A Phototypesetting Mystery

•

Carl Johnson

Hoxsie departs from its love of Albany, Schenectady and Troy history for a brief foray into offset printing history –

Someone recently reached out to a page I administer for the AM Varityper Phototypesetters, sending along some pictures of something that had been in his house since he bought it ten years ago. And at first blush, the pictures certainly looked like somewhat standard phototypesetting discs – except that they were made of metal, rather rough on the edges and in the center, and not in any way transparent. So while they looked like the discs that were used to create phototype, that’s not what they were.

He said they were 16.75 inches across and more than 1/8 inch thick. The fronts certainly look like something that would be familiar to anyone who ever set phototype. There are alphabet characters (I can’t identify the font exactly, but it’s a fairly standard style in the vicinity of Garamond). Then, there are little rectangles, in rows of 8, associated with each character, and a smaller set of two more running just inside near a central circle. Clearly, these are codes that let the machine identify where the disc is positioned – in a phototypesetter, these discs would spin quite quickly, and when the needed letter was in front of the light source, there would be a flash of light to expose the character onto the sensitized paper on which the type was set. You think digital is amazing? These things were photomechanical wonders. The back, interestingly, has a wonderful logo that just says “Imperial.” So, I can only deduce that what we’re looking at here is a master – the form from which a phototypesetting disc was made. The disc itself would have been either glass (as some of the larger ones were) or plastic film.

At first, I didn’t find much looking for “Imperial” and “phototypesetting.” Then I had the flash that may be Imperial just supplied the metal, and not what was on the other side. That turned out to be incorrect, but led me to the right company anyway. Imperial is the Imperial Type Metal Company of Philadelphia. Formed as the Imperial Metal Co. in 1913, it was renamed as Imperial Type Metal Co. in 1915. Founded by a Wilson S. Yerger, specializing in the manufacture of metal that was used in creating type, focusing on its metallurgical expertise.

But I couldn’t find anything that indicated a foray into phototypesetting – which was something a LOT of companies got into, even when they didn’t have any previous printing industry experience. Singer is one example. But I did find a couple of patents that looked promising. For instance, Imperial Type Metal (of 3400 Aramingo Avenue, Philadelphia, a site that has long since been redeveloped, but was, not surprisingly, a federal superfund site) held a trademark for Vitacoat and another for Vitacold, both registered on the same day in 1963. Vitacoat was a water solution of methyl violet dye used to develop the photographic image formed on a photoengraving plate; Vitacold was a water solution of shellac and ammonium dichromate with other ammonium compounds added, applied to a sheet of photoengraving metal as a liquid and is dried. Vitacold as a trademark also appears to have applied to a methyl alcohol solution of methyl violet dye used to develop the photographic image formed on a photoengraving plate by a light sensitive cold top. They also held a patent, registered in 1968, for a powderless etching method for etching relief images in aluminum. These are all related to photographic printing plate technologies of the type used in the photographic offset printing revolution. But still . . . they didn’t appear to have made a phototypesetter itself.

Imperial Type Disc closeup 1
Imperial Type Disc closeup 2

Then there was another flash – perhaps Imperial, experts in printing plate technology and innovative metallurgy (etching in aluminum was a challenge) just made the masters for some other phototype company? So I started by searching for Photon, the company that started it all and with which I was least familiar. A nifty little history of Photon I found mentioned a phototypesetter by Harris Intertype:

The Harris phototypesetter, called a TXT, was about eight feet long, four feet wide, and six feet tall. The fonts were on a half dozen platter-sized glass plates spinning at high speed at the ends of radial arms that rotated around when a type face change was called for. These monsters were used by newspapers well into the 1980s.

Platter-sized, you say? Intriguing – because what I was looking for, 16.75 inches across, would qualify as platter-sized. It would also appear that the Harris machine used an eight-hole punch tape for its coding (punch-tape machines used in typesetting varied between six and eight holes across for coding). Now, I just needed to find an image of the Harris-Intertype disc. What are the odds of that? Ahh, it’s the age of the internet; the odds are fairly high. And while the book I needed, David Jury’s “Reinventing Print: Technology and Craft in Typography” isn’t fully available online, its preview was enough – right there is an image of the Harris-Intertype Fototronic disc. Characters on the outside, code boxes on the inside, and a good-sized disc: looks like the Fototronic is at least a close relative of the master we were looking at. So at this point I’m pretty sure this is an Imperial master for a Harris-Intertype phototypesetting disc.

Harris-Intertype Fototronic Disc

Even among collectors of old printing equipment, this is a somewhat obscure part of the process – the device that made the device that made the type. But it was a fun mystery to try to solve!

Update!

In 2021, Stanley DePassos has been good enough to share the following with us:

“Attached here, is a photo of the inside of a Harris Fototronic 4000.  This dates from about 1982 at New England Typographic Service in Bloomfield, CT.
“You can see the five glass disk fonts mounted on a turret.  In operation, the disks spin continuously at 3,600 RPM.  Each font has two typefaces, usually roman and italics of the same weight.  To change fonts, the machine rotates the entire turret while the disks are spinning; quite a feat of engineering.  At the lower left (below disk #1) you can see a portion of the lens turret.  There were 12 lenses, one for each point size available on the machine.  The range was usually from six to 24-point type.  When a size change command is received, the lens turret also rotates to position the correct lens behind the font.  Fonts are changed by unscrewing the chrome knob at the center of each disk mount.
“As I recall, fonts cost about $900, real money in the 1970s and ‘80s.  You would not want to drop a font while changing them.
“The Fototronic was usually installed into a wall so the film could be loaded and unloaded in a darkroom or there was a person-sized closet attached to the back of the machine that was light-tight.”

