HandMade Photographic Images

Dye Transfer Materials and Processes

By James Browning

My name is Jim Browning, and I am a fine-art photographic printer, and consultant in the imaging business. I got interested in large format photography in the mid eighties in a big way. I moved my engineering career toward imaging technology as a result of my passion for photography. In 1990, I decided to leave Silicon Valley, and try my hand at printing in my native state of New Hampshire.

My interest in Dye Transfer has always been in the back of my mind, since my Uncle Don and Aunt Helen Browning had a dye lab from the forties through the eighties. I saw some great prints, Yousef Karsh portraits of the Pope, and the British royal family, and famous celebrities of the day, as well as prints of architectural images by Scott D'Arazian. The people in these prints always seemed alive, and not flat like most photographic prints of the time.

So, it was with the idea that I could make really great prints myself, and perhaps improve on the venerable old Dye Transfer process. Initially, I setup a lab to do both Dye Transfer and Cibachrome printing. In the early part of this decade, computers were still relatively expensive, and processing huge files was not something I could afford. I decided to build a laser scanner / film recorder for use in making correction masks for printing Ciba prints. I built the system, which I used to scan the original transparency, edit the image, and write out a contrast controlling mask onto pan masking film. My original intent was to extend this to a color correcting mask on transparency film, but I never started this phase of the project. I got quite good at printing the Cibas, and the laser masking system worked very well. In 199 5, the new faster Power Macs became available, so I decided to move up to high resolution imaging. I developed the equipment to scan and print onto 8x10" negative film at relatively high resolution, about 150mb. The results are really good, I was able to print images I was never able to master before, using Live Picture was a godsend to my ability to print.

I was still unhappy with the materials, either type C, or Ilfochrome (Cibachrome). My frustration lied in the limited gamut, or saturation of the dye set built into conventional materials. I remembered the quality of the prints my Aunt and Uncle used to make, their vibrancy, depth, and realness of the people's faces. The prints on conventional materials were certainly beautiful, but they did not stir the feeling of reality I had seen in a dye print, emotionally, they were flatter.

I did make some dye prints in the conventional manner. I spent quite a bit of money setting my lab up with registration equipment, and got quite good at making dye prints for myself. I quickly realized that the process could really be simplified and more control could be had by digitizing the image - dye printing does not allow simple dodging and burning controls, and this made for a certain level of complexity that I figured had to be eliminated. About this time, Kodak reaffirmed their commitment to Dye Transfer', which is code for dropping it like a hot potato. Which of course, they did.

I decided that I couldn't give up, just because the materials were no longer available. I would have to make them myself, to avoid this situation again. I also needed to make the process commercially viable, which means competing against other methods of producing prints, such as Iris. So, my goal became to make the materials, and reduce my film costs (Kodak raised the prices of the materials towards the end, to a point where the process was no longer profitable). It took two years, but I now have a great matrix film, which I use in conjunction with my laser recorder to make dye prints.

The process goes as follows: first I scan the transparency into the computer using my laser scanner. I then retouch out dust and scratches using Photoshop. I then use Live Picture to do critical color correction, and localized manipulation of tone curves, dodge and burn. I then place 8x10" Tmax film on the registration pins of the laser recorder, and expose the red, green and blue separation negatives in sequence. I develop them all at the same time. I then load up the Saltzman enlarger with the seps, and print onto the matrix film, and develop the matrices. The digital process eliminates the need to make two color correction masks, and three highlight masks. All exposures / development times, and all aspects of the transfer process are kept fixed. I only use small adjustments of the transfer to tweak the print the final bit, a very necessary thing to do to get a fin e print, and only available with the Dye Transfer process.

