Maurice Meilleur
_Research
_Patterns
_Words
_Reconstructions
_&c. (coming soon)
_About

_Jurriaan Schrofer’s IAO stamp
Courtesy Jurriaan Schrofer Archive, Allard Pierson Museum, University of Amsterdam

In 1968, the Dutch PTT (Posterijen, Telegrafie en Telefonie; the Dutch post office) commissioned Jurriaan Schrofer to design a stamp to celebrate the 50th anniversary of the International Labor Organization—in Dutch, the Internationale Arbeidsorganisatie (IAO)—in two denominations.
Courtesy Jurriaan Schrofer Archive, Allard Pierson Museum, University of Amsterdam

Schrofer drafted two options for the stamps. One was based on a photographic image, with lettering based on a constructed script design Schrofer had been using since the mid-1960s. The PTT chose instead Schrofer’s favored option, based on a grid of small letters in different weights used as image elements—pixels, though Schrofer wouldn’t have used that term—for larger letters in the design.
Courtesy Jurriaan Schrofer Archive, Allard Pierson Museum, University of Amsterdam

The larger letters, for ‘IAO Nederland 25/45c’, are fairly simple block letters. They, and the smaller pixel letters spelling out ‘1919 Internationale Arbeidsorganisatie 1969’, are variants on the same constructed script design as the one in the rejected option.
Courtesy Nationaal Archief Nederlands


Courtesy Jurriaan Schrofer Archive, Allard Pierson Museum, University of Amsterdam


Courtesy Nationaal Archief Nederlands

Schrofer’s initial plan was to vary both the horizontal and vertical strokes of the pixel letters to create the contrast used to define the larger letters. He finally chose to only vary the vertical strokes. These defined the 71 width units of the pixel grid. The larger rasterized letters in two sizes would have 11- and 2-unit vertical strokes, and 4- and 1-unit vertical counters.

Schrofer chose 3- and 1-unit horizontal strokes, and 5- and 1-unit horizontal counters, for the two sizes of larger letters. Leaving one pixel-unit row of space between the lines of larger text, and one row above and below them, yielded 28 height units for the pixel grid.



Courtesy Nationaal Archief Nederlands




Schrofer experimented with a couple different methods of varying the width of the vertical strokes of the element letters. The method he chose was based on a series of whole-number vertical/horizontal stroke ratios: 1:1, 2:1, 3:1 … 7:1. Since the columns of the element letter grid were defined by the letters’ vertical strokes, he drew them with static heartlines, in a constant horizontal space sized to ensure that the sidebearings of the glyphs were half their vertical counterspaces at all weights. Single-vertical-stroke glyphs would occupy one instead of two cells horizontally.
Courtesy Courtesy Jurriaan Schrofer Archive, Allard Pierson Museum, University of Amsterdam

To make the mechanicals for the stamps, he ordered bespoke rubdown type from Mecanorma.

Schrofer tried a number of ways to use the different weights of his script to add texture to the contrast between the larger letters in the stamp and their background, as we can see in these sketches. (The smaller letters, for ‘nederland’ and the denomination, use a simple heaviest/lightest contrast.)

Courtesy Nationaal Archief Nederlands

Courtesy Jurriaan Schrofer Archive, Allard Pierson Museum, University of Amsterdam

He chose a diagonally-offset weight modulation for the cells making up the large letters, oscillating from the heaviest weight (7:1) to the third-lightest (3:1). The offset oscillation creates a wave pattern across the letters. Using the lightest weight (1:1) for the background letters and omitting the second-lightest weight (2:1) in the field kept significant contrast between the larger and smaller letters.
Courtesy Nationaal Archief Nederlands

A strict plan for diagonally-offsetting the oscillating weight modulation would have looked look something like this: starting with the heaviest weight in the leftmost column in the second row (the first one containing the larger letters), 7-6-5-4-3-4-5-6 …

But that’s not the plan Schrofer used. Remember, he was using custom dry-transfer lettering to make his mechanicals. And for reference: Schrofer drew his glyphs 26mm tall and had the drawings photographically reduced to 1/3 scale for the rubdown lettering. That left letters 8.67mm tall.

In order to create weight contrasts between one cell in the grid and the next, he had to cut those rubdown letters in half vertically before applying them. To follow a strict plan, he’d have had to cut letters for most of the 1349 cells in the area of the field where the larger letters are. Only the single-vertical-stroke letters, and any two-stroke letters in the background, wouldn’t need this. That’s a lot of careful cutting, and evidence from his papers at the time suggests he didn’t have assistants he could hand the work off to.

For this reason, I believe, Schrofer chose to keep the vertical stroke weights constant inside any given rubdown glyph, except when transitioning from large-letter figure to the background, or vice-versa. He could have done this while shifting to the weights the strict plan otherwise would have indicated with every new rubdown glyph. That would have looked like this: again starting from the leftmost column in row 2, 7-6-6-4-3-3-5-6-6-6-6-1-1 … But this would have frozen the modulation in places, and somewhat reduced and complicated the overall effect—the interference is subtle, but it’s there.

So—again, this is my guess at a reason—in the end Schrofer refused such a strict approach and modulated weight in the rows strictly by glyph. That looks like this: again starting from the left in row 2, 7-6-6-5-4-4-3-4-4-5-5-1-1 … which matches what we see in Schrofer’s mechanicals and the produced stamps. This choice effectively turned the diagonal axis of the weight waves slightly clockwise, but made them more distinct.


 (All images not otherwise credited were generated with code, in Python with DrawBot.)