Process atlas

Between 1800 and 1900 photographers made pictures with silver, iron, platinum, palladium, gold, chromium, mercury and uranium, on paper, glass, metal and cloth, bound in albumen, collodion, gelatin and gum. The atlas has one page per process. Each explains the chemistry of image formation, the historical workflow, what the image looks like and why, how permanent it is, what it costs, how hard it is, and where the course teaches it — or, for the sixteen it will not teach, why not. The comparison fields at the top of each page are the same across the atlas so that processes can be set side by side.

!There are no plates in this atlas, and here is why

A process atlas ought to show you the process. This one does not, and it will not pretend otherwise.

On 5 September 2026 the course owner ruled that both atlases ship text-only now, and that plates are a later enhancement rather than a gate. For the process atlas the ruling is recorded as: the process atlas gives each process its comparison data, and each index page states that plates are not yet included and why.

The reasoning is the same as the troubleshooting atlas’s: a photograph of a platinum print helps you recognise one, and the sentence explaining that the image is finely divided platinum lying among the fibres with no binder over it is what teaches — and it is what lets you tell that print from the four prints it resembles. Holding the whole reference back until somebody had coated, printed, photographed and licensed forty-four examples would have cost far more than it bought.

So every entry here is written from what the course has established in its own written parts and its own reference corpus, with links back to the pages that established it. When plates come, they drop into entries that were already built to hold them.

☠Practised versus studied

28 of the 44 entries are marked practised and 16 are marked historical study. The distinction is not about difficulty; it is about whether the course will give a procedure at all.

Processes marked practised are ones the course will teach with a full procedure, at safety level A, B or C. That does not mean the procedure exists today: most of them belong to parts that are still to be written, and every such entry says so on its face, in its own Where the course teaches it section. Only the four hand-coating projects of Part V and the Level A cyanotype assignment of Part I point at a finished lab.

Processes marked historical study are classified level D: the page gives their chemistry, history, significance and hazards, and no procedure, under the Level D policy. The mercury-developed daguerreotype, wet-plate collodion, the whole dichromate family, uranium toning and the platinotype are studied here because they matter to the history and the chemistry, not because anyone should do them at home.

The comparison, and the part of it that is outstanding

Brief section 31 asks for a process comparison: one negative printed across seven processes — silver gelatin, cyanotype, salted paper, albumen, kallitype, Van Dyke Brown and platinum or palladium — compared on tonal scale, surface, image colour, sensitivity, exposure source, chemistry, permanence, cost and difficulty.

!The visual comparison is outstanding, and named as outstanding

Nobody has printed that negative. The comparison the design asks for is darkroom work, and it is gated on the parts being physically performed: the standard negative is established in Part XXI’s materials page, each process page is required to record its own measured row, and the whole is owned and completed by the Part XXV assignment the process comparison atlas. None of those pages exists yet.

That is a different product from this atlas. This page carries one row per process from the published literature; the assignment produces one row per process from one negative in one darkroom, and the two are not interchangeable. A comparison assembled from different photographers, papers, negatives and decades tells you what the processes are like; it does not tell you what they do to the same picture.

What follows is therefore the first kind and not the second, and every column in it is sourced to a written page of this course rather than measured here.

The seven, on what the course can state with sources

Process	Image substance	Colour	Surface
Silver gelatin, fibre	Developed metallic silver in gelatin	Neutral to warm, set by the developer and the rate of deposition	Emulsion over an opaque baryta layer, with a hardened supercoat
Cyanotype	Prussian blue	One blue, deepening for hours after the wash	Pigment among the fibres; no binder
Salted paper	Printed-out colloidal silver	Light to reddish brown untoned; colder and purpler with gold	Matte and sunken in, the fibre visible
Albumen	Printed-out silver in a protein layer	Rich purplish brown when gold toned	Glossy; the image sits on the paper rather than in it
Kallitype	Developed silver, reduced by iron(II)	Sepia from a tartrate developer, black from a borate one	Silver among the fibres; no binder
Van Dyke Brown	Printed-out silver, reduced by iron(II)	Warm brown, set by particle size and coating humidity	Silver among the fibres; no binder
Palladium	Palladium metal	Warm, brown, very smooth	Noble metal among the fibres; no binder

