Adjustment Curves Layer for the DN of the multi-layer gum prints?

[Dan Pavel]

I am trying to make a PS script that can be used for making the Adjustment Curves layer for the DNs intended for multi-layer gum prints. The problem with multi-layers prints comes from the fact that each gray value of each layer, even if the layers are identical, doesn't add lineary. In other words printing 2 overlaid layers each with the gray level "A" will not result in an image with the gray level "2xA".

Till now I think of taking one of the following paths in solving this problem:

1. The simpler approach could be to print a 3 layers image, for instance, then to find the curve for it as it was an 1 layer print, to make a new DN with the curve applied, print it again in 3 layers, get another curve, and so on. Each curve should be closer to the to the proper curve and after repeating the procedure 4-5 times hopefully a good curve, close enough of the correct one, will result. It will be quite time-consuming and I'm not sure that the resuted adjustments will converge to the correct one or will be divergent . However, it could work with identical layers.
2. A better approach with identical layers could be to print the first layer, measure it, then the second overlaid layer and again measure the result. By comparing the 1 layer - 2 layers results a table could be made and one could read what level of gray should be used in each layer to have a desired level of gray in the final image. Based on it a curves adjustment layer applicable to each layer could be made.
3. The adjustment curve could be split in 2 - highlights till mid tones and mid tones till black. The first printed layer should only cover highlights to midtone and everything darker than the mid tone should be printed as mid tone. The color/gum/bichromate ratio and the exposure time have to be adjusted accordingly. Being a single layer print an adjustment curve could be computed for it. The second layer should have more color and a much shorter exposure time to make it print everything whiter than the mid tone value as white. Hopefully an adjustment curve could be computed for it as well, using a table as in the "2" case. The final adjustment curve could be obtained by using the first half of the first curve (highlights) and add to it the second half of the second curve (shadows).

I am not sure about none of the paths described above. How do you think that this problem should be proper solved? Do you have other ideas?

 

[ced]

I am interested in doing tests with your software Dan, say starting with1st curve till it seems good for HL-MT then adding the MT-SH & see where it takes one.
Have to get the light tones up in density as I notice generally that there is too much contrast in the HL, the SH tends to bung up without enough DMax requiring a 3rd or even 4th layer.
It could be that the inkjet printer does not lay down enough ink in the lightest area of the neg that contributes mainly to that issue.
If some more users are interested in taking part this would be great.

 

[Bob Carnie]

I have been printing gum over palladium for a very long time and extensively..

Gum is a wash off process and IMHO any negative will work , you just need to control the wash off period. With this said I use the same curve for gum that I do for Palladium and if I feel I want to hold the gum more in areas I flatten the curve to allow more exposure to come through.. It is a hardening process and therefore I think complicating the negative thought process is not needed.

Here is a sample of a duo tone gum over the palladium. All three neg's were made with the same profile curve that was established for palladium layer.
Palladium Coating
1. Palladium image is a full scale rendition of the scene from shadow to highlight and all in between.

Gum Coating
2. Blue colour is exposed using a shadow negative - which is done in blend if and taking only the lower values and printing
3. Orange colour is exposed using a mid to upper highlight negative - which is done in blend if and taking only the upper values and printing.

Where there is no exposure in either colour or gum negs the pigment can be washed off and the exposed hardened areas are left behind. So if you think about Gum this way you will come to the same conclusion that I have is that as long as it is a negative that has the characteristics you want then it will work.

I think of the movie Karate Kid when making gum overs ... the palladium is a WASH ON process where the noble metals are introduced to the paper , and the gum layers are a WASH OFF process where the unwanted gum is washed off with water.

I have studied , all the books, my hero's are Stephen Livik , Sam Wang and Christina Anderson, and if you delve into the process deep enough you may find sage words in what I have just wrote.
Once you see this aspect of the process the faster and better you will grasp multiple negs and the Gum process.

just my two cents.

Bob

 

[ced]

The issue with multiple negs requires too much attention to the registration problem when one has a lousy punch system.
Pros, they have all that is required in house but us mortal cheapskates fiddle about.

 

[Bob Carnie]

The issue with multiple negs requires too much attention to the registration problem when one has a lousy punch system.
Pros, they have all that is required in house but us mortal cheapskates fiddle about.

I think the issue is with the operator not figuring out registration rather than the system to achieve it, which is not that difficult and certainly not expensive. I would argue after 10 miss registered prints fiddling about you have the cost of a simple punch system with the waste.

