It looks like printing grayscale negatives using an inkjet printer’s all inks works well, when the tone range of the intended photographic printing process is quite short, so that we need relatively low contrast negative. We aim for a negative with enough density to print “paper white” with standard exposure (the shortest exposure to produce Dmax, the maximum black) for the printing process in use. Not all inkjet printers can lay down enough density to produce decent negatives for some long tonal range processes (salted paper, new cyanotype, palladium…). In that case it may be possible to make colorized negatives, trying to find an ink color which acts as a color filter, blocking UV-light more than gray-tone negatives can do. With printing the negative with this particular color we can use its UV blocking density to produce paper white in our photographic printing process. For more on the subject see the sidebar link RNP Array.

This image (The HSL Array) can be downloaded from the RNP Array website and printed with an inkjet printer on transparency. We have used this image for determining our “blocking colors”, according to the instructions by Michael Koch-Schulte at the RNP Array site.

This print from the HSL Array output, made with an Epson R1800, then contact printed on Kodak Polycontrast paper, is one of our first exercises with colorized negatives. We have chosen R0-G12-B51 for the optimum blocking color here.

Another print on the same paper, printed with the same settings as the one above. The inkjet transparency was made with an Epson Pro3800. This time R242-G20-B0 was chosen for the blocking color, because it makes a nice and smooth grayscale (and we wanted to try something that is far from neutral gray).

We can notice that two given printers can build somewhat different shades of colors from the same image file, when used without any color management. Anyway, choosing any of the acceptable blocking colors should do the job equally well.

These are our colorized step wedge files, ready to be printed on transparency. Two different printers – two different colors: The R1800 (left) and the Pro3800 (right). Although their colors are very different, they should print about the same grayscale gradation on b&w photographic paper.

The contact print from the colorized negatives with the standard exposure. We chose to continue experimenting with the 3800 negative (right), because of slightly better highlight tone separation (and because we wanted to see how the red negative will manage the process).

The print highlights look quite good, but the shadows are dark and blocked. We need a correction curve in Photoshop to fix this. Measuring the test print’s density values and calculating the shape of the correction curve by hand can be laborious, but fortunately there is a computer software to do all this: ChartThrob by Kevin Bjorke is a javascript for Photoshop, it can automatically both create a step wedge to print on transparency and evaluate the print made from the transparency, and even build the correction curve needed.

The next print with the curve applied. Looks much better now, here we are getting basically a nice full tone range.

Only printing “real world” photographs will finally tell us about the usability of the chosen blocking color and correction curve for the chosen photographic printing process. This is how a red, corrected Epson 3800 negative looks.

Test Prints
A few test prints on Kodak Polycontrast II RC, made with the Epson 3800 red negatives and correction curves built by ChartThrob (you can click pictures larger):

Our first print with colorized negatives. Looks ok, except for some posterization in the deepest shadows (which is evident in the negative too), and when compared to the image on the computer screen, we wanted the lower midtones slightly darker. So, after this print the curve was tweaked a little.

Again the print is acceptable but not perfect. A little more shadow darkening was made to the curve after this print.

We’re satisfied, the colorizing / curve correction tool is working for us! This print is very close to what we see on the computer monitor.

The starting curve (left) calculated by ChartThrob, and the adjusted curve after test printing.