Parnassus is a music notation font I am working on, with the goal of capturing the style of old hand-engraved scores. It is primarily inspired by scores from Edition Peters, Breitkopf & Hartel, Belaieff, etc.. While there are many computer music fonts, some of which claim to be based on old scores, I haven’t found one that really captures all the fine points I enjoy in old editions. That’s why I started to work on Parnassus. In fact, Parnassus was meant to be a complete haskell program for experimenting with music layout algrorithms. There hasn’t been much progress though. The font is far from complete, but the basic glyphs are there: noteheads, accidentals, clefs, flags, numerals, and rests.
Here is an example of my font in action. The example was made using the opensource Musescore program. A lot of thanks for the developers of Musescore for making my font work in their program! For comparison I added the version with the default font in Musescore (the emmentaler font from the lilypond project). The music is the beginning of the second piano concerto by Liszt (piano reduction):
As you can see, my version looks closer to the Peters Edition.
I have based my font on scans from several scores, but I tried not to follow them blindly. There are a lot of choices to make, for example the width of thin and thick strokes, making sizes consistent accros glyphs, the precise form of a curve. It’s hard to derive them from a scan, so I choose to make decisions based on how I thought the designer intended the glyphs to be.
There is a lot of contrast between the thick and thin elements. For example the stem from the flat accidental above is thicker on top than on the bottom. While it is not so obvious from the scan above, the flat looks that way in the edition Peters scores. Also the bowl is much thicker, which accentuates the shape. It gives a nice, calligraphic feel to a score.
It’s a well know fenomenon in typography, that a round shape actually looks smaller than it is. Because of this reason round letters, as the letter O are often made a bit bigger to compensate, which is called overshoot. Most traditional scores use overshoot for noteheads and other elements. This brings the notes to the forground. On the other hand, in computer scores, the font is often too small, which gives an anemic feel to the page.
Blending of notes
One thing I like a lot in old scores, is how some noteheads appear to blend together. If you looks at the two half notes to the left, it looks as though they are one shape. This not only looks very nice, it is also clearer and tidier, becouse it focuses the attention on the whitespace in between them. Even if most old scores do this, I haven’t found it in any modern music font.
My font currently works only in Musescore. It is made using fontforge, an open source font design program.  The font will be available under the SIL open font licence. You can find the files on github.