I've been playing with the idea of making autochromes for about 4 years now. Making photographs 100% from scratch had always interested me, and up until this point I had only ever really worked with black and white processes. Color film never interested me very much - I think it was due to it being so ubiquitous while I was growing up. I never found it very interesting to work with. Eventually, I fell in love with two very old and very different color processes: Autochromes and Lippmann plates. They both rely on a simple black and white panchromatic emulsion to reproduce color, though their mechanisms of doing so are wildly different. I made a thoroughly amateur attempt early on using a laser printed color screen and rubber cement, managing a weak color response. When I finally had the funds and facilities to play with these processes, I eventually made more progress with Lippmann plates and they became a full time hobby. Meanwhile, my box of autochrome materials collected dust. I detail my knowledge of the Lippmann process here.
This summer I decided to pick up where I left off with my autochrome experiments, this time being a little more deliberate with my methods. I had previously achieved three complete screens with some food-coloring dyed starch, and was able to demonstrate that smashing the starch with a spoon was a very effective way to compress the grains (a necessary step that increases the speed and saturation of the plate). At the time I experienced problems the grains not adhering well to the first varnish.
Though there are still some things I need to tweak, I believe I have a decent method of producing autochrome screens.
1. The Second Varnish
It may seem a bit counter intuitive to start with the second varnish, but bear with me - one of the byproducts in this step is a necessary ingredient for the first varnish. The objective of the second varnish is to seal in the starch grains, protecting them from aqueous solutions (the emulsion and various photographic processing solutions). The varnish also is designed to match the refractive index of the starch, serving the overall purpose of increasing translucency
In "The Lumière Autochrome: History, Technology, and Preservation", the authors lay out the ingredients of the second varnish as follows:
300mL Ethyl Acetate
28.8g Gum Damar
536g Castor Oil
The castor oil addition still confuses me as to its purpose and quantity -- if the directions for this
varnish are followed directly, you will end up with an overwhelmingly oily mixture that will never dry or harden, and which slowly dissolves the first varnish and starch. I have seen other sources where, in these quantities, it calls for 4.536kg. Another source calls for as little as 5g. For the immediate future, I am omitting the castor oil, as my preliminary trials seem to indicate that it can make a satisfactory varnish layer without it.
My recipe is as follows:
300mL Ethyl Acetate, sourced from Amazon.com
28.8g Gum Damar, sourced from Etsy
38.7g nitrocellulose laquer
The nitrocellulose lacquer is presumed to be nitrocellulose dissolved in amyl acetate (a solvent that smells a bit like banana flavoring). A sample was weighed, allowed to dry over several days, and weighed again. Presuming that there are no other additives, it was found that there is 1g of dry nitrocellulose per 5.3g cellulose laquer when wet. Therefore 38.7g of wet laquer will get us our required 7.2g of nitrocellulose.
All the mixing and weighing was done outside - please note that ethyl acetate (as I'm sure you'll notice) is super volatile. The three components were added to a 500mL beaker (covered with plastic wrap to prevent excessive evaporation of ethyl acetate) and magnetically mixed for several days. Realistically, I think a few hours are satisfactory but I figured I would be safe. You'll notice a fine white solid residue that does not dissolve -- this is good. There is a component in the gum damar that is not soluble in either of the acetates, and will be used in the first varnish. After the determined mixing time is over, filter the solution into an amber bottle (500mL) and save the filtered solid.
A note on the insoluble gum damar: in my first run with this varnish, the pieces were quite large and were able to be filtered with an ordinary coffee filter. Subsequent trials yielded a solid so fine, even my smallest laboratory filters do not catch it. It seems this solution may need to sit for quite some time (1 - 3 weeks). In that time, the gum damar particles, well, "gum" up a bit and form larger solids. These larger chunks can be strained out of the solution and saved.
