[Alt-photo] suggestions about cyanotype curve

Niranjan Patel nirpat89 at yahoo.com
Tue May 23 21:18:54 UTC 2017

Hi, Bob:

Not being able to see your attachments,but my experience of using the ChartThrob which I am assuming issomewhat similar to Peter Mhrar's technique you are using (of which Iam unfamiliar) might be of value here. I encountered exact sameproblem as you on my first curve for Centennial POP after optimizingfor Dmax with colorized negative. I think these auto-curve programsare less capable to give the “correct” correction curve in theareas of extreme variations, exacerbated greatly because ofPhotoshop's inability to have points that are close together. Thereare I am sure a number of ways to circumvent this problem. They allprobably do the same thing, i.e. reduce the maximum density at the100% (0% for the negative) step. The ink reduction is a logical stepas suggested by others here. Here are a couple more:

1. Downgrade your “optimum” colorfor maximum opacity. The way I did was measure the RGB values on thecolorized negative in Photoshop of the first or the lowest step thatis all clear. Use those valuesw as the color for the “color fill”layer. Now re-do the cyanotype and the curve generation.

 2. Use a simple post -inversionstraight curve applied to your negative image below the color filllayer: Point 1 = (0,x) where x corresponds to that same lowest allclear step (inverted). For example for the 10th step atB= 10% it would be (255-230) = 25. And Point 2 = (255, 255). Redothe curve generation.    
 For my diginegs, I ended up using acombination of both of these in the end. Also I started doing thecurves manually, the old fashioned way. 

Don't know if this makes sense or willhelp your particular case. As Keith S. says, it is easier to showthan say. 


      From: Bob Cornelis via Alt-photo-process-list <alt-photo-process-list at lists.altphotolist.org>
 To: liz Brocycák <alt-photo-process-list at lists.altphotolist.org> 
Cc: Bob Cornelis <bob at colorfolio.com>
 Sent: Tuesday, May 23, 2017 11:31 AM
 Subject: Re: [Alt-photo] suggestions about cyanotype curve
Thank you Alberto, Keith and Dan for the responses! I see a consensus here. I have a couple of questions to further my conceptual understanding of why you’re making this suggestion, which I will definitely test out - forgive me if these are beginner questions, this is the first time I’m wrestling with some of these parameters. I’m also attaching my screenshots again because at least one person has said they were missing from my last email.

It might help to note that these are printed with a colorized neg - using the testing suggested in Mrhar’s book, I determine that a particular shade of orange gave the best UV blocking. Dmax was determined using a standard test of different exposure times to determine at what point the dmax under the Pictorico Ultra film matched that of the uncovered, sensitize paper. 

The file PreCurve shows a 100 step wedge printed with those parameters and no adjustment curve. Basically the range of tones is limited to about patch 70 on the dark end and 11 on the light end. Everything higher than 70 is the same max black and everything under 11 is pure white. Patches 11-70 contain all the distinct values of the compressed tonal range of the Ware cyanotype on this paper. The PostCurve file shows the same chart after adding my adjust curve create from PreCurve. You can see that steps 0-5 are white and then there is a big jump at step 6 and after that, the tonal range isn’t bad the rest of the way. 

Your suggestion is to lower ink density in the neg, which would lead to a darker print with the same exposure time. I can do this in the color density slider in the epson driver or I could use a longer exposure time, if I’m not mistaken. Are these equivalent, except for the fact that the former would result in using a slight amount less ink and be a faster exposure?

I think what the result of doing this would be that the range of distinct tonal values in the Precurve target which are currently 11-70 would shift downward though still consist of the same number of patches. For example, instead of 11-70, the range might be 5-64. There will be more pure black patches and fewer pure white patches. I think that this would lead to more “work” being needed in the last 10% of the curve representing the darkest values and less work being needed in the (currently problematic) last 10% of the highlight part of the curve. Currently this is the case and the PostCurve I end up with does much better with the darker values, so maybe a further shift in this direction is a good thing. Is that the goal? Why would the PS curve handle adjusting a wider range of pure black values to their proper tones rather than the tones in the top highlight range? Maybe it’s easier to lighten tones that are already there than to introduce tones into pure white?

Is my characterization of what would happen to my PreCurve correct? Why would this help the end result? Certainly I will test this and find out what happens and if it helps, but I’m interested in why this would be the case or if I’ve got it wrong to begin with.

This conversation is extremely helpful to me on both a practical and conceptual level, so thank you again for chiming in!




More information about the Alt-photo-process-list mailing list