Bichromate Gum Printing


Has anyone used PiezoPN for Gum Printing yet? I can’t seem to find any information regarding using it in this manner. Creating curves, etc…


We did do some gum printing at Maine Media Workshops. We created some starting GUM curves but I suggest starting w/ the Cyanotype curve and linearizing from there. Gum printing will still need some photoshop curves applied to the individual RGB layers.





How do you determine minimum exposure time for each layer?




I use a Stouffer 31 step transmission wedge to the exposure time that will give me the deepest value. In the past, I just assume that the different colors will have very close exposure times, so only test one color.


Thank you James. Which color do you test?


Whatever is handy? Usually my magenta.


Am not a gum printer but may be helping someone set up for gum in the near future at Latitude (PiezoPro inkset).

Am curious about Walker’s post #26621 about PiezoDN gum needing Photoshop curves. Are these needed because PiezoDN will not be able to linearize the CMY colors in a gum print or for another reason altogether? If so, would one linearize the gum print for K as a base, and then apply Photoshop Curves following older digital negative calibration workflows (scanning a printed gray step tablet and using Photoshop color sampler, etc.)?

I have a basic understanding of how gum works, and a lot of literature on it. I have worked a lot with PiezoDN especially for silver, but thought PiezoDN was designed so as not to have to resort to the limited data points and larger margin for error that Photoshop curves can involve.




Where would I find the Cyanotype Curve?


It’s in the Curves folder for your printer unless you have a 1430. Even so, you’ll be better off making your own since cyanotype produces wildly varying results with different formulas, papers, wash times, etc, not to mention toning which is a whole other can-o-worms.

Gum and cyanotype (classic Herschel cyanotype) are short scale processes, like silver, but with many more variables. The best practice for both is to make your own curves using the materials and procedures you will be using. If you’re not sure about those things yet, I strongly suggest doing a lot of testing using a 4x5 Stouffer 21-step to get those ducks lined up before moving on to making your digital negative curves. It may seem like a lot of extra work, but in the end it will probably save a lot of time, materials, and frustration.

I’ve been working with a student using Mike Ware’s “New Cyanotype” for a few weeks. It’s closer to Pt/Pd in scale than to classic cyanotype, and much smoother, but there is still a lot of variation among papers. I just made a successful QTR curve for Platine with OEM inks on a 3880 for my student since that is what she’ll be using. I can make one for PiezoDN on either 1430 or 3880 probably within a few days, and I’ll be happy to share it with you.




thanks very much for the answer. I have a 1500W (ie a 1430) so don’t have the curve. I am comfortable with making photoshop curves based on a Stouffler step but have to confess to not having tried making a qtr curve. Could I make a photoshop curve and then print using for example the PiezoDN Master curve, adjusting the photoshop curve accordingly?

Or should I learn how to make a QTR curve from scratch?

I have been using the PiezoDN curves for Kallitypes with great results but am now looking to start producing four colour Gum Bichromate prints - hence the question.



The 1500W is the same printer as the 1430 and there is a PiezoDN curve folder for it located at >Piezography>Curves>xxx-NON-US PRINTERS>1500-PiezoDN. Drag the 1500-PiezoDN folder directly into the Curves folder to be able to install the curve/printer.

If you look at the instructions for how to “tune” a curve with the Piezography Curve Adjustment tool you should be able to get the master to work with any alt-process fairly quickly. All you will need to do is to make a low contrast curve with the Curve Adjustment utility, then print a target with that curve and print it on your process. Then linearize.

The PiezoDN CGATS smoother has multiple different smoothing levels and cyanotype/gum may require a higher amount of smoothing than the default of “2”






Hi Charles - I didn’t mean to suggest making a standard QTR curve for your PiezoDN setup. I was just trying to make the point that cyanotype and gum are significantly higher contrast processes than Pt/Pd or Kallitype so you will need a significantly lower contrast negative to get a full-scale print. Sorry if that was unclear. Rather, I meant making your own “Master” by using the Limiter target and adjusting based on the result as outlined in the PiezoDN manual. That is basically what Walker is suggesting too, but I think in this case the difference in exposure scale is large enough that printing the Limiter target using the existing Master then adjusting with the Curve Adjustment tool to make an accurate low-contrast Master will be worthwhile.

To illustrate this in Stouffer 21-step terms, Pd typically produces 18 steps (9 stops or an exposure scale of 2.7) of visible separation, while Ware Cyanotype produces 14 steps (7 stops or a 2.1 exposure scale), and classic cyanotype produces 10 steps (5 stops, 1.5 exposure scale). There can be significant variation among papers too. This kind of difference in exposure scale cannot be accommodated by a single Master quad.

If you are using standard cyanotype chemistry - ferric ammonium citrate and potassium ferricyanide - and haven’t tried the Ware formula which uses ferric ammonium oxalate and converts potassium ferricyanide to ammonium ferricyanide in a single solution, I think you will be pleasantly surprised. Making the formula seems a little complicated the first time, but Mike’s instructions are very good. Exposure times are much shorter than the standard formula, and gradations are buttery smooth on many papers unlike the grittiness of the standard formula. At least that is my experience.

Regarding use of the Stouffer step-tablet: I did not mean to use it as a basis for making any sort of Ps, QTR, or PiezoDN quad curve. What I meant was to use it as a tool to learn the variables of the process and to have a standard by which to compare them. For example, with Ware Cyanotype I tested around 10 papers with and without Tween 20 added to the coating, some with an acid pre-soak (if they didn’t work well without). Exposure times to merge steps 1 and 2 on the Stouffer varied from approximately 5 to 12 minutes. Some papers worked well with no additives or pre-soak (Crane’s Resume, Arches Platine 145gm, Ruscombe Herschel); some were better with Tween 20 (Arches Plaitine 300gm) but others not (Crane’s, Platine 145, Ruscombe Herschel) and for some it didn’t seem to matter (Stonehenge); some responded beautifully to an acid pre-soak (Fabriano Artistico); some were very good without a pre-soak but responded well to it anyhow (Stonehenge - longer tonal scale); and some were bad in all situations (COT320, Revere Platinum, Hahnemuhle Platinum Rag, Rives BFK).

For gum the variables are even more - um - varied. The 3 dichromates that are used for gum printing - ammonium, potassium, and sodium - each have different contrast and exposure speed characteristics. Then there is the practically endless range of pigments. And of course paper. If your style is loose then much of this may not matter, and making a curve for one combination of dichromate, pigment, and paper may suffice for everything, but if you are trying to make realistic CMY or CMYK prints it’s probably worthwhile to settle on your dichromate(s), pigments, and paper first based on testing with the Stouffer and then make individually optimized curves for each layer.

Sorry for the dissertation. It’s been over 30 years since I last made a gum print but I was taught by a master, and I’ve just started doing cyanotypes again. I have no idea what your skill level is. I hope this is of some help.




That’s very helpful. Thanks.