Toning, Permanence and What the Evidence Actually Supports

A print has been toned. What are you now entitled to say about how long it will last?

That is a narrower question than it looks, and answering it honestly is the hardest thing in this part. Everything up to here has been chemistry: what the bath does to the particle, what colour comes out, what the hazard is. This page is about evidence — which of the things everybody says about toning and permanence has been measured, which is a manufacturer’s claim in a document selling toner, which is a mechanism that ought to be true, and which is the accumulated impression of people who have looked at a great many old prints. Those are four different kinds of statement. The trade runs them together, and a printer who runs them together will overclaim in good faith.

!What this page inherits, and does not repeat

Part XII owns the deterioration mechanisms, the storage numbers and the standards table; it names protective toning in one paragraph and hands the subject here. The mechanism page owns what a toner does to a particle, and the sulfide, selenium and gold and other metals pages own the individual chemistries. What is left, and what is here, is the argument about evidence.

The argument in one line, and where it is weakest

Metallic silver is attacked by oxidants. Silver sulfide and silver selenide are attacked less readily. Convert the image from the first into the second, and the image should last longer.

That is the whole case, and as a piece of chemistry it is sound. James Reilly, writing from the conservation side, gives the mechanism for the noble-metal version of it and states the physical reason it works at all: a printed-out image is made of very small, highly dispersed particles, so a large portion of the metal’s total mass is on the surface and therefore accessible to destructive chemical agents. A layer of gold or platinum “will tend to shield the silver inside from attack, especially from oxidising agents”. Notice the hedging in a Tier 1 source: a measure of protection, will tend to shield. He is describing a mechanism, and he does not overstate what a mechanism establishes.

The weakness is not in the chemistry. It is in the word convert. The sentence treats conversion as something that either has or has not happened, and in a tray it is neither: it is a reaction that starts somewhere in the tonal scale and proceeds, and that you stop when the print looks right or when the clock says so. A print described as “selenium toned” is a print that is part silver and part silver selenide, in proportions nobody measured. Every claim on this page has to survive that fact, and most of the interesting difficulty is downstream of it.

What the untoned print is up against

Briefly, because Part XII teaches this properly and this page only needs the list.

Residual thiosulfate left by inadequate washing supplies labile sulfur, which converts image silver to silver sulfide slowly and unevenly — the print yellows and loses density. Residual silver left by inadequate fixing is distributed with the unexposed halide, so its stain appears in the highlights and borders while the image looks untouched. Atmospheric oxidants attack from outside: the Library of Congress names nitrogen oxides and ozone as the two main oxidant gases, notes that ozone is also produced by some electrostatic copiers and printers, and adds that nitrogen and sulfur dioxide reacting with atmospheric water give nitric and sulfuric acids which make silver images fade. Peroxides from untreated wood, paints and varnishes do the same from within the frame. The visible results are silver mirroring and coloured microspots.

Two of those four are yours and finite; two are the environment’s and never stop. Toning addresses the second pair, by changing what the oxidant has to attack. It addresses the first pair not at all — a point the page returns to, because it is the one most often got wrong.

The standard written for exactly this question

There is one. ISO 18915:2000, Imaging materials — Methods for the evaluation of the effectiveness of chemical conversion of silver images against oxidation. First edition, published 14 December 2000, sixteen pages; adopted by BSI as BS ISO 18915:2000 in March 2001. Its existence is itself worth noticing: an international committee decided that “does converting the image protect it” was a question needing a repeatable answer, and wrote a method for producing one.

Its published scope says what it does and, more usefully, what it refuses to do.

• It describes methods for evaluating the effectiveness of chemical conversion treatments intended to increase the resistance of wet-processed silver images to oxidation. The treatment may be part of the original processing or applied afterwards — so both a combined toner-and-washing-aid bath and a separate toning tray are in scope.
• It applies to silver-gelatin images on plastic, paper or glass. A fibre-base print is squarely inside it.
• It does not recommend general or specific treatments. It will not tell you to use selenium.
• Treatment temperature, times and replenishment rates are outside its scope. It will not tell you how long to tone.
• Two test methods are described: the dichromate bleach test and the hydrogen peroxide incubation test.

Read those last three together and the shape of the thing is clear. ISO 18915 is not a specification that a print can meet; it is a measuring instrument. It gives two ways of challenging a converted image with an oxidant under controlled conditions and reading what happens. What you do with the reading — whether you call it good enough — the standard leaves to whoever is asking.