Process	Light	Exposure route	Permanence, in one line	Cost	Difficulty
Silver gelatin, fibre	Enlarger or contact	Developing out	The most controllable in the atlas, and the controls are fixing, washing and toning	££	moderate
Cyanotype	Ultraviolet, contact	Printing out	No silver and no fixer; alkali destroys it, light fading reverses in the dark	£	low
Salted paper	Ultraviolet, contact	Printing out	Turns almost entirely on the fixing and the washing rather than on the process	££	moderate
Albumen	Ultraviolet, contact	Printing out	Retained silver in every area, yellowing irreversibly as silver sulfide	££	high
Kallitype	Ultraviolet, contact	Developing out	Reputation rests on clearing and toning; every failure happens after the exposure	££	high
Van Dyke Brown	Ultraviolet, contact	Printing out	As the kallitype, plus Ware’s objection that the nitrate anion attacks the image	££	moderate
Palladium	Ultraviolet, contact	Developing out	An inert metal in the fibres; the everyday enemy is relative humidity	£££	high

Platinum is not in that last row and the substitution is deliberate. The design’s seventh process is “platinum or palladium”, and the course ruled on 4 September 2026 that students perform palladium and never handle platinum. The platinotype is in the atlas as a historical-study entry, and its own page carries the argument.

What the tables cannot tell you

They are not measured on one negative, which is the whole of the outstanding work above.

The colours are not colorimetry. Every colour statement in this atlas is a published description in words, and words for colour do not compare across sources reliably. A measured comparison needs reflection densitometry, which Part XV builds the instrument for.

The scales are not on one axis. The course can state Ware’s exposure scale of about 0.9 for classic cyanotype against about 2.2 for New Cyanotype, because one author measured both the same way. It cannot put those numbers beside a silver gelatin paper grade, because nobody has measured them together.

And the costs are bands, not prices. Dated figures live in the laboratory planner, and the cost of a noble-metal print in particular is a calculation the reader has to make with today’s metal price and their own measured drop volume.

Where the entries come from, and what is still missing

The register of everything this atlas owes is src/data/processes-planned.json, and npm run validate:atlases counts the written entries against it on every build, so the shortfall is a number rather than an impression. It currently reports 44 of 44.

That number needs reading carefully, because it counts pages rather than knowledge.

Fifteen entries were lifted from written parts. Part I establishes heliography, the daguerreotype, photogenic drawing, the calotype, salted paper, the cyanotype, chrysotype, argentotype, anthotype, the Bayard direct positive and the Becquerel variant; Part V owns the four hand-coating projects. Those four, and the Level A cyanotype assignment of Part I, are the only entries in the atlas that point at a lab a reader can do today.

The rest were built from the course’s own reference corpus — the chemical encyclopaedia, the formulary, the safety library, the glossary, the timeline and the sourced technical literature those were written from. That is how the toning, iron-silver, noble-metal and albumen entries come to be here while Parts XVIII to XXVI are still unwritten. Each of them opens or closes by naming the part that will eventually own it, so a later author knows what to check rather than what to assume.

Where the corpus could not carry a field, the field is empty and the page says so. Several entries carry stated gaps rather than plausible values: the image colour and tonal scale of the Bayard direct positive; the image substance and permanence of the Becquerel variant; the product of the second bath in toned cyanotype; the formulation of matte albumen; the plate-making practice of collotype and the Woodburytype; and almost every image field of the uranotype. A named gap is worth more than a filled one nobody can check, which is the same rule the rest of the course runs on.

The processes

Dichromate processes

• Carbon printDDate 1855Status historical studyLight uvA pigment image in hardened gelatin — expose a dichromated pigmented tissue, wash away the gelatin light did not harden, and the pigment goes with it. No silver anywhere, and therefore none of silver's failures.
• CollotypeDStatus historical studyLight uvInk carried by a reticulated gelatin surface whose own grain does the work of a screen — continuous tone in ink, printed from a plate rather than cast from a mould, and the Woodburytype's rival answer to the same problem.
• Gum bichromateDStatus historical studyLight uvGum arabic, watercolour pigment and a dichromate brushed onto sized paper — the process in which the print is most obviously made rather than taken, and in which the printer's hand during development is part of the method.
• Photoglyphic engravingDDate 1852Status historical studyLight uvTalbot's patent of 29 October 1852 on the light-sensitivity of dichromate mixed with gelatin or gum — the thing Ponton had missed, and the root of carbon printing, gum bichromate, photogravure and the whole chromium(VI) family.
• PhotogravureDStatus historical studyLight uvPhotography turned into an intaglio printing plate — a resist exposed and developed, the metal beneath it etched so that it holds ink by tone, and continuous tone in ink with no halftone screen.
• WoodburytypeDStatus historical studyLight uvA hardened gelatin relief pressed into soft lead to make a mould, and pigmented gelatin cast in it — so the ink layer really is thicker in the shadows, and the tone is a thickness rather than a pattern.

