1) I had some difficulties to write the Alt chords (for me it's based
on the superlocrian scale 1 2- 2+ 3 4+ 6- (or 5+) 7-) because the
scale has two seconds. (Note that the diminished scale cannot be
written for now with the chord notation, if you ever want to write a 8
note chord ;) ).
This an issue in chord input mode, and should be reported to bug-lilypond as
an enhancement request. For you, the limitation of only one instance of
each step in a chord is a limitation.
Well for this specific case, I used a 3- and a 4- as a workaround, so it's not a issue for me.
2) There no way to write N.C. : no chord (wouldn't the use of R, r and
s make sense in the chord mode?)
N.C. is implemented for r in chordNames context starting with 2.13.1. It
will be available in the next development version. Support for generating
N.C. with R is planned. I don't think that s should generate N.C.; s means
don't output anything.
That's how I though about it, good!
3) I had to use some tricks for chords that last on several measures
for the symbol not to repeat on each measure.
\set ChordChanges = ##t is all that it takes.
Uh, it would have saved me some time :)
4) I would have liked to use parenthesis around the column for chord
extensions like in most jazz charts. I used brackets instead.
I hope we'll be able to implement this, but if we don't, remind us again.
A parenthesis function that works like brackets would do it!
Lilypond of other programs will never be able to interpret notes as
chords (even humans can't do that because there are always ambiguities).
I think the more sensitive approach for pop and jazz is a chord
library with a string as input (maj7) and as output a notation as
markup for chord symbol, optionally as a realisation as notes or chord
diagram (could even have options to select or override the realisation).
That's what we have in essence with our current chordmode. The string has
some specific syntax, and it provides realization as notes, chord diagrams,
and chord names. But the chord names functionality isn't what we want right
The problem now is that you have to write the realization to get the chord name. Thinking the other way around is more logical from the composer point of view, in jazz at least. (or let the realization up to the interpreter).
So add me to the list of people who prefer chord names over chord tones!
I myself need only the chord symbol. Such a simple model is in my
opinion simple to use, customizable and extends easily to other
musics. For me the chord name is more semantics than just notes.
You'll find my current chord exception list below, note that i've used
the unicode chars for the flat and sharp before the extensions as it
gives a very natural layout in this case.
Thank you for this exception list. It provides a nice reference.
The major problem with an exception list is that it doesn't handle slash
chords; I hope to be able fix that.
I'm using exception with slash chords and have no problems.
However I don't think there is a way of entering polychords chords. (I don't need that -yet- :) ).