Making Commercial Color Separation Negatives of Transparencies for the Kodak Dye Transfer Process

DREW WILEY said:

Hopefully, your masking contribution with come up with something interesting and maybe new. So far it hasn't. But I'm willing to listen.

Click to expand...

This is a compilation of some of the techniques that were used in the industry. While it isn't new this was known among Kodak affiliated operators of the process. Very few people outside of the industry were aware of these methods. I am certainly interested in learning about this, as I was totally unaware of some of these techniques just a few years ago. Hopefully others will be interested to experiment with these techniques themselves, using old materials or materials currently manufactured.

Here are some curves I made myself of Separation Negative Type I film showing the self-masking effect a few people asked about. An IT-8 Q60 Ektachrome color transparency tablet was used to make the test exposures onto the Kodak Separation Negative Type I (4131) cut to 4x5 sheet film. This has step tablets of pure cyan, magenta, yellow, and neutral at the END point. When this test is performed with Super XX Pan or any other pictorial film the derivatives of the curves (-gamma in this case) all show a value < 0, indicating an unwanted contrast of the dye spectral sidebands. No other exposures were used in these tests. These were both developed by hand in trays of DK-50 with a water pre-soak, but without any surfactants added. Some uneven development was observed and this is due to the rapid speed at which Separation Negative film develops. The neutral curves on both graphs diverge and is probably due to exposure being off slightly.

On the first curve on top a Wratten #92 red filter with a spectral cutoff at around 625 nm, was used used to make a red light exposure with the IT-8 Q 60 tablet contacted with the Separation Negative Type I (4131) film. The red curve is the main absorption band of the cyan dye. The green and blue curves, are yellow and magenta dyes, and the gray and black, are the 24 step neutral scale, and the 12 step END neutral scale, respectively. Here red has a gamma of 0.63, green gives a gamma of -0.04, and yellow gamma of -0.05 (smaller R2). These were all plotted from row A to row L on each column and re-checked multiple times. The gray curve is showing is some non-uniformity development since it doesn't align with the other curves. This step tablet runs at the bottom of the transparency and is 24 steps long.

On the bottom curve a Wratten #93 green filter with a maximum transmission of 545 nm, was used used to make a green light exposure with the IT-8 Q 60 tablet contacted with the Separation Negative Type I (4131) film. The green curve gives a gamma of 0.69, and the blue curve is showing a gamma on 0.4 with a very noisy result, likely due to development uniformity error. The red curve does not show any loss of contrast of the magenta sideband of the cyan, indicating a masking effect on the yellow dye, magenta unwanted absorption band only.

Curves of the blue light exposure through several blue filters do not show any masking effects, and all three absorption bands track as one would expect.

On other tests I have performed using filters of wider spectral bandwidth, the masking effect appears less strong but noticeable. A Wratten #92 red filter is too strong an effect for this film, but I'm posting it here to illustrate this effect. This is also interesting since the masking can be controlled by filter selection.

Those are such sketchy and abbreviated curves that it's hard to tell what is really going on. How many actual points did you read for each respective curve?

The advantage of old school translucent Kodak plotting sheets is that you could overlay them on a lightbox to see how various curves compared, interchangeably.

Filters such as 92 and 93 were intended for primitive color densitometer cutoff reading usage, not for actual color separations onto film.

I got almost identical results from my 61, 29, and 47B set as from my direct narrow band RGB additive color head channels. But I never tested Color Sep Film Type 1. That's been gone a long time. Among the films I did test were Super-XX, TMX100, TMY400, and FP4. TMX was the most cooperative; but the faster speed of TMY would make it the better candidate for in-camera separations.

Those are such sketchy and abbreviated curves that it's hard to tell what is really going on. How many actual points did you read for each respective curve?

Click to expand...

The curves didn't display very well and were originally of different dimensions. I don't know how I can modify that post now. I probably should have removed the linear curves generated by the algorithm, as it is a bit confusing. They were made on LibreOffice from 12 data points for each cyan, magenta, yellow, and neutral trace. One of the curves is composed of 24 data points. Possibly I can re-post the red light curves so the 2 top sideband curves are more visible.

An R2 of over 0.99 is a very good fit to a linear function, and the derivative of the linear curve is just the multiplier on the x variable. Seeing how the linear regression algorithm plots were very close to the measured curves, shows the additive and proportional property of the sideband absorptions.

Filters such as 92 and 93 were intended for primitive color densitometer cutoff reading usage, not for actual color separations onto film.

Click to expand...

No, primitive color densitometers from the 1940's used Wratten filters 29, 61, and 47B. The more narrow band color densitometry filters came later, probably when color transmission densitometry was standardized to ISO or ASTM standards, with Status A and M.