Iron processes

• Cyanotype (classic)BDate 1842Status practisedLight uvHerschel's iron process of 1842 — light reduces an iron(III) salt, the iron(II) precipitates Prussian blue, and plain water finishes the job because the image is already insoluble and everything else already soluble.
• Cyanotype (New Cyanotype)BDate 1995Status practisedLight uvWare's 1995 redesign — the ill-characterised citrate replaced by crystalline ammonium iron(III) oxalate, four to eight times the sensitivity, an exposure scale of about 2.2 instead of 0.9, and a single-bottle sensitiser whose keeping depends on a preservative this course omits.
• Toned cyanotypeBStatus practisedLight uvA finished cyanotype bleached in alkali and given a second substance in place of the Prussian blue — brown to black with tannic acid, and a process whose first half the course can document and whose second half it cannot.

Iron–silver processes

• ArgentotypeBDate 1842Status practisedLight uvHerschel's silver siderotype of 1842 — iron does the photochemistry and silver makes the image — and the ancestor of the Van Dyke Brown, the kallitype and Ware's argyrotype.
• Gold-toned and palladium-toned kallitypeBStatus practisedLight uvPlating a colloidal silver image with a noble metal that does not tarnish — the countermeasure to the iron-silver family's reputation, and the course's stated accessibility route into noble-metal printing.
• KallitypeBStatus practisedLight uvThe developed-out iron-silver process — the same coating as a Van Dyke Brown and a developer bath instead of a long exposure, so that the developer choice sets the image colour and one sensitised sheet can give several different prints.
• Van Dyke BrownBStatus practisedLight uvThe cheapest way into the iron-silver family — ferric ammonium citrate, tartaric acid and silver nitrate coated on paper, printed out under ultraviolet, washed and fixed in a very dilute thiosulfate bath.

Mercury processes

• DaguerreotypeDDate 1839Status historical studyLight uvA mirror with a memory — silver iodide on a polished silver surface, an exposure too short to see, mercury vapour to bring it out, and an image made of reflection rather than absorption. Studied here and never performed.
• The Becquerel variantDDate 1840Status historical studyLight visibleThe daguerreotype developed by red light instead of mercury vapour — no mercury, no bromine or chlorine sensitising, about ten times slower, and still Level D because the iodine remains.

Noble-metal processes

• ChrysotypeBDate 1842Status practisedLight uvHerschel's gold siderotype of 1842 — light reduces iron(III), the iron(II) reduces gold(III) to metal, and the print is gold in the paper fibres. A commercial failure for a century and a half, and now a controllable process because a chemist worked on the ligand rather than on the gold.
• PalladiotypeBDate 1917Status practisedLight uvThe noble-metal print a reader can actually make — the same iron photochemistry as the platinotype with palladium in place of platinum, warmer, longer in scale, much cheaper, and the metal most prints sold as platinum are really made of.
• PlatinotypeDDate 1873Status historical studyLight uvThe reference the alternative processes are judged against — iron(II) made by light reduces a platinum salt to metal in the paper fibres — and the one process in this course a student is not permitted to carry out.
• Print-out platino-palladiotypeBDate 1887Status practisedLight uvThe noble-metal print that needs no developer — a soluble iron(II) photoproduct, a sheet held at controlled humidity, and the metal reduced during the exposure so that the print can be judged by looking at it.

Other

• AnthotypeADate 1841Status practisedLight visibleHerschel's flower-juice process — a plant dye painted on paper and bleached by sunlight through a positive, taking days or weeks, giving colour, and admitting no fixer at all because the light that made the image goes on destroying it.
• Heliography on bitumen of JudeaDDate 1822Status historical studyLight uvNiépce's process, and the founding member of the family in which light hardens a coating rather than reducing a metal salt — no amplification, exposures in hours or days, and the earliest surviving camera photograph.
• Uranotype and WothlytypeDDate 1857Status historical studyLight visiblePrinting processes in which the light-sensitive material is a uranium salt rather than silver or iron — the image developed out as a coloured uranium compound, and the clearest case in the course of a process excluded on hazard alone.

Printing-out processes

• Collodion print-out paperDStatus historical studyLight uvSilver chloride precipitated inside a collodion film rather than in gelatin or albumen — the last of the printing-out papers, matte or glossy, gold and platinum toned, and built on a binder the course studies and does not handle.

Silver halide in albumen

• Albumen printBDate 1850Status practisedLight uvThe dominant photographic print of the nineteenth century — the chemistry of salted paper with the silver held in a glossy protein layer on the surface instead of sinking into the fibres, and every consequence of that one change.
• Matte albumen printBDate 1895Status practisedLight uvAlbumen made matte on purpose — fresh alkaline egg white mixed with starch or another substance instead of the acidic, partly decomposed albumen the glossy papers use, and a surface that platinum toning suits better than gold.

