On Sat, Oct 24, 2015 at 3:01 PM, Noeck [via Lilypond] <[hidden email]> wrote:
Dear Abraham,
I have a question to you, but I thought perhaps lilypond-users are also
interested:
When I use smaller glyphs in a score (change clefs, cue clefs, a solo
voice above the piano part, etc.), LilyPond usually uses a different
font such that the font weight harmonizes with the other standard-sized
objects. When I choose one of your fonts, my impression is that this is
not really the case, but the glyphs are just scaled. Perhaps you can see
what I mean in this example:
\version "2.19.21"
%\paper { #(define fonts (set-global-fonts #:music "cadence")) }
{ a \clef treble a \cueClef treble a }
If the \paper line is commented out (using Emmentaler), the smaller
clefs look bolder than with Cadence.
Is that true?
Yes it is. At the moment, Emmentaler is the only music font I know of that completely supports this kind of optical scaling. I have developed some methods for applying similar principles to my fonts, but I haven't had the time to work out the kinks.
One thing to note, if you don't know this already, is that all the Emmentaler fonts are technically the same scale within the font files themselves (by design), but are designed to really only be appropriate at their respective staff-heights. LilyPond then scales them up/down to match the desired staff-height, where the "heavier" glyphs are used at smaller sizes and the "finer" glyphs at the larger sizes.
Is that intended or is it just a drawback of not using
metafont for the font creation?
Neither really. In their current state, I guess it was kind of intended in the beginning, but it really was meant to be a stepping stone to the creation of optically-scaled variants. This was mostly due to the semi-automated process I use to create the fonts in the first place. Metafont certainly can make it easier to define precisely what you intend to do as the relative "weight" of each glyph gets heavier, but this isn't the only way to do that. My method allows me to graphically create/edit each glyph precisely as I see fit and then they are just copied into pseudo-optical-variant files, where they are really identical. So, yes, the smaller glyphs will look leaner than Emmentaler's glyphs at the same size, but I have great plans to remedy this situation. Please stay tuned!
Here are some test runs of what I've done (compared to Emmentaler, using the example from the LilyPond essay). They are definitely NOT perfect, but hopefully it shows you that I'm on the right track (11pt vs 26pt):
![Inline image 1]()
I hope that clarifies what's going on and shows that I'm aware of the situation and working on a beautiful solution :-)
Best,
Abraham