!What this course has read, and what it therefore will not print

Under the course owner’s ruling of 4 September 2026, recorded in Part XII’s standards section and applied by Part XIII to film speed, this course does not buy the ISO standards and quotes no part of their normative text: it names the standard by number, states its own criterion in public where a reader can check it, and labels every figure derived under that criterion as its own. For ISO 18915 even the usual fallback failed: ISO’s own catalogue page refused every request on 6 September 2026, and the bibliographic data and scope clause above were read from two standards distributors’ product records — the ISO original and the BSI adoption — which reproduce the scope word for word and agree with each other. The BSI record gives the status as “Standard” rather than withdrawn, and the course reports that as the catalogue record’s word rather than as a check of ISO’s own database.

So no pass criterion, no reagent concentration, no incubation condition and no result from ISO 18915 appears anywhere in this course, and no page will tell you that a print “meets ISO 18915”. That is a real limit on what this lesson can teach, and stating it is more useful than a confident paraphrase would be.

The question also has a standards lineage older than the ISO number, which is worth knowing because it shows how long it has been treated as measurable. The Library of Congress’s own standards bibliography lists ANSI PH4.32-1986, “methods for evaluating processing with respect to the stability of the resultant image — black-and-white papers”, alongside ANSI PH4.8-1985 for residual thiosulfate, which is the ancestor of the residual-hypo method Part XII could not verify. This course has read neither, names them because a reader meeting the old designations should know what they are, and takes nothing from them.

There is one more thing to notice about the dichromate bleach test, and it is not a quibble. A dichromate is a chromium(VI) compound, which this course’s chromium policy excludes at every level, in any quantity, on a hazard classification that has nothing to do with photography. The standard method for demonstrating that your print is protected is therefore one you could not run in a home darkroom even if you owned the document. The hydrogen peroxide route is the other half of the standard and is a different matter, but the course has not read its conditions and will not guess them. Testing image stability is laboratory work, and this is where the boundary actually falls.

The manufacturers’ claims, examined rather than repeated

Here is Kodak’s, in full, because the wording is the thing. Toning “converts the black-and-white silver image to an inert compound, which reduces the harmful effects of intense light, ultraviolet radiation, oxidizing gases, extremes of temperature and humidity, and fumes”, and — the sentence that does all the work — “All Kodak toners will protect the image whether or not they produce a color shift.”

Taken at face value that is a claim about kind: toning is protective, full stop. But the same sheet, four pages later, is more careful than its own headline. Rapid Selenium Toner has a dilution specified for print protection: 1+20, printed there as 1:20. In the combination bath where the toner is diluted in working-strength hypo clearing agent, the instruction is to tone “approximately 3 minutes for print protection or longer for a tone change”. And the paper-and-toner table’s last row pairs the heading No Tone Shift with Rapid Selenium at 1+20 — so by Kodak’s own tabulation the protective dilution is the one that shifts no tone. It is not the one that changes nothing at all: the same page records that 1+20 or 1+40 increases shadow contrast and maximum density, which is what the selenium experiment measures. The colour holds still; the shadows do not.

HARMAN’s sheet for their selenium toner is built the same way: 1+3 for normal toning, 1+20 “where the primary reason is for protection of the image, with minimal tone change”, and protection at that dilution “will be complete in 2–4 minutes”.

⚗Why a protective dose exists at all

If every toner protected the image simply by being a toner, there would be no reason to specify a dilution and a time for protection separately from the ones for colour. That both makers do specify them is the strongest internal evidence in either document that protection is a matter of degree, reached by a particular treatment, and not a property conferred by the bath’s mere presence. The headline sentence is a claim about kind; the instructions underneath it are a claim about degree. The instructions are the part written by someone who had to make it work.

ILFORD approach it from the other end, and their wording is narrower and more useful. Prints made for display, they say, are recommended to be toned to protect them from the oxidising gases found in many environments — a named threat rather than a general benefit. Selenium is recommended because it has little effect on the image colour of MULTIGRADE FB CLASSIC; sulphide toning and silver image stabilisers are named as other protection methods. Their warmtone sheet lists which toners give a protective effect — sulphide, polysulphide, and “some metal replacement toners (gold and platinum)” — and then does something no marketing document has to do: it names toners that do not.

The Library of Congress, from outside the industry, puts it in a single clause, and puts it in context: the paragraph it appears in is about processing, and its main instruction is to insist that photographic chemical processing and development be done to ISO standards, ISO 18901. Only then: “use of chemical toners also helps to protect silver images from deterioration”. Toning is offered as an addition to correct processing, never as a substitute for it — which is the next section. IPI’s negatives booklet describes toning as a stabilising treatment and then adds the sentence that keeps the whole subject honest: “in practice this step was and is rarely performed”.

Partial conversion, which is where the argument gets difficult

Now the hard part. Nobody disputes that toning to completion converts the image. What happens in an ordinary darkroom is not that.