The Wratten #92 is a longpass gelatin filter with a stopband around 620 nm (I think), and has a narrower spectrum passband than the Wratten # 29.

The Wratten #93 is a gelatin bandpass filter with a maximum transmission of 545 nm, and is aligned with the magenta dye absorption peak. Wratten #61 has a peak transmission around 532nm and is broader. Green 61 filter also leaks a lot of red light.

There is no reason these can't be used for making separation negatives. Wratten #93 and #94 (blue at 450 nm) are also darker and would require more exposure time. For Kodak Separation Negative film, the #92 red gives too much correction, but the green #93 might be useable.

Even in the 1950's some commercial labs weren’t using the Wratten # 61 and 47B, as they don't align well spectrally with most types of Ektachromes. According to someone who worked in Evans Color Lab in 1957, an engineer was hired to test a new set of filters with their separation process that they were going to use. Ed Evans (deceased) was always very secretive about this and locked the separation filters they used in a safe. They were probably colored glass filters from Schott or Corning, since dichroic filters at the time were inferior. Corning made a number of special pigmented glass optical filters, now mostly replaced by bandpass dichroic filters.

I got almost identical results from my 61, 29, and 47B set as from my direct narrow band RGB additive color head channels. But I never tested Color Sep Film Type 1. That's been gone a long time. Among the films I did test were Super-XX, TMX100, TMY400, and FP4. TMX was the most cooperative; but the faster speed of TMY would make it the better candidate for in-camera separations.

Click to expand...

I would like to see your curves of cyan, magenta, and yellow, with their unwanted absorption sidebands. Also how did you process the Super XX pan? What kind of step wedge did you use?

IB Photochemistry said:

On the bottom curve a Wratten #93 green filter with a maximum transmission of 545 nm, was used used to make a green light exposure with the IT-8 Q 60 tablet contacted with the Separation Negative Type I (4131) film. The green curve gives a gamma of 0.69, and the blue curve is showing a gamma on 0.4 with a very noisy result, likely due to development uniformity error. The red curve does not show any loss of contrast of the magenta sideband of the cyan, indicating a masking effect on the yellow dye, magenta unwanted absorption band only.

Click to expand...

It should read a gamma of about 0.004 but very noisy. From the curve it appears almost flat. Using any other pictorial film this value (yellow's magenta sideband) will have a gamma of about approximately 20% of the main gamma of green.

I don't have much time to respond today. Early Ekta 64 suffered from red contamination in the green, which is why a special masking protocol was involved.

Step wedges potentially involve their own bias if they're old and yellowed. I used the latest Stouffer type which is clear based.

Yes 61 Green Wratten passes a tiny bit of orange and red. But these filters vary from brand to brand; and I prefer coated glass anyway. I've found a coated German one and might try that instead of the Wratten and Tiffen.

What derailed me is how you referred to Sep Film as "self-masking". Well, it's certainly not that, or anything different in that respect. What seems to be going on is that you (or someone whose notes you've adopted) managed to find a sweet spot in its spectral sensitivity and gamma nexus, "as if" it was doing something analogous to a mask in relation to just certain specific dyes. But change either the dye or the film and all bets are off. So this is really more like archaeology than R&D going forward, since that particular film is no longer made.

That's just like how I found the "sweet spot" in TMX separation exposures allowing me to develop all three together for the same amount of time and get overlapping curves. But that took some repetitious work to arrive at.

DREW WILEY said:

What derailed me is how you referred to Sep Film as "self-masking". Well, it's certainly not that, or anything different in that respect. What seems to be going on is that you (or someone whose notes you've adopted) managed to find a sweet spot in its spectral sensitivity and gamma nexus, "as if" it was doing something analogous to a mask in relation to just certain specific dyes. But change either the dye or the film and all bets are off. So this is really more like archaeology than R&D going forward, since that particular film is no longer made.

Click to expand...

No, Drew I did not find a “sweet spot” this is analogous to negative spectral sensitivity. The magenta dye's red absorbing spectral sideband, is producing a negative value of gamma, on the graph above, where the main gamma curve is positive. That is a kind of self-masking effect.

