This area of munchfonts is devoted to important aspects of typeface design that are occasionally discussed on the newslist Typo-L, and involve visual material. To subscribe to Typo-L, see the website for details (a hyperlink is available on a short step up and over to the Typo-L site).

Looking at the changes of spelling from s to x in French plurals, here is a scan of a scan of a photograph (it's really not very good) showing the s and x in contrast. Is it really an x? There aren't enough fine hairlines in the scan to show the lower left tail of x, and did the scribe consider it to be an x?

See oyseaulx, deux, glorieux; and you will notice the initial and medial long-s, and final round s in les, temps, contans. And, are those two words appetres, a petres, with a trailing s?

Les Oyseaulx

The text is in a Bâtarde current hand, from the 1400s. From Degering, Lettering plate 108.

A higher-resolution scan (218KB)

Here is the delta from Nicholas Jenson's 1472 Greek typeface, in the orthotic, upright style.


h geusews hdonhn. h gignomenhn
(With h as eta, w as omega. That u-like form with left-and-right tails is a mu.)
The written order of strokes is unclear, offering two possibilities: delta's main ascender modeled first, followed by the triangular lower part; or, the triangular lower part first, followed by the ascender. in the latter, the ascender could be made from baseline up to the left, though the dynamic would tend towards the more contemporary rightward loop of modern delta.

Illustration from Zapf's chapter, "Development of Greek Typefaces" from Greek Letters: from Tablets to Pixels

How do marks, edges, and counters relate to one another?

Marks and Counters

  1. Left two marks:
    • A visible mark has width and breadth; it's a two-dimensional shape.
    • As it is made, it separates from the ground (background).
    • What is not marked is counter to the mark.
    • Where mark and counter meet is an edge. An edge has no width or breadth, only length, along the boundary of mark and counter.
  2. Right two marks :
    • A mark that returns to itself often makes a closed mark. There is an interior counter separate from the outer counter. Many viewers imagine that the interior counter (plus the bordering closed mark) is the mark.
    • A mark that returns to its beginning has been closed and filled. The mark could as easily have been made with a bigger marker tool, or else be an enlarged portion of a smaller mark.

At ATypI Roma 2002, Linotype Library GmbH revealed their latest project: Sabon Next, a full renovation of the Sabon family of Jan Tschichold, redrawn and expanded by Jean-François Porchez. The image here is low-resolution, and of only part of the wide range of styles (yes, the italics are now free of the requirement that they be the same width as the roman) which includes new weights, alternate-sort fonts, and ornaments.

Sabon Next setting

The shaping of the German s-s ligature (ß) is a delicate and culturally sensitive matter, but can be done well. Or badly. Two variants of the shape in a design originally by Morris Fuller Benton, the first from Adobe's Franklin Gothic and the other from ITC's verison, show two attempts at the hmm, s-z ligature. The traditional variant (Adobe's) is clearly the superior sort for native readers.

In HTML, ß can be marked up as ß, referring to the two camps that posit the shape as a ligature of either long s and s, or long s and z. A version that uses the s-s style (Tschichold) is below, for comparison. Though more Latinate, and thus perhaps not completely appropriate for a grotesque such as Franklin, the third style is typical of Humanist and Antiqua ß. It is a tat too wide, perhaps, in this Bold style rendering.

ß from Adobe Franklin Gothic ß from ITC Franklin Gothic

(Above) Scans from the Adobe Catalog

ß from Linotype Ergo

Setting contact information unambiguously on a business card. Information is clustered in two ways: with a rule to differentiate personal (phone and email) information from the more general facsimile and Web site information, and with space to differentiate telecommunication from Internet communication.

The typeface is Humanist 521 Condensed (Gill Sans) in place of the (here, unavailable) Akzidenz Grotesk Condensed.
Original card and emblem by Michael Brady, RR,Typo-L.

Business Card Layout

Diferent serifings in historical hands relied mostly on what in type would be called semi-serifing; the pen makes a serif at the start and end of a stroke or letter, primarily as a way of trailing off excess ink.

From top to bottom:

Carolingian: Early Carolingian, c.800 Late Carolingian, c. 1000
Blackletter: Northern Textura, c.1200 Italian Rotunda, c.1400
Renaissance: Humanist, c. 1406 Late Renaissance, c. 1470
six historical serifing patterns

The Late Renaissance model closely relates to the models used by Jenson and Aldus. Here, the pen moves (retracts, or retraces) to the left and then right before lifting to make the arcade stroke.

Examples were roughly written by GM with about the same care that might be given one letter on a long page. They are not to be considered definitive forms.

