top of page
  • Writer's pictureJonathan Hilty

Let's talk about the Second Varnish

(Note: I originally wrote most if this in June 2019, but waited forever to post it while I continued experimenting with coatings...)

After taking a 6 month break from my autochrome experiments, I decided to take a few screens I had sitting around and coat them with some freshly made emulsion. Once the gelatin had set up, I set them out to dry overnight next to a fan as is typical. The next day when I went to go store them in a light tight container... Uh oh...

It seems as the gelatin dried, it put stress on the second varnish layer, causing it to crack and fail. The light-sensitive gelatin had even flaked off in one large piece, which was a little weird. It seems my second varnish recipe doesn't hold up, so I'm forced to go back to the drawing board!

Before I get started, I wanted to cover a few conditions that the second varnish must meet to be a good candidate. 1. The varnish must not dissolve the first varnish layer (latex). This can cause many problems, such as the starch shifting (will be discussed later on). This means a lot of common organic solvents like xylene and mineral spirits may cause problems.

2. The varnish cannot cause the dyes to bleed.

This rules out many polar solvents, which the dyes are extremely soluble in. Water soluble or alcohol based lacquers/varnishes are a hard stop.

3. The gelatin must be able to stick to the varnish after drying.

Definitely the most important -- if your gelatin won't stick, it's a no go! Often the gelatin will flake off in one piece if it won't adhere to the second varnish layer. I've only been able to determine this by trial and error.

A quick note on terminology...

Typically here in the United States, a layman like me will tend to use the term "varnish" and "lacquer" interchangeably. However, they are technically different. From Wikipedia: "A distinction between spirit-drying (and generally removable) "lacquers" and chemical-cure "varnishes" (generally thermosets containing "drying" oils) is common, but varnish is a broad term historically and the distinction is not strict."

Basically, a varnish undergoes a chemical reaction, or "curing" process after the initial solvents have evaporated. A lacquer just dries up, and can be re-dissolved with the carrier solvent.

The differentiation here with autochromes is somewhat important to consider, since varnishes will generally require significantly more time before they are ready to coat.


Polyurethane is typically sold as an oil-based varnish. It has a deep amber color. It cures via exposure to oxygen, and a typical wood application can take 2 weeks until fully cured. Polyurethane is highly chemical resistant, but can yellow with exposure to UV. In March I applied 5mL of undiluted polyurethane to several screens. The layer began to harden after a few hours. Even after several months the layer is somewhat squishy, and can be deformed with a fingernail. By September, a snowflake/window frost-like pattern had appeared, and I finally decided that I would never be able to use them. I was able to peel up a corner, and found that the inside of the varnish was still very much liquid.

Degraded polyurethane plate

Noting the long cure times, I tried diluting the polyurethane the polyurethane with a solvent to reduce the final layer thickness. Initially I used Odorless Mineral Spirits (OMS), however I found it tended to cause thousands of tiny bubbles, which would be trapped in the polyurethane after it began to cure. It somewhat scattered the light, giving the screen a somewhat hazy (but workable) look. VM&P Naptha from the hardware store worked much better. My plates were diluted 1:1. The coating takes on a very slight yellow color, though this is barely noticeable. Cure time is about 3 weeks.

The polyurethane plates performed well, and could take a good amount of abuse before they began to break down. Extended contact with water( such as leaving the plate wet for days) can cause small color shifts, likely due to trace amounts of moisture leaking through the second varnish layer. The first varnish begins to lift off the glass after a few coatings. More than 10 and nearly the whole first layer has lifted.

Overall, it's an alright candidate for the Second Varnish layer, if you can't get your hands on anything else, and don't mind waiting for the crazy long cure time.

Helmsman Clear Spar Urethane

I'm not sure what this is, exactly, but I tried it out. It was a darker amber than the polyurethane, and the coated screen was slightly yellower. It was coated with the rest of the undiluted polyurethane plates, and as of this writing it still hasn't cured either. I ended up tossing the plate after 6 months of waiting.

