Skip to content
AnalogFixerDarkroom chemistry for black-and-white film
Browse sections
Technician rolling analog film in a darkroom.
Photo by Annushka Ahuja on Pexels

Archival permanence

Washing to zero is not what the standards ask for

Shows that the standard darkroom instruction — wash until the residual hypo test reads clean — is contradicted by every document that has ever put a number on it: the US federal specification for permanent microform required residual thiosulfate to be greater than zero, and both current ISO standards warn that very low concentrations make the silver image more susceptible to oxidative attack while noting the target may sit below the detection limit of the test itself. Also traces the federal ceiling for permanent microform from 0.007 g/m² in 1980 to 0.014 g/m² today, a doubling that happened by incorporation of a different standard and was never announced as a change.

Published by AnalogFixer ·

The instruction is universal and it is wrong in an interesting way. Wash the print, run a residual hypo test, and if it shows nothing, wash a bit longer to be safe. Every darkroom guide says some version of it. No standard does. The documents that actually specify residual thiosulfate — one of them a federal specification with the force of procurement behind it — treat "as little as possible" as a failure mode rather than as the goal, and they have said so consistently for forty-five years.

Two separate things are wrong with the folk instruction. The first is that a lower number is not always better, and the standards say so explicitly. The second is that "the test shows nothing" is not a measurement of zero. It is a statement about the detection limit of whichever procedure was run, and the standards note that the concentration they are asking about may lie below it.

A federal specification required more than zero

In September 1980 the US National Bureau of Standards issued FIPS PUB 82, adopting the National Micrographics Association practice MS1-1980 for computer-output microfilm. Its section on permanence defines the archival category as records for permanent storage — "Archival is considered forever or permanent" — and then sets the processing requirement:

After film processing, the residual thiosulfate ion must not exceed 0.7 micrograms per square centimeter but should be greater than zero.

That second clause is not a rhetorical flourish or a rounding note. It is a floor, written into the same sentence as the ceiling, in a document that governed federal microform production. A laboratory reporting a genuine zero would not have satisfied it.

The specification is equally specific about how to find out. It names the methylene blue method of ANSI PH4.8, and it puts a clock on the reading: "Testing for residual thiosulfate ion must be conducted using a clear area of the film within 2 weeks of processing." Both details matter later.

FIPS 82 is no longer in force. NIST withdrew it on 29 May 1996, in the general clear-out of federal standards that had been superseded by commercial ones, recorded at 61 FR 26891. It is quoted here as evidence of what was specified, not as a current requirement — but what replaced it did not reverse the point.

Scale contrasting washing to zero thiosulfate with published measurable residual limits.

The current standards say the same thing more quietly

ISO 18929, which covers black-and-white prints in dark storage, sets one limit for both fibre and resin-coated base: prints "shall not contain more than 0,014 g/m² of residual thiosulfate". Immediately beneath it, as a note to the clause itself:

A very low concentration of thiosulfate due to excessive washing may cause the silver image to be more susceptible to oxidative attack. These concentrations may be below the detection limit of ISO 18917.

ISO 18901, the film standard, carries the same caution in materially identical wording as a footnote to its own limits table. Two standards, two material classes, one warning, placed in both cases directly against the number a reader has come for.

So there are three documents spanning 1980 to 2012 — one federal, two international, covering microfilm, film and prints — and all three say that the bottom of the range is a place you can arrive at by accident and should not aim for. The instruction to keep washing until the test reads clean is not a conservative version of the standard. It is a different instruction.

Note also what both ISO standards prohibit in the same breath. Hypo-eliminating agents that work by oxidation — the ones containing peroxides or hypochlorites — "shall not be used", because they decompose thiosulfate chemically. ISO 18901 is careful to distinguish those from hypo-clearing baths, which it describes as high ionic strength salt solutions that facilitate washing without chemically altering the thiosulfate. The two get spoken of interchangeably in darkroom advice. One is permitted and one is prohibited, and the word "eliminator" is attached to the prohibited one.

"The test shows nothing" is a statement about the test

ISO 18917 defines the three procedures the other two standards refer to, and each has a stated range rather than an open-ended sensitivity.

The iodine-amylose procedure "covers the range from 0,002 µg/cm² to 0,40 µg/cm²". The methylene blue procedure "covers the range from 0,005 g/m² to 0,34 g/m²", which the same sentence glosses as 0.5 to 34 µg/cm². The standard supplies its own conversion so there is no ambiguity about which unit is meant: a footnote in the iodine-amylose clause gives 1 µg/cm² = 10⁻² g/m², and ISO 18929 repeats the relation as 0.010 g/m² = 1 µg/cm².

Two consequences follow, and they point in opposite directions.