Wolfgang Moersch, who manufactures and sells toners, publishes the diagnostic that makes the problem visible. Tone a print, then bleach it. The bleach attacks metallic silver and leaves converted silver alone, so what survives the bleach is the part that was toned. His account: if a red-brown image remains “in which even the highlights show enough tonality”, the goal is reached; if the highlights, “maybe even down to the mid tones — appear worn out, toning was too short to give the print maximum protection from the environment”. He records that six minutes at 1+10 still leaves silver unconverted, and he has a name for the residue in the highlights — a small membrane around the grain, which he calls “hidden selenium toning”.

Put that beside his rule that selenium always reaches the higher densities first, so that “it is a question of toning time whether the mid tones are protected or not”, and the difficulty is exact:

Where a short selenium bath leaves the print, and why that is the wrong way round

1. 1Shadows: most silver, largest depositsconvert first, and are the part of the scale least at risk from oxidation because there is most of them
2. 2Mid tonesconvert next; whether they convert at all is decided by the toning time
3. 3Highlights: least silver, finest depositsconvert last or not at all — and are the most vulnerable, because a finely divided deposit has the most surface per unit of mass
4. 4The consequencea print toned briefly for colour may have protected the part that needed it least; the visible change and the protective change are not the same change

The order is Moersch's, stated flatly and irrespective of dilution or temperature. The vulnerability argument is Reilly's surface-area reasoning applied to the tonal scale, which is this course's inference and not a claim either author makes in these words.

Moersch also reports the contrast that matters for choosing a route: with sulfur toning, “even with short toning times the highlight densities receive a stabilizing effect”, and after three minutes at 1+10 the protective effect is “considerably higher than with selenium toner in a similar dilution for the same duration”, because selenium would not have reached the upper mid tones. That is a practitioner’s comparison by an interested party, made by bleaching prints and looking at them. It is evidence of a kind. It is not a measurement of protection.

⁇The study this page needs, and could not find

The central claim of this lesson — that partial toning gives partial protection, proportionally — is not established by any source this course holds.

What would establish it is a series of prints from one negative on one paper, toned for graded times in one bath, challenged under ISO 18915 and read out. That experiment is obvious, the standard for it has existed since 2000, and the course has not found it published. The corpus contains manufacturers’ protective dilutions and times with no data behind them; Moersch’s bleach-back, which shows unconverted silver but measures no stability; Getty’s statement that selenium was “recognized for its ability in increasing the chemical and environmental stability of silver particles”, with no study cited; and IPI’s remark that polyester film “together with toning” can give an extraordinarily long-lived record, whose footnote points back to the same author’s storage guide rather than to a study of toning.

So the page states it as a mechanistic inference and labels it one. Conversion is partial; protection depends on conversion; therefore protection is partial. Each step is reasonable and the chain has not been tested. If a reader knows of the study, the course wants it.

The practical corollary survives the uncertainty, and is the most useful sentence on this page. Protection and colour are different endpoints, and the protective one is the number the maker printed. If you want the picture warmer, tone by eye against an untoned reference print. If you want the image protected, use the dilution and the time the sheet gives for protection — 1+20 for two to four minutes for HARMAN’s toner, three minutes in Kodak’s combination bath — and accept that you will see nothing happen. Judging a protective treatment by eye is judging it by the wrong variable.

Toning cannot fix bad processing, and Kodak’s own account is the evidence

“Toning cannot disguise poor print quality,” says G-23, immediately after describing what toners do. That reads like a quality homily. It is not; it is a list of failure modes, and the rest of the document supplies them. What a toner actually does to a badly processed print is develop the faults.

The processing fault	What the toner does with it
Exhausted fixer, leaving insoluble silver compounds washing cannot remove	They form a dark yellow stain on meeting the toner, “especially noticeable in print borders and highlights”
Air bubbles trapped between or under prints in the fixer	Round purple stains in prints toned in selenium or sulfide toners
Fixing for longer than recommended, so fixer penetrates the base	Prints toned in selenium or sulfide toners turn yellow
A hardening fixer	Kodak state that it makes the paper emulsion less receptive to the toner solution, and recommend a non-hardening fixer for prints intended for toning
An overconcentrated stop bath, or poor agitation in its first seconds	Mottle in the base, not evident until the print is toned
Residual silver salts and traces of hypo from a short wash	Stains, uneven tones and fading

Every one of those is invisible on the untoned print and permanent on the toned one. It is why “improper fixing is probably the major cause of stains in toned prints” is the sentence Kodak leads the Fixing paragraph with, and why the ordering rule on the sulfide page matters: ILFORD’s instruction is to run the optimum permanence fixing and washing sequence first and then tone — an instruction they give for the protective toners other than selenium, selenium having a published sequence of its own in which the toner replaces the first wash.

