Woodburytype, Woodbury-process and Photo-lithophane photographic relief printing process by Nicolai Klimaszewski.
Always be careful when handling chemicals. Read the health and safety instructions.
I. General overview of 19th century Woodbury relief processes
This first concept is easier to demonstrate than to explain. Carefully flatten a pile of mashed potatoes, being sure to impart an interesting variety of fork marks. Then prepare a light and transparent gravy and proceed to pour it onto the potatoes. If you carefully scrutinize the results, you will find that the gravy has in fact produced many shades of brown. It is darker where the fork went deep, and lighter in the high spots. The methods of imaging that follow all use this basic concept: that by varying the thickness of a single colorant, you can produce a wide range of tonal values.
- PRINTS; Woodburytype images are made by pouring a translucent mixture of pigments suspended in warm gelatin onto a relief surface, then transferring this pigment layer onto paper. In this case, the thick and thin areas of gelatin, along with the white of the paper produce a wide range of beautiful tones. I state that the tones are beautiful because the thickness and translucency of the pigment layer provides a remarkable kind of color that a conventional flat deposit of ink cannot achieve.
- RELIEF; Emaux Ombrant or Ceramic Intaglio describe the same process. A white ceramic slab, usually porcelain, contains the surface relief. A fluid translucent glaze is melted on this surface to produce a range of tones created by the varying thicknesses of glaze. Most of the nineteenth century examples of this technique were not photographically produced, even though the images appear to be photographic. Skilled craftsmen hand-carved the molds by working from photographs.
- TRANSMISSIVE; Ceramic Lithophane images are made from translucent porcelain and require viewing by backlight, such as against a window. The thick and thin areas of clay produce dark and light tones as light passes through it. Again, most nineteenth century examples were executed by hand-carving the original art. The viewing of an original lithophane, particularly a large one that is skillfully done, is exciting. Not only is the tonal range remarkable, but also the idea that you are looking through clay is strange and magical. For those of you who have scrutinized the ridges on your antique records, wondering how the sound gets out, the lithophane offers yet another opportunity for puzzlement. As you hold a lithophane, depending on how you position it, you can transition from viewing an odd jumble of surface relief to experiencing a spectacular range of photographically articulated forms.
- ALL OF THE ABOVE METHODS; To the three groups of historical methods mentioned above, I would like to add Woodbury-process. In 1994 I began to use digital technology to carve photographic relief surfaces. To describe my work, I use the term Woodbury-process so as to credit Walter Woodbury with the concept, and to also distinguish it from his original patented chemical process, the Woodburytype. Producing the carved surfaces digitally offers many advantages over the chemical method originally used. The main one is that the amount of surface relief produced can be exactly controlled and exactly matched to the needs of a particular colorant. All three groups of historical methods mentioned above can be successfully explored with the Woodbury-process method.
This chart summarizes the categories:
| Prints: | Woodburytype up to 4 x 5 inches, pigmented gelatin only.
Woodbury-process up to 16 x 20 inches, a wide range of translucent materials can be used for imaging on a variety of substrates. |
| Relief: | Woodburytype up to 4 x 5 inches, the very low relief requires a particularly thin and fussy layer of colorants.
Woodbury-process up to 16 x 20 inches, wide variety of materials can be used for colorants and substrates. |
| Transmissive: | Woodburytype in porcelain not recommended, some plastics might be OK.
Woodbury-process up to 8 x 10 in porcelain, larger in plastics, epoxies etc. |
II. Making a woodbury-process print
First, a quick introduction to the concept of XYZ carving. If you hold a pen to paper and move left to right along the top edge, you are moving in the X direction. If you run your pen along the left side of the paper, top to bottom, you are moving in the Y direction. As youlower your pen to the paper, you are moving in the Z direction. With these three motions, a machine with three positioning motors can carve a detailed photographic surface, line-by-line, pixel-by-pixel. This is called raster carving.
Unfortunately, 99% of the equipment commercially available is for vector carving, because vector can do almost everything faster and better than raster, except that it can’t carve continuous tone photographs. I am still using the two raster machines I assembled from parts back in 1994, and if you are considering building your own, I will end this paper with a technical section of machine specifications. Somewhere in a lab at your local university there is a group of students building fighting robots. Show them the specifications list; it would be an easy project for them.
Recently, sign-carving machines with raster ability have come on the market. While you do give up some control by outsourcing the carving to one of these shops, it does avoid the large initial investment in equipment. One shop I know of is located in Lubbock, Texas, USA. The owner is Grady Robertson, phone 877 863-3935, or woodgraphix.com. He can do photographic carving of molds and the cost is based on machine-time. His estimate for an 8 x 10 image is between $50.00 and $150.00, depending on the amount of detail, which of course influences the carving time. My recommendation is to first try to find such a shop within driving distance of your home because communication is easier with materials in hand. If not, Mr. Robertson provides a very good alternative. While I have not used his services, he seemed very knowledgeable, patient, and helpful during my phone conversation with him.
Steps to make a woodbury-process print
1Create a photograph in any manner. In Photoshop convert it to grayscale and size it. I have found that 200 dpi is sufficient for most images and is a reasonable compromise between speed of carving and detail produced. Generally, I stretch the shadow areas of the image and compress the highlights a bit. I always include a white border of 1/4 inch in the image file to help with printing later. Save it in TIFF format, and be sure to write down the pixel dimensions somewhere.
If you are outsourcing the carving, skip directly to (STEP 5) below.
2Bring the image to a second computer that is connected to the carving device. The good news is that a cheap old computer will work just fine here. Set the image parameters in the software. You will specify things such as XYZ starting position, size, dept of carving, speed of motors, lines per inch (dpi), etc.
3Clamp the material to be carved to the table in the machine. I use a urethane sign foam made by “foamular” in an 18-pound density, 3/4 of an inch thick. There are many other choices out there.
4When everything is set, I start running the machine with the Z motor intentionally set too high to touch the surface. I learned this the hard way. Because math is involved, I destroyed a few parts by miscalculation. After I have verified that all things are moving as they should, I start over with the Z depth positioned correctly, so as to actually cut. Depending on things like resolution and the amount of detail present, my low-tech home-made set-up carves at about one square inch per hour. An 8 x 10 image will take 80 hours. In the software, you can elect to turn off all motors when the file is done carving, so there is no need to be present during this time.
5When this original urethane mold is finished being carved, some choices must be made. Here is a partial list of how the workflow can go:
Note: When referring to molds, I use the term “positive” to indicate a mold in which the dark areas are cut deeper and the highlights protrude. “Negative” would be the opposite.
A. TO MAKE A PRINT WITH DARK COLORS ON LIGHT PAPER:
Original urethane negative > second mold plaster positive > print.
Original urethane negative > second mold rubber positive > print.
Original urethane positive> second mold rubber negative> third mold plaster positive> print.
B. TO MAKE A PRINT WITH LIGHT COLORS ON DARK SUBSTRATE:
Original urethane positive > second mold plaster negative > print.
Original urethane positive > second mold rubber negative > print.
Original urethane negative > second mold rubber positive > third mold plaster negative > print.
Notes on choices:
Pigmented gelatin can be cast from plaster or rubber molds. Plaster is easier and faster for small work, but as the work gets larger, rubber is easier to remove without tearing the gelatin.
6Prepare the paper. If you desire bright, clean highlights, then the paper must be sized to reduce the absorption of colorants into the paper. A few coats of clear gelatin can accomplish this. If you’re not a purist, white spray primer can be used. Also, some art papers and some inkjet papers have enough sizing to work as-is.
7Prepare the pigmented gelatin. I use regular unflavored gelatin from the grocery store, follow the directions on the box except make it four times as concentrated. Water-soluble dyes are a joy to use, but can fade over time. If you use pigments, grind the pigments in a small amount of hot water and strain. Getting the right amount of pigment into your gelatin mixture is critical and will require lots of testing and good record keeping. I average about 10 prints before I get it right. When you change colors, you must do the testing all over again. My opinion is that “complex” colors – made up of several pigments- are more visually interesting than single-pigment colors.
8Prepare the mold. Silicone rubber does not require a release agent. Other rubbers require a spray application of silicone. Plaster releases well with multiple applications of a cooking spray called “Pam”. Keep spraying until the surface develops a low gloss. In both cases, polish the release agent with a soft bristle brush to prevent any accumulation in the low areas of the mold. If you work in a cold studio, warm the mold to around 75-85 degrees F.
9Cast the image. Pour the warm pigmented gelatin into the mold until there is an excess of material in the mold. To reduce trapped air bubbles in the print, bend the paper slightly so that the center of the paper touches the center of the mold first, then carefully flatten toward the edges. Place a cold, rigid, smooth piece of material (like glass) on top of the paper with some weight on it. Gelatin will ooze out of the edges of the paper, so layers of newspaper under the entire assembly are recommended. In winter, I place this assembly to a cold room so that the gelatin can “set”. In summer, I use a refrigerator. Do not allow it to freeze. Rubber is a pretty good insulator, so the mold may need to chill for 6 to 12 hours.
10Remove the Woodbury print. Start at the edges to release as much of the paper as you can, then slowly move inward. The gelatin will tear if: (A.) You work too fast. (B.) The mold was not sufficiently coated with release agent. (C.) The gelatin is not fully chilled. (D.) This is only your ninth try.
11Upon removal, notice the high amount of surface relief present. When the gelatin looses all of its water to evaporation it will become almost flat. However, if you were to cast in another material, such as pigments in epoxy, this relief would remain intact.
If the print displays uneven tones, it is likely that the paper was not rigid enough and it collapsed slightly into the mold. You can try a heavier paper, or laminate the paper to a rigid substrate such as “masonite”. With rigid substrates, it is more likely that you will trap air bubbles.
12Dry the gelatin on the print. Heat will cause the gelatin to slump. Keep it cool with a gentle change of air. If it is dried too fast some curling and distortion can occur. If it is dried too slow, way too slow, you might get some mold. On large works, you might pin or tape the edges of the paper to minimize curling.
13In (STEP 1.) I stated that a 1/4 inch of white space was included in the image to be carved. You can now see that just past this clean 1/4 inch area is a thick mass of gelatin that’s been squeezed out. I trim this off with scissors when the print is dry, thus leaving a nice border around the image that can aid with matting the print later. That’s it, you’re done!
III. Steps to make a carved photograph in plaster or epoxy
1-5: Same as Woodbury-process PRINT.
6Make the original urethane mold positive (dark areas of photograph are carved deeper, highlights of photograph protrude in the mold.) Cast a second mold in rubber (negative), using silicone release agent.
7If casting plaster, no release is necessary in most rubbers, and since plaster is white, no colorants need be added. If casting epoxy, use a silicone release agent on the rubber, and load up the epoxy with enough colorants to make it a light value (but not necessarily white). I have experimented with applying gold leaf and silver leaf to the carved surface prior to adding the pigment layer, and that can be interesting.
8Removing a rubber mold is easy, release the edges first, then take one side and peel off. At this point, your relief surface is done, but the image doesn’t look photographic because the topography is “illusionary”. If you had a picture of a person smiling, their white teeth are protruding beyond their lips. You need to apply colorant to the surface to force the highs and lows of the relief to read as an image. If you used plaster without a release, go to (STEP 9). If you used epoxy and a silicone release, you must scrub the image surface repeatedly with hot detergent to remove the silicone residue.
9With both the plaster and epoxy substrates, you must make sure the panel is very level, or the colorant will run off to one side. Next, pour pigmented epoxy or pigmented liquid urethane directly on the surface for an effect much like a glazed porcelain tile. If you mix pigments in varnish instead, or if you thin your epoxy with lacquer, some amount of solvent will evaporate as the finish dries. This will have the effect of shrinking the thickness of the coating, thereby revealing more of the surface relief while at the same time producing a photographic presence.
IV. To make ceramic relief panels and lithophanes
1-5: Same as Woodbury-process PRINT.
6When this original urethane mold is finished being carved, some choices must be made. Here is a partial list of how the workflow can go:
A. TO MAKE A CARVED CERAMIC SURFACE (Emaux Ombrant)
Original urethane positive > second mold plaster negative > press clay positive.
D. TO MAKE A CERAMIC LITHOPHANE
Original urethane negative > second mold plaster positive > press clay negative.
Original urethane positive > second mold rubber negative > third mold plaster positive > press clay negative.
Note: Clay should always be cast or pressed into plaster molds because they are very hard and absorb moisture. Select the appropriate plaster. Plasters for slip-casting are porous and somewhat weak, while plasters for press-molding are much harder and less absorbent.
7For ceramic relief tiles, press or slip-cast white clay into the plaster mold. Dry and bisque fire. For ceramic lithophane, press-mold very translucent white porcelain into the mold, then use a taught wire and metal straight edge to thin the panel down. Dry and fire the clay to maturity. The lithophane is finished at this point, although historically some were augmented by hand-coloring in transparent oils on both the front and the reverse sides.
8For ceramic relief tiles, prepare and apply a translucent glaze to the bisqued clay panels. Much testing is required to formulate glazes that will flow out like water, tolerate thick applications, and possess the correct amount of transparency. Historically, leaded glazes accomplished this best. If you formulate a great glaze, but cannot get it to achieve a sufficient thickness, you can re-carve the original mold with less relief. In other words, working digitally offers the freedom to “work backwards” to accommodate the characteristics of a particular glaze that you like.
NOTE: This brief instruction set has not covered the details of working with plaster molds, rubber molds, mold-making, or casting. If you have no experience in these areas, please read the appropriate sections in any “sculpture” text.
V. Technical specifications and details
1. The carving set-up if you are making your own
X and Y axis travel will determine the size you can carve, such as 8 x 10 inches. 6-inch positioning stages are common and not expensive. The stepper motor and lead screw combination should provide at least 800 steps per inch.
Z axis determines depth of cut and adjustment range. 2 inches is plenty, more is OK. This stepper motor and lead screw combination should provide at least 4,000 steps per inch, and 16,000 is better. Arrick Robotics provides a 1:4 drive converter to enable this second higher number.
I use and recommend stepper motors and drivers from Arrick Robotics. He includes software for the drivers and good instructions.
A computer must be dedicated to the carving set-up. It must be able to run DOS and have two parallel ports out. (This is an easy modification on most machines).
Software: DOS 4 is the first layer. Software provided by Arrick Robotics is the second layer. The third layer that I use is a program written in BASIC that controls the interface with the user. If you get this far, and want a copy of my program, contact me.
Carving spindle: I started out buying the deluxe Dremel tool with ball bearings and a five-year guarantee. I would return one of these every month or so because the long run times would wear it out quickly. I now use the indestructible Precise Super 30 motor only available on e-bay (about $150) because they were made in the 50’s and 60’s. I have one that has gone through about 10 pairs of brushes (still available from Precise Corp.!).
Carving burr: I buy mine from a jewelry supply company. I use either the pointed burrs or the pointed diamonds. They last for 5 to 10 carved images if you are working in sign urethane.
Material to be carved: most sign shops now use a dense urethane board such as “Foamular”. A 15 pound density is OK, 18 pound is a little tighter. I have carved wood, Corian, and wax, but the urethane is the most versatile. I use it in the 3/4 inch thickness.
2. Details about outsourcing the carving
If you intend to make Lithophanes, the depth of relief after firing should be around 1/10 of an inch, depending on the characteristics of your particular porcelain. If you assume shrinkage around 12%, that would put the carved depth of your original mold at around 1/8 of an inch.
If you intend to print from molds, 1/16 of an inch of relief, or a bit less, works fine. Smaller work will tolerate less relief. Larger work will require more relief. Keep in mind that if you are translating 255 levels of grayscale into this short space, every level change is around 1/5,000th of an inch.
If you intend to make glazed ceramic surfaces, 1/16 of an inch of relief is a place to start testing. Your glaze behavior will determine if you need more or less.
If you intend to make photographs with visibly carved surfaces, or apply colorants directly onto the work, then the range of relief will be from about 1/16 of an inch up to 1/4 of an inch. Again, larger work will require more relief than smaller work. As you exceed this 1/4 inch threshold, the illusionary nature of the relief starts to overcome the power of the colorants to render the photograph understandable.
I have two Precise Super 30’s for sale if you need more Nicolai or if anyone is interested in duplicating this project!! dave@dshillingphoto.com
In response to Shawn’s questions: the chemical method (dichromated gelatin) produces a very low amount of surface relief. Even if epoxy rather than lead is used for the mold, the printing process is extremely fussy (difficult) because of it. This is one of the reasons why Woodburytype prints are found only in very small sizes. Walter Ford modified this chemical process in an effort to produce more relief for the purpose of working with images in ceramic materials, but those images lacked crispness. With xyz carving you can accurately control the depth of relief, and tailor the characteristics of the relief to your specific applications. This allows you to go as large as you want, along with providing the widest choice of materials.
Why is it that you suggest XYZ carvings? Why not just make a relief in dichromated gelatin, and, if necessary make a casting using epoxy? I’d think you could even bake the gelatin relief after washing out – or even just use as is.
Am I missing something?