I don't want to get into a highly technical discussion of silver halide spectral sensitivity. From a purely sensitometic point of view a red separation for example, will primarily record the cyan dye peak absorption (at about 650-660 nm), producing the greatest contrast. However it will also record the small amount of red absorption in the magenta dye which is about 8-10% of the cyan dye's contrast, because the magenta dye's spectral absorption curve absorbs some red light as well. This unwanted absorption reduces light exposure to the separation film (whatever film one is using, Tmax, Super XX, Sep. Neg. etc.) where magenta appears on the transparency, so according to basic sensitometry less exposure should be produced on the film where there is magenta on a red light exposure. Films responding only to the light exposure intensity will only produce a positive gamma for all of the three curves on the red separation. The self-masking effect appears to “sense” the magenta spectral sideband from the Ektachrome magenta in the red light exposure and add rather than remove density, a rather curious and interesting effect.

If people still have old stock of Kodak Separation Negative film, my suggestion is to use that and not try to reinvent something since it works well for this purpose. In fact using an isolation mask on the red separation negative is very important. My Separation Negative film was made in 1995. Agfa made a color separation film called P330P until 1998-2000, for separating color records on transparencies.

You might be correct with respect of the viewpoint of the dye itself, so to speak - the "net effect". But referring to the film itself as self-masking is misleading. Switch the exposing filter or the dye, or the film itself, and all bets are off.

Spectral sensitivities of film also change with reciprocity failure. That's something which has plagued DT workers.

Who on earth is going to have a stockpile of Sep Neg Film? Anything that old is likely to frill off the emulsion if one is not extremely careful, unless they stockpiled and froze the last of it. At least Efke matrix film appears to still be perfectly viable if frozen. I have about 90 sheets of that left (20X24). Don't know if I'll ever had the time to actually use it, however.

Today I was simply making some black and white 8X10 internegs from master color duplicates, for sake of ordinary silver printing. That was one my own specialties - precision master dupes and internegs with quite a few masking controls already built in. But that itself is no longer feasible. By far the best duping sheet film ever was Astia 100F. Portra 160 is an excellent interneg film, provided contrast masking is also used; but that won't help DT prospects, and has now itself gotten absurdly expensive in 8X10.

You might be correct with respect of the viewpoint of the dye itself, so to speak - the "net effect". But referring to the film itself as self-masking is misleading. Switch the exposing filter or the dye, or the film itself, and all bets are off.

Click to expand...

I'm not sure what you mean here. The dye is made by oxidized CD-3 color developer reacting with various couplers. Dyes for Ektachrome, Fujichrome, and Agfachrome [no longer available] are very similar structurally, with very close spectral properties.

For any separation negative made from a color transparency the silver image contrast is a function of, the spectral composition of the exposing light source, the transmission of the isolation filters, the spectral density of the dyes in the transparency itself, spectral sensitivity of the separation film itself, and any wavelength dependent scattering effects. Commercial labs used a dye based step tablet made on the film they intend on separating. Silver based step tablets are intended for white light sensitometry only. Using these for through color filters can produce erroneous results.

Spectral sensitivities of film also change with reciprocity failure. That's something which has plagued DT workers.

Click to expand...

Most commercial DT labs used Xenon flash lamp powered enlarger heads, so matrix reciprocity failure wasn't a problem. Individuals using low powered enlarger heads often ran into reciprocity failure and latent image stability issues, since most didn't know what they were doing very well.

Who on earth is going to have a stockpile of Sep Neg Film? Anything that old is likely to frill off the emulsion if one is not extremely careful, unless they stockpiled and froze the last of it. At least Efke matrix film appears to still be perfectly viable if frozen. I have about 90 sheets of that left (20X24). Don't know if I'll ever had the time to actually use it, however.

Click to expand...

It turns up on eBay occasionally. I have used various expiry dates without any emulsion problems. I believe it was frozen.

Have you tested your Efke matrix recently? It doesn't store as well as the old Kodak matrix film.

Today I was simply making some black and white 8X10 internegs from master color duplicates, for sake of ordinary silver printing. That was one my own specialties - precision master dupes and internegs with quite a few masking controls already built in. But that itself is no longer feasible. By far the best duping sheet film ever was Astia 100F. Portra 160 is an excellent interneg film, provided contrast masking is also used; but that won't help DT prospects, and has now itself gotten absurdly expensive in 8X10.

Click to expand...

Have you used an IT8 Q60 tablet or other color transparency target to compare Astia 100F vs. Kodak and Fuji transparency duplication films? I know you have said on this site many times that Astia 100F is a better duplication film, however it was never designed as a duplication film, it was a pictorial film. Duplication and internegative films have internal isolation masking to partially correct some unwanted spectral absorptions, from E6 transparencies.

There are color isolation masking procedures for making far more accurate color duplicate transparencies, using transparency duplication films, that were known in the industry and selfishly guarded. This would be where a nearly identical duplicate was needed for some reason that was visually indistinguishable from the original. It closely relates to making separation negatives.

