How I make cyanotypes
(17 November 2014)
This article outlines my search for a passable cyanotype. I will not go into the history of the cyanotype process (it's been around since 1842) or the chemistry behind it. There are plenty of places that have documented this much better than I could.
Instead, I will focus on the steps I took to arrive at a process that repeatably makes cyanotypes with acceptable tonality from any digital file, whether of purely digital origin or from digitized film. I will describe the materials I use and the general methodology.
The overall process is as follows:
- Coat the paper
- Prepare a digital negative
- Expose the paper using the negative
- Develop and dry the print
1. Coat the paper
I use a standard cyanotype mix consisting of 1 part (weight) potassium ferricyanide and 2.5 parts (weight) ferric citrate. The potassium ferricyanide comes as red granules and I paid about € 8 for 50g. The ferric citrate comes as a light green powder and has a lower weight/volume ratio. The volume ratio of both chemicals (in solid form, non-dissolved) works out as roughly 3.5(ferric citrate):1(potassium ferricyanide). I worked this out by roughly measuring the water displacement of 50g of each of the chemicals. In practice, this turns out as close enough to produce usable results.
I dissolved both chemicals in water with a volume ratio of ca. 1:10 for the potassium ferricyanide and ca. 3.5:10 for the ferric citrate. The sensitizer can then be mixed from equal parts (volume) of both solutions. I store the solutions separately and only mix when I coat my paper, and I only mix what I need for one coating session. 4 teaspoons of the sensitizer solution (so 2 tsp from either bottle) allows me to coat ca. 13 sheets of A5 paper, or 6 8x10" sheets.
The paper I use at the moment is Schut sketch paper, 160gm/m2, with a rather coarse texture (although it says 'medium fine' on the blocks of 70 sheets that I buy). It costs about €0.10 per sheet in the arts supply store. I use a no-brand el-cheapo paint brush intended for cheapskates who intend to repaint an old, second-hand table to place in their shed. It's perfectly adequate for my purposes as I like the 'artisan' look of the rough edges and artifacts it leaves. If you are a neat little boy or girl, use a foam brush or any kind of decent brush.
Coating the paper is done in the kitchen (north-facing window, so little UV radiation) on a piece of cardboard, under a couple of fluorescent lights. I notice no problems with fogging. After coating, I put the papers to dry for a few hours, usually overnight, in a dark room. I then store the coated papers in a light-tight cardboard box. I usually apply a single coat, which yields a dark enough shade of blue for my taste. Double coating intensifies the blue a bit further, but I find it unnecessary for most prints I make.
2. Prepare a digital negative
As an illustration, we'll start with this image, shot on an early October Saturday evening in Metz, France, with a Canon 7D and an EF-S 15-85 lens on ISO 2000.
The digital negative is what I spent most of the time on figuring out how to make a passable print. I tried printing regular negatives, with varying results between the three films I tried:
- Fomapan 100 printed quite nicely, although high-contrast scenes would result in no shadow detail or no highlight detail. Exposure times in the range of 4-10 minutes under a tanning lamp (see below under 3.) depending on density.
- T-Max 100 virtually didn´t print at all due to their UV blocking. Print times > 40 minutes.
- Ilford HP5+ printed like Fomapan: fairly fast, but with insufficient contrast control.
So I decided to go for digital negatives, which I had anticipated from the start.
I use an Epson 3880 printer with a Cone Color ink set, although I've used the Epson yellow ink (and the Cone Color black) so far. I expect no big differences once I switch to Cone's yellow ink. I use cheapish Avery transparency sheets, which I thought at first were crap (but this was due to the matte ink at higher densities; see below), but that turn out to be quite usable.
I started by following the Digital Negative Tutorial on Freestyle and I used the test pattern that can be downloaded through that page. I found and chose the following:
- Basic exposure time for my paper/coat/lamp combination is 3m30s
- I use photo black, as matte black tends to wrinkle on the transparencies much sooner than photo black. The wrinkling results in an interesting cracked pattern that renders the negative completely unusable.
- I use full yellow as explained in the Freestyle tutorial
- The profile I use is 2.2 gamma as outlined in the tutorial above.
- Printing lots of transparencies of this kind frequently results in clogged nozzles, so I regularly perform an automatic nozzle check to clean them out. I also try to print a regular photo after printing a transparency to remove the ink that collects at the head.
- I got the best highlight reproduction when setting the color density (CD) to -18%. This brings pure white just within range of the process, but still allows for well-discernable differences between 0%, 5% and 10% black.