 

[Andrew O'Neill]

I use four thumbtacks, and my registration is spot on.

 

[adelorenzo]

I use an office hole punch and some screen printing registration pins. Total cost about $25 and it works perfectly.

 

[nmp]

I am trying to make a PS script that can be used for making the Adjustment Curves layer for the DNs intended for multi-layer gum prints. The problem with multi-layers prints comes from the fact that each gray value of each layer, even if the layers are identical, doesn't add lineary. In other words printing 2 overlaid layers each with the gray level "A" will not result in an image with the gray level "2xA".

Till now I think of taking one of the following paths in solving this problem:

1. The simpler approach could be to print a 3 layers image, for instance, then to find the curve for it as it was an 1 layer print, to make a new DN with the curve applied, print it again in 3 layers, get another curve, and so on. Each curve should be closer to the to the proper curve and after repeating the procedure 4-5 times hopefully a good curve, close enough of the correct one, will result. It will be quite time-consuming and I'm not sure that the resuted adjustments will converge to the correct one or will be divergent . However, it could work with identical layers.
2. A better approach with identical layers could be to print the first layer, measure it, then the second overlaid layer and again measure the result. By comparing the 1 layer - 2 layers results a table could be made and one could read what level of gray should be used in each layer to have a desired level of gray in the final image. Based on it a curves adjustment layer applicable to each layer could be made.
3. The adjustment curve could be split in 2 - highlights till mid tones and mid tones till black. The first printed layer should only cover highlights to midtone and everything darker than the mid tone should be printed as mid tone. The color/gum/bichromate ratio and the exposure time have to be adjusted accordingly. Being a single layer print an adjustment curve could be computed for it. The second layer should have more color and a much shorter exposure time to make it print everything whiter than the mid tone value as white. Hopefully an adjustment curve could be computed for it as well, using a table as in the "2" case. The final adjustment curve could be obtained by using the first half of the first curve (highlights) and add to it the second half of the second curve (shadows).

I am not sure about none of the paths described above. How do you think that this problem should be proper solved? Do you have other ideas?

I am not sure I fully understand....are you trying to do a multi-layer single color (like black and white) or a multi-color print like a full RGB image. The two cases will have to have completely different approaches, no?

:Niranjan.

 

[Dan Pavel]

A multi-layer single color print, Niranjan, but using the same DN for all the layers.

 

[Dan Pavel]

I have been printing gum over palladium for a very long time and extensively..
... It is a hardening process and therefore I think complicating the negative thought process is not needed.

Yes, the gum-palladium combination works quite nice using the same DN and carefully choosing the colors and the process parametres, too. I prefer to do it the other way around - palladium ove gum. It involves some additional steps but this way the details of the palladium prints are better preserved. One sample here:

The problem is with the DNs for the gum-only multi layer prints (a single DN for all the layers). I don't know any script that can produce an adjustment curve for it.

 

[Dan Pavel]

I am interested in doing tests with your software Dan, say starting with1st curve till it seems good for HL-MT then adding the MT-SH & see where it takes one.
Have to get the light tones up in density as I notice generally that there is too much contrast in the HL, the SH tends to bung up without enough DMax requiring a 3rd or even 4th layer.
It could be that the inkjet printer does not lay down enough ink in the lightest area of the neg that contributes mainly to that issue.
If some more users are interested in taking part this would be great.

Thanks ced!
I haven't yet made the script as I wasn't yet able to decide which is the most rational approach. I hope that people answering this thread will help.
The split-curve approach is the most appealing to me, too. I think that a script to find the correct curve for 2 layers (HL-MT & MT-SH) could be done. The additional 3-rd layer for deeper shadows could be a short-exposure darker pop-up layer, if necessary, but not covered by the curve. Dealing with the calculations for a curve to cover more than 2 layers seems too complex for me.

 

[nmp]

A multi-layer single color print, Niranjan, but using the same DN for all the layers.

Dan, OK. So it is a monochrome gum print using multiple iterations of (coat/expose/process) steps, using the same negative.

Raises more questions for me:

1) what is the purpose of multiple layers - is it to increase the shadow density incl. Dmax? Like what Irving Penn did with his pt/pd's.

2) why the same negative? You still need to align the negative every time regardless of whether you use same negative or not. Then why not tailor each layer with its own negative.

Am I missing something here?

