There are a lot of things still wrong with this leaf. The veins don’t match the creases in the 3D leaf, the veins need to taper off from thick to thin, the green leaf color is off, BUT this is the first time I have been able to get a transparent leaf with two sides with different colors and with the graphic detail showing through enough. What I think I have found is the method, now I just have to get the photo to match the mold and get the right color.
The key to getting this kind of two-toned cast is to use a thicker fabric such as Cotton Sheeting from Jacquard. Print the front side on the Cotton Sheeting, which because of the thicker material, keeps the ink from going all the way through. This inserts a white barrier between the front print and the back, so what is printed on the front doesn’t overpower the back. The back side was printed on Polyester Tafetta, but it may work to print both sides on the Cotton Sheeting.
I need to learn Photoshop so I can make the prints how I want them and not keep pestering Sally in Graphics
This was a good day.
I think I have finally stumbled upon the way to get what I want with making accurate, translucent leaves. Earlier, I wrote about matching photographs of leaves with plaster molds made from the same leaf and then using inkjet prints of leaves on a scrim fabric with the associated mold to get a good replica (blog entry: July 2013). This was close to perfect except, I went down the wrong track by trying to make vacuform casts from the molds and adhere the fabric to the vacuform leaves. This week, close to the June 15th date of the Shoreline diorama, I went out and collected more red osier from a nearby lake. I photographed the leaves and then made plaster molds, this time, of both the front and back.
Plaster molds (front and back) of red osier leaves.
From the photographs, Sally made prints off the inkjet printer on a fabric from Jacquard called Habotai 10mm. This fabric is close to a scrim material and therefore the colored print shows through front and back. By the way, I spent a day or so getting accurate color from the actual leaves-both front and back. Sally then printed several leaves with different saturations of different colors. We ultimately found a green that is almost an exact copy of the actual leaves (in daylight).
Print of leaf on Habotai fabric by Jacquard
Five minute epoxy is mixed with a bit of green oil pigment made from ultramarine blue and cadmium yellow light. The epoxy is spread over one side of the mold (light vaseline used on leaf surface, heavier on the periphery) and the fabric tamped down into it making sure the middle vein lines up with the mold’s vein. The second half of the mold (also with vaseline) is given a coating of epoxy and carefully laid over the first half of the mold. A spring clamp compresses the two halves. A good cast is insured when epoxy oozes from the seams. In five minutes, the epoxy is set and the cast can be lifted out.
Cast leaf in half of the mold.
The cast can be fully removed from the mold immediately or it can rest in the mold for an hour until the epoxy hardens more. The flashing is then cut away with scissors and the result is extremely lifelike! There is no need for any extrinsic painting or waxing. The shininess is controlled by dusting with fumed silica.
I will have two interns working for me over the summer and we are going to go into production. I have one outlying concern. Thin epoxy over time does tend to droop and yellow. The fabric may keep the drooping from happening, but I will be making tests of this and watch for color changes. A tapered wire might need to be cast in along the midline for support.
The new sprig has larger leaves and more insect damage and is right in front.
I installed the red osier sprig into the Shoreline diorama. I was happy that the leaves look less artificial than what is in there now, but the translucency was not that apparent and the graphic quality of the leaves is not easily seen from outside the window. So, I have a new set of problems. I am going to have to pump up the graphics a bit more-especially the veins of the leaves so they will be visible. I also think the wax (or something else?) has to go on thinner. The detail of the vacuform leaf is lessened by the wax.
I think for now, I have decided that the first test with printed fabric may have been given the best results so far. Now, I have to find a suitable substitute for the “poly silk” material. I have purchased a test roll of a fabric called Habotai (pronounced habit-eye) from Jacquard inkjet fabric systems in CA. Sally is running a test with this material in our Epson inkjet printer. If it still isn’t right there are seven more with names like Cotton Lawn,Drepe de Chine, Poly Taffeta,Charmeuse, Fuji Broadcoth, etc…
This project has thrown me a lot of curve balls. Sometimes work on a problem is like that, but if I can work out the bugs, it will save an enormous amount of time for leaf production in the future. I remain hopeful!
More problems with printing on acetate or in this case on a matte medium skin. I painted 3-4 layers of matte medium on an 8 1/2 X 11 sheet of polypropylene. InkAid was the final surface over the matte medium. This gets sent through the printer (set the feed to the highest setting so it won’t get stuck-mine did). After drying, the matte medium surface is peeled off the polypropylene and adhered to the leaf with more matte medium. The InkAid surface has to be white to get adequate saturation of color. That leaves a white surface on the back. So, the problem with this experiment is that either the back has to be painted, or both the front and back need a print attached to the vacuform leaf. The fabric prints make this unnecessary because the ink sinks into the fabric and colors both sides. This leads me to think the first idea of printing on fabric is the best.
Matte medium print-note the skid marks from getting caught in the printer. Still, this print showed me that the distortion problem can be handled with a low-tech solution in photoshop.
Back side of matte medium print. The white is unacceptable.
