Experiment: Exhaustion, Capacity and What a Used Developer Actually Does

“The developer is exhausted” is one sentence covering three different chemical events, and they do not happen at the same rate, do not show the same signature on a curve, and do not call for the same remedy. One of them you can compensate with time. One of them you cannot. One of them happens whether or not a single frame of film goes near the bottle.

ILFORD’s own process-control introduction lists all three in six lines, without ceremony: processing a film changes the pH, uses up the developing agents, and puts waste products — mainly bromide — into the bath; and separately, developers react with air, which breaks some of the active ingredients down and reduces activity. This session takes one mixing of one developer, divides it four ways before anything is done to it, and measures which of those processes does what.

Purpose

To separate the three processes the darkroom lumps together as exhaustion by running three parallel treatments of one mixing of one developer, reading a sensitometric strip from every cell on the instruments Parts XIV and XV built, and turning the resulting drift into a capacity figure that carries, on its face, the tolerance that produced it. The single variable is which of the three ageing processes the developer has been through; the control is fresh developer from the same mixing; and the hypothesis predicts a different signature for each treatment. All three are set out in full below.

Hypothesis, with the three signatures it predicts. That the three treatments are distinguishable at this bench’s resolution, and that each moves a different pair of the four measured quantities:

• Sequential use lowers contrast index steadily and lowers maximum density late, while base plus fog barely moves. Prediction with a size in it: the contrast index at the manufacturer’s stated capacity is lower than the fresh control by more than 0.03 and less than 0.20 — the lower bound being roughly the resolution the earlier pages of this part establish, the upper being Kodak’s own control limit for a process that has gone too far to print.
• Added bromide, at the concentration a used bath is estimated to reach, lifts the toe and costs speed while contrast index holds longer than the sequential arm’s does. Prediction: the speed point moves further, relative to the contrast index, than it does in the sequential arm at the same nominal loading.
• Aerial oxidation with no film through it at all raises base plus fog and eventually collapses activity, and does so with a knee rather than a slope.

How it fails. If all three arms lie inside one another’s uncertainty bands, this apparatus cannot separate them, which is a finding about the apparatus and is reported as one. If the bromide arm reproduces the sequential arm, the common claim that a used developer is a bromide-restrained developer survives a test it usually never gets; if it does not, the difference is the size of everything else that accumulates.

The control. Fresh developer from the same mixing, drawn from a full capped bottle, run as a strip at the start of the programme and again at the end. Two strips because the programme spans a fortnight and the control has to bracket it: the difference between them is the drift of the film, the instruments and the reader, and everything else is measured against it.

The one variable. Which of the three ageing processes the developer has been through. Film, emulsion batch, exposure batch, wedge, lamp, dilution, temperature, agitation script, volume per strip, fixer, wash and reader are all held.

!What this page is not, and what it borrows

It is not a repeat of Part VIII’s aerial-oxidation experiment, which watched colour in a tube and tested activity with a fogged strip. That page asked whether the colour you can see reports the developing power you cannot. This one puts the answer on a density scale.

It is not a re-derivation of what a restrainer does. Part IX and the curve-families page have already measured a bromide family at 0, 0.49, 0.98 and 1.96 g/L, and this page’s third arm is read against that family rather than instead of it.

And it is not an argument about whether to reuse developer. Both makers have already made it, in writing: ILFORD recommend one-shot working “when image quality, reliability and consistency are more important than economy”. The question here is what the alternative actually costs, in numbers.

Learning objectives

By the end of the session and the arithmetic that follows it you will be able to:

• name the three processes that a used developer undergoes, say which measured quantity each one moves first, and design a treatment that isolates one of them;
• estimate the bromide concentration a developer reaches after a stated loading, from published silver content and published film area, and state every assumption the estimate rests on;
• read a manufacturer’s capacity figure as what it is — a working practice with a compensation attached — and set two of them side by side when they disagree by a factor of three;
• convert a measured drift into a capacity, by choosing a tolerance in advance, and show that three defensible tolerances give three different capacities from one set of data;
• describe what a replenisher does chemically and why a process-control chart is the only honest way to run one, without running either;
• write a discard rule for your own darkroom in terms that another person could check.