Dye Transfer works by creating a relief image in gelatin. The thickness of the gelatin on the matrix is proportional to the amount of exposure the area receives. This is accomplished by exposing the matrix through the base. A yellow dye is incorporated in the emulsion which absorbs the blue light to which the film is sensitive. The exposure proceeds to a greater depth into the emulsion with greater exposure. The film is developed in a pyro tanning developer which cross-links the polymers of the gelatin in exposed areas, and 'hardens' it, or makes it insoluble in water. The film is then washed in very hot water, and the unexposed gelatin washes off. The matrices are then soaked in dye baths, and the dyes migrate into the gelatin relief image on the matrix. The matrix is then rinsed, and then rolled into contact with the receiver sheet. The dye transfers from the matrix to the receiver. This is repeated three time, Cyan, Magenta, and Yellow transfers create a print, with a very strong and neutral black D-Max, and a D-Min (highlight) at the paper white. Many controls can be applied during the transfer process by varying the chemistry of the solutions - color balance and contrast are fine-tuned at this time. You can even paint dye onto the matrix (mat) by hand, and transfer again to beef up the color in a specific area. The prints are easily bleached, and retouched - an important aspect of a dye print.

The results reflect the process. I use the best aspects of the digital techniques, combined with the totally smooth continuous tone characteristics of the Dye Transfer materials. The special qualities of the prints can vary from extremely subtle rendering of delicate pastels, with all of the nuances of tonality preserved, to vivid and bold images which are only available with the pure dyes evenly co-mingled in the paper. The process avoids the harsh appearance of an ink jet print, and has no patterning or grain imposed on it. The shadow D-Max can be very high, but with sufficient light, detail is seen right down to D-Max, similar to a well made silver gelatin print. Since the dyes can be transferred to any gelatin coated smooth surface, I use a variety of high quality fiber surfaces, and have transferred to gelatin sized Arches papers - without loss of image quality.

My work these days is concentrated on identifying and testing dyes selected for specific purposes. I plan to offer prints with a dye set matched to the image, since no single set of dyes will render all colors optimally.

This is a complete description of how to make your own Dye Transfer printing materials, including a formulation for the matrix film, coating techniques, formula for a developer, and preliminary dye formulations. I am making this public in the spirit of furthering the cause of the Dye Transfer printing community. Please feel free to distribute this information.

Notice: Use caution when handling the chemicals listed. In particular, use gloves, goggles, and have adequate ventilation when handling the Silver Nitrate, Gold Chloride, Oxalic Acid, and Pyrogallic Acid, and Acetic Acid. These chemicals are hazardous, and should be used with proper techniques.

Browning / Adams Matrix film Formulation

The following matrix film was developed by me (Jim Browning) with the generous help of Rae Adams. This is a conventional Iodobromide emulsion, which has been adjusted for moderately high contrast, but without sacrificing tonal linearity. This emulsion works quite well. You will have to develop your seps to a higher gamma than when working with the Kodak materials. The film is about the same speed as the old Kodak matrix film.

Making the emulsion requires a system for heating a five liter container (Stainless), and maintaining the temperature accurately. A burette suspended over the container is used to slowly drip solution A into solution B over long periods of time. A propeller stirrer is also mounted over the reaction vessel, and run at slow speed.

The following formulation (Trial # 19) is for four Liters of emulsion:

Solution A:
Potassium Bromide 168 g
Potassium Iodide (5% solution) 62.4 ml
Inert Gelatin 160 g
Distilled Water 3500 ml
Solution B:
Silver Nitrate 160 g
Distilled Water 500 ml
Solution C:
Sodium Thiosulfate (0.1% solution) 10 ml
Gelatin 20 g
(gelatin added directly to the heated emulsion)
Solution D:
Potassium Bromide (1% solution) 10 ml
Manganous Sulfate (1% solution) 10 ml
Acid Yellow Dye # 23 (Tartrazene) 2.0 g
Saponin (1% solution) 2.5 ml
Distilled water 50 ml

Emulsification / Physical ripening

Add B (at 55 deg C) to A (at 55 deg C). Use a burette over a heated beaker holding solution A at 55 deg C. Stir the solution using a paddle mixer. (approx. 200 rpm). Temperature must be controlled to 1 deg. C using a temperature controller and hot plate.

Addition as follows:

Add 10 ml of B to A in 5 seconds.
Wait 1 minute
Add 245 ml of B to A over 4 minutes.
Wait 10 minutes
Add 245 ml of B to A over 5 minutes.
Ripen additional 15 minutes

Immediately chill the emulsion using an ice bath. Chill until the emulsion is very solid, whack the side of the container, there should be a distinct 'jiggle' feeling.