:Niranjan

 

[Dan Pavel]

Well, when testing a strip with 11 levels of grey (0% to 100%) a single layer of gum only covers max. ~7-8 distinct steps. I was never able to conclusively get more than that. The needed black is hard (I'd say impossible) to obtain with a single layer of gum. If the color is too concentrated in the sensitiser you get pigment stain. That's why you have to add more gum to the solution and therefore you'll have a diluted black. The usual approach is to make a 2-4 layers print - usually one layer for highlights (longer exposure, greater dilution), one for midtones and one or two for shadows (very short exposure with a bigger concentration of color in the sensitiser). To balance the color/gum/bichromate ratio and get the correct exposure for each layer is quite difficult if you want the gum print to resample the gradations of your initial image. And an adjusted DN is, IMO, mandatory.

Of course, the other approach is to creatively use the gum process and never intend to resample your original image gradations. I'd even say that this is the most appealing way to use the gum process.

Why a single DN? Well, because I think it's possible and it simplifies the process. Some experienced gum printers achieved this goal with some empirical-made adjustment curves but there is, AFAIK, no script to automate the work.

I am really curious how some of the more experienced members of this forum manage to master the multi-layers monochrome gum process when the intention is to correctly match the initial image tonalities.

 

[jim10219]

I use registration marks in the four corners printed on my negatives and don’t have registration issues. It doesn’t cost me a dime. If I’m printing straight from a camera negative, I use the four corners of the negative. If you don’t want them to show in the final print, and keep a feathered edge, just place them in the far corners and drop a dab of emulsion out there, not connected to your feathered edges. Then trim them off when you’re done.

As for controlling the density of multiple exposures, I use the same curves for each pass and adjust the chemistry and exposure times to adjust my density. I find that is a better approach for controlling highlights and shadow density independent of the mid tones. Double the pigment and half the time for shadows, and half the pigment and double the time for highlights, as a rough guide to get you started. You really have to experiment to figure out your system.

 

[ced]

At present I like the single neg approach as the neg never leaves the glass till complete.
The paper after preshrink gets two double side alu tapes and punched (office punch) and this gives pretty good registration so introducing multiple negs for me is a pita.
I think getting a 1st curve more or less close for HL-MT adding the 2nd layer and then measureing the result for fine tune of 1st neg. curve might be a direction.

 

[Bob Carnie]

Yes, the gum-palladium combination works quite nice using the same DN and carefully choosing the colors and the process parametres, too. I prefer to do it the other way around - palladium ove gum. It involves some additional steps but this way the details of the palladium prints are better preserved. One sample here:
View attachment 210052
The problem is with the DNs for the gum-only multi layer prints (a single DN for all the layers). I don't know any script that can produce an adjustment curve for it.

I am sorry but are you suggesting the pt pd after a thick layer of pigmented gum.???? so how does the pt pd penetrate into the paper, is it sitting on top of the gum?? enquiring minds need answers.

I have tried multiple coatings and using multiple negs with single process before with disastorus results , I need to see more of your prints and workflow to understand what you are trying to say here.

I use different methods to get the effects I want but I certainly do not think one negative can do what I do with two , three or even sometimes four.

 

[Bob Carnie]

At present I like the single neg approach as the neg never leaves the glass till complete.
The paper after preshrink gets two double side alu tapes and punched (office punch) and this gives pretty good registration so introducing multiple negs for me is a pita.
I think getting a 1st curve more or less close for HL-MT adding the 2nd layer and then measureing the result for fine tune of 1st neg. curve might be a direction.

I make a lot of negatives and am completely familiar with it and I am still trying hard to understand how a single negative with a wash off process will give you what you want.. It is a hardening wash off process and what
you are suggesting or Dan is suggesting makes absolutely no sense to me. I print gums every month and am quite aware of the effect of water on the gum pigment Ammonium Dichromate mixture after it has received exposure to UV light.

 

[Dan Pavel]

I am sorry but are you suggesting the pt pd after a thick layer of pigmented gum.???? so how does the pt pd penetrate into the paper, is it sitting on top of the gum?? enquiring minds need answers.

.

Yes, pure Palladium (not Platinum!) works on gelatine sized papers. After the gum is processed and dry I apply a new thin layer of gelatine, let it dry, then the Palladium layer. It needs a longer exposure than usual. It is quite fragile and needs slow agitation in the processing baths. Washing needs attention, as well - slow agitation in multiple baths, but in the end you'll have Palladium over gum. When it is dry you can apply a protective gelatine or gum layer on top of it, as well.
If the Palladium layer is not dark enough you can apply another Palladium layer one on top of it.

