Category Archives: optical scaling

More on optical scaling

My last post (4/23) explained that some digital type designers today are interested in the way Morris Benton’s fonts, and indeed all the metal types produced by the American Type Foundry in the early 1900s, were optically scaled. Optical scaling was easily accomplished at ATF by adjusting certain settings on Linn Boyd Benton’s matrix engraving machine. Linn Boyd Benton explained in an essay he wrote in about 1906:

The adjustments are such that the operator is enabled to engrave the letter proportionately more extended or condensed, and lighter or heavier in face, than the pattern. All these variations are necessary for the production of a properly graded modern series containing the usual sizes. In fact, on account of the laws of optics, which cannot be gone into here, only one size of a series is cut in absolutely exact proportion to the patterns.

The illustration of optical scaling reproduced below was made in 1989 by ATF’s successor, the Kingsley/ATF Type Corporation. At the time, Kingsley/ATF was embarking on a program of digitizing typefaces, including the optical scaling characteristics of the original metal types. Ultimately, the company went bankrupt in the early 1990s, but that’s another story.

The illustration uses the capital M from Morris Benton’s Wedding Text, designed in about 1901. In the earlier “metal type” days at ATF, the set of Wedding Text patterns, one image for each letter (these particular patterns, by the way, are now part of RIT’s Benton Collection), were used to produce matrices for every type size. According to the handwritten “daybook” of general engraving machine settings for cutting the matrices for 228 ATF typefaces, no size of Wedding Text was cut exactly proportional to the pattern. Instead, the matrices for each type size were either condensed or expanded in relation to the pattern. (In most other typefaces, one size, usually somewhere in the middle of the range of sizes for that face, was “normal,” i.e., the letters were cut exactly proportional to the images on the patterns, not condensed or expanded.)

To generate this illustration, Kingsley/ATF photographically enlarged these three sizes of a Wedding Text capital M to a uniform height, so that customers could then easily compare them. Notice that the smaller the size, the more expanded the character. This is necessary simply for legibility, although in the days of metal type mechanical parameters also dictated that smaller sizes be expanded.

In addition to the expansion or condensation of the letter, the “set width” of letters in different sizes also had to be adjusted for good optical scaling. The set width is the total amount of horizontal space (width) on a piece of metal type. In order for the eye to be able to read very small type, more white space is needed around each letter, so the type needs to be relatively wide.

Kingsley/ATF produced the following illustration, also in 1989, to show its customers this aspect of optical scaling. Because enlarging this sample will perhaps also distort it, I’ve left it at its original size; I apologize for the very small 6-point example. But hopefully it is understandable. I’ve re-typed the Kingsley/ATF caption to this illustration below it in case the original caption is too small to decipher.

Kingsley/ATF’s original caption: “Notice the difference between a true 6-point type enlarged to 24 points and a true 24-point type. The true sizes were created using Optical Scaling. Typeface: Wedding Text”

More later …


Why a book about the Bentons?

Why would anyone today care about what the Bentons were doing with metal type more than 100 years ago? (These days I’m in the last stages of tweaking my book about them, and I must confess that sometimes this question still haunts me.) But in fact, I’ve found that plenty of people do care, especially those involved with designing or improving digital type.

Just the other day a group of digital type designers came to RIT who were pursuing the train of thought begun in a 1912 type legibility study, one that Morris Benton worked on!

Type designers today have to deal with what came naturally to the punchcutter when he cut punches by hand. [For anyone new to the subject, the punch was a raised, reversed (“wrong-reading”) image that determined the shape of a letter. It was pounded into a brass bar which became the matrix, with a sunken, right-reading image. The matrix, fixed in a type mold, was then used to cast individual types, again wrong-reading so that they would appear right-reading when printed.] What came naturally to the punchcutter was that letters must be slightly distorted as they become larger or smaller. Today we call this phenomenon optical scaling.

In my book I explain the term like this:

Briefly put, because of the inherent limitations of the human visual system, as the size of a letterform gets smaller it needs to be expanded, and as it gets larger it needs to be condensed. “Traditional punchcutters and scribes made such proportional changes in order to optimize legibility,” RIT Professor Charles Bigelow wrote. “Recent research in visual perception suggests that such proportional changes are necessary because the human visual system has non-linear sensitivity to visual features of different spatial frequencies.”[1]

Benton understood that slightly distorting the characters as they went up or down in size enabled the basic design to retain its integrity, and so he built this capability into the matrix engraving machine. For example, according to a hand-written ATF reference book, each size of Morris Benton’s Wedding Text was expanded or condensed relative to the pattern for the typeface. Many years later, ATF’s successor, the Kingsley/ATF Type Corp., coined the term “optical scaling” to describe this phenomenon of distorting letters.