Cut the emulsion into 'noodles' 1/4" crossection. Force the gelatin through a 1/4" grid constructed from fishing line. Wash using distilled water for 4 hours. Change water frequently. Use at least 1 gallon of distilled water.

Note - In order to match the speeds of several sheets of film, it is necessary to sensitize enough emulsion for the full batch. Multiply the Solution C and Solution D quantities by the number of sheets to be coated, and sensitize and final prep the entire batch at one time. Filter the emulsion two times, once with 40 um filter paper, and again using 5 um filter paper. Use a vacuum filtration system. Divide the emulsion into separate 500ml batches, and pour into one liter stainless beakers with covers. Refrigerate until fully gelled. To coat, remelt the emulsion in one container, and coat. Follow the same procedure for all sheets.

Sensitization (Digestion or Chemical Ripening)

Remelt the emulsion, heat to 60 deg. C. Add the 20 g gelatin to the mix, stir until fully dissolved. Add Solution C, mix thoroughly. Stir rapidly for 1 hour while maintaining temp at 60 deg C.

Control temp to 1 deg. C. This step should be stirred vigorously, initially moving the paddle stirrer around in the emulsion to thoroughly mix. Cover the emulsion with aluminum foil while stirring to prevent fogging from t he safelight. The emulsion's speed increases 1000 times during the sensitization process.

Final Prep

Add solution D, mix. The Tartrazene dye is used to absorb blue light to cause the depth penetration exposure effect, and to minimize scattering. A wetting agent (saponin) is added to promote even coating.

Coating: Coat 500 ml over a 30" X 40" area. At 100 deg f.

Setting: Set at less than 50 deg F for 10 minutes to gel emulsion. Note: the emulsion will reticulate at this point if it is too thick.

Drying: Dry in a dust free enclosure for 2 hours at 70 deg F, 50% R H, moving to 85 deg. F 30% RH for 6 hours.

Use care when making the emulsion. Completely scrub the mixing vessel, beakers, stirring paddle, stirring rods, etc. Use soap, and an abrasive scrubber. Rinse thoroughly, final rinse with distilled water. Filter the emulsion with 5um filter paper using a vacuum filter before coating. A filter must be installed in the coater, placed in a position before the emulsion passes through the slot. Thoroughly clean all mixing vessels, and other utensils as you go, and carefully clean the coater after a coating run to prevent buildup of gelatin which would re-melt into subsequent emulsions. The coater must be repeatably rinsed with hot water between uses. All operations are best carried out using the light of a sodium safelight, which provides adequate light for working for long periods of time.

About Gelatins

The gelatin greatly affects the sensitization of the emulsion. In the past, less refined gelatins which contained sulfur compounds were used. These are termed 'fast' gelatins, and they sensitize the emulsion without the need to add a thiosulfate sensitizer. The problem with this is that each batch of gelatin will have a different effect. In modern emulsions, either a highly refined inert gelatin is used, or the gelatin is artificially synthesized. This allows a controlled sensitization by addition of sulfur and gold compounds to the mix. I find that the lot controlled Kind and Knox photographic gelatins work very well. They harden well in the tanning developer, withstand vigorous washoff, have low fog characteristics, and absorb and transfer the dyes readily. Use of other gelatins will probably yield very different results!

The Dyes I am using are as follows:

HL reducer: 20 g Sodium Hexametaphosphate in 1.0 L Dist. Water.

Acid Blue # 25 1 g
HL Reducer 10 ml
Set pH to 4.80
Acid Red # 289 1.5 g
HL Reducer 100ml
Set pH to 7.00
Acid Yellow # 23 3.2 g
HL Reducer 10 ml
Set pH to 4.00

Mix the dyes with about 100 ml of distilled water and heat to near boiling. Add a few crystals of Thymol as a preservative. Add most of the remainder of water. Titrate the solution to the specified pH using Acetic acid and Trieth. Be careful not to let magenta dye solution become more acidic than pH 4.0 at any time, as this will ruin the dye. Top off with distilled water to make 1L of dye. Filter before use. Check the pH frequently, and adjust. Filter frequently during use. You may make up a more concentrated replenishment dye, which is added as the dyes are removed from the working solution. Thanks to Andy Cross for determining these dye mixtures.


