The Making of a Dye-Transfer

[logan2z]

Interesting video on the making of Dye-Transfer prints by the couple making the prints for the exhibition 'William Eggleston: The Last Dyes' at David Zwirner Gallery in LA.

 

[Milpool]

Interesting video on the making of Dye-Transfer prints by the couple making the prints for the exhibition 'William Eggleston: The Last Dyes' at David Zwirner Gallery in LA.

Thanks for posting this. Charles Cramer also has some good primer stuff on his website (although he switched to inkjet quite a while ago).

It was a very involved process and I suspect relatively few people were very good at it. It was just so impractical unless you were regularly spending all day on it.

 

[Pieter12]

Thanks. I was impressed by the show, now even more by the process. I knew the basics of dye-transfer printing, but had never witnessed it. I was taken back a little by the amount of dust swirling around the top of the enlarger, though. Obviously, it made no difference, the prints are pristine.

 

[Sharktooth]

Thanks for posting this, it was very interesting. What surprised me most was that the separation negatives were much smaller than the final matrix images. This means that there are two enlargement steps, the first from 35mm to about 8x10, and then from 8x10 to the large matrix image for the final paper transfer.

I assumed that the matrix sheets would be a low sensitivity film that would need a very bright light source, and not suitable for exposure with an enlarger. If that's the case, then the enlarger light source must have a very special light source indeed.

 

[Pieter12]

Thanks for posting this, it was very interesting. What surprised me most was that the separation negatives were much smaller than the final matrix images. This means that there are two enlargement steps, the first from 35mm to about 8x10, and then from 8x10 to the large matrix image for the final paper transfer.

I assumed that the matrix sheets would be a low sensitivity film that would need a very bright light source, and not suitable for exposure with an enlarger. If that's the case, then the enlarger light source must have a very special light source indeed.

I think he said it was pulsed xenon. I am still astonished after seeing the show that the originals are 35mm, the prints are stunning.

 

[Pieter12]

The follow-up video is quite interesting, too:

 

[DREW WILEY]

Milpool - DT was THE primary method for high quality color printing for decades, and hundreds of people had to be competent at it. In the bigger operations, there were assembly-lines, with different workers specializing in specific tasks of the overall workflow. There were mulitiple high-volume manufacturers of the needed supplies : Eastman Kodak, Color Corp of America, the Technicolor Corp, the US Army, plus overseas, Agfa at one time, I believe. Add to that many amateur as well as serious pro individual printers. Now I'd be surprised if there are even twenty people still active worldwide. Only a single commercial lab still exists, or rather, about one and a half labs, since one of them is inevitably going to drift into retirement pretty soon.

Sharktooth - one of the advantages of dye transfer is that you can expose the necessary matrix film using an ordinary enlarger (not quite ordinary, because you need a precision registered film carrier as well as vac easel). It can be done by contact too; but it doesn't require UV exposure or any great amount of light. Matrix film certainly exposed way faster than Cibachrome, for example.

The current linked Eggleston show was done in a very simple lab with a garden variety ordinary lamp enlarger, except for the tweaks I just mentioned, pretty much an antique Beseler enlarger in fact. If a pulsed xenon colorhead was optionally used, it would have been the low power amateur-style flashtube Beseler/Minolta one, with separably operable RGB channels. That would make sense for at least generating the color separations and their masks from a color slide via enlargement. The next step of enlarging the color separations themselves onto matrix film doesn't require any tricolor light option or special filtration; that medium is essentially blue-sensitive or slightly orthochromatic, and maybe around ASA 2 or so in terms of speed. It doesn't need a particularly bright light.

More serious sequential RGB pulsed xenon heads were once sometimes used for direct enlarging onto offset printing plates, and ran so hot that they ideally needed water cooling jackets, and would probably blind you if you looked into one. Those massive enlargers were built like tanks and cost about a hundred thousand dollars - whole different animal.

 

[Sharktooth]

Sharktooth - one of the advantages of dye transfer is that you can expose the necessary matrix film using an ordinary enlarger (not quite ordinary, because you need a precision registered film carrier as well as vac easel). It can be done by contact too; but it doesn't require UV exposure or any great amount of light. Matrix film certainly exposed way faster than Cibachrome, for example.

That's interesting. I'm assuming that the matrix film must be silver halide based to have enough sensitivity. If so, how is the gelatin hardening/softening achieved? Would it be possible to make homemade matrix film, since it no longer seems to be commercially available?

 

[Andrew O'Neill]

Very similar to Carbon Transfer. I own a couple of Ctein's dye transfer prints.

 

[Patrick Robert James]

I'm surprised they didn't punch the paper.