Prerequisites

Development kinetics and restrainers and antifoggants own the mechanism and the published reuse tables; this page measures what those tables compensate for and does not restate them. Aerial oxidation and Part VIII’s oxidation experiment own the third arm’s chemistry. Fixer capacity and exhaustion owns the parallel question in the other bath. The development time series is a hard prerequisite twice over: its contrast-index-against-time curve is the ruler that turns a contrast loss into an equivalent time, and its three control strips are where your resolution sentence came from. Both instrument certificates on the bench, per Part XIV and Part XV.

Safety classification

Level A. Standard home-darkroom controls, on a session that opens no jar of powder.

Every solution here is drawn from something already made: the developer stock from Part VIII’s mixing lab, the bromide from Part IX’s 10 % w/v potassium bromide stock, the fixer and wetting agent from bought concentrates. Under the course’s rubric that is gloves, eye protection whenever a solution is poured or agitated, dedicated labelled vessels and the ordinary ventilation of a wet bench.

One raised step, if you cannot avoid it. Mixing 2.5 L of stock from raw chemicals is a Level B operation: weighing metol, hydroquinone and a quarter of a kilogram of sodium sulfite means fine powders that must not be inhaled and two skin sensitisers. The rubric prefers such a step avoided rather than declared: mix on another day inside Part VIII’s session, which already carries the controls, and bring a labelled bottle to this one. If you must weigh on the day, that step takes Part VIII’s controls in full — weighing over a tray, a particulate mask where the safety data sheet asks for one, splash goggles and an apron — and the session returns to Level A when the jar is closed.

What is not a hazard here, and why. The deliberately oxidised arm sounds worse than it is: an oxidised developer is a weaker reducing agent, not a more aggressive substance, and its hazard profile is the fresh bath’s — metol and hydroquinone in solution, contact rather than airborne. Nothing becomes volatile, nothing gasses, and the brown colour is quinone and its condensation products rather than a warning. Nothing is heated, and no concentrated alkali is opened, because D-76’s alkali is borax already dissolved at 2 g/L. A fixed, washed, dried strip presents no chemical hazard at all: its silver is metallic and locked in hardened gelatin. None of those is a general statement about the substances; each is about this operation, which is what a hazard assessment is.

What the session does add is a half-full open bottle standing somewhere for a fortnight, which is a spill and a mix-up waiting to happen rather than a toxicological problem. Label it, date it, stand it in a tray, and keep it where it cannot be mistaken for working solution.

Hazards

Hazard	Where it arises	Control
Skin sensitisation from metol and hydroquinone	Pouring, agitating and emptying about four litres of developer across the programme, over two sittings and a fortnight of aliquots	Single-use nitrile gloves for every wet operation, changed if one is splashed inside. The glove page has the compatibility argument. A sensitisation, once acquired, does not go away
A part-full open container of developer standing for a fortnight	The oxidised arm, which is deliberately stored badly	Label it with contents, concentration, date opened and the words experimental, do not use, per the container SOP. Stand it in a tray, in the dark, where nothing will be knocked into it and it cannot be reached for by mistake

| Splashes to the eye, and long wet sessions in darkness | Loading and unloading twelve strips and ten films | Eye protection from the first pour to the last emptied vessel. Bench laid out in the light and not moved; one hand wet and one dry; the light switch found by touch before it is needed | | Miscounting the load | The whole sequential arm, whose x-axis is a tally | ILFORD name this risk explicitly. Keep the tally on paper beside the bottle and mark it as the film goes in, not afterwards |

Required PPE

Single-use nitrile gloves for every wet operation, donned and removed per the PPE SOP, because metol and hydroquinone are skin sensitisers and this session handles more developer over more sittings than any other page in the part. Eye protection from the first pour to the last emptied vessel. And clean dry hands for the film, which protects the strips rather than you: a fingerprint on a step is read as density and cannot be undone.

Respiratory protection has no role in this session, for a specific reason rather than a reassuring one: nothing here is a powder, nothing is heated, and an aqueous developer at 20 °C produces no vapour that a mask would filter. The particulate control belongs to the mixing session, which is deliberately held on another day.

Ventilation

Ventilation is not among the controls this page relies on, because nothing in it evaporates or aerosolises at room temperature: the developer is a dilute aqueous solution of an aminophenol salt, a phenol, a sulfite and a borate, and the fixer a thiosulfate at working strength. What the room needs is ordinary comfort ventilation for a two-hour sitting, which Part XVI’s darkroom build solves with a light-tight vent. The place where the oxidised bottle stands should be ventilated in the ordinary domestic sense and not warm: ILFORD note that high temperatures speed oxidation up, so a warm airing cupboard is a different experiment from the one you designed.