Expose the film through the base (Emulsion down), and develop in the following tanning developer:

Solution A:
Benzoatriazole 2.0 g
Oxalic Acid 8.0 g
Metol 28.0 g
Pyrogallic Acid 30.0 g
Water to make 4 L
Solution B:
800cc by volume of Sodium Carbonate to make 4L of liquid.

Mix 1 part A to 2 Parts B for normal contrast, develop for 2 minutes @ 68 deg F. Make sure you presoak the film for 1 minute. Rinse film in cold water for 30 sec, and fix in a non-hardening fixer for 5 min. Wash off unhardened gelatin using four of five vigorous rinses at 120 deg. F. Dry. Soak matrices in 120 deg. F water for 1 minute prior to soaking in dye for at least 5 minutes. Transfer the image for at least 5 minutes for the cyan and magenta, and 2 minutes for the yellow.

Condition the paper in paper conditioner for about 15 minutes before transferring the image. This is Bob Pace's formulation, I haven't tried it.

Triethanolamide  60 ml
Glacial Acetic Acid 19.4 ml
Ethylene Glycol 100 ml
Water to make 4 L
Check pH and adjust to 6.0.

For a complete discussion making separations, exposing and developing the matrix film, and rolling the prints and retouching, please refer to the books mentioned. This area is where DT printing really shines, and printers develop their own unique methods for controlling the print during the rolling process.

Some of the controls you can add to the first rinse are:

Acetic acid  increases contrast of print.
Sodium Acetate  decreases contrast of print.
Sodium Hexametaphosphate  Reduces amount of dye in the highlight (Highlight reducer)

Also, you can increase or decrease the pH of the dye baths for large changes in contrast. Some people actually paint the dyes onto the matrices, and re-transfer to increase saturation in a local area. The opposite of this is to squirt the Sodium Acetate solution at the matrix to leach out some dye to remove unwanted color. The print is easily retouched, one of the best features of a dye print.

About Coating

I have built a sheet coater which is very simple in operation. I had a large piece of aluminum precision ground jig plate (50" x 34" x 1") machined with vacuum channels, and anodized for corrosion resistance. This plate is VERY flat. This is called the platen, and it holds the film flat for coating horizontally. The platen is mounted on SS screw leveling feet, placed in a SS 36 x 60" sink. I use a machinist's level to get the plate very level. The slot coater consists of a 34" x 2" x 4" aluminum piece which has a 3/4" slot milled out of the center, which forms a chamber to hold gelatin. On the bottom of this body two triangular 'jaws' are mounted, which are adjusted using a feeler gauge to set a 5 mil gap running a full 32 ". On the top of the coater I have mounted three solenoid valves, and three funnels. The coater is mounted on four precision wheels, and the assembly is driven across the platen by timing belts, sprockets on shafts, and a DC motor which pulls on both side s, for a very smooth motion.

In operation, I first cut a 50 x 34" piece of 0.007" Melenex 583 (ICI polyester film with a special coating which accepts aqueous coatings), place it on the platen, turn on the vacuum, and use the DT roller to roll out any air pockets. Previous to mounting the film, I use a sprayer system placed in the sink beneath the platen to spray hot water (100 deg. F). This raises the temperature of the platen so that the emulsion will spread evenly. I then load the funnels with equal amounts of emulsion (three funnels, back, center, and forward are used to feed the emulsion evenly into the coater chamber). I then close the dust cover over the sink, which has a positive pressure supplied by a small HEPA blower. I then push the button which opens the solenoids, which pours some of the emulsion into the coater, I then start the coater moving about 1" per second. I watch the bead as the film coats, and add more emulsion when the bead starts to thin. This control s the flow well. When the coater reaches one end, I reverse the coating direction, and 'doctor' the emulsion with the coater blade formed by the slot jaws coming to a point 0.015" above the film. This tends to even the emulsion further, and remove any bubbles (rare). I then turn on cold water to the sprayer (50 deg F) for about 10 minutes, which chills the platen, and causes the gel to set. I then turn off the vacuum pump, open the dust cover (make sure you use full body tyvek overalls to limit dust), lift the sheet out of the sink, and carefully tack it onto a drying frame, and place it horizontally inside a laminar flow HEPA filtered dryer. I use a Clestra Cleanroom Super II 2' x 4' HEPA filter blower module, which feeds an enclosed area 2' x 4' x 4'. O n the front of the dryer cabinet is a light tight door, which is made from light baffle panels which allow the laminar flow air to move out of the cabinet without restriction. Drying takes a long time, usually 6 to 8 hours, and I dry 8 sheets at a time.