There is a second, sharper version of the same point, and it belongs to the bleach-and-redevelop route. Moersch: a print in which not-yet-toned metallic silver has been bleached to silver bromide is no longer archival, because that silver bromide would tone uncontrolled over time by gas action. It must either be dissolved by fixation or converted by toning again. A half-finished conversion is not a half-protected print; in that particular case it is a print in worse condition than the one you started with.

📓What has to be true before a permanence claim is worth making

Run the two tests in the residual hypo and silver SOP on a control sheet carried through the same session, against Part XII’s stated criterion, and record the result. A toned print with residual fixer in it is not a protected print; it is a stained one waiting to declare itself. The wash-testing lab is where the technique is taught.

Comparing the three protective routes, as far as the evidence goes

Sulfide, selenium and gold each carry a protection claim. Here is what actually stands behind each one, with the kind of evidence named — which is the point of the table.

Route	The claim	Who makes it	What kind of statement it is
Sulfide (sepia, polysulfide, brown)	Where great permanency is required, prints should preferably be toned to a silver sulfide image	Wall, 1924 — and he says why: “experience having shown this form of silver to be one of the most stable”	Accumulated trade experience, explicitly labelled as such by its author
Sulfide	Even short toning gives the highlights a stabilising effect, unlike selenium	Moersch, from bleach-back tests on his own products	A practitioner’s observation by an interested party; no stability measurement
Selenium	Recognised for increasing the chemical and environmental stability of silver particles; entered archival processing for the art and museum market	Getty Conservation Institute	A conservation body’s historical account of why the practice was adopted; no study cited
Selenium	Enhances the archival stability of prints; 1+20 for 2–4 minutes	HARMAN	A manufacturer’s claim with a specified dose and no data
Selenium	More effective routes exist — gold, sulfur, Sistan — though weak selenium toning beats none	Moersch	A maker arguing against his own product’s main selling point
Gold	Provides print protection while changing image tone only slightly (GP-1)	Kodak	A manufacturer’s claim
Gold	Coats the silver with a layer “quite impervious to attack”; albumen prints so treated show little evidence of fading	Ware, on the Fading Committee’s recommendation and on surviving objects	A historical recommendation plus a survivorship observation
Gold and platinum	A measure of protection, by partially replacing and enclosing the silver; the layer “will tend to shield” what is inside	Reilly	A mechanism, hedged by its author

Read the right-hand column downwards. Not one entry is a controlled comparison of toned against untoned samples under a stability test. That is the honest state of the question in this course’s corpus, and it does not mean the practice is unfounded — a mechanism plus a century of trade experience plus two conservation bodies’ endorsement is a reasonable basis for doing something. It means the practice is not quantified, and that a printer who says “selenium toning doubles the life of the print” has made a number up.

🕰Reilly's 85 per cent, and why it is a model of honesty

Reilly writes that approximately 85 per cent of extant albumen prints made after 1860 show moderate to severe highlight yellowing, and probably 95 per cent of those from the 1850s. Then he adds: “These figures are not based on any formal statistical sampling, merely on the accumulated experience of the author in examining prints and through discussions with curators and collectors.” And then he specifies the study that would settle it — reflection densities in image and non-image areas through red, green, blue and visual filters, on a statistically significant sample across several collections, rechecked at two, five and ten year intervals.

A number, its provenance, and the experiment that would replace it, in one paragraph. That is the standard this course holds itself to when it prints a figure it did not measure.

Reilly supplies the one comparative claim in the whole table that rests on more than mechanism, and it is a historical one: separate toning in alkaline gold chloride, followed by fixation in fresh strong hypo, “allowed albumen paper to attain a much better record of resistance to fading than had been accomplished with the older plain salted papers”, because the alkaline toners “deposited more gold than the other toning methods”. Even here he immediately complicates it — the albumen layer itself also protected the silver — so the improvement cannot be attributed to the gold alone. And he closes the chapter by warning that “much research has been done concerning the permanence of gelatin-based printing papers, but there is no certainty of the applicability of these results to albumen-based materials”. The transfer problem runs in both directions, and it is why a nineteenth century albumen result is not evidence about your fibre-base bromide print.

Why accelerated ageing settles less than it seems to

Any stability test that gives an answer this year is an accelerated test, and IPI — who build their institutional life on such tests — are the most explicit about their limits. Of their acetate storage predictions they write that the numbers “are based on extrapolations from accelerated aging” and “should be seen as a convenient way to quantify and express how good or bad a storage environment is for preventing the vinegar syndrome, not as literal predictions of how long a collection will last”. For the numbers to be read literally, a collection “would have to reproduce exactly the circumstances under which the aging experiments were done”. Their analogy is an actuarial table: it says nothing about any individual and a great deal about the population.