Materials

Item	Quantity	Note
Test strips from one exposure batch	12	135 mm strips, one emulsion batch, exposed in one sitting per the exposure-session SOP. A 36-exposure cassette yields ten or eleven, so this is two cassettes and the balance is spares
Load film	800 square inches, about 10 films’ worth	Any one film type, of constant coating, uniformly flashed. It does not have to be new: outdated stock, fogged short ends and bulk-loaded remnants all carry silver, and silver is what the load is for
Transmission step wedge	1	The Stouffer T2115 or equivalent, the same one used for the whole batch
Opaque batch tin, card label slips, negative sleeves	1 tin, 12 slips, 6 pockets	The eight-field slip specified by the test-negative lab. A strip without one is not data
Storage bottle, 1 L, brown glass or HDPE, full and capped	1	Holds the sequential arm between films, per ILFORD’s own instruction that the used portion is returned and mixed with the unused
Storage bottle, 1 L, wide-mouthed, kept half full and open	1	The oxidised arm. Labelled as a hazard control, not as a convenience
Storage bottle, 500 mL, filled to the neck and capped	1	The reserve the two control strips are drawn from. It must be full, because Kodak’s own storage table makes a part-full bottle a different material
Measuring cylinders, 250 mL, dedicated to developer	4	One per cell in the bromide and control arms
Graduated syringe, 5 mL, reading to 0.1 mL	1	For the bromide additions, which are 1.5 and 3.7 mL
Developing tank and reels	1 tank, 2 reels	The load film and its strip go through together, on separate reels
pH meter with buffers at 4.01 and 7.00, or narrow-range papers	1	Calibrated per the pH-meter SOP. A named gap in the course’s price file
Lab notebook sheets, curve-plotting sheets and the formula version record	1 set + 11 + 1	One plotting sheet per strip; the version record names the mixing all four arms came from

Chemicals

Nothing is weighed in this session. Every solution is diluted or drawn from a stock made earlier, per the SOP for mixing from a stock.

Chemical	Quantity	Form
D-76 stock from Part VIII’s lab — metol, sodium sulfite, hydroquinone, borax	2.5 L	Solution at stock strength, one mixing, divided four ways before anything is done to it
Potassium bromide	5.2 mL of the Part IX 10 % w/v stock, which is 0.52 g	Solution, kept from Part IX
Rapid fixer, ammonium thiosulfate type	1 L at 1+4	Fixed for twice the clearing time, per the clearing-time SOP
Wetting agent	1 L at 1+200	ILFORD’s figure for ILFOTOL is 5 mL per litre
Plain water rinse at the process temperature	about 12 L	ILFORD note that a water rinse may replace a stop bath but increases the risk of processing marks and stains

The silver bromide in the emulsion is not a substance you handle: it is the reagent the developer reduces, and the whole subject of the third arm is the bromide that comes out of it.

Equipment

The sensitometer from Part XIV with its certificate; the densitometer from Part XV with its certificate and its terminal logging to a file, checked per its SOP; a changing bag; one thermometer for the whole programme, checked per the bench-check SOP; a stopclock; a water bath that will hold 20 °C for two hours; and a spreadsheet or short script that can solve one equation by trial, because the contrast-index construction is a one-unknown root find. None of it is consumed, which is why none of it appears in the table below.

Estimated cost

££ on the planner’s bands, and the band survives only because of one design decision: the load is area, not pictures. The twelve test strips must come from one emulsion batch. The 800 square inches of load film need only be one consistent film type — outdated stock, fogged short ends, bulk remnants — and if you buy ten fresh cassettes for it at retail this session is dearer than its band and the table below will tell you so.

Estimated consumables cost

One run of the whole three-arm programme:

