Choosing Colours and Curves: The HSL Array

[dwross2]

Very interesting. I'm glad this topic popped up. The interplay of variables is exactly what makes problem-solving both frustrating and irresistably rewarding. I never would have guessed so much could be going on with gum. I'm looking forward to seeing the answers to all those questions.

Katherine: I was having problems with a popup-blocker, but finally got a chance to see your steptablet. Thank you for posting it. It's probably because I'm a photographer, but pictures always make things clearer!

 

[Katharine Thayer]

I think you've nailed it or are very close. My next post was going to be something to the effect that maybe we're seeing the interaction of two variables. The question is, if it is a pigment load issue, what's the actual mechanism that's causing the stain? Exposure or lack thereof seems to be part of it. Why isn't the gum wiped clean off? Does a longer soak get rid of the stain eventually? Would we, or do we, see the same stain in an unexposed paper treated with that emulsion?
~m

Hi guys, I've had company for several days and am just now catching up. I'll do my best with the questions:

Q: "If it's a pigment load issue, what's the actual mechanism that's causing the stain?... Why isn't the gum wiped clean off?" A: The unhardened gum does get wiped clean off, but the pigment doesn't wash off with it, because the pigment has sunk into the paper as stain. This happens, as Demachy explained more than 100 years ago, when the pigment/gum ratio is too high. As I say somewhere on my site, it may help to think of the difference between watercolor painting and gum printing. Both create images using pigment in a gum arabic medium, (never mind for the sake of this discussion that gum printing involves a light-sensitive chemical and exposure to light, because IME that's immaterial to the creation of stain); however, in watercolor painting, the idea is to have the pigment/gum ratio high enough so that the pigment sinks into the paper as a stain. The more gum you add (the lower the pigment/gum ratio) the less likely the pigment is to stain the paper, or any other surface. There is a myth that pigment stain occurs only on unsized or badly sized paper and it's not an issue for well-sized paper or on metal, glass or plastic; this is simply a myth. I have examples on my site of pigment stain on glass and on plastic, caused by overpigmentation. A watercolor painting, then, is essentially made of pigment stain. In gum printing, we use a lower pigment/gum ratio so as not to stain the paper; if you used the same pigment/gum ratio for gum printing as for watercolor painting, you'd just get a paper that's stained all over, rather than an image. In other words, most examples of pigment stain are not affected by exposure at all. See my page on pigment stain for examples of this. The stain in what's called "tonal inversion" isn't so bad as to obscure the image, only bad enough to to stain the unexposed areas, for reasons that I don't understand. But this density explanation just doesn't work for me. Gum printers have been printing 21-steps, which according to the Stouffer spec sheet go up to 3.05, for decades without getting this tonal reversal thing on the upper steps. As I said, I've only seen it a few times, and those times I knew for a fact that the mix was overpigmented.

Q: "Does a longer soak get rid of the stain?" A: No. Pigment stain is permanent, and can't be removed by soaking or treating.

Q: "Would we, or do we, see the same stain in an unexposed paper treated with that emulsion?" Yes. In fact I did this experiment at one point and I think I posted the results for the alt-photo list, so it's probably on my site somewhere. It's the exact same stain on the completely unexposed paper as on the unexposed areas of the exposed paper. This might indicate a possible interaction, as you say, for this subspecies of pigment stain, but I don't see how that interaction operates, if it does, and as you say, it's a footnote at best. Although for folks who are plagued with this tonal inversion, it's more important than a footnote.

What I haven't seen yet is proof that changing the color of the negative to more perfectly match the ES of the emulsion will eliminate the "inversion," and in fact it seems clear that that solution wouldn't work with the inversion shown in my example above, because the inversion exists across the entire spectrum.

k

 

[Katharine Thayer]

What I haven't seen yet is proof that changing the color of the negative to more perfectly match the ES of the emulsion will eliminate the "inversion," and in fact it seems clear that that solution wouldn't work with the inversion shown in my example above, because the inversion exists across the entire spectrum.

k

After I sent that, I realized what I'd just said, and when I tried to edit it, it put me into an endless loop. When I said that picking a different color wouldn't make any difference, I was thinking only of hue, but of course changing the lightness, luminance, or whatever the word is in HSL will make a difference.