That passage is about cellulose acetate and a hydrolysis reaction, not about toned silver and an oxidation reaction, and this course’s extension of the caution to toning is its own inference. But the logic is not specific to a failure mode. An accelerated ageing result ranks treatments under one set of conditions; it does not convert into years on your wall. This is exactly why the ISO 18915 scope reads the way it does. It is a comparison method. It was never going to give you a lifetime.

The counter-example: toning that changes the chemistry without improving the stability

If toning protected because it is toning, every toner would protect. ILFORD say plainly that some do not: “Other metal replacement toners such as blue (iron) and red (copper) toner may not give extra protection and the image might fade. Dye toners do not give extra protection.”

Two Tier 1 manufacturers, one writing that all its toners protect whether or not they shift the colour, the other naming toners that may not protect and may fade. That disagreement is not a detail; it is the reason this page exists. On this course’s reading it is ILFORD whose statement fits the chemistry, because iron-blue and copper toning are not conversions of silver into a more stable silver compound at all — they build a new coloured compound where the silver was, with its own stability, which is a different question from the silver’s.

For iron-blue the answer to that different question is unusually clear, and it is worse than neutral. Prussian blue is destroyed by alkali: the gold, iron and other metals lesson carries the chemistry, Ware’s three destruction pathways and Holtzman’s figure — a buffered solution at pH 9.4 decolourises it in one to ten minutes, and pH 9.4 is the pH of saturated calcium carbonate, the buffer put into conservation board on purpose. So an iron-blue toned print is not merely unprotected. It is incompatible with the standard archival kit: with buffered mount board, with a carbonate washing aid, with alkaline tap water and a long wash. A protective decision taken for the silver print becomes a destructive one for the blue one, which is the sharpest illustration in this part of why “archival practice” is not a single recipe.

📷The other half of permanence, referenced not repeated

Even a perfectly toned print spends its life in an environment, and the environment is the half that Part XII and Part XIX’s mounting page own between them: relative humidity and temperature, pollutants, light, and the materials in contact with the sheet.

Two numbers from the Library of Congress are worth carrying here, because they bear directly on the threat toning addresses. Permanent display is not recommended; exhibition light should be 30 to 100 lux, with ultraviolet not exceeding 75 microwatts per lumen. And the enclosure question has an answer with a number on it: the Photographic Activity Test, ISO 18916, puts a material in prolonged contact with two detectors, one screening for the oxidation and reduction reactions that cause image fade, silver mirroring and red or gold spots, the other for chromophores that yellow the support. ISO 18902 requires the PAT and a pH between 7.0 and 9.5, at least 2 per cent calcium carbonate and lignin-free stock. IPI are explicit that passing only the PAT, or only the pH requirement, does not make a material photo-safe — and that photo-safe covers chemical reactivity only, not abrasion, creases or tears.

Does “archival” mean anything?

It is worth asking, because the word is on every second product in a photographic supplier’s catalogue and it is the word a printer reaches for when writing about their own work.

IPI answer it directly. “Acid-free”, “archival”, “museum-quality” and “conservation board” are, they write, “not standardized or legal terms but are simply marketing terms”. The Library of Congress say the same about enclosures: many are labelled archival or acid-free, and what to look for instead is the claim that the material meets ISO 18902, which includes the PAT. IPI’s framing guide goes further and notes that “some ‘acid-free’ products may still fade or yellow a photo”.

So: “archival” is not defined by any standard this course can find, and a claim using it is unfalsifiable. A claim naming ISO 18902 or ISO 18916 can at least be checked with the manufacturer.

IPI make the same move about a different product in a sentence worth carrying across to toning. Some enclosures, they note, “advertise additional protection from airborne pollutants and/or excessive humidity, although standards to evaluate their effectiveness are lacking”. That is precisely the position toning was in before ISO 18915 existed, and — because the course could not find the standard applied to a published comparison of toners — it is close to the position toning is still in.

!The course's own use of the word

This course uses “archival” in exactly one way and it is worth being explicit, because otherwise the paragraph above convicts several of its own pages. It never means “will last a long time”. It names a published procedure: ILFORD’s “optimum permanence” fixing and washing sequence, which is what Part XII’s archival processing sequence is built on, and Getty’s “archival processing”, which is that same class of practice as adopted by the art and museum market. Note that ILFORD’s own phrase is optimum permanence, not archival, and that they define it as a sequence rather than as an outcome — the standard recommendations give “excellent print permanence for all commercial needs”, and the longer sequence is offered “when optimum permanence is needed”. Two named tiers, each a procedure, neither a promise in years.

