Darkroom Math

There are a number of formulae that come in handy for film photography and darkroom printing. I'm gathering the most significant formulae I use below, and will update the list when I have more to add:

Convert Filter Factor to f-stops:

Adjust Exposure Time by Number of f-stops:

Fomatone Warmtone Papers - Toning Effects

Fomatone has a very good reputation as a Lith-capable paper. It's also an excellent warmtone paper that can have fascinating color shifts when toned. I've experimented with different toners and would like to show the effects here. I'll update this post with any future toning experiments related to this paper.

Fomatone 131, print without toning.

Same print on Fomatone 131, fully toned
in 1+9 Selenium.

 

Fully Analog B&W Trichromes

Time to take my proof of concept for analog trichromes to the next level.

I shot these photos on an expired roll of Ilford Delta 400. It was about 20 years expired, and I metered at 200ASA. I didn't change the development time at all, under the assumption that overexposure alone would be sufficient to get good negatives.

The initial digitally-made trichrome, for
reference.

Scans of the three negatives I'm using, from left to right: Red filter, green filter, blue filter.

This process is simple enough to create on a computer, but I wanted the challenge of creating a fully-analog color photograph. This particular roll of film gave me trouble printing because it still had a tendency to curl even after drying. But the result was good enough as a proof of concept.

For printing with cyanotypes, the RED filter is used for the CYAN layer, the GREEN filter is used for the MAGENTA layer, and the BLUE filter is used for the YELLOW layer.

The yellow layer came first, and the exposure seemed good. All details were clear. This layer was exposed for 120 minutes.

The magenta layer, in this case using cuprotype instead of cyanotype, developed a little stronger than I wanted, overpowering the yellow quite a bit. The exposure time for cuprotype is much longer than for cyanotype, so I think I can reduce the exposure time by a fair amount, hopefully at the same time increasing contrast. I did mess up the alignment here though, my bad... This layer was exposed for 60 minutes.

The final cyan layer was the trickiest to balance. Too much exposure and the cyan is way overpowering. I think I got pretty close here but the magenta/red is still very strong. This layer was exposed for 40 minutes.

Overall it's pretty obvious that the magenta/red layer completely overpowers the print. The pure white (or light grey) on the front of the cap is far too red. But aside from alignment issues, I'm pretty happy with how this turned out. At least it proves the concept, and leaves room to make adjustments for the next attempts.

Trichrome Cyanotypes

Hopefully those reading this have seen my previous post about my Bicolor Cyanotype technique. Calling this next step "Tricolor Cyanotype" is a little misleading, because one of the color layers is in fact not cyanotype at all, but rather cuprotype. I've found that other attempts at creating a red layer don't satisfy my "purist" view on creating a fully analog process, or at least one that can be recreated using easily-obtained raw materials.

The chemistry of cuprotypes is similar to that of cyanotypes, where a solution of copper ions reacts to ultraviolet light similar to how an iron solution is used for cyanotypes. The chemistry is remarkably similar too, with ferric ammonium citrate comprising one part of the cuprotype mixture, the other part being copper sulfate. I use a third component, thiourea, but this is not strictly necessary when making cuprotypes. There are many resources available online to learn about cuprotypes, so I won't describe the process here.

Exposure for this technique is a little tricky, as the magenta/red (cuprotype) layer can very easily overpower the other two layers. Additionally, too much exposure on the final cyanotype layer can completely destroy the pretty green tones from the mix with the yellow layer. In general, exposure for the cyan layer is roughly one-third the exposure for the yellow layer. I'm experimenting with exposure for the magenta layer, the exposure above I believe was roughly equal to the exposure for the yellow layer.

Another post will follow soon where I use this technique to make a trichrome print from negatives.

Silver Recovery from Exhausted Photographic Fixer

I've been saving my spent fixer for the past 2 years, after learning how much silver it contains. Rather than sending valuable silver to the hazardous waste facility and gaining nothing from it, I decided to make an attempt at extracting the silver myself.

Whereas there are several methods for precipitating solid silver from the solution of silver thiosulfate, I opted for the cheapest method, the iron reduction reaction. I used 0000 steel wool in a flask of fixer solution on a magnetic stirrer to precipitate the silver.

Left: the flask of fixer with steel wool being added under stirring. Right: the flask after 2 minutes of stirring, with a noticeable amount of silver already having precipitated.

This is approx. how much steel wool I used for
1 L solution. The piece had a mass of ~1.5 g.

After allowing the flask to stir overnight, I placed a magnet onto the side of the flask to collect any remaining iron filings that didn't react. Luckily there were very few, and I was able to decant the silver precipitate into a beaker for washing. I didn't take photos during this process, but I washed the silver several times until the water remained clear between rinses. At first the water was turning yellow during washing, but the yellow color went away after 3-4 rinses. I suspect there were still iron salts in solution.

I collected the dry powder and melted it in a crucible with a little borax for flux. My oxy-propane torch made quick work of it! The silver piece weighed in at 3.17 g.

I'm quite happy with this yield; it seems as though I extracted nearly all the silver from the exhausted fixer. I'm still curious what the theoretical yield is for a 1 L fixer solution, so I think I will make another post in the near future with some calculations to figure out those numbers.