On Jan 9, 2013, at 1:07 PM, Werner LEMBERG wrote: > > Hello Eben! > > >> One of the things to be clear about for yourself and the designer >> you work with is what the medium you want it to work in will be. >> >> For example is this mainly a font for printing on paper? > > Yes. > >> Is it for the web too? > > No. > >> Do you want it to work well in fine printing? Do you only care >> about laser printing? > > What's the difference between `fine' and `laser' printing?
It is a question of the level of detail. If you are making a font for laser printers you have to avoid certain characteristics in shapes. The shapes have to be more gross or brutal to reproduce well. If you know you want something to look great in fine print ( digital and offset can have very similar resolution now ) then you want to be sure to build in extra nuance to take advantage of the medium. Probably you would want to emphasize one over the other and then test to see that the less favored medium produces results that are still acceptable. With laser printing and fine print the distance is far less than print and web and so a nice compromise is very reachable. > >> Maybe you want web and fine print. If this is the case you probably >> want to look at having two related fonts made. Oddly enough if you >> have two designs they can be made to appear far more similar than if >> you have only one. > > I don't understand this. Please explain. Because of the differences in reproduction and resolution the impression the a single font will give in the two mediums will vary a great deal. For instance it is often true that the x height that is optimal and looks normal on screens will feel excessive in print. There are many many others. But if you find you have a good result in print you can adapt this and make changes until the web version feels like the print version. The typeface Turnip is a great example of this. Here is the web version
<<inline: PastedGraphic-1.png>>
Here is the print
<<inline: PastedGraphic-3.png>>
Notice that the web version is larger in the UPM, The x height is bigger. The shapes are wider. Bothe examples are meant for small sizes which is why the details look so crude. A display version meant for large sizes would have more elegant details. > >> What glyphs do you really need? If the glyph set is smaller then >> the designer can remain more focused. It may not be obvious what >> glyphs will be most helpful. > > The font should cover the most important languages used for > `classical' vocal music, especially operas. This includes Italian, > German, French, Czech (e.g. Dvořák), Russian (in Cyrillic), English, > probably Hungarian (Bartók). Today it's common that the original > language is typeset in upright shape, and a translation in italic, but > sometimes it's vice versa. > > I've also seen a transliteration (using IPA) instead of a translation, > so covering the IPA characters for the above languages would be useful > also. Wow. This is a large project in terms of glyph coverage then. :-) You will want someone who is sensitive to Cyrillic to Latin and who can manage the difficulties of the dense use of diacritics in Czech set which are significant. Is Polish relevant? What about small caps? Do they matter? > >> Do you want the font to be relate well to an existing font for >> setting music? > > It's not clear what exactly you mean with `font for setting music'. > The font used for lyrics essentially doesn't interact with any other > font. Normally, you only see upright and italic. What I mean is not that the two fonts should mix but if there is a font that tends to be used for the score it will have characteristics too and so it will be a good idea to design in relationship to it. If there are 3 music fonts that are most likely to be used then the new text font should be tested next to all 3. It is a question which is a bit like does this lamp look good next to this chair? The two are separate but if you know what the chair is like you can better choose the character of the lamp that suits it. I hope that makes sense. > >> It would also be important to give the designer examples of how you >> want to be able to use the type so that they can test it in those >> contexts and be sure it works well for you. > > There are thousand of pages available for closer analysis in the > scanned vocal scores found in the Petrucci library (http://imslp.org). Cool! > >> I would suggest raising the figure Dave mentions to $9,000 for one >> style so that the designer can afford to test the design extensively >> and can refine the design. If you do that you will be able to have >> something of enduring value made for you. > > OK. Most important is very good kerning to get optimal legibility > with the highest possible density, since lyrics usually need more > horizontal space than the associated note heads, so the smaller the > needed horizontal extension, the better the layout for the music. Right - this is why you want a somewhat condensed design as well. It might even be a good idea to stipulate that the font should still perform well even if it is tracked negatively to -5 or -10 or -15 or whatever is typical. Ideally of course a narrower font won’t need that because it will be more efficient. But it is good to set the performance standards at the start! > > > Werner
_______________________________________________ lilypond-user mailing list lilypond-user@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-user