Sally Pallatto, Peabody’s graphic artist, did some low-tech morphing of the image in photoshop using a distortion tool. She pulled one side one way and the other side the other way. The result was a print that sucked down over the leaf with the printed veins matching the vacuform veins. It is not scientific, but it worked. Ken Lovell told me he was going to do some R&D with photoshop this summer at the DMCA to try to figure out a more reliable way to distort the image.
Next step is to try fabric again and figure out a better way to get the vacuform to work better.
One thing that goes with this territory of building and maintaining dioramas is that sometimes things don’t go as smoothly as you’d like. I am now running into this with the leaf making. Here is a list of the problems as they stand right now:
1. Ken had difficulty creating a matrix on the scanned mold. I have to see if he can come up with another way to do this. If there is not an easy way, my prints won’t fit well into the molds.
2. I experimented with the opaque white Inkaid substrate and found that the 50-50 and 75-25 mixes with transparent substrate were too thin to get good saturated color. I airbrushed full strength white substrate and even that didn’t give me the saturation I got in the fabric. It might be that fabric absorbs more color into the weave and therefore is innately more saturated than acetate. I will spray out a double coat of the white and see what I get. The single layer of white still gives me adequate translucency.
fabric on top, acetate on bottom. The photo doesn’t show the lesser saturation of the acetate as it actually is.
3. There is not as much strength in the vacuforming as I had wanted and this is because of problems I am having with the 10 mil acetate and the need to do one leaf at a time. If I use a heat gun on the printed acetate, the ink discolors or burns easily. You have to be very careful for this not to happen. I found that if I cut the leaf to the contours dictated by the print, lay it down and register it on the mold, cover it with a larger piece of acetate and heat all of it, I can get good suction and a leaf with good contours, but not with the veins and subtle detail. The lack of full saturation of color is not going to derail this, but if I can’t figure out a way to get better detail, I’m not sure I will find this method as workable as I’d hoped. I guess I could let the print give a 2D representation of the detail, but I want the form of the leaf to show it too.
Vacuforming method for single leaves.
4. The non-print side needs attention because it is white.
So, there are a lot of things to consider before this is really ready to go into production. Stay tuned!
Leaf mold to be scanned
I dropped off a leaf mold today at Ken Lovell’s lab (Yale Digital Media Center for the Arts). He is going to make a 3D scan of the surface of the mold, create a matrix that will be laid over the prints of the leaves. From this, he will morph the prints so they should vacuform right down onto the mold perfectly. As Ken said, in theory! I haven’t had a chance yet to use a milky subsurface on the acetate that will run through the inkjet printer.
Pabric with printed leaves
One other note, I purchased some “Pabric” fabric that can be run through the inkjet printer. I made a test and wasn’t happy with the saturation of the color in the fabric (not saturated enough) or the ability to glue it to the vacuform leaf (the material is too heavy). The fabric printing still is a strong plan B. Ken mentioned today that he can get a printable polysilk (cheaply) that can also be run through the printer. The only problem with doing it with fabric is I lose some or all of the detail of the form of the leaf I can get with the vacuforming.
I found 10 mil PETG from a company called Sabic Polymershapes. They cut the plastic to my specs so it is ready to use in my vacuform machine. Now, I have what I need to test the printing on the vacuform sheet itself. I also found the material, Vivak, that I was given at the AMNH is also 10 mil PETG.
This Friday, I went to Yale’s Digital Media Center to work with Ken Lovell. Ken demonstrated rolling out a couple of layers of a print-friendly substrate called Inkaid on the PETG, letting it dry, and then, running it through the inkjet printer. The ink from the printer adheres to the Inkaid substrate just like it was paper. Here’s a photo of the results.
Leaf prints on 11X17 PETG. Color saturation good with white background.
I inquired with Inkaid about the archival quality of the material and they assured me that it would last as long as the Epson inks, which are calculated to last for over 100 years! This is very good news for making leaves for the dioramas. I brought the prints back to my lab with the molds. I set up a vacuform box so I can test this process, one leaf at a time.
THE TEST: I was able to pull a pretty good cast, but getting the printed veins to line up with the veins in the mold was almost impossible. I had to try to move the hot acetate into place and had trouble with the plastic folding and sticking to itself. There is just so much the PETG can be moved around while in its melted state. Another problem is that printing on the clear acetate makes the inks appear less saturated. When printing on fabric, the white surface reflects the light back through to color in such a way that the colors appear about twice as saturated as on the plain acetate. This could be a problem, but Ken suggested I try a mixture of clear and white Inkaid substrates to produce a milky surface on which to print. It might be possible to get the same translucency and reflectivity of the fabric. Then, I would have to work on getting the print to line up in the mold. Ken thinks the shape of the 2D prints could be changed to fit the 3D shape of the mold by 3D scanning the molds and using some morphing software he has in the lab.
Ken also suggests I try printing on a subsurface made up of several layers of acrylic gloss medium or matte medium mixed with white acrylic to get a milky translucent “skin”. The acrylic medium is built up, layer-by-layer on a sheet of plastic, covered with another layer of Inkaid and the whole thing is run through the printer. The dried acrylic medium/Inkaid “skin” can then be peeled off the surface of the plastic and adhered to the vacuform leaf. This is esssentially another cheaper version of the fabric printing method. More work to be done!