Choosing Colours and Curves: The HSL Array

Up until now I've just been using the HSB model to calibrate negative colours. But I've been making some gum prints recently and thought it was time to try out the HSL-Array. This is the colour density gauge that is modeled on the HSL Double Cone. n.b. It's a useful approach because it also shows you the "path" the colour will take as it fades to black in the upward direction.
View attachment 82
I've attached a print out of a test gum print which used red (Quinacridone). The question I want people to answer (or guess at) is: Since I have literally 360 colour paths to choose from on which to base my curve, which would give the gentlest curve? Or would there be the benefits of different approaches?
View attachment 83
My first instinct said use point "A" (120 degrees) which is the "least" dense and is the closest to the top. Point "B" (240 degrees) is the furthest away, uses the most black, and has a longer path to take, but, seems to have a gradual mid-tone area which might give a straighter/gentler curve.

~m

seems like if you choose point B your negative would have the best chance to have full tonal range...hmmm.....? I could be wrong


Miles

Hi Michael,

Can't help about the color (anyway, I would choose point B - instinctly...) but looking at your test print it seems that you've overexposed it. I don't know if this is because of my screen but I can't see any tonal difference below step 6-7... Also, you may try to use a less contrasty sensitizer (= more dichromate) since reds tend to give high contrast due internal filtering effect. (Note that you'll need to further adjust your exposure time with stronger dichromate...) Less contrasty sensitizer is better in hiding printer's ugly dithering pattern + I feel the curves are also less drastic then...

Regards,
Loris.

More Dichromate = Less Contrast

Myles: If I choose "B" I've got a lot a space to "fill" though with a tiny range. I think what I'll do is curve both to see what the difference is. BTW, the horizontal line in the centre represents an RGB step wedge (e.g. PDN or my "down and dirty" RNP-RGB Step Wedge).

Loris: You're right. Red -- has been giving me problems. My range is barely .70 logD. The attached sample was shot at 6 minutes (the one in the sink from last night is 3:30 seconds. I'll post it after breakfast ) To compare this to another colour, I've been shooting my yellow plate at 14 minutes and getting 1.1 logD, but the red was problematic to say the least. What is this "internal filtering" you are speaking about? When speaking about an emulsion with "less" contrast, it has a longer tonal range -- like a grade #1 silver paper, right?
I'm using Potassium Dichromate. So should I "double" the dichromate? Triple?
Thx.

~m

Red will not allow much UV to reach the bottom of the emulsion layer because it filters it very effectively (think of rubylith masking tape - it's red!) and therefore it will increase the contrast of the coating solution. Emulsion w/ less contrast is indeed like a grade #1 paper against grade #3 paper. If you're already using saturated (around 10-15%) Potassium Dichromate you have two options to decrease coating solution's contrast:
a) Decrease the amnt. of pigment (so that less useful light is filtered out)
or
b) Switch to Ammonium Dichromate - which you can mix solutions up to 25-30% (double what you get from Potassium Dichromate) sensivity greatly increases (and contrast greatly decreases) with high strenght AD. My exposure times with a conventional exposure unit (bank of 8 BL tubes) and 25% AD is 4 - 6 minutes with non-oiled plain paper negatives.
You have to fiddle with parameters to find the optimum balance (for your environment/working procedures) between them...

Regards,
Loris.

edit: BTW, switching to 25-30% AD may be a double edged sword since - besides increasing sensivity / decreasing contrast - it will filter more UV light than 15% PD because it's more intensely colored. You have to test for the optimum balance...

mkochsch said:

Loris: You're right. Red -- has been giving me problems. My range is barely .70 logD. The attached sample was shot at 6 minutes (the one in the sink from last night is 3:30 seconds. I'll post it after breakfast ) To compare this to another colour, I've been shooting my yellow plate at 14 minutes and getting 1.1 logD, but the red was problematic to say the least. What is this "internal filtering" you are speaking about? When speaking about an emulsion with "less" contrast, it has a longer tonal range -- like a grade #1 silver paper, right?
I'm using Potassium Dichromate. So should I "double" the dichromate? Triple?

Click to expand...