Interestingly all chromogenic color materials, E6 films, RA-4 papers, color reversal papers, Ilfochrome, and color negatives have internal color masking for correction of their own dyes. For example, Ektachrome uses 6 internal linear color correction masks to maintain an Equivalent Neutral Density at 5000 K in addition to other types of inter-image and inter-spacial effects.

When making separations, the dye responses for all the various categories of chrome films differ significantly. Little resemblance between old Ektachrome and new. With Fuji, there are significant differences between the three major categories of Velvia, Provia, and Astia. Then there was Kodachrome, and all kinds of Agfachromes, including pre E6.

Big secrets with the dupe trade? Ha! I could make far better duplicates any day of the week using off the shelf Astia. 100F was the best. And I have tested multiple generations of Fuji CDU and And Kodak EDupe. CDU was basically just tungsten-balanced Astia to begin with. Sure, if you were going to make a big quantity of 35mm slide dupes involving flashing instead of true supplementary masking, then the dedicated slide duplicating product had its commercial advantages, but optimal repro quality wasn't one of them.

Labs simply didn't have time to go to that level of fuss; and nobody would pay the price if they did. I never duped slides, although I sometimes made enlarged 35mm or 120 film dupes onto sheet film. More often, it was sheet film contact dupes, mostly 8x10.

I made master chrome targets for each specific type of chrome film I routinely shot, plus master color negs in that case. Those are valuable regardless of the print medium involved, whether chromogenic RA4 paper, Ciba, DT, etc.

Pulsed xenon sequential RGB enlargers were an uncommon offset printing trade item used for plate exposures. They were expensive, ran hot, and the bulbs themselves were very expensive. I don't see any reason why they would appeal to dye printers in relation to matrix film, which exposes relatively fast and doesn't need anything that powerful.

DREW WILEY said:

When making separations, the dye responses for all the various categories of chrome films differ significantly. Little resemblance between old Ektachrome and new. With Fuji, there are significant differences between the three major categories of Velvia, Provia, and Astia. Then there was Kodachrome, and all kinds of Agfachromes, including pre E6.

Big secrets with the dupe trade? Ha! I could make far better duplicates any day of the week using off the shelf Astia. 100F was the best. And I have tested multiple generations of Fuji CDU and And Kodak EDupe. CDU was basically just tungsten-balanced Astia to begin with. Sure, if you were going to make a big quantity of 35mm slide dupes involving flashing instead of true supplementary masking, then the dedicated slide duplicating product had its commercial advantages, but optimal repro quality wasn't one of them.

Click to expand...

Looking at the spectral properties of the dyes themselves they really are not that different. Most of the difference in appearance in new vs. old E6 Ektachrome for example, is due to the inter-image effects, not the dyes spectral properties.

There was a former commercial lab operator who would make color corrected duplications, using transparency duplication films with special masks. He has been retired for a while now, but wont disclose exactly how he did it. He says they were far superior to using the duplication films by themselves. I don't think this was offered very much as this kind of custom duplication would be much more expensive than a traditional duplication process. The only thing he can say is that using 3X3 matrices (linear algebraic matrices) of the dyes spectral components, one can calculate the masking needed. I have a pretty good idea of how it was done.

Matrix algebra of 3X3 matrices are also used to calculate the masking and separation gammas used in making DT color separation negatives, masks, and matrix films. A source that covers some of the basics of matrix algebra in making of color separations is: The Reproduction of Colour (6th Edition), R.W.G. Hunt, Print ISBN: 9780470024256, online ISBN: 9780470024270, chapter 28 "Practical Masking in Making Separations", p.504-518. This source doesn't have a lot of examples nor does it go into how one would make DT separation negatives, but it is a good source for those looking to learn more about this topic.

I think Frog Prince used at least 1 isolation mask to get better colorimetric reproduction when they used Super XX pan, and 3 isolation masks when they used separation film (built in isolation masking). Some labs used all 6 isolation color correction masks, to get a far better match to the E.N.D. point of the transparency. Its possible to match a transparency to another transparency or a transparency to print (possible in some cases with DT but cyan is a weak point), within the Luther-Ives condition, but not the original scene to a transparency.

I strongly doubt that Drew can use Astia 100F to obtain better results than a duplication film without extensive masking. The published spectral sensitivity curves for Fuji CDU II and Fuji Astia 100F are not identical in curve shape. Usually tungsten sensitivity will just make the red about roughly 1/3 as sensitive as the blue region and roughly 2/3 for green sensitive region. Its sensitivity for red and green is linearly scaled in relation to the blackbody spectral distribution curve. Take a look at the blue sensitivity region on the graph of CDU II, other pictorial films do not have that curve shape. That is a masking function that partially corrects some of the dyes in the transparency being copied.