Consumed	This session	Sourced price	Cost this session
35 mm film for the test strips, one emulsion batch	12 strips of 135 mm; a cassette yields ten or eleven, so two cassettes with the balance kept as spares	£6.37–£11.40 per 36-exposure roll	£12.74–£22.80
Load film, any consistent type	800 square inches, about 10 films’ worth	None matching. The route the page recommends is outdated or fogged stock, which the price file cannot price; ten new cassettes would be £63.70–£114.00 on the row above	—
D-76 stock from Part VIII, at stock strength	2.5 L	Costed in Part VIII’s mixing lab	—
Potassium bromide, as the Part IX 10 % w/v stock	5.2 mL, which is 0.52 g	£23.00 per 250 g	£0.05
Rapid fixer concentrate	200 mL, to make 1 L at 1+4	£21.05–£25.98 per 1 L of concentrate, diluted 1+4 for film	£4.21–£5.20
Wetting agent	5 mL, for 1 L at 1+200	£28.70 per 1 L of concentrate, diluted 1+200	£0.14
Water for rinses and washes	about 10 L	Metered supply; the planner prices no water	—
pH buffer sachets at 4.01 and 7.00	2 sachets	None. A named price gap in the price file: no pH meter, electrode or buffer standard is priced	—
Negative sleeves and card label slips	6 pockets, 12 slips	None. A named price gap: sleeving that passes the Photographic Activity Test	—
Squared paper for the drift plots	4 to 6 sheets	None. The planner carries no stationery line	—

The priced rows come to £17.14 to £28.19, at the retail ranges read on 5 September 2026 and recorded in the laboratory planner. That is a floor, not a total: 6 of the 10 rows carry no dated price, so they count as nothing here and are certainly not free, and the largest physical quantity in the session — the load film — is deliberately the row with no number in it.

The film row dominates it, and it dominates it for a bad reason: a strip is 7.32 square inches and a cassette is priced whole, so twelve strips buys twenty-two and throws ten away. A bulk tin is the cheaper route and the price file cannot cost it, as the time series also found. If the strips come out of an existing batch, the floor falls to £4.40 to £5.39.

Waste streams

Four, kept apart, labelled per the container SOP and routed per the general waste SOP.

1. Spent developer, about 2.5 L, alkaline, carrying metol, hydroquinone, sulfite, borate, their oxidation products and the bromide released by about eleven films’ worth of area. All of it is waste: nothing on this page is kept for reuse.
2. The oxidised arm, which goes to the same stream and is worth naming separately only because it is the one bottle that must not be poured back into anything.
3. Spent fixer and the first rinse after it, acidic and silver-bearing, to the silver stream. This session’s litre takes about eleven films’ worth of area against ILFORD’s rating of 24 per litre, so it is inside capacity and it is the one stream worth recovering.
4. Rinse and wash water: the rinses after development go with stream 1, the first change after fixing with stream 3, the rest with the general stream.

Streams 1 and 3 are never combined — one is a sulfite solution at a borax-buffered developer’s pH and the other is acidic — because mixing them spoils the silver recovery.

Alternative route

The session needs darkness for two operations only: taking the strips from the tin and notching them, and loading each strip and its load film onto reels. Both happen inside a changing bag, and everything else — developing in a closed tank, rinsing, fixing, washing, drying, reading and every piece of arithmetic — happens in room light. This is the page in the part that needs a darkroom least.

If you have no changing bag either, a windowless room at night, proved rather than assumed with the blackout leak test, does the loading.

If you have neither and cannot borrow either, the wet half cannot be run and the honest thing is to say so rather than substitute something that is not it. Two pieces still stand alone. The bromide estimate below is a calculation and you can do it for your own film and loading tonight. And the capacity-as-a-decision analysis can be run on somebody else’s data, including on the manufacturers’ published compensation ladders, which are themselves claims about drift and can be plotted, compared and found to disagree. That analysis, with the disagreement drawn and the missing measurement named, is worth more than a capacity figure taken on trust.

Preparation

About 90 minutes the evening before, plus the exposure session, plus a fortnight of elapsed time that is not time on task. The 150 minutes is the sequential arm’s wet run and the bromide cells; the oxidised arm’s aliquots are two twenty-minute visits a week apart, and the reading session is a separate sitting once every strip is bone dry.

One mixing, divided four ways before anything happens to it

This is the design’s whole economy and it is worth doing slowly. Mix — or draw from Part VIII — 2.5 L of D-76 at stock strength, in one operation, and give it a single formula version code from the versioning SOP. Then, before a single strip is exposed, divide it:

Portion	Volume	Vessel	What it becomes
U, sequential use	1000 mL	1 L bottle, filled to the neck, capped between films	The arm that meets film
O, oxidised	500 mL	1 L wide-mouthed bottle, half full, cap left off	The arm that meets air
B, bromide	500 mL	Two 250 mL cylinders, made up on the day	The arm that meets a restrainer and nothing else
C, control	500 mL	500 mL bottle, filled to the neck, capped	The reserve the two control strips come from

Every arm is the same solution on the day it is divided. Anything they differ by afterwards is what the treatment did, and that is only true if the division happens first. A control drawn from a second mixing is not a control; it is a second experiment.