Ammonium Dichromate Plan "B"

I too am thinking path "B" now...let me explain the reasoning behind this. I've been looking at the rate of change path "B" has. When underexposed "B" starts out near the top in the same range as path "A" (Saturation of 30 and brightness of 100) but as exposure progresses it dramatically races down to the bottom where it eventually flattens out again. The Hue seems to be right on 240 degrees which is also interesting. This is right between 180 and 300 degrees (read: right between the cyan and magenta channels). Coincidence?! The other interesting thing is my yellow Azo pigment does exactly the same thing. So a blocking colour of 240 degrees, 100 Sat. and Brightness value of 40, or in RGB terms 0,0,102 if you can believe that. It may be too rough (if black ink is a factor) for Pt/Pd but I'm going to curve it for gum and see what happens...

View attachment 84

p.s. I'm now getting better separation on my step wedge. Partly this is because I've set the exposure back to where it should be and also upped the dichromate. Thanks for the tip. Potassium seems the same as Ammonium, but Ammonium is three times faster and needs to be sized.

ChartThrob Colours

So just so everyone can visualise how different the colour choices are...

View attachment 85

The one of the left is point "B" and on the right "A".

Thanks for sharing your findings - it's indeed helpful.

At the same strength PD and AD are almost equal in speed; AD will have 16% more dichromate ions per volume when compared to PD. For instance, 10% AD is equal to 11.6% PD - decide yourself if it gives a meaningful difference... But since you can up AD concentration almost 3x compared to PD, you gain speed accordingly.

I use Fabriano Artistico (Traditional White) and don't have to size when using AD for 3 layer prints. Why do you think you need to size?

Regards,
Loris.

mkochsch said:

p.s. I'm now getting better separation on my step wedge. Partly this is because I've set the exposure back to where it should be and also upped the dichromate. Thanks for the tip. Potassium seems the same as Ammonium, but Ammonium is three times faster and needs to be sized.

Click to expand...

Loris Medici said:

Thanks for sharing your findings - it's indeed helpful.

At the same strength PD and AD are almost equal in speed; AD will have 16% more dichromate ions per volume when compared to PD. For instance, 10% AD is equal to 11.6% PD - decide yourself if it gives a meaningful difference... But since you can up AD concentration almost 3x compared to PD, you gain speed accordingly.

I use Fabriano Artistico (Traditional White) and don't have to size when using AD for 3 layer prints. Why do you think you need to size?

Regards,
Loris.

Click to expand...

I/we don't have to size but I find that both PD and AM (AM a little more so) will leave a slight off-white colour in the paper making it look just ever so slightly beige. Which is fine, actually I think it warms up the print a bit, but when matting you have to take this into account and not use a white-white matt but an off-white instead. If you want a cooler/bluer look in the print then I would say size. It should depend on the image really.

~m

mkochsch said:

I/we don't have to size but I find that both PD and AM (AM a little more so) will leave a slight off-white colour in the paper making it look just ever so slightly beige. Which is fine, actually I think it warms up the print a bit, but when matting you have to take this into account and not use a white-white matt but an off-white instead. If you want a cooler/bluer look in the print then I would say size. It should depend on the image really.

Click to expand...

I see, have you tried to clear in sodium or potassium metabisulfite (when the print is finished)? That would cool the warm tint considerably....

Regards,
Loris.

Loris Medici said:

I see, have you tried to clear in sodium or potassium metabisulfite (when the print is finished)? That would cool the warm tint considerably....

Regards,
Loris.

Click to expand...

Nope. I've read that though. I thought it was for bad bad staining....hmm....need to find my wine making supplies I think that's where my metabisulfite is....thx.

~m

Life Below the Equator or Splitting the Third Dimension

So I ran the colour test and got some good news and bad news. First, the bad news. As indicated below I got A LOT of white space at the top of the chart. More than usual on the "B" chart (the blue one) and not as bad on the "A" chart (the green) one. The HSL Array gets you in the neighbourhood lets say, but doesn't always knock on the door for you. Luckily, the good news, printing the ChartThrob 101 Step Chart knocks on the door -- the gradations are much finer -- it pin points the starting block quite accurately. At this point you could just ignore everything I've written below and let the curve fix everything for you. Or, you can do what I did and go back to the original coloured charts and measure the colour of the block where density starts. You have to print a second set of charts but at least you know you're starting at right density --- but read on....