The Eggleston video cracks me up. He is the FIGJAM of photography.

 

[gordrob]

That's interesting. I'm assuming that the matrix film must be silver halide based to have enough sensitivity. If so, how is the gelatin hardening/softening achieved? Would it be possible to make homemade matrix film, since it no longer seems to be commercially available?

Pretty much all you want to know about the Dye Transfer Process can be found here including how to coat your own matrix film. There is also a lot of information from Charles Cramer , David Doubley, Bob Pace and Ctein. I use a 200 or 300 watt bulb in a Durst L138 enlarger to expose my matrices so a light source like a pulsed xenon is not needed.

http://www.dyetransfer.org/Site/Dye_Transfer_Resources_files/DyeTran.pdf

 

[koraks]

I'm assuming that the matrix film must be silver halide based to have enough sensitivity. If so, how is the gelatin hardening/softening achieved? Would it be possible to make homemade matrix film, since it no longer seems to be commercially available?

Yes, it's a silver halide film. The hardening is used by using a tanning developer; it's basically a pyro developer as is used by many photographers, but for other reasons.
And yes, it's technically possible to DIY a matrix film. Here's pretty much all you need to know: https://www.processreversal.org/public/text/Browning_dye_transfer_materials.pdf (which is a mirror of the pdf linked to by @gordrob above.)
It requires a considerable commitment in time and materials to produce something remotely usable. It's a nice project if you're capable of focusing on one thing, and one thing only for a year or two, minimum.

At that point, you'll still have to learn the actual dye transfer process and the masking involved, of course. Another challenge would be to get hold of the dyes, or source alternative dyes with similarly attractive properties to the dyes that were originally used. There's also the matter of the receiving paper, which needs to have a mordant incorporated for the dyes to stick (chemically) in the target emulsion. This may or may not be feasible by soaking an existing gelatin-coated medium in a mordant solution. I think the original receiving paper simply had the mordant incorporated into the gelatin layer that was coated on it. That would be the alternative route, but obviously more challenging and laborious.

At a conceptual level, I see a couple of alternative routes that could use the dye imbibition principle and that are more feasible to do at home. I've done a tiny bit of experimentation in that direction and the result was somewhat promising, but I've not yet pursued it further.

As to the light source: in principle, any white-light/broad-spectrum light source can be used, but if you're going to build something today, LED is the obvious choice to do the separations.

 

[Andrew O'Neill]

I'll stick with tri-colour gum

 

[DREW WILEY]

Andrew - If I lose too many more teeth, I'll be gumming it too.

Per DT - obtaining dyes is the easy part. Kodak offered distinct kits with their own dye buffer sets; but that can easily be DIY concocted too. With the DT process, you need a pH meter anyway. Making your own matrix film would be highly involved (as evident in Jim Browning's example), though it is feasible. Having matrix film coated in volume for you would require the interest of some wealthy entrepreneur or public grant with a philanthropic heart, willing to lose a lot of money on the project.

It would be difficult to replicate Kodak's pre-mordanted receiver paper, which was dual mordanted, including a mildly radioactive agent (thorium nitrate). But fixed-out silver gelatin paper can be DIY mordanted shortly before usage. The whole process is fairly complex, and needs blocks of time. I just couldn't find enough time to get sufficiently deep into the process, though it is quite interesting and compelling.
It's also becoming particularly expensive because you need a lot of sheet film per image, which is obviously getting quite pricey too.

You also need a full set of precisely matched punch and registration equipment. Also doable if you have machine shop skills, or simply by ordering up a fresh set from a graphics house, or finding something used or surplus.

Light sources for color separations are also easy. Any enlarger head emitting "white light" can be employed in conjunction with separate
R,G,B glass or gel filters under the lens. You don't need a programmable additive colorhead, although I do use one of those for color separation as well as masking work. Dye transfer and related kinds of prints were even commercially made for decades before true additive heads even existed.

I'd be a bit cautious, however, about LED output, because you'd basically be pioneering an untested route where all your masking and separation protocol would have no specific track record in previous literature as per aim points. The one seriously remaining DT lab does it completely differently, and generates digital color separations for sake of a wholly different kind of customized matrix film matched to blue laser exposure, somewhat simplifying the overall process. So various hybrid options are indeed possible, in that case, dependent upon keeping multiples of those old scanners, lasers, and computers still running by cannibalizing parts from spares if necessary. Not simple.

My own experimental tweak was more a modernized reversion to the original Eastman Washoff-Relief Process, the direct predecessor of Kodak Dye Transfer, but still kept alive parallel to that by a number of practitioners, including some large commercial labs. It's easier to control, being less rushed per the development steps, making even drum development realistic, and can use common developers like HC110 or DK50, rather than the expensive regular formula.