 

[Katharine Thayer]

Katherine: I was having problems with a popup-blocker, but finally got a chance to see your steptablet. Thank you for posting it. It's probably because I'm a photographer, but pictures always make things clearer!

Hmm, I'm not sure why a popup blocker wouldn't like my site, since I shouldn't have any pop-ups (if pacifier puts pop-ups on my site that are visible to others but not to me I'd like to know about it). I don't use any of the fancy stuff, no flash, no java, nothing but old-fashioned basic html, which should be friendly to all.
kt

 

[Katharine Thayer]

Q: "Does a longer soak get rid of the stain?" A: No. Pigment stain is permanent, and can't be removed by soaking or treating.

On reflection, I might qualify this answer somewhat. I would still answer the question "no," and say that pigment stain can't be eliminated by soaking or treating. (The type of stain some gum printers recommend long soaking to remove is the other kind of stain, dichromate stain.) For example, when I printed those prints of the HSL array the other day that showed the "tonal inversion," I took one of them and soaked it for about six hours to see if soaking would lighten it; it had no effect on it whatever.

But I might back off a bit from the assertion that in all cases pigment stain constitutes a permanent indelible stain, because I've seen cases on well-sized paper and on glass, where both normal pigment stain and this "tonal inversion" kind of pigment stain can be easily wiped off with a rag or a finger, even if long soaking doesn't budge it. And don't ask me how pigment can stick to the surface if there's no hardened gum to hold it there and when it hasn't formed an indelible stain in the paper. Some have posited some electrostatic mechanism, but we don't really know.

Anyway, Denise, you may be sorry by now that you asked. Yes, gum is incredibly complicated, and there's a lot going on that hasn't been well explained yet. I've sometimes likened discussions among gum printers to the fable about the blind men describing an elephant by feeling the different parts of it: "It's like a tree trunk!" "No, it's like a big hose!" "No, it's like the leaf of a banana tree!" "No, it's like a piece of rope!" "No, it's like a wall!" That's about where we are in understanding how the gum process works.

 

[sanking]

Tonal inversion to take place with most processes, including in my own experience carbon, kallitype, palladium, platinum/palladium, and gum. And it is certainly not uncommon.

The mechanism with every process appears to be more exposure than necessary to obtain Dmax with the specific process.

Sandy King

 

[Katharine Thayer]

Tonal inversion to take place with most processes, including in my own experience carbon, kallitype, palladium, platinum/palladium, and gum. And it is certainly not uncommon.

The mechanism with every process appears to be more exposure than necessary to obtain Dmax with the specific process.

Sandy King

Just for clarification, are you saying that in your experience the mechanism that produces tonal inversion is overexposure?
kt

 

[sanking]

Just for clarification, are you saying that in your experience the mechanism that produces tonal inversion is overexposure?
kt

Yes. There may be other causes, but from my experience the primary mechanism is overexposure.

Sandy

 

[Katharine Thayer]

There you have it, Denise; as I said, it's like the blind men and the elephant: "It's underexposure!" "No, it's excess density in the negative!" "No, it's overpigmentation!" "No, it's overexposure!" I'm personally starting to think that the light source might be a significant factor.

Of all the proferred explanations, I think overexposure is the least likely for the rare appearance of this phenomenon in gum. Sometimes mechanisms can be successfully generalized across processes, but more often, they can't.
kt

 

[sanking]

Of all the proferred explanations, I think overexposure is the least likely for the rare appearance of this phenomenon in gum. Sometimes mechanisms can be successfully generalized across processes, but more often, they can't.
kt

You may believe tonal reversal is rare in gum, and it may be in your work, for whatever reason. However, I don't believe it is really all that rare. I looked at some five dozen test strips in gum made by Sam Wang, in cyan, magenta and yello, and every single one of them showed tonal reversion with over-exposure. Exposures were very carefully controlled with a light integrator with a NuArc platemaker, and exposure was the only variable.