Golden MSA Acrylic Varnish

I had tried a few acrylic varnishes with no success. Although they provided a good seal, the gelatin does not like to stick to the acrylic after it dries, which was very disappointing. However, based on a recommendation from an old post, it seems the Golden varnish allows for good adhesion to Liquid Light. I figured I would give it a shot.

The first thing you notice is that this stuff is THICK. Applying 5mL to a 4x5 takes a few minutes and a lot of guiding with your fingers to ensure full coverage. This is in contrast to any other varnish/lacquer I tried, where coating the plates is somewhat similar to that of wet collodion. After failing with a few hardware store solvents (xylene, OMS, etc.) I was able to thin the stuff with Golden's official mineral spirits solvent product. This makes the plates much easier to coat. However, the solvent seems to affect the first varnish somewhat and can cause it to drift and crack. Often these plates dry with big obvious wrinkles.

Screen plates with thicker, undiluted coatings seem to have issues with a limited viewing angle (when viewing them at any angle other than exactly head-on, the plate appears to be greyscale). After some experimentation, I've found that the varnished can be thinned with artists's grade turpentine. I dilute 1:1, and apply 2.5mL of the solution to each 4x5 plate. It dries overnight, but takes about a week to fully cure.

Coating technique is really important here, and I'm happy to say I've finally figured out the major cause of the wrinkling that occurs when applying varnish (this type, at least)! After coating plates indoors one night, rocking the plate on my fingertips, I noticed that a lot of these "wrinkles" seem to sprout out from over where my fingers were. It seems the heat from my fingers was causing it in some way -- although I"m not certain what the exact mechanism of this is. Now, I coat by pinching the corners of the plate, and the wrinkling problem has all but vanished. I've found that Golden's MSA is one of the best performers. One of my oldest screens was coated with undiluted varnish, and even after >10 coatings it's only just now showing slight signs of wear. There almost no tension on the starch as it dries, and the grains don't appear to spread out at all. One last note -- after using this as my primary Second Varnish choice for about two months, I am noticing that many of the plates seem to be losing their color due to the gelatin shrinking. This varnish works really well in the short term, but may not be great in the long term!

Amsterdam Acrylic Gloss This varnish has most of the positive qualities of the Golden Acrylic Varnish, but is slightly thinner and can be used straight without dilution. It seems to have more issues with lifting off the plate, and artifacts start to show up after 5 or so coatings. I'm not certain yet if this suffers the same degree of desaturation as the Golden varnish does, long term.

Alkyd Varnish

The alkyd varnish did not seem to perform very well. The first screen was water tight, but failed when gelatin was poured on it and allowed to dry (the varnish flaked off). A thicker coating was more resilient, but ultimately the gelatin did not seem to stick to it very well. Still smells kind of weird even after several weeks.

The varnish delaminated from the plate as it dried.

Testors Clear Coat

5mL of the clear coat was added to the screen and allowed to dry (which occurred fairly quickly, within an hour or so). The coating took on a barely noticeable blue tint. The plate was allowed to cure for a few additional days, just in case. Upon contact with water (during the submersion test), the coating became very cloudy and never cleared after drying.

Dupli-Color Automotive Clear Coat

5mL applied to several plates during testing. The lacquer dried very quickly, within an hour or so. The plates continued to gas out over the course of a few days. Initial tests were very positive -- the layer appeared to be water tight and the gelatin stuck very well. The solvent was a massive migraine trigger for me, so a gas mask was necessary if I planned on getting within 20 feet of the area I was working in. As tests were carried out, it became apparent that the coating had a tendency to allow water in when the plate was being handled, and would rapidly curl and deform if even a fraction of a drop managed to find its way in. Eventually, the sides were sealed with epoxy to prevent water from getting in. This seemed to work well, and otherwise the plates have held up fairly well over several coatings. The epoxy has a tendency to discolor from the erythrosine in the emulsion and the potassium dichromate in the bleach.