I would like to see curves of the cyan, magenta, and yellow with the spectral sidebands plotted for both the original transparency and the duplicate transparency. My strong suspicion is that colorimetrically there is no improvement using a pictorial film, in what one would achieve from using a duplication film.

DREW WILEY said:

Pulsed xenon sequential RGB enlargers were an uncommon offset printing trade item used for plate exposures. They were expensive, ran hot, and the bulbs themselves were very expensive. I don't see any reason why they would appeal to dye printers in relation to matrix film, which exposes relatively fast and doesn't need anything that powerful.

Click to expand...

I'm not asking your opinion on Xenon heads; I'm trying to bring some clarity to this topic from my own research into this. Xenon heads were commonly used to expose Kodak Matrix film in many commercial labs, such as CVI, Frog Prince, Berkey K&L, Pace Color, etc. The reason for this is its ability to maintain reciprocity better, when large numbers of matrix exposures were needed. Sometimes over 100 sheets of film were used to make a set of mats, with multiple images and graphics.

Thanks for the reply. But if you find anyone anywhere who could make a cleaner dupe from a chrome than I could, that would be an interesting research project! Yes, expert supplemental masking was involved. That gave me much better control than flashing, or than any commercial dupes I've ever seen. Take it the other direction, and CDU with correct daylight filtration would come out looking just about the same as the same generation Astia itself when used as a taking film. In fact, that's what I sold off my reserve of 8X10 CDU for. EDupe was more disappointing because it sold so slowly that it nearly always already had highlight crossover issues.

Some dupe films had restructured curves; that's all. There was no "self-masking" involved or any deep secrets requiring James Bond to infiltrate Eastman Kodak. Astia had the most latitude of any regular chrome film, and really didn't need more wiggle room. It also had excellent recip characteristics for a chrome film. I've done third-generation dupes with it, with no shift in color or saturation. I trained color matching pros, so know a thing or two about critical evaluation. The huge advantages of 100F over all the CDU series was the distinctly finer grain, along with a dimensionally stable PET base (I just had to re-punch on old CDU 8X10 dupe this morning because the inferior triacetate base had shrunken over a sixteenth inch over the past decade).

Back to Color Sep Film Type I. The only advantage it seemed to have was in its extended red sensitivity which allowed all the orange mask to be separated from red itself with an extreme no. 70 cutoff filter. That would only apply to rather uncommon interpositive separations from CN film, or hypothetically to Pan Matrix film work; but the only person I know who professionally worked with that used an ordinary 25 Red.

I've seen a fair amount of the "big lab" production work. Rarely as good as the more dedicated small DT shop work,
or "hired guns" as I called em. They charged more of course, and limited their clientele to just a few select individuals.
If someone asked me to make a precision color interneg or dupe for them, the minimum charge would have been $500 apiece, and a lot more if multiple masks were involved.

Photo Comp work is more an offset printing trade thing, where xenon units might have been around anyway for other applications. Regular Halogen enlargers were far more common. Either way, matrix film exposes way way faster than heavily masked Ciba work. The separations themselves would have been made either ordinary halogen etc or with an image setter if by contact. Xenon heat and light would be brutal on filters. Gels would fade fast, and dichroic filters shift its spectral sensitivity with excessive heat. I've been down that route.

Repeatability with halogen colorheads is easily accommodated with a built-in shutter, like the later Durst 8x10 units had.
Or one could simply synch a Compur electronic shutter for the lens itself; easy enough.

I enjoy discussing some of these things in relation to your own viewpoint. But I must admit, the more I do, the more it becomes apparent that you're looking at things awfully narrowly. There were all kinds of alternative routes to DT and wash-off relief procedure.

But good luck to you. It seems to be more an antiquarian project than any new pathway forward.

DREW WILEY said:

Frog Prince was just one lab among many, and a rather small one at that. There were different methods. And of course, all the films in question are no longer made. And exposure data points aren't of much value unless someone else has exactly the same outdated equipment.

You are completely wrong about old versus new films. Better separations as well as contrast masks can be made any day of the week using a modern film like TMX100. No need for Super-XX or Pan Masking Film. The question nowadays is more about the very high cost of sheet film of all types.

All DT workers had a hard time with specular highlights. One strategy was to just go ahead and overprint the highlights with some kind of inevitable color tinge, then selectively bleach that away afterwards.