If you look in the literature I think you will find that Mike Ware has already offered an explanation for this type of tonal reversal in gum. Mike has a web site so you might contact him for more information.

Sandy King

 

[Katharine Thayer]

You may believe tonal reversal is rare in gum, and it may be in your work, for whatever reason. However, I don't believe it is really all that rare. I looked at some five dozen test strips in gum made by Sam Wang, in cyan, magenta and yello, and every single one of them showed tonal reversion with over-exposure. Exposures were very carefully controlled with a light integrator with a NuArc platemaker, and exposure was the only variable.

If you look in the literature I think you will find that Mike Ware has already offered an explanation for this type of tonal reversal in gum. Mike has a web site so you might contact him for more information.

Sandy King

Sandy, thanks for the data from Sam Wang; that's interesting, especially since it seems to provide nonsupport for the theory that tonal inversion is a function of underexposure, which has been a widely held belief in some quarters of the gum universe.

One of the indirect underpinnings of that belief was a speculation from Mike Ware (which I am assuming must be the "explanation" you're referring to here, because I've not heard of other explanations from Mike Ware that would touch on this) that crosslinking takes place mainly at the paper surface rather than at the gum surface, due to some special attraction that dichromate has for paper that makes the dichromate travel down through the gum and attach to the paper, in other words that because of this hypothesized affinity of hexavalent chromium for paper, crosslinking may start at the bottom rather than at the top, and so even the most minimal exposure will result in hardened gum being present in the paper and serving as a resist to repel stain, and it's only where the gum doesn't get this minimal exposure that stain can occur.

After hearing for several years on the alt-photo list that Mike Ware had "explained" this phenomenon, I did have some correspondence with him about it a couple of years ago, but it seemed to me from our correspondence that (1) it was just a speculation, a hypothesis, and would have to be tested before it could provide any explanatory power, and (2) depended on assumptions he held about gum printing that weren't accurate, for example that unreduced dichromate is extremely difficult to remove from paper after printing. In fact there's nothing easier than removing unreduced dichromate; in my experience it passively washes out of the print without effort in the first minute after the print goes into the water. But at any rate, as Mike informed the alt-photo list last year, his speculation was simply a speculation and shouldn't be taken as an explanation for anything.

But it puzzles me how that speculation could possibly be used as support for the idea that overexposure causes tonal inversion, since in the past it's always been used to support the idea that tonal inversion is a function of underexposure, and I'm not seeing a way how it would work the other way. But maybe it's just too early in the morning yet.

I hope you weren't taking my remark about the blind men and the elephant personally; I was including myself in that remark as well. None of us are seeing the whole picture, or we'd all be seeing more the same thing, was my point.

But as far as why I don't see this all the time as you report Sam seems to, I don't know. The last few days, I've started to wonder if the light source might not be a factor.
kt

 

[sanking]

Sandy, thanks for the data from Sam Wang; that's interesting, especially since it seems to provide nonsupport for the theory that tonal inversion is a function of underexposure, which has been a widely held belief in some quarters of the gum universe.
kt

OK, I may have mis-represented Sam's results. Now I am wondering if what I saw was not tonal reveral with underexposure. I will ask him about this and get back later on what I find.

Frankly I can not think of any good reason why there would be tonal reveral with a dichromated colloid process like gum with overexposure. I definitley do not get it with carbon, though I do get less Dmax with very long scale carbon than short scale. This may be what is going on with gum.

However, tonal reversal with the iron processes is well-documented.