View attachment 86
On the left:
"B" -- Hue 240, Saturation 34, Brightness 66. (RGB112,112,169)
On the right:
"A" -- Hue 120, Saturation 53, Brightness 100. (RGB 120,255,120)

Above are the returned values. At first the results of "B" really baffled me, "how could I be so far off?" Then I went back and measured the colour where density actually started -- What!? I was no longer on the "skin" of the RGB double cone, I was inside the cone. Check it out: HSB :240,34,66 (RGB 112,112,169) is IN not on the cone. The reason: the HSL Array (most density wedges in fact) measures density on the outside only -- remember the shape of the model -- two ice cream cones stuck big end to big end.
View attachment 89
When you pick a colour on or above the equator (green line) as the negative reduces the saturation it follows on the skin until it reaches white, BUT if you're below the equator (the blue line) it takes the shortest path to white short-cutting through the model! (Remember no matter what colour you pick it always moves toward the north as it desaturates.)
Imagine it this way...you just landed on the planet HSL-Array in the Chaos system and you've landed at the North pole of this planet, where everything's white (the south pole is completely black BTW), and you're carrying a laser pointer so you can signal the ship. It's a very strange planet in that it is the shape of two cones stuck together and it's also translucent gelatin in the core. You decide to go exploring and start walking toward the equator. Once and a while you point your laser back at your comrades waiting at the space ship, they lift their glasses to you and smile back. The laser is always going straight back to the ship through the planet's atmosphere up to the point that you start walking south of the equator. Lucky for you though the planet is translucent so even though you're now standing 23.5 degrees south of the equator on the tropic of Capricorn (i.e. colour "B") your friends can still see you and your laser signal because it's able to cut right through the gelatin (unhardened by glyoxal, an in joke) like substance and still puts a shiny red dot on your space ship.
OK, so you're asking why don't a see a little bit of colour at 0 per cent on the ChartThrob grid? The answer is, sometimes you will and sometimes you won't depending on if you over or underestimate which HSL value you start with. If you did manage to catch a piece of colour it might fade out to white and then fade back in again before it started its progression towards white. It kind of makes me think the core of the model is denser than the skin all things being equal, hard to say without plotting out more points.

So what have we learned. One, using colour at or above the equator (i.e. "A") of the model probably gets you closer to where you want to be, initially. Two, while path "B" is perfectly valid option what you see on the HSL Array doesn't really reflect the gradation you will end up with because you're no longer on the surface.
I counted about 31 discernable steps in the Chartthrob test of "A" and about the same in "B" so as far as one colour giving you an advantage in tonality over the other, I don't see it. To me, the obvious choice, if it is available, is "A". In fact maybe a wedge based on the hue of 120 degrees would be useful in this case. A single green with fine gradations which runs from the equator to the North pole. Ho ho ho!

~m

Attachement

Here's the result's they seem to get lost on the last post?!
View attachment 91

Miscellaneous questions and comments

Michael, this is very interesting, thanks for the tests and explanations. I'll probably have more questions as I get back into running some more tests myself.

As Loris said, it's always hard to judge from jpegs, but to my eye, from my screen, this color looks a little wimpy for quinacridone red (PR 209). I need to go back and redo my color patches (or start over with your new arrays) for PR 209, because as I mentioned at the time I ran all those color patches on different pigments, the PR 209 was inadvertently underpigmented because I hadn't realized I was out of PR209 until it was too late. So I need to do that color again to be sure, but my guess at the time was that it came out different from all the other colors not because it was red, (other reds behaved the same as all the other pigments) but because it was undersaturated in comparison to the other pigments. But I do need to do that test again with a properly saturated PR209 to be sure. Your statement that your Azo yellow behaves the same way as your PR 209 seems to provide some support for the idea that pigments behave more similarly than differently.

As to your dilemma about sizing, there are two different kinds of stain. The kind you're getting, dichromate stain, tends to be unrelated to sizing (or if anything, is related in the opposite way: some combinations of sizing and paper tend to be more prone to dichromate stain than the paper by itself without the size); if you're getting dichromate stain, sizing isn't likely to help. But as Loris suggested, even light dichromate stain will be removed by a quick rinse in bisulfite after the print has dried, and I'd recommend that, unless you really do like the warm tone and want to keep it as a picture element.
Katharine

wow Micheal you have got to get outside and photograph and breath some fresh air...no more STARTREK for about 2 weeks...ok?....


No, just kidding...great explanation...I suppose I do enjoy learning from someone who must be a generation or two away... .