 

[nmp]

Well, when testing a strip with 11 levels of grey (0% to 100%) a single layer of gum only covers max. ~7-8 distinct steps. I was never able to conclusively get more than that. The needed black is hard (I'd say impossible) to obtain with a single layer of gum. If the color is too concentrated in the sensitiser you get pigment stain. That's why you have to add more gum to the solution and therefore you'll have a diluted black. The usual approach is to make a 2-4 layers print - usually one layer for highlights (longer exposure, greater dilution), one for midtones and one or two for shadows (very short exposure with a bigger concentration of color in the sensitiser). To balance the color/gum/bichromate ratio and get the correct exposure for each layer is quite difficult if you want the gum print to resample the gradations of your initial image. And an adjusted DN is, IMO, mandatory.

Of course, the other approach is to creatively use the gum process and never intend to resample your original image gradations. I'd even say that this is the most appealing way to use the gum process.

Why a single DN? Well, because I think it's possible and it simplifies the process. Some experienced gum printers achieved this goal with some empirical-made adjustment curves but there is, AFAIK, no script to automate the work.

I am really curious how some of the more experienced members of this forum manage to master the multi-layers monochrome gum process when the intention is to correctly match the initial image tonalities.

Thanks, Dan for the elaboration. I get the scenario now, sort of. I am still learning.

Are you familiar with works of Keith Taylor (of Cy DeCossy fame)? They are the most photo-realistic gum prints that I have come across. Wonder how he does it. He uses C, M, Y layers apparently but no idea of multiple applications of the same color to get those dark shadows:

https://www.cydecosse.com/collections/san-ambrosio-market-color/

:Niranjan.

 

[Bob Carnie]

Yes, pure Palladium (not Platinum!) works on gelatine sized papers. After the gum is processed and dry I apply a new thin layer of gelatine, let it dry, then the Palladium layer. It needs a longer exposure than usual. It is quite fragile and needs slow agitation in the processing baths. Washing needs attention, as well - slow agitation in multiple baths, but in the end you'll have Palladium over gum. When it is dry you can apply a protective gelatine or gum layer on top of it, as well.
If the Palladium layer is not dark enough you can apply another Palladium layer one on top of it.

Interesting , does the gelatin sit as a layer like carbon transfers do or does the emulsion sink into the gum layer.

 

[Bob Carnie]

Thanks, Dan for the elaboration. I get the scenario now, sort of. I am still learning.

Are you familiar with works of Keith Taylor (of Cy DeCossy fame)? They are the most photo-realistic gum prints that I have come across. Wonder how he does it. He uses C, M, Y layers apparently but no idea of multiple applications of the same color to get those dark shadows:

https://www.cydecosse.com/collections/san-ambrosio-market-color/

:Niranjan.

He does three layers with cyan, magenta and yellow as most tricolour gum printers do.

 

[nmp]

He does three layers with cyan, magenta and yellow as most tricolour gum printers do.

Thanks Bob. No need for a K layer then, as the conventional printers do.

I love your work too...

 

[Dan Pavel]

Interesting , does the gelatin sit as a layer like carbon transfers do or does the emulsion sink into the gum layer.

The thin gelatine layer applied on top of the gum layer sticks on it and is not perceived as a separate layer.
If you want to try it I'd advice to first test Palladium on gelatine sized paper till you get it right and then the whole Pd over gum process. If the Pd sensitizer doesn't apply evenly on the gelatine sized paper (there are many kinds of gelatine) just moisten the gelatine layer with a brush, absorb well the excess water with some absorbent paper, let it stay for 3-5 minutes and then apply the Pd sensitiser. It will come out evenly (at least it works for me...).

 

[Dan Pavel]

Are you familiar with works of Keith Taylor (of Cy DeCossy fame)? They are the most photo-realistic gum prints that I have come across. Wonder how he does it. He uses C, M, Y layers apparently but no idea of multiple applications of the same color to get those dark shadows:

https://www.cydecosse.com/collections/san-ambrosio-market-color/

:Niranjan.

Yes, I know his work and I like it very much.

Tony Gonzalez is another color gum printer I like.
A description of his techick :
http://tonygonzalezartist.tumblr.co...-bichromate-journal-tweaking-the-shadow-curve
And a layer-by-layer time-lapse of one of his work :

 

[Bob Carnie]

Thanks Bob. No need for a K layer then, as the conventional printers do.

I love your work too...

I actually use a palladium layer for the detail and use cmy on top, I suck out density on the palladium layer in colour regions so full colour can come in.