Sandy King

Sandy

 

[dwross2]

Well...this blind woman is about to step in - and I can hardly wait. Yesterday, I finished ordering the last of what I need to start Gum-n-Ag printing. (it wasn't much - for better and for worse, I'm the typical multi-media artist packrat). I'll report in - with what is bound to be excruciating detail - my findings/opinions. I'm not very good at taking anyone's word for gospel until I've mucked about myself. In art, there are just too many variables - as has been hinted. The staining of the individual pigments, the quality of the dichromate, the paper... If you add film vs. mylar, light quality and the gods know what else, what a wonderful, puzzling stew. This is the reason that we do this, right? Otherwise, we'd turn over all our creative variables to Epson.

Muck boots at the ready,
Denise

 

[Katharine Thayer]

I'll report in - with what is bound to be excruciating detail - my findings/opinions. I'm not very good at taking anyone's word for gospel until I've mucked about myself.

This is the only sensible approach to take with gum. I'll be looking forward to those reports.
kt

 

[Katharine Thayer]

More on gum tonal inversion

A discussion occurring in a different thread seems to have some relevance to my present musings about the various explanations for "tonal inversion" in gum.

It occurred to me, while puzzling about this, that I didn't remember seeing a tonal inversion in the 21-steps I printed to determine the exposure I used for the print of the HSL array posted earlier in this thread which, as Michael pointed out, did have a tonal inversion. I went down to the darkroom and dug out the test strips, and sure enough, that set of test strips, ranging roughly from 20% underexposed to 20% overexposed, like all the other test strips in the pile, are pristine white from the end of the tonal scale (7-8 steps for this particular emulsion) all the way through the 21st step. This is what I routinely see when printing step prints.

We've already established, I think, that it can't have been underexposure that caused the tonal inversion, because the array was well exposed. And I'm not inclined to be convinced by the underexposure explanation anyway, after running a series of tests using exposures ranging from 50% underexposed to 50% overexposed, in which I was not able to demonstrate any effect of exposure on stain in general or on tonal inversion in particular. If a paper was going to stain, it stained equally across all of the exposures and across all densities within each exposure; if it wasn't going to stain, it was equally stainfree across all exposures and all densities. If Sam's tests show differently, it would be lovely if he'd share them so we could all have the benefit of the information.

But then there's the question of excess density, which I think is a somewhat different issue, although the two do overlap to some extent. But that explanation also doesn't seem sufficient to me, because like many other gum printers, I've followed for many years the time-honored practice of printing from a negative with longer DR than matches gum's ES, and printing the negative twice, once for the shadows and once for the highlights. In other words, the first printing prints the lower end of the DR of the negative, and the second printing prints the higher end of the DR of the negative. When printing the shadow end first, I've always got pure white under the denser part of the negative, not a tonal inversion. And, as I've mentioned before, I almost invariably get pure paper white all the way to the end of a Stouffer 21-step.

So why would I get a tonal inversion printing the HSL array when I didn't get it printing the Stouffer 21-step? According to Stouffer, the 21st step represents an (optical) density of 3.05. But then the question of uv density vs optical density comes in, and also a different question, that was addressed in a different thread in this forum, about spectral sensitivity.

The conventional wisdom about the spectral sensitivity of gum seems to rest on assumptions rather than actual data, but as I probably wrote in that thread, the only research I've seen that actually looks at the spectral sensitivity of dichromated gum compared to the spectral sensitivity of other dichromated colloids, found that gum's spectral sensitivity was moved up into the visible range compared to whatever other dichromated colloids it was compared to, which were more where people assume the spectral sensitivity is for dichromated colloids in general. So I don't even know for sure whether a uv densitometer is appropriate for gum. And while there's a question hanging there about whether perhaps the squares toward the bottom of the HSL array are "denser" to whatever wavelengths gum is most sensitive to than the densest steps of the Stouffer 21-step are, and that's caused the tonal inversion, (although then, as I asked before, why wouldn't it happen every time?) I'm not sure there's any way to answer that question, since there are too many unknowns in the equation.