Miles

Questions

Michael, did you try printing without a color imposed? I guess, having a statistical/research methods background, I need to see something compared to nothing, to see what difference it makes. The reason I'm asking, is that your chartthrob charts with the colors applied, either of them, don't seem an improvement over the chartthrob charts I made last fall, without any color applied, at the same time I was being told that one MUST be applying a blocking color in order to be doing the digital negative thing "properly."

All my test prints are still in boxes from a move of two weeks ago, so what I can show you is something I scanned last fall to send to someone who wrote me privately, at the time ChartThrob was posted, to ask me "What 'should' the initial ChartThrob chart look like, in gum?" I wrote back and said, "I don't know what they 'should' look like, but here's what mine look like," and grabbed a fresh one to scan.

This happens to be PV 19 gamma (quinacridone rose) and it answers a question for me. Looking at your PR 209 in isolation, it seemed so pinkish that I was wondering if maybe you were using Daniel Smith's "quinacridone red" which isn't PR209 at all but PV 19, but now with the PV 19 next to it, I can see that the hue of your sample is right for PR 209 (which has a slight yellow cast in comparison to PV19, which is very blue) but just much less saturated than I'm used to printing it.

I'm afraid my questions will seem annoying to someone to whom it all seems so obvious, but it's still not obvious to me what the relationship is between the blocking color and the density range, especially since this blocking color seems to vary depending on the inkjet printer used. For example, it seems like most gum printers use a green color, but that doesn't work for my printer at all; my best blocking color (at least based on the 2-color patches you sent me last fall) is R 255 G 0 B 64, or something close to that). If there is a relationship between the blocking color and the density range, then why wouldn't the blocking color be the same for people using a similar emulsion, regardless of the printer? (and as I reported last fall, I found the same blocking color worked for all the pigments I tried).

I keep asking "What IS the relationship between the blocking color and the density range?" and I get vague answers like... well, it's like using a filter for black and white film.

Well, okay, but what's the relationship? When I ask "What's the relationship?" I mean, I suppose, a mathematical equation. Something like: Blocking color in wavelengths= density range in log units, that would show how the blocking color actually relates to the density range. So far I'm not grasping what this relationship is, and until I do, nothing about this color thing is going to make sense to me. I mean, I can see that with the red color imposed on the negative, I get a better gradation of tones all the way through the 100-step patch of ChartThrob, but I need to understand the principle that rules the observation. Thanks to anyone who can explain this in a way that would help me understand it.

Katharine

Printing without a colour imposed, by that you mean using Black ink, which needless to say is a combination of colours. Yes this is quite possible. In fact with certain driver combinations on the Epson I've come very close to pegging the density (on the top row) depending on the process. I think it was Ware's Cyano (1.8-2.0 logD) which worked for me.
If you want to read more about light/filter "attenuation" read the link below. This will give you a taste for the physical mechanics behind the phenomena. You can read the math:

http://cord.org/cm/leot/course01_mod06/mod01-06frame.htm

But the short answer is: more blocking equals less density on paper. There's tons of high school science projects on the web that show this using photosyntheses and growth in plants in place of density. Those experiments show remarkably similar results to results expressed here -- green-yellow-red are the big UV blockers/filters causing plant to grow slower....

Yes my gum tests are a little inconsistent at the moment....I started out using a measured line on my pipette for adding my components and I found that the results varied quite a bit so now I am counting drops: 10 drops dichromate, 10 drops Gum, 10 drops pigment, 10 drops water. Also, that "smoothing" phase where I'm brushing in the mixture seems to be another variable because there's a "blotting" factor (absorption into the smoothing hake brush) which means sometimes it's sucking up more or less of my emulsion causing variances in the final density. It's a more complicated balance than some other processes...so still getting my coating technique is one of my problems.

~m

p.s. Miles I haven't watched TV seriously since they pulled the plug on "Studio 60" and prior to that "Six Feet Under".

Well, no, I didn't mean black ink necessarily; I don't print with black ink when I print greyscale negatives; I print them in color. But I take your answer to mean that an uncolored greyscale negative (or an unsaturated RGB neg) can sometimes work well for a particular density range, which is what I've always said.

I read the link you pointed to, and I understand the math, but it doesn't answer the question I'm asking. There's nothing that relates blocking wavelength to density range, which is the relationship I'm trying to understand. In reading the different websites for people working in this area, and following the discussions here and elsewhere, I gather that the blocking color for a particular density range is different for different printers, and even that very different colors can work well for the same density range for the same printer, and there are other considerations (smoothness of gradation, distance of path) that enter into the selection of one color over another.