Silver halide in collodion

• AmbrotypeDDate 1851Status historical studyLight uvA wet-plate collodion negative on glass, deliberately exposed and developed thin, then backed with black so that it reads as a positive — and a failure that looks like fading and is not.
• Tintype (ferrotype)DStatus historical studyLight uvA collodion positive on a sheet of iron lacquered black — the cheapest photograph the nineteenth century made, unbreakable, flat enough to post, and containing no tin at all.
• Wet-plate collodion negativeDDate 1851Status historical studyLight uvThe process that made photography ordinary — a glass plate flowed with collodion carrying soluble halides, dipped in silver nitrate, and exposed and developed before it dries, because a dried collodion film is impervious to the developer.

Silver halide in gelatin

• Dry plate and lantern slideBDate 1871Status practisedLight eitherA gelatin silver emulsion on glass, dried and usable at any time afterwards — the material that ended the wet plate, turned the sensitive layer into a manufactured product, and is coated by hand in Part V on a levelled bed.
• Gold toned silver gelatin printBStatus practisedLight eitherPlating the image with a metal that does not tarnish — three different jobs from one salt, a bath that both deposits gold and dissolves the silver it is plating, and a colour shift that can be almost nothing.
• Hand-coated bromide paperBStatus practisedLight uvThe first washed emulsion — precipitate, set, rice, wash in six changes of ice water, ripen and coat — and the project where the sulfur digestion is moved out of the procedure and into the experiment because no published home formula performs one.
• Hand-coated chlorobromide paperBStatus practisedLight uvThe same five grams of silver as the chloride project with 6 mol per cent of the halide as bromide — and the controlled series that shows halide ratio setting speed and scale while the developer sets the colour.
• Hand-coated silver chloride contact paperBStatus practisedLight uvThe first emulsion the course makes — five grams of silver nitrate whisked into salted gelatin in three portions, ripened for half an hour, and coated unwashed on wet watercolour paper, so that everything the make left behind stays in the sheet.
• Iron-blue toned silver gelatin printBStatus practisedLight eitherThe cyanotype's pigment made inside a silver print — a ferricyanide bleach, an iron(III) salt, and Prussian blue deposited exactly where the silver was. A colour decision, and by the manufacturer's own statement not an archival one.
• Selenium toned silver gelatin printBStatus practisedLight eitherThe only toner in common use that can be run for permanence alone on a paper that will barely change colour — so the print looks untouched and is nonetheless a different substance.
• Silver gelatin developing-out print on fibre baseAStatus practisedLight eitherThe standard black-and-white print of the twentieth century — a developed latent image in a gelatin emulsion over a baryta layer on a paper base that soaks, which is why it takes an hour to wash and why it can be made to last.
• Silver gelatin developing-out print on resin-coated baseAStatus practisedLight eitherThe same emulsion on a paper core sealed both sides with polyethylene — titanium dioxide doing baryta's job, a four-minute wash instead of an hour, and a permanence argument that is different rather than simply worse.
• Sulphide (sepia) toned silver gelatin printCStatus practisedLight eitherBleach the developed image back to silver bromide, then convert it to silver sulfide — a toner that changes the image substance rather than plating it, and whose colour was decided before the print reached the tray.
• Uranium toned printDDate 1865Status historical studyLight eitherA silver image turned red-brown and intensified at the same time by a uranium ferrocyanide deposited where the silver was — a print whose colour you can wash off, and photography's one radioactive reagent.

Silver halide on plain paper

• Arrowroot and albumenised salted paperBStatus practisedLight uvThe two ways of putting a binder into a salt print without making it glossy — a starch cream carrying citric acid because starch is not chemically active, and a thin coat of egg white on plain salted paper.
• Bayard direct positiveBDate 1839Status practisedLight uvSilver chloride paper blackened all over, soaked in potassium iodide and exposed moist so that light bleaches instead of darkening — a positive straight out of the camera, on paper, and unprintable for exactly that reason.
• Calotype negativeBDate 1840Status practisedLight uvThe first developed latent image on paper — silver iodide excited with gallo-nitrate of silver, exposed for a second, brought out with more gallo-nitrate, and worth about a hundredfold in speed over printing out.
• Photogenic drawingBDate 1834Status practisedLight uvTalbot's first process, and the one whose secret is an apparent mistake — a deficiency of salt, so that every crystal of silver chloride sits in a bath of unreacted silver nitrate, which is what lets the darkening continue instead of stalling.
• Salted paper printBDate 1839Status practisedLight uvThe first silver print on paper and the ancestor of every print in this course — silver chloride formed among the fibres with an excess of silver around it, printed out by inspection, gold toned after 1847 and fixed in thiosulfate.