HOWEVER, I am forced by the data to concede that pigment concentration can't be a sufficient explanation; if it were, then I should have seen the tonal inversion in the 21-step as well as in the HSL print that was made with the same emulsion on the same paper with the same exposure, and it wasn't there.

Michael's got to be right; it has to be some kind of complicated interaction between variables. My 2cents worth,
Katharine

 

[smieglitz]

Coming late to this thread trying to figure out the digineg thing with a new printer, but since tonal reversal in gum and clearing gum in bisulfites have been mentioned, here's a file illustrating both:

The pigments were bone black and cobalt violet (both Daniel Smith IIRC- it has been a few years since I've printed any gum). Exposure was with a NuArc mercury exposure unit. Reversal is more pronounced with less exposure in the bone black and seems to be related to pigment choice and pigment stain as well as exposure. I thought the interval between coating, printing, and processing might have some effect, but that doesn't seem to be the case.

Make of it what you will. It has me baffled.

Joe

 

[Katharine Thayer]

Thanks, Joe. I think it's a good illustration of several things: (1) the already noted fact that it's difficult to isolate one variable that "causes" tonal inversion (2) that using the same amount of pigment to gum regardless of the pigment results in different color saturations for different pigments, as I've pointed out, (3) that bisulfite clearing targets dichromate stain rather than pigment stain, as was pointed out earlier in this thread.

But I'm not "viddying" any important difference between the 400 unit exposure and the 75 unit exposure for bone black, in regard to tonal inversion. There's obvious tonal inversion in both; while the lightest step in the middle of the inversion is very slightly darker for the greatly overexposed strip, the value of the darkest stain is the same in both, which makes the "tonal range" of the inversion slightly less in the case of the overexposed strip, if one can talk about stain in terms of a tonal range, but doesn't change the fact. The overexposure runs everything a little farther up the scale, but doesn't change the basic observation of stain and tonal inversion.

I've never used cobalt violet to speak of, so can't judge by the sample how much more one could saturate the color of this pigment by using more pigment, or how much would be required to induce pigment stain. By the way, when I talk about the relationship between pigment concentration and pigment stain, I am always talking about a relative relationship; within a given pigment/brand there's a point at which you cross the line to overpigmentation, but that point is not the same for every pigment/brand. What I don't mean, and I hope everyone understands this, is that if 1 gm pigment/10 ml gum stains in one pigment/brand, then the same proportion of pigment/gum should stain in another pigment or another brand. That is exactly what I don't mean.

My general impression of cobalt violet is that it is a rather weak pigment, as well as a not very dark-valued one, which is a different issue. Since pigments are all different on the variables of strength, intensity, and value range, one can't of course draw any conclusions about the influence of pigment amounts on pigment stain or tonal inversion by using the same amount across different pigments. What might be more useful is a comparison between more pigment and less pigment using just the bone (ivory) black or just the cobalt violet, or even better, both (a multivariate table).

Just more thoughts for the collective mind to chew on, FWIW.

P.S. Your 4x5 step prints always make me want to throw away my tiny little T2115s and get a bigger step tablet.

Katharine

 

[Katharine Thayer]

While we're sharing observations about tonal inversion, here's another one for consideration, since a theory came up earlier in the thread that the cause of tonal inversion is excess density in the negative. A couple weeks ago when I was running some test prints and the printer settings got screwed up, I made some very dense greyscale separations for tricolor gum printing, much denser than my usual greyscale negatives (printed with color inks in both cases), because the printer settings were other than the usual settings.

I printed the cyan separation from this set and got a pronounced tonal inversion; I thought at the time, okay, here's my first observation of tonal inversion as a result of too much density in the negative. But that speculation had to be discarded when I made another set of separations with the density adjusted to more closely match the gum scale and a curve applied, and got tonal inversion with that separation (printed on the same emulsion) as well. Below, the print from the "too dense" negative is on the left (never mind the brown spot from an air bubble), the one from the "not too dense" negative on the right.