Jim Browning is an electrical engineer who holds patents on the Chromira printer system. He designed his own registered 8x10 film recorder, and it's apparently unique. Jim isn't hoarding anything. He'd be willing to make separations for others; but the whole problem is that these need to be tailored very specifically to the end user, who would naturally do certain things differently than he does, and would have a different registration setup anyway. He can hypothetically also output digital camera files onto sheet film. Up till recently he used 8X10 TMX100, but due to its increasing cost, switched over to Delta 100 once he figured out how to tweak its files.

Using D100 traditional darkroom style would be a lot more challenging; FP4 would be a much better alternative. Andy Cross has been using FP4 for separations for a long time. HP5 would be relatively worthless for that application. Fomapan pseudo-200 would be nightmare; it does have a long straight line, but poor batch to batch quality control, and can't be developed to a high gamma. But I did test it once just out of curiosity.

I have no interest in writing a separation manual. The practical demand is zero. One needs to orient their separations to their specific output media, anyway; and a lot of people like gum printers just aren't that picky.

The last pan film used for highlight masks was Tech Pan. There was a Euro substitute for that which was liquidated about 20 yrs ago, but had questionable quality control, perhaps from being out of date. There are workarounds for that issue too. For highlight sparkle in just certain areas, all it takes is a fine point Sharpie pen on a registered sheet of frosted mylar; I've done that many times for Cibachromes. Image setter film is down to residuals; very few people still use it.

But yeah, I've seen 16 sheet DT results that weren't any better than just 3 sheet separation results. Just depends on the methodology. Everyone seems to have had their own tricks. The process would be horribly expensive to revive at this point.

The ideal would be to completely re-invent the matrix film itself for sake of a longer straight line. That would solve a lot of problems, but isn't going to happen either.

Click to expand...

What advantage is this method assuming it was still readily available?

DREW WILEY said:

Thanks for the reply. But if you find anyone anywhere who could make a cleaner dupe from a chrome than I could, that would be an interesting research project! Yes, expert supplemental masking was involved. That gave me much better control than flashing, or than any commercial dupes I've ever seen. Take it the other direction, and CDU with correct daylight filtration would come out looking just about the same as the same generation Astia itself when used as a taking film. In fact, that's what I sold off my reserve of 8X10 CDU for. EDupe was more disappointing because it sold so slowly that it nearly always already had highlight crossover issues.

Click to expand...

Sensitometrically they are not the same. Those of two very different emulsions. So far you haven't provided any empirical data that supports your conclusions about Astia 100F vs. CDU II vs. EDupe. Saying one looks the same is qualitative, what I'm interested in is what sensitometric and colorimetric data do you have to support your conclusions?

Some dupe films had restructured curves; that's all. There was no "self-masking" involved or any deep secrets requiring James Bond to infiltrate Eastman Kodak. Astia had the most latitude of any regular chrome film, and really didn't need more wiggle room. It also had excellent recip characteristics for a chrome film. I've done third-generation dupes with it, with no shift in color or saturation. I trained color matching pros, so know a thing or two about critical evaluation. The huge advantages of 100F over all the CDU series was the distinctly finer grain, along with a dimensionally stable PET base (I just had to re-punch on old CDU 8X10 dupe this morning because the inferior triacetate base had shrunken over a sixteenth inch over the past decade).

Click to expand...

How do you define self-masking? My definition is the same as used in peer review papers describing this type of internal color correction that is performed inside the photo sensitive material itself. So yes THERE IS INTERNAL MASKING IN SEPARATION FILM, COLOR TRANSPARENCY FILM, COLOR NEGATIVE PAPERS, and OTHER CHROMOGENIC MATERIALS! This is well-known in the literature. I can cite papers on it. They do use the term “masking”, “mask”, and sometime “self-masking”.

DREW WILEY said:

Back to Color Sep Film Type I. The only advantage it seemed to have was in its extended red sensitivity which allowed all the orange mask to be separated from red itself with an extreme no. 70 cutoff filter. That would only apply to rather uncommon interpositive separations from CN film, or hypothetically to Pan Matrix film work; but the only person I know who professionally worked with that used an ordinary 25 Red.

Click to expand...

It was widely used to separate color transparencies, not color negatives. Nearly every DT lab with a roller transport processor used it. And it can be repeated no professional lab used Tmax 100, although some claimed to, what they actually did was totally different. I'm not interested in what these individuals with little to no sincerity claim, I'm interested in what they actually did. Some small commercial labs used Super XX pan, with the extra isolation hold back correction masks, because it was easier to process in trays. Super XX Pan doesn't have extended red sensitivity either.