•

6 responses to “A Phototypesetting Mystery”

  1. William Shipley

    I have a Harris disk from a Fototronic 1200 typesetter. I wrote software to interface with one in the late 1970’s. Each disk had two fonts, the inner and outer. Typically it was regular and italic or regular and bold. Mine is Helvettica regular. The inner square holes are the encoding of the character widths.

    The device had a pentagram, holding 5 disks. It would rotate the selected disk into place in front of a strobing light source. There was also a rotating disk of lenses to allow for changing size. So to set a character you would make sure the disk was rotated into position with the right lens and then it would flash the light through the spinning disk at the precise time to image it on the photographic paper and advance the amount for the next character.

    Reply
  2. Stanley De Passos

    I used the Fototronic TxT and 4000 models in the 1970s and ’80s. I kept one of the glass disc fonts and it is 9.75″ in diameter. Since the discs were manufactured photographically, I don’t see how your metal disk could be used to make fonts. The font disks appear to have a photographic emulsion on one side but I don’t know if it was exposed in a contact frame or via some type of projection. Later fonts did away with the square character codes in favor of a second row of slits: one for timing and one for to identify the character.

    William Shipley’s description of the machine is correct. The font turret was driven by a belt and the fonts rotated at 3,600 rpm. Typesetting speed on the later model machines was in excess of 100 newspaper lines per minute. It was also possible to purchase high-speed fonts that contained two copies of the most common letter in each row (e, t, a, i, etc.). These were specifically for newspapers, attaining even higher speeds. Many commercial printers and trade typographers adopted the Fototronic machines. They were beautifully engineered, much more reliable than their competitors. Unfortunately, Harris lacked the authentic type library of Linotype which limited their use in ad typography.

    I have photos of the inside of a 4000 model which I can send if you are interested.

    Reply
    1. Carl Johnson

      Thanks, and sorry it took so long to approve the comment; I didn’t get a notification on it. I’d be glad to post your pics!

      Reply
  3. Andre Bormanis

    My father has a glass Photon Inc Disc 2F in a wooden box. The disc includes eight fonts.

    Reply
  4. Don Walters

    I too have a TxT font disc leftover from my days running the TxTs and 480s.

    The 480 was the “little brother” of the 4000 … it had only two font discs rather than five.

    The TxTs were more modern machines controlled by an actual computer, in our case a Varian brand–and I think you could get a different controlling computer.
    These computers had NO hard disc storage. Every time you changed a font disc, you had to also load a “width tape” to tell the computer the widths of each character for the disc you just loaded. Failure to do so, all the characters would be spaced wrong and the resulting type was unusable.

    The 480 had a “dual head” paper tape reader, about 3 inches apart.
    The paper tape was loaded such that the first head would read the tape, then using the square “tv” shaped windows on the disc, calculate the width of each character so it knew how to justify the line.
    The 2nd reader would then read the letters and “flash” each character in turn onto the RC paper on the back of the machine.
    The TxTs had an actual computer, so one read head was enough, it could read the line, calculate the justification, then expose the characters directly out of memory.

    On the 480/4000 the RC paper would move back and forth on a carriage, much like the way the paper moves on a typewriter.
    On the TxT machines, the characters were flashed first to a servo mirror, which would turn back and forth to shine the letters where they were to be positioned on the RC paper. There was a curved lens between the servo mirror and the paper to “flatten” the image since the characters would be distorted otherwise.

    As a teen (in the 1970s) I ran the TxTs as my job, for $3.25 per hour in Los Angeles area, working for a large typesetter producing type for magazines. And yes, dropping a font and breaking a $1200 disc was a firing offense. Both because a grand was a lot of money back then, but also because we would have to re-order the font, and couldn’t typeset anything in that font until the new disc arrived–potentially losing a customer who would have to have their type set somewhere else in the mean time (at least).

    The fonts on disc came in different “layouts,” like a 1000 layout or 2000 layout … which dictated which characters were on the disc in what positions.
    There was a layout called “TxT 652” … which repeated popular characters like e t a o i n two or three times which reduced the “spin time” to get to the next instance of that character, and yes, then you could approach 100 newspaper lines per minute of output.

    Reply
  5. Stanley De Passos

    Don Walters’ comments add more details to setting type using the Fototronic machines. It sounds like his used internal programs to justify (and, perhaps, hyphenate; “H&J”) text lines. Many of these phototypesetters were driven by “front-end” computers with editing terminals that performed the H&J and output pre-justified paper tapes coded for the Fototronic or other devices. Minicomputers from DEC and Data General most often ran software from companies such as Penta, CSI, Magna, and RayPort. At RayPort, we hired an engineer to build an interface to the Fototronics to eliminate paper tape.

    Don also reminds me of the field-flattening lens in the machine just ahead of the photographic film or paper. As he describes, an oscillating mirror would write out letters across the film and the flattening lens would make them appear flat and square. This was a much better solution than some other machines which actually curved the film to keep it equidistant from the mirror. The Fototronics also used pin-feed film so forward and reverse movements were accurate.

    Andre Bormanis, in another post, mentions the Photon font disc. These machines worked in a similar manner to the Fototronics. Both these, and many others, worked on the stroboscopic effect: If you flash a light of very short duration at a moving object, it will appear to stop the motion. This was first applied to typesetting by two engineers at AT&T France in 1947. Their invention was first commercialized by the Deberny & Peignot type foundry, also in France. They sold the rights to Photon which developed it in the US. Being first to market meant that Photon had to bear the development costs and the company was never profitable. Their machines were popular especially with newspapers but competitors soon overtook them.

    These electromechanical marvels enjoyed a relatively short lifespan; from about 1955 to about 1985. By the late 1980s, several companies introduced CRT-based typesetting machines. Fonts were stored in digital formats on disk and letters were written to a video screen which exposed the film.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *