Designing an Inclusive Audio Guide Part 2: Tactile Reproductions

By Leah Morelli

This is the second post in a series about the development process of The Warhol’s new audio guide. 

You might not think a museum would offer a meaningful experience for a person with visual impairment. But why not? At The Andy Warhol Museum, we’re working on changing that perception, and we’re providing experiences to visitors with the widest possible range of abilities.

We are very excited and proud to introduce three tactile reproductions of Andy Warhol’s artwork on the 7th floor of the museum. Visitors will be able to get a sense of the texture, shape, and composition of Warhol’s artwork through touch. The beauty of tactile reproductions is that they are beneficial to all visitors. We can all learn something about Warhol’s artwork by experiencing it through the sense of touch.

I first learned about tactile diagrams at the 2013 Leadership Exchange Arts and Disability (LEAD) conference in Washington, DC, and I was excited to share my ideas with the rest of the museum team. After that, we began working on a mobile app with visual descriptions for an audio guide, and a raised map of the gallery spaces so that visitors could have a sense of space before they peruse the galleries.

During this time of testing and research, I also discovered something called a “swell form machine.” The swell form machine uses a specific kind of paper and a carbon ink pen that allows you to draw on the paper and put it through the machine, which heats up the ink. The ink swells on the paper giving specific physical definitions that are raised, juxtaposed against the smooth paper. We experimented with many kinds of raised reproductions and got a positive response from visitors, as well as some feedback. For example, a consultant named Sina Bahram, when using a larger-scale tactile reproduction made from high-density polyethylene (HDPE) material, advised us to mount the reproductions on a 45-degree angle on pedestals so that they were more accessible to touch for everyone.  

The Warhol is working with David Whitewolf of Tactile Reproductions, LLC, who is making the three tactile reproductions that will soon be on display at The Warhol. I asked David a few questions about the fabrication process, because it’s fascinating.

Leah Morelli: How did you get involved with making tactile reproductions?

David Whitewolf: I first got involved with crafting tactile reproductions when I received a call from a friend of mine, Anika Wilcox, who was working with The Warhol at the time. She asked if I would meet with the education team and discuss what might be necessary to create tactile reproductions of some of Andy Warhol’s artworks, which they could take on educational road trips. Several months later we had finished four pieces, two of which were full-scale reproductions and were cut using a machining process, and two of which were smaller, scaled-down, hand-held pieces created using a high-resolution 3D printer.[1] This was the very beginning of our process, and through it we developed a basic understanding of the software workflow, the tooling and time requirements, and the overall philosophical reasoning driving our choices.

LM: From the beginning, explain the materials that you started experimenting with and how you have come to the current material that you are using to create the tactile reproductions.

DW: In the beginning we were looking at materials that might cut easily, and therefore quickly using a computer numerically controlled (CNC) router, and yet be robust enough to be handled by many people over several years. This was a crazy prototyping period, which had two big constraints—the budget and access to the machinery required to make these objects. Those two things go hand-in-hand because of the time required to create a single piece. CNC machine time can be expensive, especially when a piece might take up to 70 hours of continuous operation to create, so we had to make new friends who were interested in our project and who donated time on their machines to support it.

Our first pieces were made using two common light-weight materials: medium density fiberboard (mdf) and rigid foam insulation. Both of these materials are fairly lightweight and easy to machine quickly, but, as I soon discovered, they require extensive finishing to make them robust enough for handling. For the mdf there was a lot of careful sanding involved and many, many coats of paint and clear enamel in order to create a pleasant tactile experience. With the insulation I painted a layer of acrylic latex paint over the surface and then attempted a two-part epoxy coating, which did not leave a very satisfying result. In fact, I failed with that coating the first time, which meant I had to cut the entire piece again!

A far simpler process was the use of a selective laser sintering machine (SLS) to 3D-print two of the pieces. The original works are larger (approx. 1 meter square) so they had to be reduced to fit in the machine, which only has a 12” x 12” x 12” build volume. The pieces came out of the oven complete and only required a coat of enamel in order to keep them clean. As we’ve discovered, this is fine for smaller works, or expanded sections/details of works, but the loss in definition when reducing a large work makes it almost impossible for a patron to distinguish important areas. Also, the enamel coating over the sintered nylon isn’t as smooth and pleasant an experience to touch as you might imagine.