Writing an honest permanence statement

A print you sell, exhibit, accession or give away leaves your control, and whatever you say about it travels with it. A permanence statement is what you are entitled to say, given what you actually did. It has four parts, and they must not be allowed to blur into one another.

The four parts of a permanence statement, strongest first

decreasing certainty

1. What was donematerials, processing sequence, toner, dilution, time, temperature, washfact
2. What published procedure it followsnamed by publisher and document, not by the word "archival"fact
3. What was tested, and against what criterionthe residual hypo and silver tests, the criterion used, the resultmeasurement
4. What is inferencethat the toning protects; that the print will last; anything about yearsclearly labelled

The order is the course's own. The rule it encodes is that a reader must be able to tell which sentences they could check and which they are being asked to take on trust.

Here is a statement built that way, for a fibre-base print toned for protection. Every figure in it comes from a cited source or from a test the maker actually ran.

Printed on ILFORD MULTIGRADE FB CLASSIC and processed in ILFORD’s optimum permanence sequence with selenium toner as published in the paper’s technical information sheet: RAPID FIXER at 1+4 for one minute, toning in HARMAN SELENIUM TONER diluted 1+20 in working-strength ILFORD WASHAID (the sequence names the diluent but not the ratio, the toner sheet gives the ratio but not the diluent, so pairing them is my own reading of two documents), ten minutes in WASHAID with intermittent agitation, and a final wash of thirty minutes in running water. The 1+20 dilution is the one HARMAN specify where the primary reason is protection of the image; toning was for four minutes, the upper end of the two-to-four minutes they state for protection at that dilution. A control sheet carried through the same session was tested for residual thiosulfate and residual silver by the course’s stated criterion and passed. Toning was carried out for protection against oxidising gases, which is what the manufacturers recommend it for; the degree of conversion achieved was not measured, and no claim is made about how long the print will last. Recommended framing: materials meeting ISO 18902, glazing that blocks ultraviolet, display below 100 lux, no permanent display.

And here is the same print oversold, which is what most statements look like:

Museum-quality archival fibre print, selenium toned for permanence. Guaranteed to last 100 years.

Four failures in fourteen words. “Museum-quality” and “archival” are marketing terms with no standard behind them. “Selenium toned” does not say at what dilution or for how long, which is the only part that bears on protection. “For permanence” states an outcome where the evidence supports a mechanism. And the hundred years is invented — no source in this course’s corpus, manufacturer or conservation body, attaches a number of years to a silver-gelatin print on the strength of its toning, and the one standard that measures conversion effectiveness explicitly declines to.

✕The four sentences to strike from your own writing

“Archival.” Name the sequence instead. “Museum quality.” Meaningless; the museums say so. “Will last N years.” Nothing in the literature supports a figure for a toned silver-gelatin print. “Fully toned.” Only if you bleached a test print and looked — and if you did, say that, because it is the most informative sentence you could write.

The assignment at the end of this part asks for exactly this document alongside the print, and it is marked on whether the four parts stay separate.

The conservator’s view, and what identification tells them

The last thing to understand is that your print will eventually be read by somebody who did not make it, and who has instruments.

Toning is identifiable, and mostly by the metal it leaves behind. Getty’s atlas sets out what X-ray fluorescence sees: gold in a gold-toned print, whatever colour it produced — including the bright reds and the surprising blues that gold toners can give on developing-out paper. Uranium by two peaks at 13.64 and 17.22 keV, confirmable in seconds with a sensitive radiation meter, the example print reading about 117 microroentgen per hour at a centimetre against a background of about ten: well above background, and, Getty are careful to add, not high enough to matter in occasional handling. Prussian blue by a higher iron concentration in the Dmax area than the Dmin. Copper, in a red copper-toned print, alongside the iron of the ferricyanide it was precipitated with.

Two cases are harder, and both are instructive. Sulfur toning is the awkward one: the sulfur signal is dominated not by the silver sulfide of the image but by the barium sulfate of the baryta layer under it, which is almost always present, so distinguishing the source is, in Getty’s words, “extremely difficult”. A sepia-toned print may not be identifiable by its chemistry at all — only by its colour, which is exactly the evidence a conservator distrusts. Selenium is the easy one, and surprisingly so: XRF is sensitive enough to detect selenium “even in photographs toned for a very short period of time as part of archival processing”, so a treatment that changed nothing visible still leaves a signature. Getty add the observation that properly done archival selenium toning caused no major colour change and only a slight increase in contrast and brilliance — the same effect the selenium experiment measures on a step wedge.

Why does the identification matter to whoever holds the object? Because it changes the storage decision.