A simpler method of coating should work well with smaller sheet sizes. Make an aluminum blade longer than the width of film to be coated. The blade can be rectangular in crossection, with a triangular point, or it could be a simple cylinder. Wrap a wire around the blade tightly over the full length. By adjusting the gauge of the wire, you can control the amount of emulsion passing through the gaps between the wraps, thus controlling the emulsion thickness. The emulsion and film should be hot enough to allow the emulsion to flow after doctoring it with the blade. Pour a supply of emulsion in front of the blade, and draw it through the emulsion, and over the full length of the film. The ideal thickness of the emulsion will b e enough to allow heavy exposure of the matrix film without having the image appear appreciably during development. A too thick emulsion will reticulate when chilled or during drying, which will pattern into the image. A too thin emulsion will 'print through', and the rough surface of the film will texture the image, as well as limiting the DMAX of the print. Start with 100 ml of emulsion for a 16x20" film.


The dyes transfer very well, but as with all matrix film transfers, you need to maintain the temperature at least at 70 deg. F, with higher temperatures working better. The magenta dye does have a tendency to stain the fil m base slightly, but this should not affect the image. It also may be necessary to use an ammonia / calgon matrix cleaner on the magenta matrix to remove the last bit of dye. highlight clarity may be controlled by using a small amount of the highlight reducer (Sodium Hexametaphosphate) to get the clearest highlights, near paper white. Use a good quality fiber based paper - I have used Ilford Multigrade FB, fixed out with a non-hardening fixer. I have also used Arches hot pressed paper which I treat with a silic one water repellent, and then size with gelatin, and harden. You may mordant the paper with an aluminum salt, but the image transfers well, is sharp, and has good water fastness without the use of a mordant.

Materials Cost

The cost of materials is about $8 per sheet of 30x40" film, most of the expense is in the silver, I buy the inert photographic gelatin from Kind and Knox in minimum 25 LB lots costing $ 10 / LB, and the polyester film cost about $ 0.50 / foot of the 50" wide rolls. The Silver Nitrate is from First Reaction at $ 277.00 / Kg.

About making Seps

I am creating my separations digitally using a laser based film scanner / film recorder I designed and built a few years ago. I typically use 150mb files, which make 20x24" prints which are completely sharp, and show no digital artifacts. The recorder both scans the originals (Up to 150 l/mm) and records onto 8x10" film (EPN, VPS, TMX). The film is held on registration pins, which allows exposing three seps in perfect register.

Alternately, seps can be made by exposing TMX (Tmax-100) film using red, green, and blue filters. Red and Green exposed color correction masks should be used when making the seps. Develop the seps to a higher gamma than the old Kodak film, Dmax should be about 1.50 above base fog. Make a highlight bump mask with about a 0.30 density which records only the highlights. Print each sep with it's associated highlight mask onto the matrix film. Develop in the tanning developer. The masking for this dye set will be different than the old Kodak dyes, but I haven't yet determined the correct masking for this dye set. I'll supply this information when I have it.

Some Phone Numbers

You should try to buy the book: The Dye Transfer Process by David Doubley. Condit used to sell this book.

If you need some advice about DT printing, or matrix film coating, please feel free to contact me:

Digital Mask
187 Stevens Rd.
Lebanon, NH 03766
(maintained by justme@well.com)