After these experiences we decided to move to a more robust material that didn’t require as much finishing, would be easy to clean, and would be easy to machine. It would also need to be cost effective. I made the decision to use HDPE, not only because it met these requirements, but also because it was so pleasant to touch. By this time, we had acquired our own CNC router. The first piece we cut was fairly small, but cut full size, and came off of the machine nearly complete. Only minor deburring, or getting rid of unwanted pieces, around the edges was required. The next piece was larger, however, and we noticed a significant deformation during the cutting process.  After a slight modification to the data and checking the material for defects, a second iteration was attempted. This also failed with the piece cupping and flexing as it was being cut, and cupping severely after being removed from the machine. The learning curve got a little steeper at this point as I learned about the molecular and structural composition of the material. It turns out that HDPE was actually not the best material for this process!

We are now using a copolymer with the trade name Acetal. This is a dense material, made for machining, which remains dimensionally accurate even under fairly extreme circumstances. However, because of its molecular structure and density, the Acetal feels rougher than HDPE, which is often described as greasy. We’ve learned that the same data cut on both materials gives patrons completely different experiences. In order to make sure that we are communicating the intended information of the piece, and not creating an unpleasant experience for the patron, we are modifying our datasets in software. This work is ongoing.

LM: Would you describe your process and any considerations you have when transforming a two dimensional image into a relief?

DW: The primary purpose of a tactile reproduction is to make the visual information created by the artist accessible to a patron who cannot experience it visually. Because of this, we’ve taken the approach of providing as much of that information as possible to the patron with limited modification. In other words, it is important to us that our tactile reproductions do not add information to the works. Rather, we make every attempt to provide an accurate experience by simply differentiating one feature from another. After all, no one knows what the back of Mona Lisa’s head looks like, so we don’t make any claims like that in our reproductions.

The process itself is quite time-intensive due to the need to do accurate work at a very high resolution and also because of the craftsmanship required. I have to look at each piece both as an engineer and artist in order to make choices about how best to bring out the information for the viewer. Each piece is completely different in this regard, but the tools I use are common. I utilize a typical image editor, which allows fine resolution manipulation of pixel color and provides standard filters for blurring (this allows gradations between areas so that we are not creating unpleasant transitions). Once I have created the grayscale image, I use another software program to build a 3D object from the image data. In this program, black areas are “farthest away” from the viewer and white areas are closest, with varying shades of gray falling somewhere between.  When this process is complete, the dataset is brought into another software package in order to generate toolpaths for the CNC router. Finally, this g-code is loaded onto the router’s computer and the part is cut. This process, not including the machining time, can take up to 80 hours.

LM: What are some things that you learned from meeting our consultant, and working with users, and how did that inform your process?

DW: Meeting with Sina Bahram (of Prime Access Consulting, Inc.) was a lot of fun! He is a very intelligent and pleasant person and has been studying and working in this field for quite a while, so his notes on the pieces we created were concise and direct and very helpful.  

Because Sina’s primary interest is in universal design, he spoke with us about all aspects of presentation, from the works themselves and their audio descriptions to how the works are mounted and the placement of signage. One very interesting thing I learned as we went over the pieces was that line drawings require more verbal description than other types of work. This is a little counterintuitive for people who are able to experience these pieces visually. We might tend to think that the work is less busy or more precise due to the inked line, but, in fact, the opposite appears to be true for those who are sight-impaired. The precision of the line means there is less information defining the various areas of the drawing. Whereas I may clearly see the outline of an arm leading to a shoulder and neck, a blind person wouldn’t be able to differentiate those outlined areas as objects, especially if, say, the arm was draped across the chest in the figure. This is where a more detailed audio description is required.

Working with Sina also helped to confirm many of the choices I had intuited when building the pieces and let us know that we were on the right path. This may not sound like much, but it really helps clear your head so you can get to work!

LM: Is there anything else you want to add?

DW: I would like to add that working with The Warhol to create these pieces has been a fun, engaging, and professional experience. It is very satisfying to work with people who really do want to reach each and every member of our community. We are happy and thankful to be a part of this process and to help to share it with others who might do the same in their own communities. And, to anyone who is interested, come to Pittsburgh and touch the art!

LM: I am excited that The Warhol will have three tactile reproductions on the 7th floor of the museum in time for the opening reception for the LEAD conference, which is in Pittsburgh this year, and opens the first week of August. [2] Come check it out!

Accessibility initiatives at The Warhol are generously supported by Allegheny Regional Asset District, The Edith L. Trees Charitable Trust, and the FISA Foundation in honor of Dr. Mary Margaret Kimmel.

[1] The first four tactile reproductions we made were based on the following Warhol artworks: Campbell’s Soup I: Cream of Mushroom, 1968; Female Fashion Figure, 1950s; Marilyn Monroe (Marilyn), 1967; and Self-Portrait, 1986.

[2] The three tactile reproductions that are on view beginning August 2016 include representations of the following Warhol artworks: Coca-Cola [2], 1961; Reclining Male Torso, 1950s; and Female Fashion Figure, 1950s.