)
staff with smaller string staves above them?
7. The staccato point is rather far from the note head.
8. Slurs and ties are often pointing down when I think they should
point up, e.g. the upper tie in the upper harmonium staff in my
example.
9. When I tie a note with accidential to a note in the next bar,
can I avoid displaying the accidential there again? This is
common practice.
10. The accidentials of a chord collide sometimes (i.e. they are
not moved horizontally). Look at bars 35-40.
11. Slurs/ties and dynamic marks (<, >) get printed on top of
eachother.
12. Multi-voicing. In the harmonium part, I tried to enter
some notes using two voices. I don't know if I understand
how to correctly using voices: I use a context with a voice
where I didn't give a ydirection value (hx and lx). Then, when
I want to determine the direction of the stem, I nest another
voice (ha and la, hb and lb). These two voices however seem to
'forget' sometimes the ydirection property! Especially in the
low staff, I had to respecify this property often. Am I doing
something wrong?
13. How can I move the rests in the low staff of bar 90 down,
after the eighth note?
Well, these really are a lot of remarks, and I hope they are
useful in some way to improve lilypond. I intend to immerse
deeper into this nice program, but that will take some time.
I really would be glad if someone comment on this!
Now, here is my first try, the Dvorak Bagatelle (warning--ugly!
[not the music, beware, but the code!]). It should be old
enough to be copyright free, I hope.
Dirk Lattermann
baga.tar.gz