A note on alternatives: In light of the fact that the dye from my blue grains was being dissolved into the ethyl acetate, I searched for a more modern replacement. The best material I could find was some "UV Cure" epoxy, which could be thinly spread across the starch with a finger and allowed to sit (it would slowly level out somewhat, though never completely) before exposing it to UV. The epoxy sealed the starch from water well enough, but I believe the layer it creates is ultimately too thick, allowing the light to spread out too much before it hits the the emulsion. I think this could work well if the epoxy could be diluted with some solvent to be less viscous, but my tests with ethanol, acetone, toluene did not pan out.
2. The First Varnish
The first varnish is a bit more simpler to create. The purpose of the first varnish is to provide an optically clear, sticky layer on the glass that gives the colored starch something to stick to. I tried many different materials here, but unsurprisingly the good ol' Lumières' formulation worked the best and most consistently. It consists of the following:
To source the rubber, you don't have to go much farther than your local general store - rubber cement happens to be latex that is already dissolved in toluene (sometimes xylene). From my calculations, it seems there is about 1.5g of dry latex dissolved in 13.3g of the glue. I found it difficult to find a source of toluene locally, but I believe for all intents and purposes xylene can be used here interchangeably with toluene. My formulation of this recipe is as follows:
I had some noticeable difficulties getting the insoluble damar 'leftovers' to dissolve in xylene. Eventually I obtained toluene, though it proved to be just as difficult to work with as the xylene. Ultimately, my solution was to leave the solution in magnetic stirring for a long time, 24+ hours. The solution was filtered and added into the existing rubber/xylene solution. The damar solution adds a 'stickiness' to the rubber which allows it to remain tacky after the xylene has evaporated.
Notes on other varnishes: I played around with the UV Cured epoxy, as well as a solution of Canada Balsam of varying dilutions in xylene. The epoxy seemed like it would be an alright choice for a first varnish, but ultimately lifted off the glass after the application of the second varnish (ethyl acetate). I played a bit with Canada Balsam after a bit of success, but it was hard to keep it 'tacky'. After the solvent had dissolved away, the layer of Canada Balsam quickly hardened and would not take up stray starch particles. I never tried Canada Balsam with the gum damar residue dissolved - that may be something worth trying in the future.
3. Dyeing the Starch
In my initial tests, I have used 5 common clothing dyes to dye the starch granules. I have not yet been able to achieve a fully neutral screen with these ratios; however I've gotten pretty close.
Each batch of starch was mixed in a 50mL beaker with a magnetic stir bar and stirred for 30 minutes - 1 hour. Dye was slowly added to the stirring water and allowed to dissolve for a few minutes, before the starch was added via small scoops. After mixing is complete, the starch is vacuum filtered and allowed to dry.
It is fairly important to wear a dust mask while mixing dyes, or you will likely end up with dark and multicolored snot.
For the future, I will be performing a secondary (and possibly tertiary) wash in water or alcohol to remove excess deposited dye from the starch. In retrospect, there were a few instances where I should have noticed that the excess dye was causing some issues. Most notably, both the previously listed blue dyes are soluble in ethyl acetate, meaning the dyes were bleeding into the second varnish.
After each of the batches of starch have dried, I break the dried chunks apart thoroughly with a mortar and pestle. The starch is filtered and added to a container. Larger chunks removed by the filter are added back into the mortar and pestle and broken down more until no more large chunks remain. All the starch is added to the same container and mixed gently. If done correctly, the starch should be a very dark grey, most likely with a slight color cast.
While I experiment with these dyes, I have been able to source all dyes necessary for the original autochrome plates and are currently waiting for them all to arrive. I'll probably switch gears to those dyes while I knock out the problems elsewhere in the process.
