Chiaroscuro watermarks are considered the pinnacle of the papermaker’s craft. The process of creating them is difficult and technically demanding. It is both the technical aspect and inherent beauty that I find so alluring. This remains as true as it did when I expressed these exact sentiments in my 2012 Hand Papermaking article on the subject.1
I have had a decades-long fascination with light-and-shade watermarks. I created my first chiaroscuro watermark—a portrait of Dard Hunter—in 2001, and since then have explored several novel methods of creating them with varying degrees of success. My latest method employs the use of a 3-D printer to create the entire papermaking mould with the watermarking surface built in. Before I explain how this is done, it would be instructive to understand how watermarks are created during sheetforming and the traditional method of creating the embossed wove facing.
The principle of how these watermarks are formed is straightforward. The bronze wove mesh or facing that makes up the mould’s surface is pressed between mating dies creating a three-dimensional surface. During sheet forming, areas that are deeper collect more pulp, resulting in darker areas of the watermark image, while areas that rise above the surrounding mesh collect less, corresponding with lighter image areas. Between these extremes lie all the midtones that combine to create the finished chiaroscuro watermark.
The process of creating the embossed mould facing is complex. It begins with a semi-translucent slab of wax, sometimes tinted, about ¼-inch thick. Wax is used because its translucency mimics paper and both are viewed using transmitted light. Using an upright light table, an engraving artist traces the intended image onto the wax surface and then selectively carves away portions of the wax to produce thinner, and therefore lighter areas of the image.
Once the wax engraving is complete, it must be converted into male and female dies utilizing the process of electroplating. The surface of the wax is first coated with a layer of electrically conductive graphite powder. It then goes into an electroplating bath where a thin layer of metal, usually copper, is deposited on the surface. The wax is then removed, leaving a thin, fragile copper shell that is strengthened by filling the backside with molten lead to create a die of the required strength. Following the same process, this die is used to make a second mating female die.
Next a piece of 60 mesh brass screen is annealed using heat, which reduces the internal stress and hardness of the weave making it easier to form. This mesh is then pressed between the male and female dies creating the finished embossed facing which is sewn to a specially prepared mould frame whose ribs have been altered to fit the recessed areas of the screen. The entire process is lengthy and technically demanding requiring specialized equipment, which puts this technique out of reach of most hand papermakers.
In the early 1990s the UK company Portals, which in the past made the security paper for British banknotes, developed a computer program that would reference the grayscale values in a black-and-white photograph, assign them a corresponding depth (deeper for areas that are darker), thereby creating a virtual 3-D model of the male die that could then be machined from brass, removing the need for a wax engraving. This technology eventually spread throughout the security paper industry.
I came across this novel method of creating embossing dies in 2001 while doing a Google search, which revealed patents describing the process. I found a local machinist able to create a male die using this method when creating my first light-and-shade watermark of Dard Hunter. Since then, I have experimented with other methods to bypass entirely the need to make male and female dies, including the use of laser engraving technology to carve polyurethane foam into a 3-D mould surface.2 The results were promising at first, but the sponge-like mould surface remained problematic in terms of image quality and long-term integrity of the material.
In the 2010s, around the same time as I was laser engraving foam, I became interested in the emerging technology of hobbyist 3-D printers. The process begins by creating a virtual 3-D model in a 3-D modelling program, which the 3-D printer then prints in plastic. I purchased the first available 3-D printer kit from the UK and started 3-D printing papermaking moulds and deckles as well as the mould’s laid facing including integral line watermarks. Unfortunately these printers, known as Fused Deposition Modelling (FDM) printers, cannot print at a resolution fine enough to print a wove facing or mesh. In 2020 I read about a new kind of hobbyist 3-D printer that uses UV light to harden layers of photosensitive liquid resin to print the 3-D model. Known as stereolithography (SLA), these printers create 3-D prints at a much higher level of resolution. My first thought was, can I now 3-D print a wove surface and if so, possibly a chiaroscuro watermark?
In January 2021 I purchased my first SLA printer and began experimenting. I started by creating a 3-D model of a plain wove mesh. Rather than 3-D modelling a true woven mesh in which wires weave over and under each other, I perforated a thin flat sheet with square holes to simulate the interstices of a woven mesh. I also made it 5 millimeters thick because I knew I was going to be removing some of that height when I carved into it to create the undulating surface. In addition I added coarser layers of mesh simulating a backing and sides to strengthen it.
After 3-D modelling a virtual plain wove mesh, I created a virtual die using a computer program called Photomesh that is designed to create lithophanes. Dating from the early nineteenth century, lithophanes were traditionally made of fired porcelain clay that is translucent so it can be backlit to reveal an image, very much like a watermark. Today lithophanes can be 3-D printed by converting a grayscale image into a virtual 3-D model for 3-D printing. Instead of using the 3-D model to print a lithophane, I use it to create a virtual die, the virtual equivalent of the physical machined brass die.
The steps I described above to create a traditional chiaroscuro watermarking mould is all done virtually within a computer program: the virtual die is pressed into the virtual wove mesh to a depth of 4 millimeters, then the program subtracts the virtual die leaving its impression on the virtual wove mesh. The virtual 3-D model is then 3-D printed, bypassing the need for a wax engraving, male and female brass dies, the embossed wove mesh, and even the mould frame itself as the entire finished mould is 3-D printed as one piece.
Now, with the watermarking mould ready, we turn our attention to the furnish. The type of fiber, typically cotton, and how it is processed in a Hollander beater are as important as the mould itself. Even the highest quality Fabriano chiaroscuro watermarked mould will yield a poorly defined watermark if an improper furnish is used. The most important factor in preparing a furnish for chiaroscuro watermarks is fiber length, which must be extremely short. The second factor is the rate of drainage during sheet forming. A slow-draining furnish is the enemy of a good watermark. The goal is a short, free stock. This is accomplished by significantly increasing the pressure between the roll and bedplate from the beginning of the beat in order to favor a cutting action versus bruising and fibrillation (both of which are still required but not as much as cutting). I achieve this with my Valley beater by adding a 15-kilogram weight to the lever arm from the beginning of the beat.
While the quality of these watermarks does not compare with the finest examples from Fabriano, this is a new technology with factors that I do not yet fully understand, so I continue to experiment and make incremental improvements. I am confident I can further improve quality as the technology improves. Like smart phones, new models of 3-D printers that feature ever-increasing resolution are released every year. My hope is that this new technique—far less costly than producing traditional chiaroscuro watermarks—opens the door to papermakers and artists to explore this rich and fascinating field.
Brian Queen, “Computer-Driven Laser Cutter for Chiaroscuro Watermarks,” in Hand Papermaking vol. 27, no. 1 (Summer 2012).
For more on laser engraving polyurethane foam, see my 2012 article, cited above in endnote 1.