Readers' respons to a critique of glyph scaling in text justification

TEXT JUSTIFICATION

Readers’ respons to the article “InDesign, the hz-program and the secret of Gutenberg”

By Torbjørn Eng

Following the publication of this article in 2007 I have received a few comments, and it has also caused some debate on the Internet. Below you’ll find some extracts from these.
Debate within the TeX community “I wonder if some of you are aware of this article... I have a hard time believing Zapf and Bringhurst are completely wrong; I am almost sure they have had occasions to check the B42 by themselves in museums and libraries.” DW - February 2008 “The point is that font expansion shouldn’t be visible. In practice this means that expansion factors of 5% are visible, but only by TeX users. Expansion factors of up to 3% are usually not visible at all. As far as I know the default value in the microtype package is 2%. You will not see any difference but, though the value is ridiculously small, your paragraphs are formatted much better. The example text provided by Karow is a bad example. pdftex will never produce such bad output, unless you set the expansion factor to a very high value. I suppose that this had been done by Karow in order to demonstrate the hz-algoritm. Of course, with expansion factors of 10% or 20%, you can easily prove the whole concept wrong. But it isn’t. There are a few interesting points in the article you mentioned, though. The hz-algorithm implemented by URW changes the width of glyphs but retains the width of vertical stems. This is not a bad idea but even if it’t supported, expansion factors greater than 3% are undesirable and you’ll see no difference. The authour of the article you are talking about assumes that the hz-algorithm is a bad thing and tries to prove his assumption, which is definitely wrong.” RK - February 2008 “Here are my (perhaps biased) opinions: - The article made objections to some of the points claimed by URW, Hermann Zapf and others. I more or less agree with the objections in the article. - However I do not agree with main intention of the article, ie to show/prove that glyph scaling leads to reduced quality. The objections in the article are not convincing enough to me. It only showed that there are some inaccuracies in what URW/Zapf/Adobe claims. - Like other techniques, glyph scaling is also a double edged sword: it can help and also hurt. The users should be aware of that. - I use glyph scaling myself and find it useful, especially when you want to give a bit more flexibility to paragraph formatting and there is no other possibilities to do so. The limit I find useful is +/- 2%, but when looking closely I can also see the effect of glyph scaling at that limit. I guess -1% to +2% is a better choice.” TTH - February 2008		Glyph scaling in text justification means that it is permitted to make the widths of letters wider or narrower on separate lines of justified text. In InDesign this also means changing the width of the vertical stems of letters. This is done in order to reduce word spaces and the number of hyphenated words at the end of lines. My objection to this is that changing letterforms leads to irregular visual appearance of lines and that it gives pages or columns uneven “colour”. This results in less than optimally harmonious typography and may disturb the reader. I also believe that the references to Gutenberg and Zapf in order to defend this technique are misleading. – TE The debate on the left hand side took place on “TUGboat – The Communications of the TeX Users Group.”

Another approach: make variants of those characters where you can't see it “I attach a scan of a specimen of the antiqua of the Bremer Presse. As you can see, f, r, s, t and also the ligatures ch and ck have different variants with different widths. I'm sending this scan to you because the Bremer Presse used this antiqua to produce “De civitate Dei”, the finest book of the 20th century where these matters are concerned according to Zapf. So there is another approach to produce a page as even as the ones of the B42. Just make variants with different widths of those characters where you can't see it. Of course not m, n, o, l, i. But if I remember rightly, the Subiaco type by the Ashendene Press has different a's and e's. And I don't think this would be too hard for designers to realize and it was a way that Han The Thanh followed from the beginning. Glyph scaling in general is crude of course but it can be the lesser evil. There are two facsimile pages of “De civitate Dei” in “Wo der Buchstabe das Wort führt” by Kurt Weidemann, Curt Weidemann und Cantz Verlag 1994, pp. 256-257.” Email from MV - August 2008		Click on the image for a larger version.

Man muss mit das Werkzeug richtig umgehen “Bei Dehnungen und Stauchungen der Wortzwichenräume arbeitet TeX allerdings mit etwas anderen Methoden als den von Dir beschriebenen Regeln. Ich weiss nicht, ob das vielleicht daran liegt, dass TeX aus den amerikanischen Raum stammt. Allerdings hat Knuth mit Hermann Zapf zusammengearbeitet, so dass ihm diese Regeln, wenn sie denn spezifisch deutsch sind, vielleicht auch bekannt waren und er sich aus anderen Gründen dagegen entschieden hat. Von Zapf stammt wiederum ein Algoritmus namens HZ, der versucht, die oben beschriebene Gutenbergsche Vorgehensweise zu imitieren, indem die einzelnen Lettern gestaucht und gedehnt werden (was aber nicht das gleiche ist). Eine vereinfachte Variante von HZ wird heute im satzprogramm Adobe InDesign angeboten.” - S - Oktober 2008 “Buchstaben stauchen und dehnen hört sich nach Verbrechen an. Was ich wiederum mit Meister Zapf nicht zusammendenken kann. Wie sieht denn das aus? Kann man das irgenwo anschauen?” – MZS - Oktober 2008 “Vom originalen HZ-Algorithmus weiss ich nicht viel, die Welt offenbar auch nicht :-) Auf jeden Fall wird das Glyph-Scaling auch heftig kritisiert (in “InDesign, the hz-program and Gutenberg’s Secret”). – S – Oktober 2008 “Der Text von Torbjørn Eng ist sehr erhellend, vielen Dank nach Stralau! Vielen Dank auch an Chris für den Link zu Zapfs Artikel. Den lese ich bald, ich bin aber durch die Probe schon überzeugt. Die abschreckenden Beispiele von Torbjørn Eng sprechen nicht gegen das Werkzeug, man muss mit ihm eben richtig umgehen, wie mit allen Werkzeugen. Das Auge den Typografen entscheidet, das Werkzeug scheint mir aber grossartig zu sein. Und in InDesign ist das also Schrott, wie T. Eng schreibt? Immer noch?” – MZS - Oktober 2008		Debate on the blog of a Berlin letterpress workshop (http://www.druckerey.de/) – http://www.blog.druckerey.de/ Dehnungen = expansion Stauchungen = condensation “To make letters narrower and wider is a crime.” “I am convinced by the test. The horrifying examples of Torbjørn Eng do not go against the tool, one has to handle it correctly, as with all tools. The eye of the typographer rules, but the tool seems great to me.”

Please contact me for your comments on glyph scaling.

To the article “InDesign, the hz-program and Gutenberg’ secret” (in English) To a collection of opinions on glyph scaling by well-known typographers (Robert Bringhurst, Olav Martin Kvern, Ari Rafaeli, The Thanh Han) (in English) To main page of “Typografi i Norge”. Til artikkelen «InDesign, hz-programmet og Gutenbergs hemmelighet» (på norsk) Til noen kjente grafiske formgiveres/teoretikeres meninger om glyph scaling (Robert Bringhurst, Olav Martin Kvern, Ari Rafaeli, The Thanh Han) (på norsk) Published for the first time 2008-12-21. Last updated 2008-12-25.