So I'm left to conclude that even though people are always invoking a relationship between blocking color and density range, there really isn't such an identifiable relationship. It's more a tautology: the color that works is the color that works for any individual printer and practitioner and emulsion, but there's no rule that governs the relationship between the color that works for me and the color that works for you and a particular density range, or more generally, no rule that governs the wavelengths of blocking colors to the density ranges of emulsions. That's not very satisfying to me, but I suppose it must be that the different inks behave so differently, that the opacity of a particular ink to UV has little or no relationship to the color of the ink.

Ah, I think maybe I just finally answered my own question. Thank you for your patience,
Katharine

Katharine Thayer said:

I suppose it must be that the different inks behave so differently, that the opacity of a particular ink to UV has little or no relationship to the color of the ink.

Click to expand...

Funny, I was smarter a couple of weeks ago when I was explaining to Doyle that color and opacity of inks/pigments vary independently, and that different inkjet inks and different pigments used for gum printing behave idiosyncratically, so that the fact that a particular yellow inkjet ink is very opaque to visible light shouldn't be used as the basis of a rule that says yellow pigments in gum printing will be opaque. . In other words, the opacity of a particular ink or pigment has little or no relationship to the color of the ink or pigment. But for some reason I never thought of translating that knowledge to this other situation: obviously the same "principle" of independence of variables applies here as well. I don't know why that was so hard for me to grasp, but it all makes sense to me now, thanks....
kt

mkochsch said:

....I started out using a measured line on my pipette for adding my components and I found that the results varied quite a bit so now I am counting drops: 10 drops dichromate, 10 drops Gum, 10 drops pigment, 10 drops water.

Click to expand...

Then I'd say it's the added water that's diluting your color so much. Just out of curiosity, why do you add so much water? When the ambient humidity is so low that it makes coating difficult, then it helps to add a little water to the gum/pigment/dichromate to help the coating go on more smoothly, but that's a lot of water you're adding, and it can't help but dilute the pigment.
kt

mkochsch said:

Printing without a colour imposed, by that you mean using Black ink, which needless to say is a combination of colours. Yes this is quite possible. In fact with certain driver combinations on the Epson I've come very close to pegging the density (on the top row) depending on the process. I think it was Ware's Cyano (1.8-2.0 logD) which worked for me.
If you want to read more about light/filter "attenuation" read the link below. This will give you a taste for the physical mechanics behind the phenomena. You can read the math:

http://cord.org/cm/leot/course01_mod06/mod01-06frame.htm

But the short answer is: more blocking equals less density on paper. There's tons of high school science projects on the web that show this using photosyntheses and growth in plants in place of density. Those experiments show remarkably similar results to results expressed here -- green-yellow-red are the big UV blockers/filters causing plant to grow slower....

Yes my gum tests are a little inconsistent at the moment....I started out using a measured line on my pipette for adding my components and I found that the results varied quite a bit so now I am counting drops: 10 drops dichromate, 10 drops Gum, 10 drops pigment, 10 drops water. Also, that "smoothing" phase where I'm brushing in the mixture seems to be another variable because there's a "blotting" factor (absorption into the smoothing hake brush) which means sometimes it's sucking up more or less of my emulsion causing variances in the final density. It's a more complicated balance than some other processes...so still getting my coating technique is one of my problems.

~m

p.s. Miles I haven't watched TV seriously since they pulled the plug on "Studio 60" and prior to that "Six Feet Under".

Click to expand...

Michael,

I really know what you mean... ,


Miles

Michael,

If your purpose is to test the color array would it not make sense to use a process that will give more consistent results than gum? Figuring out what is going on by combining data from the array and gum printing might drive you crazy, even if you watch nothing on television but the weather channel.

This is not meant to knock gum, only to point out that widely variable results can be predicted unless you go to great lengths to control conditions and materials.

Sandy King

sanking said:

Michael,

If your purpose is to test the color array would it not make sense to use a process that will give more consistent results than gum? Figuring out what is going on by combining data from the array and gum printing might drive you crazy, even if you watch nothing on television but the weather channel.

This is not meant to knock gum, only to point out that widely variable results can be predicted unless you go to great lengths to control conditions and materials.

Sandy King

Click to expand...