 

[Sharktooth]

Thanks for the info, folks. As Drew says, the matrix film doesn't look to be something the average Joe is going to be able to do themselves. Out of interest, does the gelatin in the matrix get transferred to the paper, or is it just the dye that gets transferred? If it's just the dye, then can the matrix be reused to make multiple prints?

It seems that with all these digital imaging advances, creating a large printed color image is still very difficult and expensive to achieve for a DIY'er. Inkjet works for large images, but the printers and inks are too expensive for occasional home use. The same applies for RA4 type photo images. It really only makes sense for a commercial operation where there is enough volume to cover capital outlay.

There are several alt processes that can be done by a DIY'er for B&W, but there's a quantum leap to use those for color images, which again only makes sense in a commercial process.

Digital processing of scanned film images creates so much potential, but it's that final step to get a printed color image that continues to remain just out of reach.

 

[Samu]

Unfortunately, buying of potassium dichromate is illegal here in Lithuania, unless you possess a government-issued license for handling toxic chemicals. These regulations are getting a true nuisance for anybody interested in alternative processes.

 

[koraks]

Out of interest, does the gelatin in the matrix get transferred to the paper, or is it just the dye that gets transferred? If it's just the dye, then can the matrix be reused to make multiple prints?

Just the dye, and yes, the matrix can be reused many times.

The same applies for RA4 type photo images.

RA4 is very accessible to a home user, actually. But it's more challenging if you want to make really big prints.

Unfortunately, buying of potassium dichromate is illegal here in Lithuania, unless you possess a government-issued license for handling toxic chemicals. These regulations are getting a true nuisance for anybody interested in alternative processes.

Dichromate is not part of the classic dye transfer process. I'm also not aware of any practitioners of a similar dye imbibition process that involves dichromate, although it's definitely conceivable. And even then, there are alternatives; most notably DAS. So it's not really a valid concern. The real challenges in dye transfer are in other areas.

 

[DREW WILEY]

Sharktooth - Yes, matrix film is silver based, with a very thick emulsion, and only the dye is transferred, and that's why it's called, the dye transfer process. It's based on a pH differential between the dyes and the paper, plus the mordant in the paper to anchor the dyes. Slight changes in pH are used alter the saturation of the dyes, along with many other potential tricks to achieve exactly what one wants. It's an extremely malleable medium, with many possible alternate routes.

That being said, spending ten years or so just learning to master it, while expending a great deal of time and money (let alone having to revive the materials themselves in this day and age) is pretty hard to justify when just about anyone can learn to make decent inkjet prints in a matter of days, provided they have suitable scans or direct digital capture. Chromogenic RA4 printing tends to have a different look, unless one has seriously mastered that itself; but at least its basic steps are easy to learn too.

I still have enough DT supplies for five years of fooling around perhaps. But I realistically have to give priority to what I already know how to do quite well instead, in terms of both color and black and white printing, while I still can. I did obviously have the option to take up carbon printing instead of DT, which is far more straightforward in terms of acquiring needed materials at least; but I was rather paranoid about all
the dichromate involved - I've known people who got awfully sick due to it.

A small amount of ammonium dichromate is used in certain DT workflows; but there are also workarounds to it. I tested both methods, and found no difference in the final outcome. What bothers me more with DT itself is all the acetic acid vapor one is exposed to. Even with an excellent fume hood, the cumulative amount of constant exposure to even 1 or 2 percent solutions is still potentially quite irritating to the respiratory system over the long run. You've got tray after tray of that involved, and not just a single very weak solution like ordinary black and white stop bath requires. Enormous amounts of distilled water are also needed; DT is a very sensitive process.

 

[koraks]

just about anyone can learn to make decent inkjet prints in a matter of days

Hell no!

 

[DREW WILEY]

Millions of people do it, Koraks. I didn't state they all necessarily know how to do it well ! I stated that in relation to their own expectations of a "decent print". A few days of practice, and they very well might achieve prints as good as what a street corner photofinisher provides. It might not be yours or mine expectation. In fact, the best inkjet printers I know often spend an entire week fiddling around with the files for a single given image before they get what they want. But inkjet isn't a quantum leap from what basic home and office printers do; and most people these days can easily reorient to it. That's why its gotten so ubiquitous.

Even with black and white printing, a person can attend a darkroom class and learn the basic steps in a single afternoon, yet need the rest of their lifetime to progressively refine their vision and technique. I learned the basics of Cibachrome in a single afternoon, but then spent the next three decades fine tuning it to the highest level.