Separation Negative Type I and Type II can be used for blue and green separations of color negatives. It does not have extended red sensitivity! David Doubley published a spectral sensitivity curve of Separation Neg. Type II, and you can see it mostly cuts off at 600 nm. He obtained this from someone who worked at Kodak, since this curve was never published by Kodak. I have measured the sensitivity of the Sep. Neg. type I material myself and can't good exposure above 600 nm. Most C-41 color negatives use an indoaniline dye formed by oxidized CD-4 color developer, that has an absorption peak around approximately 690-710 nm. On some films the sideband of the cyan dye goes in the 600 nm range, but this creates a problem with sep. neg. film, since most color negative films do not and the density may be too low.

Technical Pan can be used for the red light separation to make a gamma 1 interpositve of a color negative, since it does HAVE EXTENDED RED SENSITIVITY. Yes, I'm aware of the E-81N Kodak disinformation paper. Processing to a gamma of 1.5 is not easy to control and there is no discussion of masking of the printing dyes. Kodak decided to omit discussion of masking when printing color negatives and wanted people to believe that the orange mask in the color negative, somehow corrected the dyes you were printing to as well.

Pan Matrix does have extended red sensitivity and uses a type of masking function. It designed to be flashed otherwise there will be halation effects with insufficient color correction with the Kodak dyes. So far I have never found a publication mentioned this, and it was very selfishly, shamelessly and jealously guarded. I will have more information on this soon. I believe Lachlan Young asked about this a few years ago on here. Its not a Super XX pan type emulsion.

DREW WILEY said:

Photo Comp work is more an offset printing trade thing, where xenon units might have been around anyway for other applications. Regular Halogen enlargers were far more common. Either way, matrix film exposes way way faster than heavily masked Ciba work. The separations themselves would have been made either ordinary halogen etc or with an image setter if by contact. Xenon heat and light would be brutal on filters. Gels would fade fast, and dichroic filters shift its spectral sensitivity with excessive heat. I've been down that route.

Click to expand...

Xenon heads were used without color filters for exposing mat film. Dichroic filters made in the last 40-50 years or so can have good attenuation and remain withing tolerance up to 100 C. In fact I have used dichroic filters on Xenon laboratory arc lamp housings without problems, although most of them were colored glass long pass filters.

So it is possible some additive lamphouse color enlarger heads used Xenon flash lamps or Xenon arc lamps. Those are mainly of interest for Ilfochrome printing or other kinds of graphic arts separation processes. For making enlarged separations an incandescent light head was normally used.

DREW WILEY said:

I enjoy discussing some of these things in relation to your own viewpoint. But I must admit, the more I do, the more it becomes apparent that you're looking at things awfully narrowly. There were all kinds of alternative routes to DT and wash-off relief procedure.

But good luck to you. It seems to be more an antiquarian project than any new pathway forward.

Click to expand...

I'm interested in what can make the best DT/dye imbibition prints. My criteria is they need to meet or exceed the results of color processes currently available. Using the kind of masks I discussed it should be possible to beat most digital inkjets except for cyan, where Kodak dyes are the weakest. Its great if others want to try other things. Since the system that was used by commercial labs is proprietary it can't be replicated unless you know exactly how it was designed. Separation making for DT was not known outside the industry, and was a trade secret. There are a number of people interested in cutting corners on this, by omitting sensitometry, using inferior dyes, using films not designed for the purpose, etc. While some of this might lead to useful results, my problem with it is people are unaware of what was used in the industry.

Alan Edward Klein said:

What advantage is this method assuming it was still readily available?

Click to expand...

I'm assuming you mean using multiple types of masks. Its possible to match the transparency to the print very well using DT. These array of masks extend the useable color space of the process, and allow for control of highlights and shadows.

I just can't put up much anymore with all this "trade secret" talk. Different practitioners developed somewhat different methodologies, and in commercial cases, some of it was privately held. But figuring out how to make separation negatives was known all along, and was inherent to any kind of tricolor printing. As film options evolved, so did specific details. I personally encountered zero "secrets" in that respect. But one does have to tailor their separations and masks to specific usage.

Much about what's "best" in DT prints depends on the subjective controls during the actual matrix stage. There's so much flexibility there that the individual themselves, and their own level of experience, is really the key. Some worked with basic dye kits, some kept a large selection of dyes on hand. Masking itself is more of a huge tool box of potential options rather then just a single fixed set of techniques. And now there are many hybrid options too.

Xenon was useless for Ciba printing. There was an amateur xenon flashtube colorhead marketed by Beseler way too weak for serious Ciba work. Very powerful halogen units were built by Durst, Devere, and under the ZBE Starlight brand. Excessive heat was their achilles heel. That's why I designed my own unique 8x10 narrow-band additive colorhead, which ran way cooler, but had somewhat fussy electronics; it's also huge and heavy. The key is to make the bandwidth ideal for the specific print medium itself.