The use of different typefaces to mark the speakers of dialogue was used by Massin in the Grove Press edition of Eugene Ionescu's The Bald Soprano.

Bald Soprano detail

A full page from Ionescu's The Bald Soprano (36K gif)

Two Devanagari typefaces, one traditional, the other a project by Adrian Frutiger.

The sense of the pen angle in the calligraphic is strongly of Northeast to Southwest light; in the Frutiger, it is more strongly North to South.

Frutiger Devanagari

A larger image with a continuous text (~100K gif)

A comparison of custom Mac bitmaps for screen reading to Verdana TrueType at 12, 10, and 9 points.

The custom 9pt is problematical.

The 12pt is the normal browser size, and at 82dpi renders at about 10.5pts physical.

Customized Mac bitmapsVerdana TrueType aliased

An ornamental gif of Really Greek Medium
(TDC2 2001 Type Design Competition Winner)

Really Greek Medium, Alpha to Omega

Jean-François Porchez has provided a PDF that details comparisons of his Le Monde Journal typeface and of Times New Roman.

Le Monde

The PDF is composed of typeface font text converted to outlines, and can be opened in Illustrator, etc. for examination of the curves.

The usage of ornamental faces with Blackletter (Gothic) came up. Here's a modern sample of the use of Lombardic Uncials as a heading for an English Textura Quadrata text. Typically, the usage alternates red/blue/red/blue. This is preferable to the use of all-capital Blackletter characters whose forms and ornamentations drastically impede legibility -- not that Lombardics are as legible as lowercase, but they are certainly more legible than Blackletter capitals!

GM Londinium Ornaments

A while ago, another list was looking for a logo. Idly, I knocked up a prototype; and put it away. On Typo-L the subject of the Golden Mean (Section, Ratio; Phi) arose; and so here:

What goes around.

Q: What happens to glyphs placed in the Space character slot?

A: On this Mac, PS1 fonts will show the glyph. In justified text, extra space is added to the right of the glyph; it isn't centered.

QuarkXPress PS1
PageMaker PS1

TT fonts don't show the glyph at all.
QuarkXPress TT

PageMaker TT

First lines are flush-left, rag-right; second lines are justified. Upshot: for reliable setting of the raised period, use that glyph (Shift-Opt-9); the two rasterisers (TT & ATM) seem to disagree on how to show the space glyph.

Q: What is the relationship of the strokes to each other in sans serif typefaces? Are they all the same?

mirroring a typeform reveals the actual relation of the strokes to each other

Here, you can see that the strong strokes from upper-left to lower-right are distinctly wider than the weak strokes from lower-left to upper-right.

This differentiation is directly derived from the way the angled edged pen will place the weight in those positions.

That all great sans benefit from this compensation of weights -- rather than having their strokes all physically equal -- points to either a possibly culturally-induced bias towards the weighting of edged-pen letters, or else towards a natural optical illusion that causes lower-left to upper-right lines to appear wider and heavier, and therefore necessitate compensating for the illusion in order to appear to be equal; which latter argument positions the edged-pen as the premier instrument for judging the amount and position of such compensation.

Q: What does a blackletter Euro look like?!

A: Might look like this, in Rotunda:

Euro currency in Blackletter

Here is the illustration of the earlier message on the design of the Cyrillic de.

Top to bottom:

1. A Trajan-like typeface project.
2. PT Lazurski.
3. PT Academy (inspired, in part, by Cheltenham).
4. PT ITC Garamond.
5. PT ITC New Baskerville.
6. PT Didona.
7. PT Baltica (a Cyrillic version of Candida).
8. PT ITC Beesknees.
9. SU Evangelie.

Sorry for the low resolution (but I hope you get the idea).


Display of Cyrillic De from; various typefaces

...how essential are the extensions on the lower bar of de? My gut reaction is that, since they confine a counter, they are essential parts of the letter, and are not mere ornament; though I can see the evolution from serif to feature -- like the horizontal of many sans G's.

They are *very* essential, Gary. In the equilateral variety of the de this is precisely those hanging terminals that make the glyph look Cyrillic, not Greek. They are pretty prominent in the earlier, pre-Petrine, styles (ustav, polu-ustav). As a rule, in more "standard" romans (old styles, transitionals, moderns, slab serifs) the shape of the terminals should be coordinated with the terminals of the bar of the T, and the hanging terminals of the tse and shcha. Exceptions exist, though. In some more flamboyant, idiosyncratic styles the terminals of the tse and shcha can go pretty wild (in that sense they are comparable to the queue of the Q).

to munchfonts main site
to munchfonts annex