The control bottle must be full. Kodak’s own storage table is the reason: D-76 stock keeps six months in a full tightly closed bottle and two months half-filled, and their stated reason is that partially filled bottles allow some oxidation of the solution. A half-full control bottle would age in the same direction as the oxidised arm over the fortnight, quietly narrowing the difference the session exists to measure.

The film, the developer and the control time

The course’s reference pairing is ILFORD FP4 Plus in Kodak D-76, and this page uses it at stock strength for a reason that is not a preference. Both makers forbid the alternative in writing: Kodak say D-76 diluted 1 to 1 is diluted just before use, discarded after one batch and neither reused nor replenished, and ILFORD say they do not recommend reusing diluted developers at 1+1 or 1+3 at all. A dilute developer has no capacity to measure, because its published working practice is a single use. Reuse is a stock-strength question and the experiment has to be run where the question lives.

ILFORD’s FP4 Plus sheet gives Kodak D-76 at stock as 6, 8 and 9 minutes for EI 50, 125 and 200 in a spiral tank with intermittent agitation. At the film’s own EI 125 that is 8 minutes at 20 °C, and that is the time every cell in this experiment runs, from the first to the last.

!Why no cell gets the manufacturer's time compensation

ILFORD’s ladder — the second film at N+10 %, the third at N+20 %, the tenth at N+90 % — is not an instruction this page disobeys out of carelessness. It is the page’s hypothesis.

A compensation is a claim: this much extra time restores what that much use took away. Applying it would cancel the very drift the session is trying to measure and leave you with a set of strips that all look roughly alike, which tells you nothing about how large the drift was or whether the compensation is the right size. Run every cell at the published N, measure the drift, and then compare it with the compensation: convert your measured contrast index at each loading into an equivalent time on your own time series, and see whether the extra time your curve asks for is 10 per cent per film, or 4, or 25.

That comparison is the result. The compensation ladder is a testable claim, and almost nobody tests it.

The loading ladder, and where the published capacities fall on it

The x-axis of the sequential arm is films per litre, where a “film” is the makers’ own unit: Kodak define it as 80 square inches, which is one 135-36 roll, one 120 roll, four 4 × 5-inch sheets or one 8 × 10-inch sheet. That definition is what lets a load be assembled out of whatever film you have.

Read a strip at 0, 1, 2, 4, 6, 8 and 10 films per litre — seven strips, spaced so that the early part of the curve, where the published capacities live, is sampled more finely than the tail. And notice where the published figures fall inside that range, because they are startlingly far apart.

Developer	Films of 135-36 per litre, unreplenished	Compensation published with it	Source
Kodak D-76, full strength	4 (16 per US gallon)	+15 % after every four rolls per gallon	Kodak J-78, and the same figure in the E-103CF chemicals table
ILFORD PERCEPTOL, stock	4, with the fifth marked not recommended	+10 % per successive film	ILFORD powder-developer sheet
Kodak XTOL, full strength, as the chemicals table gives it	5 (19 per US gallon)	none stated in that table	Kodak E-103CF chemicals table
ILFORD ID-11 and MICROPHEN, stock	10	+10 % per successive film, N+90 % at the tenth	ILFORD powder-developer sheet
ILFORD ILFOTEC DD, working strength	10	+10 % per successive film	ILFOTEC DD sheet
ILFORD ILFOTEC DD-X at 1+4, reused	10 (and 16 one-shot from a 1 L bottle)	+10 % per successive film	ILFOTEC DD-X sheet
Kodak XTOL, full strength, as its own data sheet gives it	15	rolls 1–5 at N, 6–10 at +15 %, 11–15 at a further +15 %	Kodak J-109, the XTOL sheet itself
Kodak T-MAX Developer at 1+4	about 12.7 (48 per US gallon)	1–16 at N, 17–32 at +1 min, 33–48 at +2 min	Kodak J-86

⁇Kodak's own two documents give XTOL a capacity of 5 and of 15

Two rows of that table are the same product from the same maker, and they differ by a factor of three. The E-103CF chemicals comparison sheet prints XTOL at 19 per gallon and 5 per litre. The XTOL technical data sheet, J-109, says the capacity with normal unreplenished processing is approximately 15 rolls per litre with time compensation, and instructs the user to discard after 15. Neither document mentions the other.