A person can learn oil painting almost instantly with a paint by numbers kit, and be proud enough of themselves to frame it. But that doesn't make them Rembrandt.

 

[IB Photochemistry]

Milpool - DT was THE primary method for high quality color printing for decades, and hundreds of people had to be competent at it. In the bigger operations, there were assembly-lines, with different workers specializing in specific tasks of the overall workflow. There were mulitiple high-volume manufacturers of the needed supplies : Eastman Kodak, Color Corp of America, the Technicolor Corp, the US Army, plus overseas, Agfa at one time, I believe. Add to that many amateur as well as serious pro individual printers. Now I'd be surprised if there are even twenty people still active worldwide. Only a single commercial lab still exists, or rather, about one and a half labs, since one of them is inevitably going to drift into retirement pretty soon.

Sharktooth - one of the advantages of dye transfer is that you can expose the necessary matrix film using an ordinary enlarger (not quite ordinary, because you need a precision registered film carrier as well as vac easel). It can be done by contact too; but it doesn't require UV exposure or any great amount of light. Matrix film certainly exposed way faster than Cibachrome, for example.

It should be noted that Kodak Dye TransferTM (DT) process was a highly proprietary Kodak process. Kodak only published very general information and a few other publications and books contain very basic information, that would not allow one to be successful for making prints for highly demanding commercial clients. If I understand correctly, the Dye Transfer training program at Kodak was run by the Eastman Kodak Marketing Education Center. This would be where Kodak customers and their sponsored affiliates would be trained in using the system.

The Dye Transfer has some unusual and unique properties, not found in current imaging processes. It is very capable of producing a very long accurate neutral grayscale, but also much more vivid colors compared to conventional color chromogenic, and silver dye bleach processes like Ilfochrome, and even the popular digital pigment inkjet printers used now. The color gamut is very large comparable to an earlier iteration (E2, or E3) Ektachome transparency. Using the Kodak dyes one can produce a Dmax of 3.0 on a baryta semigloss/gloss print without color contamination, and giving excellent tonal separation. While it was used in the industry for five decades, it was improved slightly over time. It was used because mainly it gave better control compared to other processes of the time. DT's could be retouched easily which was a huge advantage. Contrast and color control, and the ability to combine multiple images were also highly desirable to the printing industry.

There were more companies than those few, who made dye imbibition transfer materials. I'm not sure what you mean by “high volume manufacturers”; volumes of this kind of material varied, but Eastman Kodak officially claimed it was the lowest volume product they ever made. I have reason to doubt this however. Technicolor had matrix and blank stock made by Eastman Kodak since the early 1950's. I have heard the DOD and US Army had a use for DT. I have heard somewhere the DOD had matrix and paper coated but I don't know much about this. Please let us know your source of this information, I would be very interested in what you know about this. I have even heard DT was used in other branches of the DOD as well, perhaps for aerial reconnaissance imaging, but a lot of that may still be classified.

There are about 4 or 5 commercial DT or dye imbibition transfer operations in the world now. I can name them, but not all of them will print for random people off the street. One, who I will not name, will claim he is busy for several years or insist on cutting corners with quality (fixed out paper, etc.). The $1000 for the first print isn’t enough of a motivation for him, since he has absolutely no passion or interest for the sake of the craft, yet his published material can be found online.

Yes, Guy Stricherz does make dyes now; he just no longer has enough Kodak matrix film for making new matrices. There are probably at least 100 people in the US and Europe with old stock Kodak, DTC, or Efke matrix film that can theoretically print DT or a variant of it.

Drew, I don't know your point about one of the labs “inevitably going to drift into retirement pretty soon”. Are you trying to imply that interest in DT is simply a function of mass commercialization, since the mainstream use in commercial advertising is functionally obsolete, therefore all interest should follow a decline? I take issue with your statement, that is why I'm decided to comment on some of your recent posts here. There are others of a younger generation, who are interested in DT or variants of if for making fine art color photographic prints and in continuing its use.

 

[IB Photochemistry]

The current linked Eggleston show was done in a very simple lab with a garden variety ordinary lamp enlarger, except for the tweaks I just mentioned, pretty much an antique Beseler enlarger in fact. If a pulsed xenon colorhead was optionally used, it would have been the low power amateur-style flashtube Beseler/Minolta one, with separably operable RGB channels. That would make sense for at least generating the color separations and their masks from a color slide via enlargement. The next step of enlarging the color separations themselves onto matrix film doesn't require any tricolor light option or special filtration; that medium is essentially blue-sensitive or slightly orthochromatic, and maybe around ASA 2 or so in terms of speed. It doesn't need a particularly bright light.