• A print identified as iron-blue toned must not be buffered. Everything a conservator would normally reach for — alkaline-reserve board, a carbonate-buffered enclosure — is the specific agent that destroys it.
• A print identified as selenium or sulfide toned is a silver-gelatin print whose image silver is partly converted, and is handled as one; the identification tells the conservator that a yellowing they observe may be the original toning rather than deterioration, which is the difference between leaving an object alone and treating it.
• A print identified as uranium toned is recorded as such and, on Getty’s measurement, handled by standard conservation procedures.

AIC’s photographic materials catalogue makes the same point from the identification side: both silver-gelatin printing-out and developing-out photographs “can be toned to different colors with the addition of a toning process”, using selenium or gold among others, so the colour of a silver-gelatin print is never by itself a process identification. IPI’s Graphics Atlas goes further in practice and catalogues toned prints as object types of their own — a polysulfide-toned postcard, a selenium-toned stereoview, a sulfur-toned studio portrait each get their own entry alongside the untoned examples. The mirrored copy this course holds is the tour index rather than the object pages, so that is as far as it can be pressed, but the cataloguing decision is itself the point: to a conservation institution a toned print is a different kind of object, not a variant finish.

📓What you can do that no instrument can

XRF will tell a future conservator that your print contains selenium. It will not tell them the dilution, the time, the temperature, whether the print was toned before or after the washing aid, or what paper and developer preceded it. You are the only source for that, and only if you write it down and keep it with the print. Record it in the lab notebook and, for a print that leaves your hands, put the processing record on the reverse or with the permanence statement. It is the cheapest contribution to photographic conservation available to anybody in this course.

Alternative route

This page needs no darkroom to read and prescribes no procedure, so there is nothing here to route around on facilities grounds. But two of its obligations can be met by a reader who never tones a print, and one cannot, so it is worth being exact.

What is fully available without a darkroom. The whole of the evidence argument, and — more practically — the entire second half of permanence. Enclosures, board, framing, glazing, light level, relative humidity and pollutants are decisions made in a living room, not a darkroom, and by IPI’s own account they matter at least as much as anything in the trays: IPI’s Media Storage Quick Reference states that enclosures cannot overcome deficiencies in the storage climate, and that while providing more inert enclosures is desirable, improving the climate is more effective overall. A reader with prints they did not make — bought, inherited or given — can apply every storage and framing paragraph on this page and in Part XII today.

What is fully available without toning. Reading a permanence claim critically — on a print you are buying, on a supplier’s enclosure, on a gallery’s condition statement — needs no equipment at all. That skill is the point of this page more than the toning is.

What genuinely is not available. You cannot verify a protection claim about your own print. Nor can anybody else in a home darkroom: the standard method needs a laboratory and, in the dichromate case, a reagent this course excludes outright. The honest position is not a workaround but a limit, and the response to it is the one this page has argued for throughout — state the treatment, not the outcome. A statement that says what was done can be checked by anybody. A statement that says how long the print will last could not be checked by you even with the darkroom.

Recap

• The argument is sound and the evidence is thin. Metallic silver is attacked by oxidants; silver sulfide and silver selenide less readily; converting the image should therefore help. That is a mechanism, and mechanisms are not measurements.
• ISO 18915:2000 exists and is the standard written for this exact question. It describes two ways of challenging a converted image — a dichromate bleach test and a hydrogen peroxide incubation test — and explicitly recommends no treatment and prescribes no time or temperature. It is a measuring instrument, not a specification a print can meet. This course has read its scope and not its normative text, and prints no criterion or condition from it.
• The manufacturers’ headline and their instructions say different things. Kodak’s “all Kodak toners will protect the image whether or not they produce a color shift” is a claim about kind; the 1+20 protective dilution and the three-minute protective time in the same document are a claim about degree. HARMAN’s sheet is built the same way. ILFORD name a specific threat, oxidising gases, and name toners that do not protect at all.
• Conversion is partial and unmeasured in ordinary practice. Selenium works from the shadows up, so a short bath may protect the deposits least at risk and leave the highlights — the finest, most vulnerable silver — least converted. Moersch’s bleach-back makes it visible; nothing in this course’s corpus quantifies it. No published study of partial versus complete toning was found, and the course states the proportionality as an inference.
• Protection and colour are different endpoints. Use the maker’s protective dilution and time and expect to see nothing. Judging a protective treatment by eye is judging the wrong variable.
• Toning develops processing faults rather than hiding them: exhausted fixer as a yellow stain in borders and highlights, trapped air bubbles as round purple spots, prolonged fixing as overall yellowing, an overconcentrated stop bath as mottle. Test the wash and the fix before toning, not after.
• The three protective routes rest on trade experience, manufacturer claims, historical survivorship and mechanism — not on a controlled comparison. Wall’s sulfide, Getty’s selenium, Reilly’s and Ware’s gold are all worth acting on and none is quantified.
• Accelerated ageing ranks conditions; it does not predict years. IPI say so of their own predictions, and this course extends the caution to toning as its own inference.
• Iron-blue and copper are the counter-example. They change the image’s chemistry without improving its stability, and Prussian blue introduces a new incompatibility with the buffered board and carbonate wash aid that protect a silver print.
• “Archival” is a marketing term with no standard behind it, on IPI’s and the Library of Congress’s own account. Name the sequence, name the standard number, name the test, and label the inference as one.