They are not both wrong. They are answers to different questions, and neither document says which. The E-103CF row sits in a column headed 8 × 10 sheets per gallon (litre) alongside tray and tank working, and carries no compensation ladder; J-109’s figure is explicitly with time compensation, and the compensation compounds to about 32 per cent extra development by the fifteenth roll. A bath worked to a fixed time and a bath worked to a rising time have different capacities, and the difference is roughly this size.

The course states the observation and not a resolution. What it draws from it is the page’s central claim: a capacity figure without a working practice and a tolerance attached is not a measurement of anything. If Kodak can print 5 and 15 for one developer, the figure on a datasheet cannot be a property of the bottle.

The bromide arm, and where its two concentrations come from

The third arm needs a number nobody publishes. Part VIII already said so plainly: the makers publish a compensation rather than a bromide concentration, a pH drift or an agent consumption. So the course computes a route, from documents that were written for an entirely different purpose — silver recovery — and marks every step of it as its own arithmetic on published inputs.

🧮Estimating the bromide a litre of developer reaches, from Kodak's silver-recovery data

The stoichiometry is the easy half. Every silver ion reduced releases one halide ion into the bath. For hydroquinone in an alkaline developer, and for a bromide emulsion:

C₆H₆O₂ + 2 AgBr + 2 OH⁻ → C₆H₄O₂ + 2 Ag + 2 Br⁻ + 2 H₂O

Two silver ions reduced, one molecule of agent oxidised, two hydroxides consumed and two bromide ions released into the bath

So bromide released, in moles, equals silver developed, in moles. Everything else is finding out how much silver that is.

Three published inputs, all from Kodak’s own silver-recovery literature.

• Kodak’s area table gives a black-and-white 135-36 roll as 592 square feet per 1000 units, that is 0.592 square feet or 550 cm² each.
• Kodak’s worked medical example gives KODAK T-MAT RA Film at 14.5 troy ounces of silver per 1000 square feet. One troy ounce is 31.1035 g, so that is 451 g per 1000 square feet, or 0.451 g per square foot.
• The same example assumes an average film exposure of 50 per cent, so that half the silver is removed during processing and half stays as image silver.

The arithmetic.

0.451 g/ft² × 0.592 ft² = 0.267 g of silver per 135-36 roll

Silver in one roll

At 50 per cent developed that is 0.134 g of silver reduced, which is 0.134 ÷ 107.87 = 1.24 mmol, and therefore 1.24 mmol of bromide released. Expressed as the mass of potassium bromide that would give the same bromide-ion concentration — which is what the third arm actually adds — 1.24 mmol × 119.0 g/mol = 0.147 g of KBr-equivalent per film.

Loading, films per litre	Bromide released, as g/L of KBr-equivalent	What it corresponds to
1	0.15	one film through a litre
4	0.59	Kodak’s published capacity for D-76
10	1.47	ILFORD’s published capacity for ID-11
15	2.21	Kodak’s published capacity for XTOL in J-109

Five things this estimate is not, and every one of them matters.

1. T-MAT RA is a radiographic film coated for X-ray screens, not the pictorial film you are using. It is here because it is the only black-and-white film for which Kodak print a silver figure anywhere in this corpus: they state that silver content varies from emulsion to emulsion and between makers, and direct anyone wanting a product figure to Kodak Environmental Services rather than publishing one.
2. The 50 per cent is Kodak’s assumption for a hospital’s workload, not a measurement of your negatives. ILFORD make the same point in words: a film of night shots develops out relatively clear and uses little agent, a bright beach scene develops out blackened and uses much. That is why the load film here is uniformly flashed — so the load per film is at least constant.
3. Not all the released halide is bromide. A modern film is a silver iodobromide, and Part VIII’s restrainer lesson records Kodak’s 1928 statement that accumulated bromide and iodide restrain, and the argument that iodide restrains harder per mole because silver iodide is the least soluble halide. A KBr addition therefore under-restrains at equal moles, by an amount this course cannot size.
4. Carry-out is ignored. Each film takes 250 to 300 mL out of the litre on its reel, which ILFORD’s reuse instruction handles by having you pour it back; where it is not poured back, the litre stops being a litre.
5. It is an estimate and the plots say so. Every figure derived from it carries the word estimated, and the third arm’s axis is estimated equivalent loading.

