About this last thought, I have recently studies Joe Pass's chord book and realized that Ignatzek's notation is wildly insufficient to label Pass's chord diagrams. It stops short of describing a chord voicing and seems to be sufficient to only do a real book-style kind of functional notation with the occasional color specified. Pass's editor did something to note down chord names (in some parts of the book), but I think it could be done better.
Thinking about this, it seems to me the issue at hand is that there are two needs in play here: one is giving a name to a set of notes, a different one is placing a symbol on a music sheet to convey a functional harmony message to the reader. Many voicings have several names, obv example might be <b, d f a> is this a Bm7b5 or a rootless form of G9? (Or really a Dm triad plus a natural B in the bass, because you're just in a Cm melodic passage?). The chord namer knows nothing of this, obviously, and it will insist that's just a Bm7b5 and move on. But a good editor might have deep disageements with it, Pass's book is a brilliant example of this: very very few of his chords would be named right by the lilypond namer. Another interesting source to look at to see these shortcomings would be Galbraith's book. There you see it also that the symbol does not capture enough of the voicing as you'd want to see on the page, but a choice is forced whether to pick a harmonic meaning or a name for a chord. Now, these books, per-se, have an analytic and didactic nature, it's study material as opposed to published music. But I have asked myself if one were to typeset/engrave these books with lilypond how would you proceed... editorially, I guess. I am not sure how much this is more guitarist's perspective than it is a pianist's perspective, but still it seems like it's something worth discussing L On Mon, 31 Aug 2026, 14:33 Werner LEMBERG, <[email protected]> wrote: > > > So really it is Lilypond trying to make sense of something that does > > not quite fit into the concept (it is also kind of problem with the > > design chord spec -> actual notes -> guess chord spec -> text when > > Lilypond could very well remember the chord spec used and directly > > use that one. > > Patches to make this work in a better way are highly welcomed! > > > Werner >