Now that I mostly print color RA4, that particular monster of an enlarger is sometimes just too much over the top in terms of contrast and saturation, and sheer lumens; so I tend to use an ordinary 8X10 commercial Durst CMY rig instead. I have a separate RGB enlarger for making precision dupes, internegs, separation, and masks, and well as smaller format work.

If DT is ever to catch on again, the way forward might be bypassing the rising cost of sheet film itself and going direct from digital capture files to output onto the matrix film. That's already being done on a limited scale by some.

IB - Unless you have something of substance to contribute, rather than just more accusations of people hiding this or that, or casting aspersion on their own expertise, count me out from this point onward.

DREW WILEY said:

I just can't put up much anymore with all this "trade secret" talk. Different practitioners developed somewhat different methodologies, and in commercial cases, some of it was privately held. But figuring out how to make separation negatives was known all along, and was inherent to any kind of tricolor printing. As film options evolved, so did specific details. I personally encountered zero "secrets" in that respect. But one does have to tailor their separations and masks to specific usage.

Click to expand...

My reasoning behind this is Kodak had a training program for professional dye transfer operators, that regularly covered topics like the different types of highlight masks, isolation hold back color correction masking, exposure factors, use of matrix algebra to determine masking gammas, the actual Equivalent Neutral Density aim-points for the dyes, etc. There were certainly variations in techniques from lab to lab, but they were working from information that was regarded as proprietary by Kodak. This was not easy to derive these methods unless someone really knew the underlying phenomena of film masking, color science, physics, chemistry, and optics. That qualifies it as a trade secret. The commercial success of DT was reliant on trade secrets. Given that most professional lab operators didn't have the time nor interest to research how to make separations beyond what exists in published literature, I believe the public has a right to know about this arrangement, since it was designed to monopolize the commercial use of DT by Kodak affiliated labs only.

DREW WILEY said:

Much about what's "best" in DT prints depends on the subjective controls during the actual matrix stage. There's so much flexibility there that the individual themselves, and their own level of experience, is really the key. Some worked with basic dye kits, some kept a large selection of dyes on hand. Masking itself is more of a huge tool box of potential options rather then just a single fixed set of techniques. And now there are many hybrid options too.

Click to expand...

What I meant by “best” in DT prints, is how well one can fit the gamut of a transparency into the print gamut. Its possible to even make improvements on the transparency itself in the separation making process. It offers a medium that is unique, in that a large amount of color detail can be placed on the print medium itself. And numerous creative controls and possibilities exist. Its important to note that commercial labs were not artists, so they rarely entertained many of the various highly creative uses of it. In fact many commercial lab operators would laugh at anyone claiming photography is an art form or especially Dye Transfer. At least Guy Stricherz (deceased 1948-2025) would acknowledge this, which is a step in the right direction.

I didn't publish the method to make specular highlight masks yet, this is due to my information on this area being incomplete. I do have a few more details about this type of mask. From my notes it appears that the specular highlight mask receives 0.5 stop less exposure than the highlight masks do (N - 0.50), assuming they are made properly. The masks are developed in D-11 for 2 ½ minutes at 20 C. This is usually made on Kodalith Ortho (or other high contrast ortho film still available) as a white light exposure, or possibly on panchromatic lithographic film, but I don't have any specific examples of this.

At least one specular highlight mask is needed for a set. There were different approaches to this. A highlight mask is essentially an exposure on the toe of the high contrast film to raise the contrast of the toe of the negative being printed. The specular mask just records the specular highlights, so its a smaller portion of the toe curve on the high contrast masking/litho film. Some have made specular highlight masks from normal highlight masks using a Farmers reducer solution. I don't know if this method was ever used in commercial DT separations.

The super specular highlight mask is a high density specular highlight mask. I believe there were two ways of making this. One method I believe, is to take the specular highlight mask and make a gamma 1 positive of it then over exposing it to another negative, thereby increasing the density of the image. Another method is to duplicate the specular mask onto a black and white reversal duplicating film by underexposing it.

Some commercial DT labs used super specular masks, to get pure white highlights. I'm not sure how common this was. I know Taratro Color Lab used this type of mask regularly.

Interesting. I'd describe that as, How to make a simple task overtly complicated, but can understand its premise. Otherwise, I've often fine tuned various masks with dilute Farmer's Reducer. That also helps to clear any residual fog, which can be quite a problem with Ortho Litho style films.