I had some casting epoxy laying around, which uses a mixture of 1 parts solution A and 1 parts solution B. I added some acetone to the epoxy to thin it down considerably. Ultimately, my epoxy mixture was: 30mL A + 30mL B + 15mL acetone

5mL of this solution was added to a screen and allowed to cure. It was extremely smelly, though apparently safe. It took about 24 hours to cure completely. The final layer was a bit wavy and somewhat thick. Possibly a different choice of thinner (MEK?) might yield a better layer. The epoxy provided a good layer, and the gelatin was able to stick to it readily. The epoxy plate was never tested however, as I accidentally dropped it and broke the plate before I could expose it. It most likely would have suffered from some parallax viewing problems, since the layer was so thick.

2k Automotive Urethane

The compound was mixed as directed and 5mL applied to several plates. The coating cures in a few hours, but will continue to gas off strongly over the next few days. The smell never really went away even after months, though it is very faint at this point, leading me to believe the bottom of the coating is still curing very slowly. The coating is clear and the gelatin sticks to it readily. The only downside is that, after a few coatings, the areas between the first and second varnishes start to separate, creating hazy artifacts. 10+ coatings and the screen is probably toast. If you're confident that you can get a good picture in just a few coatings, this is a great candidate. It's generally very easy and forgiving to work with.

Artifacts appearing after several coatings of gelatin

The Original Lumiere Formula

With my very first attempts, I tried to recreate this formula with a few shortcuts. Most notably, I was worried about ordering pure nitrocellulose in the mail, and attempted to substitute it with an already formulated nitrocellulose lacquer. This ended up being disastrous, as much of the butyl acetate in the lacquer never seemed to gas out on its own, so that the varnish remained tacky and malleable. Over time, the gelatin seemed to pull on it, causing heavy wrinkling (something that was described in a previous blog post). This time, I decided not to take any shortcuts, and make it as faithfully as I could. This effort is still ongoing, and is worthy of a blog post in and of itself. However, I'll share my results so far.

The recipe is as follows:

  • Ethyl acetate 300ml

  • Dammar gum 28.8g

  • 7.2g nitrocellulose

  • Castor oil 4.536 g

Typically, the nitrocellulose is dissolved first, then the damar gum. Damar gum is made of two major components, alpha and beta resene. The beta resene is a waxy substance, and is mostly insoluble in the ethyl acetate. After filtering off the wax, the castor oil is added and you're good to go... except not. I've had a lot of problem a white precipitate forming in the varnish upon coating the plate. My theory is that there is a small amount of the beta resene that is able to dissolve, which falls out of solution as the solvent evaporates. This leaves big, white, opaque chunks on the plate that block all light passing through. I believe this can be avoided by a more thourough "de-waxing" process, and have a few attempts. As results with other varnishes picked up, I became too busy to keep pursuing this frustrating varnish. There is another major issue with excessive wrinkling with this lacquer. I'm not sure if this has the same cause as the wrinkles from the Golden Acrylic attempts or not. In my first version of the varnish, these wrinkles did not seem to occur, since after the highly volatile ethyl acetate had all evaporated out, there remained a good amount of butyl acetate that evaporated much more slowly.

One major positive thing I can say here is that, 6 months down the line, these plates do not have the same problem with desaturation as other Second Varnish choices have had. I think this is worth continuing to pursue, and I plan to do that when I get my fume hood installed.

Wrapping Up

Right now, my favorite choice is a diluted form of the Golden Acrylic MSA Varnish. It coats without stretching out the starch, can be easily made, can be coated indoors without special ventilation, gelatin sticks readily to it, and the screen plate can be used over and over without any significant degradation. However, after using it primarily for two months, I am noticing problems with the gelatin shrinking, causing a loss of color on the plate. This does not seem to occur in other surviving plates, notably the "Dupli-color" automotive clear coat, the diluted polyurethane plates, and the Lumiere nitrocellulose lacquer. I will have to explore this problem further to see if a better choice in protective top coating can help prevent this, or if things will be "back to the drawing board" with the Second Varnish...

Plate #106, using Golden Acrylic MSA Varnish

713 views0 comments

Recent Posts

See All

Small update - Magazine & "How To" Guide

Hey everyone, This is just a teeny tiny update on a few relevant things. First off, one of my authcromes was recently featured on the cover of The Hand Magazine, Issue #28. If you're interested in

bottom of page