More serious sequential RGB pulsed xenon heads were once sometimes used for direct enlarging onto offset printing plates, and ran so hot that they ideally needed water cooling jackets, and would probably blind you if you looked into one. Those massive enlargers were built like tanks and cost about a hundred thousand dollars - whole different animal.

I believe Guy is using a flash tube light source on the L184 condenser matrix enlarger. It might be made by Macbeth or NuArc. I think they ranged in power output from 1.0-6.0 kW.

Xenon flash lamp systems were generally preferred by professional labs for exposing matrix film, and some custom Cibachrome printers for better tonal separation. The reason is multiple masks are exposed different amounts of time on the same piece of matrix film. Kodak matrix film has a small reciprocity failure when using long exposure times. Xenon flash lamp illumination can offset this reciprocity failure, making the mask exposures and curve shape more linear.

 

[IB Photochemistry]

Pretty much all you want to know about the Dye Transfer Process can be found here including how to coat your own matrix film. There is also a lot of information from Charles Cramer , David Doubley, Bob Pace and Ctein. I use a 200 or 300 watt bulb in a Durst L138 enlarger to expose my matrices so a light source like a pulsed xenon is not needed.

http://www.dyetransfer.org/Site/Dye_Transfer_Resources_files/DyeTran.pdf

Old instruction manuals and books were not very useful and were greatly inadequate. I've read a number of publications on the subject and haven't found anything well written on DT. This was a highly proprietary process where even sensitometric data on the materials were not published by the manufacturer but known to some through other Kodak channels. I have found a few masters thesis' that explore DT but nothing very detailed on what was actually being used in the industry.

We are planning to publish a comprehensive method of making separation negatives from chromogenic color transparencies, complete with the proper equations and theory. It should help you, Drew and others interested in making excellent DT prints. Without the proper knowledge one will not make DT's of very high quality. This was a largely considered a trade secret and selfishly guarded. This DT secrecy continues to this very day. THIS WILL END SOON! I am not well liked by most of these people, because I see a need to share these methods to create significantly more interest in this process. Unfortunately I probably won't be able to post most of my information here, since this forum may not permit this.

 

[IB Photochemistry]

Andrew - If I lose too many more teeth, I'll be gumming it too.

Per DT - obtaining dyes is the easy part. Kodak offered distinct kits with their own dye buffer sets; but that can easily be DIY concocted too. With the DT process, you need a pH meter anyway. Making your own matrix film would be highly involved (as evident in Jim Browning's example), though it is feasible. Having matrix film coated in volume for you would require the interest of some wealthy entrepreneur or public grant with a philanthropic heart, willing to lose a lot of money on the project.

Obtaining which dyes are easy? Finding good dyes is difficult and requires a lot of research and chemical synthesis. The dyes in that digital technophile's manual, can be found in numerous patents and publications, I can name them, and are not as good as the original Kodak dyes. The Acid Blue 45 is the same cyan dye that was used for the kit they sold. Assuming you already have Kodak dyes then they will work fairly well with the Kodak paper. The dyes were far from the best that was available. We have found some dyes of superior characteristics that I am planning on disclosing at a future time. Ideally one should use desalted dyes, which behave more consistently and are replenishable.

As far as resurrecting it, it can certainly be done. I am aware that materials for this type of process have been recently manufactured in early 2020 for Bettina and Egbert Haneke in Germany. They supposedly had matrix and transfer paper coated at the plant in Switzerland previously used for coating Ilfochrome, at great cost to them. This material isn't available to the public, and they really consider this their own trade secret process and discourage anyone from doing this. A small independent manufacturer certainly can produce dye imbibition materials, but the attitude of secrecy needs to change on this. They will likely not respond to inquires about this, and I'm not sure if there is anyone who knows about their materials will speak in public about them.

 

[IB Photochemistry]

It would be difficult to replicate Kodak's pre-mordanted receiver paper, which was dual mordanted, including a mildly radioactive agent (thorium nitrate). But fixed-out silver gelatin paper can be DIY mordanted shortly before usage. The whole process is fairly complex, and needs blocks of time. I just couldn't find enough time to get sufficiently deep into the process, though it is quite interesting and compelling.
It's also becoming particularly expensive because you need a lot of sheet film per image, which is obviously getting quite pricey too.

No its not difficult to replicate a good mordanted receiver paper. Numerous suitable organic mordants exist and can be used. Printparency process used Chitosan as a mordant. The Printparency (the CCA “dye transfer” process) process had a fiber based receiver paper using the same mordant.