And there is a book that would probably settle it. Haist’s Modern Photographic Processing is the standard reference on developer chemistry and would very likely say whether a bromide addition reproduces a used bath. This course does not hold it, quotes nothing from it and will not paraphrase a figure it has not read. If you have it, write the chapter and page beside your result and say whether it agrees with what you measured.

The two bromide cells therefore run at 0.59 g/L and 1.47 g/L, the estimates for Kodak’s and ILFORD’s own published capacities, made up from Part IX’s 10 % w/v stock into 250 mL cells:

Cell	Estimated loading it stands for	KBr	10 % w/v stock into 250 mL	Read against
B-4	4 films per litre	0.59 g/L	1.5 mL	Part IX’s B1 at 0.49 g/L
B-10	10 films per litre	1.47 g/L	3.7 mL	Part IX’s B2 at 0.98 and B3 at 1.96 g/L

A 5 mL syringe reads to 0.1 mL, so B-4 carries about ±7 per cent on its addition and B-10 about ±3 per cent. Write the volume you actually drew, and put those percentages in the budget rather than in a footnote.

The oxidised arm, and how long a fortnight is

The oxidised arm has to be stored badly enough to show something and not so badly that the result is a curiosity. The makers’ published keeping figures are what to size it against, and they agree with each other on the direction and disagree on the magnitude:

Solution	Full and tightly capped	Half full, or worse	Source
ILFORD ID-11, MICROPHEN, PERCEPTOL stock	6 months	1 month half full and tightly capped; 4 months in a deep tank with a floating lid and 1 month without one	ILFORD powder-developer sheet
Kodak D-76 stock	6 months	2 months half-filled	Kodak J-78
Kodak XTOL, mixed	6 months	at least 2 months partially filled	Kodak J-109
Kodak T-MAX Developer, working strength	6 months	2 months half-filled, 1 month in a covered tank	Kodak J-86
ILFORD ILFOTEC DD-X concentrate	24 months	4 months half full	ILFOTEC DD-X sheet

Every pair is a factor of three to six from headspace alone, and Kodak give the reason in a clause: partially filled bottles allow some oxidation. ILFORD’s floating-lid row says it twice in one line — a deep tank keeps four times as long with a lid as without, and the only difference is the area of solution touching air.

The tightest published figure for a half-full container is ILFORD’s one month, and that is for a container tightly capped. This arm’s is half full and open, worse in both of the ways the makers name, so a fortnight should show something well inside a figure written for a better-stored bottle. Read it at day 7 and day 14, keeping the remaining third of the 500 mL as a spare.

If nothing has moved at day 14, that is a result and not a failure: at 100 g/L of sulfite the sacrificial preservative is still ahead of the oxygen. Report it with the storage conditions in full — temperature, light, headspace and whether the room was warm.

Run order, the data table, and the version code

Draw the run order from a hat. The sequential arm is the one arm whose order cannot be randomised — film 6 must follow film 5 — and that is exactly why everything else must be. Randomise the order in which the bromide, control and oxidised cells are run within their sittings, and randomise the reading order separately and completely, because reading order and loading order would otherwise be the same axis, and a densitometer drift across the reading session would arrive looking precisely like exhaustion.

Draw the data table in full before the session, because a table drawn afterwards acquires exactly the columns that turned out to be interesting. Eleven rows — C-start, U-1, U-2, U-4, U-6, U-8, U-10, O-7, O-14, B-4, B-10, C-end — and columns for cumulative load in films per litre and in square inches, developer pH and its measurement temperature, developer colour against a stated reference, volume remaining in the bottle, temperature at start and end of the cell, and then base plus fog, speed point, contrast index and maximum density after the reading session.

Procedure

Five stages across a fortnight, of which two are sittings at the bench.

Stage 0 — The exposure batch and the load

1. Expose all twelve strips, plus spares, in one sitting on the sensitometer from one lamp at one distance through one wedge, per the exposure-session SOP. Every strip carries the masked patch the test-negative lab specified, because base plus fog comes from each strip’s own patch.
2. Flash the load film to a uniform, stated exposure. The load must draw a comparable amount of developing agent per film, and the only way to make that true is to expose it evenly rather than to pictures. A safelight-free room, a bare lamp at a stated distance and a stated time will do; so will a sheet of white card at the enlarger. Record the exposure, and check one clip in the fixer that it develops out to a strong, even density rather than to maximum black — a film developed to maximum black draws the most agent it possibly can and is a load, not a typical one.
3. Cut and divide the load into ten equal portions of 80 square inches, and label them 1 to 10.