Check your understanding

Question 1. A supplier advertises a fibre-base print as "archival, selenium toned for permanence, guaranteed 100 years". Which single claim in that sentence could in principle be checked?

• "Archival", because the term is defined by ISO 18902
• "Selenium toned", because the presence of selenium can be detected by XRF even after a very short treatment
• "For permanence", because the manufacturers state that toning confers permanence
• "Guaranteed 100 years", because accelerated ageing tests produce life expectancies in years

Show the answer and why

Answer: "Selenium toned", because the presence of selenium can be detected by XRF even after a very short treatment

Getty record that XRF is sensitive enough to detect selenium in prints toned for a very short period as part of archival processing, so the bare fact of a selenium treatment is checkable on the object. Nothing else in the sentence is. IPI and the Library of Congress both state that "archival" is a marketing term neither standardised nor legal; "for permanence" states an outcome where the evidence supports only a mechanism; and no source in this course attaches a number of years to a silver-gelatin print on the strength of its toning. Note what XRF still cannot tell anyone: the dilution, the time, or how much of the image was converted.

Question 2. Two prints from one negative are toned in the same selenium bath, one for 45 seconds and one for four minutes at the protective dilution. What does this course say about their relative protection?

• They are equally protected, because the manufacturer states that its toners protect the image whether or not a colour shift is produced
• The four-minute print is better protected, and the course can say by how much from published measurements
• The four-minute print has probably converted more of the scale, including mid tones the short bath would not reach, but the course found no study quantifying protection against degree of conversion and labels the proportionality an inference
• Neither is protected, because protection requires complete conversion to silver selenide

Show the answer and why

Answer: The four-minute print has probably converted more of the scale, including mid tones the short bath would not reach, but the course found no study quantifying protection against degree of conversion and labels the proportionality an inference

Moersch states that selenium always reaches the higher densities first, so toning time decides whether the mid tones are converted at all, and his bleach-back shows silver still unconverted after six minutes at 1+10. That establishes that conversion is partial and time-dependent. It does not establish how protection scales with it, and the course found no published study that does — which is why the page states the chain (conversion is partial, protection depends on conversion, therefore protection is partial) as a mechanistic inference rather than as a measured result.

Question 3. A print is toned in an iron-blue toner and the maker wants it to last. Which framing decision, standard for a silver-gelatin print, would actively damage it?

• Glazing that blocks at least 97 per cent of ultraviolet energy
• A mat board buffered with at least 2 per cent calcium carbonate
• Keeping display below 100 lux
• An enclosure that has passed the Photographic Activity Test, ISO 18916

Show the answer and why

Answer: A mat board buffered with at least 2 per cent calcium carbonate

The alkali reserve is the problem. Prussian blue is destroyed by alkaline hydrolysis, and Holtzman's figure quoted by Ware is that a buffer at pH 9.4 decolourises it in one to ten minutes — pH 9.4 being the pH of saturated calcium carbonate, which is the buffer ISO 18902 requires in conservation board. The other three decisions are neutral or beneficial for a blue-toned print. This is why ILFORD's note that iron and copper toners "may not give extra protection and the image might fade" understates the case: for the blue toner the standard archival kit is itself the hazard.

Question 4. Why does this page argue that ISO 18915 existing is important, even though the course has not read the standard?

• Because a print that passes it can be described as archival
• Because it fixes the toning times and dilutions that protect an image
• Because it shows that the question is answerable by measurement rather than opinion, while its scope also shows what it deliberately refuses to settle
• Because conservation bodies require prints in collections to have been tested to it

Show the answer and why

Answer: Because it shows that the question is answerable by measurement rather than opinion, while its scope also shows what it deliberately refuses to settle

The scope is the informative part. An international committee wrote a repeatable method for evaluating whether chemical conversion increases a silver image's resistance to oxidation, which means the question is empirical and not a matter of taste. In the same clause the standard declines to recommend any treatment and puts treatment temperature, times and replenishment rates outside its scope. So it is a comparison instrument, not a specification: nothing in it licenses the word "archival", and nothing in it tells a printer how long to tone.

Check my answersClear