Thanks, Tim and Jakob, for the responses!
> The flat in the Eb chordname looks disproportionately large to me in > both examples and I wonder if it should have a slightly higher > baseline, closer to that of the alphanumerics it modifies, as well > as being a little shorter. The text version of the flat glyph has exactly the same dimensions (i.e., the same bounding box) as the text version of the sharp glyph. I think the appearance is OK. You might have a look at the regression test pages https://lilypond.gitlab.io/-/lilypond/-/jobs/13401522643/artifacts/test-results/index.html to see more examples (pressing the 'before' buttons allows for a blink comparison). > Is it an inconsistency that the + in the final A augmented triad is > on the baseline with the alphanumeric; should it be raised like the > diminished symbol in the chord preceding it? No, it isn't. In the Ignatzek system, the plus sign is not raised. Additionally, the circle symbol should be much bigger and less raised than what LilyPond uses currently. > Now that I am scrutinizing, it is somewhat visually inconsistent > with the major chord symbol and the diminished symbol not being in a > music font and having different line weights. I suppose there are > no music font glyphs specifically for either of those and that they > are drawn from the text font. The triangle is created with the `\polygon` markup, while the circle is an enlarged degree sign from the font. It is impossible to make this fit for all text fonts people are using. However, it's easy to adjust that by setting the variables `\whiteTriangleMarkup` and `\whiteCircleMarkup` to proper values. > This example also made me realize that part of why I like the > Lilyjazz font, and use it exclusively, is that the typeface style of > the accidentals and the alphanumerics is visually consistent and > also consistent with the style of the notes, ties, slurs, barlines, > clefs, etc. Please bear in mind that Lilyjazz is very specific; it has a clear use-case, while LilyPond has to be as generic as possible. > In my opinion, there's a kerning issue with the the A diminished 7 > chord. This is something that cannot be improved on the LilyPond side by default, unfortunately. The amount of kerning depends (a) on the selected font, and (b) on the letter-circle combinations. The same holds for your suggested negative kerning between the circle and following digit. If you have a font that supports all those nifty symbols needed for chord notation *as text glyphs* (like the abovementioned Lilyjazz font), you can add a large bunch of entries to `ignatzekExceptions` to completely replace LilyPond's chord name handling. These exceptions are then handled directly delegated to the Pango text rendering library, which can correctly apply the kerning available in the font automatically. Werner