At the low end, a result of "none detected" from the iodine-amylose procedure means below 0.002 µg/cm², or 0.00002 g/m². That is a real number and it is seven hundred times below the print limit. Reporting it as zero, and then treating zero as success, is precisely what the notes in both ISO standards were written to discourage.

At the high end, the arithmetic looks alarming at first and then resolves. The iodine-amylose ceiling of 0.40 µg/cm² is 0.004 g/m², which is below the 0.014 g/m² limit that ISO 18929 mandates that same method to adjudicate. The standard is not contradicting itself, though: the procedure includes a documented route out. If the absorbance falls below 0.090 — which happens when there is a lot of thiosulfate, because thiosulfate is what discharges the colour — ISO 18917 instructs the analyst to re-extract using a smaller sample and rescale the result by the sample area. The quoted range is the range at the nominal specimen size, not a hard ceiling on the method.

That is worth stating plainly because it would have been an easy and satisfying error to make. The two figures genuinely do not match as printed, and it takes reading the procedure to the calculation step to find the correction that reconciles them.

Which method is mandatory depends on the material, and the reasons are physical

The three procedures are not interchangeable, and the standards are unusually direct about why.

For prints, ISO 18929 permits only one. All three methods, it says, are "considered sufficiently reliable to determine thiosulfate concentrations at the 0,014 g/m² level, but only the iodine amylose method has been found to give reliable results with all types of fibre-base and RC prints. Therefore, it is the only method chosen for this International Standard." The exclusion of the second is explicit and reasoned: "The methylene blue method shall not be used because it fails to detect residual thiosulfate in RC-base prints whose unexposed emulsion layers contain a considerable amount of developing agent in order to provide very rapid development during processing." The silver sulfide method is rejected separately for over-reporting, since it responds to polythionate decomposition products as well as thiosulfate.

For film, ISO 18901 permits any of the three and adds that the methylene blue method "is considered to be reliable for thiosulfate concentrations of 0,007 g/m²" — a note about how far down it stays trustworthy, not a limit.

So the method FIPS 82 prescribed for microfilm in 1980 is the method ISO 18929 forbids for prints today, and the reason is not a change of opinion. It is that resin-coated paper carries developing agent in the emulsion, and microfilm does not.

The federal ceiling doubled, by reference

The current requirement for US federal records is 36 CFR § 1238.14(c). Agencies "must process microforms so that the residual thiosulfate ion concentration will not exceed 0.014 grams per square meter in accordance with ISO 18901", and must use the processing procedures in ANSI/AIIM MS1 and MS23.

Set that beside FIPS 82. Both govern microforms held permanently — the 1980 document's archival class, the current rule's permanent and unscheduled records. The 1980 ceiling was 0.7 µg/cm², which is 0.007 g/m². The current ceiling is 0.014 g/m². For the same material and the same retention case, the permitted residual thiosulfate is twice what it was.

The continuity makes it easier to miss rather than harder. The current regulation still cites ANSI/AIIM MS1, the direct descendant of the NMA MS1-1980 that FIPS 82 adopted, so the paperwork trail looks unbroken. What changed is that the limit now arrives by incorporation of ISO 18901 rather than being written out, and the number that arrived was a different one. No document announces the change, because from the drafter's point of view nothing was changed: a reference was updated.

This is the same failure mode as the capacity figure in what fixer exhaustion actually means — not an error by anyone, but a number that means something different from what its continuity implies, and no cross-reference telling the reader so.

What to do instead

  • Stop treating the residual hypo test as a pass at zero. Both ISO standards attach a warning to the bottom of the range, and a federal specification once required the result to be above it.
  • Note that a "clean" result is a detection limit, and which detection limit depends on which procedure was run. For iodine-amylose it is 0.002 µg/cm²; for methylene blue it is 0.5 µg/cm², two and a half orders of magnitude apart.
  • Wash to a sequence you can repeat rather than to an appearance. Both standards specify the concentration in the material, and neither offers any way to infer it from how long the water ran.
  • Keep hypo-clearing baths and hypo eliminators apart. The oxidising kind is prohibited outright by both standards; the salt kind is what most people mean and is not.
  • Test within two weeks of processing. FIPS 82 required it, and both ISO standards repeat it, because the thiosulfate concentration in anything but a minimum-density area changes over that period.
  • Quote the limit with the standard and material it belongs to. 0.014 g/m² is the print figure in ISO 18929 and the microform figure in 36 CFR 1238.14; ISO 18901 tiers film limits by class and life expectancy, and they are not all this number.

The uncomfortable summary is that the most widely repeated washing advice in photography optimises for a quantity the standards treat as having a floor, measured by a test whose "nothing" reading nobody quotes a value for.

References