- For a good reproduction of the shadows, I found that a pure white negative was just too much; it turned out that I needed to offset the white point of the negative (so the black point of the print) considerably in order to retain shadow detail. I could have solved this by just making for a more aggressive curve, but I chose a solid black overlay layer in Photoshop instead. I use a black fill layer with the fill set to 32%.
- I found that a really steep, exponential curve made for the best (to my eye, most linear) reproduction of grey (blue) tones. This is the curve I currently use:
When making a digital negative, I perform the following steps:
- Flip the canvas horizontally (this is a contact print, so mirror image!)
- Convert image to grey scale (desaturate, convert to grey scale, or using the channel mixer)
- Convert the image to the 2.2 gamma profile
- Perform regular adjustments with curves or levels, possibly using adjustment layers - to taste
- Overlay an invert layer to make a negative
- Overlay the curve above using an adjustment layer
- Overlay a black fill layer with 32% fill
The digital negative file looks like this:
Then print, using the settings outlined above: CD=-18%, photo black, full yellow toning. Quality level set to 'quality' (not 'speed') and all the rest as indicated in the Freestyle tutorial. After printing it, I dry the negative by putting it directly under my UV light source for 5 minutes. This prevents the ink from sticking to the print in the next step.
The printed negative, with the settings outlined above, looks like this:
3. Expose the paper using the negative
I experimented with exposing using the sun, but in the late autumn in The Netherlands, I found it virtually impossible to get any kind of acceptable exposure, let alone anything reproducible. So I picked up a second-hand Philips face tanner that uses 4 15W UV fluorescents (so total 60W). I used this to construct the curve above and to determine my basic exposure time for full development, which is 3m30s.
I place the negative ink-side down on the coated part of the paper, position to taste, on a wooden board. This is covered with a sheet of glass from a cheap photo frame and clamped with 6-8 cheap and fairly big clamps. The tanner goes directly on top, so the light source hovers at maybe 8cm above the glass. I usually turn it on before placing it on top of the print, and remove it as soon as the exposure time has passed. The tanner I use has a timer, but it isn't precise enough for my standard exposure time.
When the exposed print comes out from under the glass, the fully exposed shadows/blacks are brownish grey, the shadows are blue-green, the mid-tones are a paler blue-greenish and the highlights and whites are greenish yellow.
The example picture looked like this when it was just exposed, but not yet developed (excuse me for the bad lighting in the following shots):
4. Develop and dry the print
I develop my prints in a shallow dish with about 700ml of water and a dash of cleaning vinegar (about a shot glass; I never measure it). I leave the print in to soak, agitating now and again, for 3-6 minutes. I note very little effect of variations in the 'development' time; a shorter stay in the vinegar bath just leaves more yellow unexposed chemicals on the paper, which rinse off anyway. I then rinse under the tap and place the print in a dish filled with tap water for another 5 minutes.
Here's what the image looks like as it has just hit the water for a couple of seconds:
The fully areas in the margin and the deepest shadows are still grey as the water hasn't fully soaked the paper.
A minute or two into development, it looks like this:
The highlights haven't completely cleared and the mid-tones are still green as well, but the paper is now almost completely soaked.
Towards the end of the acid bath, it starts to get its final look:
At this point, the image was soaked in tap water for a few more minutes, resulting in this:
Then I diverge from common wisdom, as I'm usually impatient, and I noticed that it's quite easy to cut a corner here. You could leave the prints to dry on a clothesline, but it will take hours depending on the paper used. Me, being impatient, I carefully dab the print dry using a clean towel and then carefully iron it, by pressing the iron on its highest setting on the backside of the print for a few times, then ironing the front and the back side, alternating between both sides. This can cause the paper to mottle, but it can be pressed into form later on. Usually, it comes our fairly straight though. The only drawback I have found so far is that this method leaves a subtle gloss on the higher areas of the printed side. If you don't like this, then dry the regular way and wait it out. Incidentally, ironing the print brings out most of the final color right away instead of after a few hours.
This is the example image right after ironing:
The blues are still a bit brown-grey, but it intensifies over the following 24 hours as the paper soaks up some moisture from the air and the Prussian blue oxidizes further.
Here's a scan of the final image that we have just made (scanned after 24 hours):
As you can see, the blue color is now fully saturated and contrast is fully developed.
Done! Now scan, frame, or discard according to your taste.