4. Initial Plate Coatings
I start by lining the outside of a glass plate with 1/4" - 1/2" masking tape - this leaves spaces on the border to allow the second varnish to overlap the starch and first varnish. I apply the first varnish in the same manner as one would while coating a plate with collodion -- allowing a larger amount of fluid to coat the entirety of the plate area, before tipping a corner into the bottle and pouring the excess back into the bottle. The plate is allowed to dry. It should be noted that the enthusiastic amateur may 'test' the varnish by pressing their finger on a corner and ensuring it is sticky. One should take into account that the rubber coating will be very tacky at first, lesser so while the xylene evaporates, then much stickier again when the solvent is fully evaporated.
The starch is deposited (I use two small scoops) and is gently brushed around the plate for about a minute. Tap the plate on a hard surface to remove excess starch, and gently blow. Next, add carbon black (I use artist's grade charcoal here) and brush it around in the same manner. Brush to remove and blow gently. Finally, add a good amount of talc to the plate's surface and brush again in the same manner. The talc will turn grey; this is good. The talc is used here to remove excess charcoal from the plate. After the plate is brushed sufficiently, remove the excess talk and brush the plate while gently blowing on the surface. After your brush strokes stop leaving behind noticeable marks, your plate is now ready for compression.
I think step may be one of the most daunting for amateur photographers that are looking into recreating the process. As many sources will point out, the press applied 2500 pounds of pressure per square inch to the plate. There's a lot of force to do that.... right? They key takeaway that anyone with an engineering background should have noticed is that pressure is a measurement of force / area. If the area is significantly reduced, the amount of force needed here can also be significantly reduced.
This is a step I'm still working on, but I've been able to do a very satisfactory job compressing the starch with a simple steel castor roller ball. My particular castor came lubed with oil, which needs to be removed (otherwise it will coat the starch). With the castor in hand, the ball is pushed back and fourth on the plate, leaving a trail of compressed starch behind it. It's much like coloring in a drawing with a crayon. Make your way across the plate slowly, occasionally observing the plate via transmitted light to ensure you're not missing starch. Performing this step of the process in this manner is incredibly tedious and tiring, and leaves behind noticeable "pencil stroke" patterns on the plate even when the most care is taken to avoid making them. Also, my particular castor has begun tiny metallic chunks into the starch. It's really noticeable unless the plate catches the light at a certain angle.
Last week I attempted to make autochromes to see if I was able to achieve any sort of color response with the screens I was making. Unfortunately, I was unable to achieve anything more than a grainy black and white negative.
The most prominent problem here is the emulsion -- it's fogged. I've made hundreds of Lippmann plates and have that ritual down, but with anything faster. I've tried in the past to make faster silver gelatin emulsion, which also resulted in fogging. I think my addition of ammonia in this run may have caused the fog, and will be omitting it in the future.
Under a microscope, it seems there are grainy chunks of exposed emulsion that correlate to the starch, but they are all shifted somewhat. I think that may be indicative of the second varnish layer being too thick. In this run, I used epoxy due to the aforementioned blue dye dissolving in the ethyl acetate. The epoxy remains somewhat viscous as it is spread on the plate, and may be just too thick, allowing the light to spread out before it hits the emulsion. I've obtained the dyes from the original autochrome recipe for the violet grains, so the next round of plates will be using the actual second varnish next time.
One final problem I noticed - parts of the emulsion remain purple after processing. This is due to the sensitizing dyes, erythrosine and pinacyanol. I think this is probably indicative that my emulsion is waaaay too thick. In the future, I'll probably coat these things the same way as I do the Lippmann plates - measured with a syringe, with the glass plate resting on a perfectly level combination square.
Over the next few weeks, I'll try my hand at making a better starch with the new violet grain recipe. In the meantime, I've been able to source all the original autochrome dyes and am currently waiting for them to arrive in the mail. Really, the only dye I've yet to be able to get my hands on is the green sensitizer, Orthochrome T, though that may be a bit of a pipe dream at this point.
On further review, there definitely was a marginal color response in some more pure colors. I think the emulsion is too heavily fogged to be able to reproduce them in any noticeable manner.