You are sooo right. Up until recently I hadn't use the HSL Array at all. I was interested in an emulsion which required very little density so I could 'play' with the top half of the Array. It was a mistake to choose gum when I still have to get my process under control (variables like 'spray' developing are a huge factor). Silver with a #5 filter or, Classic Cyano might be more appropriate. I haven't been in the studio for a couple of days (busy seeding) but I've got some time coming up this week again...I probably would have put together the whole notion of the 'colour equator' at some point, so it did lead to that realisation.

~m

Katharine Thayer said:

I keep asking "What IS the relationship between the blocking color and the density range?"
Katharine

Click to expand...

I think I'm closer to bridging the gap between your thinking and mine...

Blocking colour of the negative has no effect on the 'density range' of the emulsion. The emulsion is the emulsion, its range is its range. DR changes depend on changes to the chemistry, application and media applied.

The use of a particular blocking colour is for the sake of the negative and used to find an equal to the emulsion's highlight property that's all. e.g. Emulsion XYZed prints highlight at density 1.05, ink combo R255, G0, B64 also prints UV highlight at 1.05....therefore they match. The technique of choosing a colour to so that you don't make negative better suited to pt/pd than gum.

I don't know if it would make sense to settle on a blocking colour which was at the density limit of gum and then just use exposure as the main variable to print your highlight or not. Dichromate, gum or pigment could all be 'main variables' I suppose.

~m

mkochsch said:

I think I'm closer to bridging the gap between your thinking and mine...

Blocking colour of the negative has no effect on the 'density range' of the emulsion. The emulsion is the emulsion, its range is its range. DR changes depend on changes to the chemistry, application and media applied.

The use of a particular blocking colour is for the sake of the negative and used to find an equal to the emulsion's highlight property that's all. e.g. Emulsion XYZed prints highlight at density 1.05, ink combo R255, G0, B64 also prints UV highlight at 1.05....therefore they match. The technique of choosing a colour to so that you don't make negative better suited to pt/pd than gum.

I don't know if it would make sense to settle on a blocking colour which was at the density limit of gum and then just use exposure as the main variable to print your highlight or not. Dichromate, gum or pigment could all be 'main variables' I suppose.

~m

Click to expand...

Hmm... I actually had thought the gap had closed entirely when I had my little revelation that the ability of a particular combination of inks to block UV has little or no relationship to the color the inks produce, but this post widens it again for me.

I certainly wasn't suggesting that I expected the blocking color to have any affect on the DR of the emulsion, which as you rightly say is what it is. So apparently you're still not understanding my earlier puzzlement, which as far as I'm concerned is completely resolved and off the table.

What I was wrongly inferring from what you were saying about the blocking color being related to the density range was that there must be some actual relationship between blocking color (wavelength of the color) and the density range of the emulsion, and I was trying to understand exactly what that relationship is. But as I said before, I see now that there is no such mathematical relationship. It's a tautology; the blocking color that "matches" empirically, as you say, IS the blocking color, but there's no rule that says that in general, across printers, say, a red blocking color matches a particular density range, and a blue blocking color matches a different particular density range, and so forth. Different printers yield different blocking colors for the same density range, and the same printer may even yield several very different blocking colors that all seem to match the same density range equally well, and the choice of ink color is based on other considerations. So I'm content, as I've said, to accept that there's no general relationship between blocking color and density range, only a one-to-one individual match that has to be determined empirically by trial and error for each separate emulsion, and by extension, that the opacity of a particular combination of inks to UV has little or no relationship to the visible color that the inks produce. That makes complete sense to me, and resolves my earlier confusion.

But what you're saying above makes very little sense to me; especially the last paragraph. Are you saying that you don't know whether it makes sense to use a colorized negative matched to the density range for a gum emulsion? If so, why wouldn't it? If not, perhaps a clarification?

Yes, all those variables can be varied to change the behavior of a gum emulsion, and that's how I've controlled my gum printing for decades, by varying printing variables rather than varying the negative. But I'm not arguing for, nor am I arguing against, using any particular method of control in gum printing; as I'm always saying, there are many roads to the kingdom.

I don't think we're having an argument; at least I'm not having an argument; I'm just trying to understand this stuff. I know gum, that's all I know, and now I'm trying to understand how to translate what I know about gum to this different way of thinking about printing gum, and to see if it can be useful to me. But it can't be useful to me if I don't understand it. Thanks,
Katharine