[1] Charles Bigelow, “On Type: Form, Pattern, & Texture in the Typographic Image,” Fine Print 15, no. 2 (April 1989): 77.

Optical scaling is as important today as it ever was. Digital type on a computer screen especially needs to be as legible and comfortable to read as possible.

Type designers may ask: Can we systematize optical scaling? Is there an algorithm that we can use to tweak the letters at various sizes for optimum results? This is the same question Benton faced in the early 1880s in Milwaukee as he worked on his pantographic matrix engraving machine. Theo Rehak told me this past February (see my 2/14 and 2/22 posts) that he was sure there was such an algorithm. But, as he explained in Practical Typecasting, Benton’s “Engraving Factor Tables” were lost after he died. Today Rehak is not in the business of producing the full range of sizes of his metal type offerings at the Dale Type Foundry, so he can manage without such an algorithm.

When I Google “optical scaling Benton,” there are about 183,000 results in 0.28 seconds; it’s not a dead subject. It seems to me that anyone who admires Morris Benton’s optically scaled typefaces would do well to study how his father’s matrix engraving machine actually distorted the letterforms on the patterns it used for making type in different point sizes. My book would be a very good starting point for such a study.

A trip to see an original Benton matrix-engraving machine

This Saturday I’m going to visit Theo Rehak in Howell, New Jersey, to see his two working pantographic matrix-engraving machines, invented by Linn Boyd Benton. I think one of them dates all the way back to Benton’s Milwaukee type foundry, from about 1886. Theo and I have been corresponding for years, at least since 1993 when he read and edited an article I wrote for the American Printing History Association Journal about Linn Boyd Benton and his son Morris Fuller Benton.

At that time, I didn’t have a clear, detailed photograph of a Benton engraver for the article, so I traced one from a nine-year-old photocopy of a magazine article about how the American Type Founders Company (ATF) made type in the early 1900s using the “Benton system.” The article had appeared in the American Machinist magazine for December 16, 1909. I couldn’t find an original copy of that magazine anywhere in my home town of Rochester, New York. I had even traveled to the Syracuse University science library to photograph its copy of the article, only to find that those exact pages had been cut out of the bound volume of American Machinists from 1909!

So I struggled with a very dark photocopy of a picture of the Benton engraver, and later found out that Theo didn’t like my tracing! Maybe I left out some important part of the machine. I’ll ask him about that.

I had actually visited the old ATF headquarters in Elizabeth, New Jersey in November 1984 with Richard Marder, the grandson of one of the founders of ATF. He spent the better part of a day explaining many things to me, which I recorded in a notebook that I still have. Mr. Marder helped me to understand how the Benton engraver worked and told me what he remembered about Morris Benton. At the time I was researching the Bentons for my master’s thesis in Printing Technology from RIT.

The day I visited ATF, Theo Rehak was there working, although we didn’t meet each other. The company was struggling to stay in business; it was now a tenant in the building it had formerly owned. Mr. Marder introduced me to George Gasparik, who gave me a tour of the facility. We had to move the plastic off of several Benton machines so that I could photograph them—only one or two were actually being used. The photographs I took weren’t particularly detailed.

In 1984 I didn’t get to see how the Benton machine was adjusted for optically scaling the letter patterns it used to produce matrices for different sizes of type, but this Saturday I will. I’m bringing along copies of about ten or so pages of the ATF “Day Book,” which gives instructions for adjusting the machine for the various sizes of specific fonts. I’ll also take copies of the “cutting slips” for Morris Benton’s Freehand that Theo donated to the Benton collection at RIT’s Cary Library.

Mr. Marder read my completed thesis in the summer of 1986 and made a cassette tape of comments about it for me, which was very helpful for revising the manuscript. My expanded story of the Bentons, with many illustrations, is going to be published by the RIT Press. I’ve modified the original thesis considerably so that it will be understandable to readers who have no background in type.

This has been a very long process for me, and I guess I’ve been preparing for this trip for years. I’ve invited my two sons to come along on Saturday and see something they’ve been aware of their whole lives (one is 22 and the other is 19). Luckily they both live near Theo’s type foundry, Roger in Manhattan and Gus in the Bronx as a student at Fordham, so it will be an interesting diversion for them (I hope). In any case, Theo has mentioned several times that we’ll “do lunch,” which sounds good to me.

Tomorrow, Friday, I’m going to see Jan Siegel, the Rare Book Librarian in the Rare Book and Manuscript Library at Columbia University. Columbia inherited ATF’s extensive type library the first time the company went bankrupt (in 1936), and so all of the original type specimens and books that Morris Fuller Benton studied for his type revivals and legibility studies should be there. This ATF Collection has an original copy of the American Machinist article, as well as several other original documents I want to photograph. Actually, the main reason for my going to Columbia is to show Jane the 21 references I’ve made to Columbia in my book, to make sure they are properly documented.

More to come …