Stage 1 — Preflight

4. Check the thermometer against a reference per the bench-check SOP and write the offset down. Calibrate the pH meter at 4.01 and 7.00 per its SOP, and record the slope. Bring the water bath to 20 °C and give it twenty minutes to settle.
5. Measure and record the fresh developer’s pH and colour from the 2.5 L mixing, before it is divided. This is the zero of two columns that will be read nine more times. ILFORD publish a pH table for their own fresh stock solutions and then advise users to make their own control measurements from their own accurately mixed fresh solutions; that advice is this step.
6. Divide the mixing four ways as the table in Preparation sets out. Fill, cap and label. Stand the oxidised bottle where it will spend the fortnight and note the room’s temperature and whether the spot is dark.
7. Make up one litre of fixer at 1+4 and the wetting agent at 1+200, and measure the clearing time on a scrap of the load film per the clearing-time SOP; the fixing time is twice it. Measure the clearing time again at the end of the session — the litre will have taken about eleven films’ worth of silver against ILFORD’s rating of 24 per litre, which is inside capacity and is not the same thing as unchanged.

Stage 2 — The control and the bromide cells, first sitting

8. Draw 250 mL from the full control bottle, bring it to 20 °C, and develop C-start for 8 minutes on the written agitation script. Re-cap the control bottle immediately and top it up with a marble or a smaller bottle if the level has dropped: a control that oxidises is not a control.
9. Make up B-4 and B-10 in their cylinders from the B portion and the bromide stock, and develop them for 8 minutes in the drawn order. Record the volume of stock drawn to 0.1 mL.
10. Rinse, fix for twice the clearing time, wash and hang.

Stage 3 — The sequential arm, second sitting

11. Take 300 mL from the U bottle into the tank; load load film 1 and test strip U-1 on separate reels; develop for 8 minutes at 20 °C on the written script.
12. Pour the used developer back into the U bottle and mix it with the unused part before the next film goes in. This is ILFORD’s own instruction for reuse, and their stated reason is that it gives more consistent results by reducing the risks of solution losses and of the restraining effect of the by-products. It is also what makes the x-axis mean what it says: without it, the twelfth film meets a different solution from the one your tally describes.
13. Run load films 2 to 10 the same way. A test strip goes in with films 1, 2, 4, 6, 8 and 10; films 3, 5, 7 and 9 carry load only. Cap the bottle between films.
14. At every strip cell, before the developer goes into the tank, record the pH, the temperature, the colour against your stated reference and the volume remaining. Four columns, thirty seconds, and the entire mechanistic half of the analysis depends on them.
15. Rinse, fix and wash each film and strip as they come off. Ten films through one fixing bath is a lot; if the clearing time on a scrap has doubled by film 7, change the fixer and write down which strips were fixed in which bath.

Stage 4 — The oxidised arm, day 7 and day 14

16. On day 7, swirl the open bottle once to mix it, draw 250 mL, bring it to 20 °C, record pH, colour, temperature and headspace, and develop O-7 for 8 minutes. Return nothing to the bottle.
17. On day 14, repeat for O-14 from what remains.
18. Immediately after O-14, draw the last 250 mL from the control bottle and develop C-end. Running the closing control on the same evening as the last treated cell is what makes the fortnight’s drift measurable rather than assumed.

Stage 5 — The reading session, on another evening

19. When every strip is bone dry — not before — run the densitometer’s warm-up and daily check per its SOP, recording the calibration step before the first strip and after every third.
20. Read all twenty-one steps and the masked patch of every strip, in the separately drawn reading order, logging to a file rather than transcribing. Read three of them twice, lifting each off the stage and replacing it between readings: that replaced-sample spread is your Type A term.

Expected observations

What follows is what the design predicts, together with the observations that would show the prediction wrong. None of it is a statement about the numbers you will get: an experiment whose page has already announced the answer is not an experiment.

In the bottles. The U bottle darkens slowly and the O bottle darkens faster, and both go the same way: colourless to straw to amber to brown. The pH of both falls. ILFORD list the falling pH first among the three effects of processing a film, and the O arm shows that it is not only films that do it.

In the trays. The U-10 strip should be visibly thinner at the top of its scale than C-start, and the difference should be concentrated in the dense steps, because that is where the agent is asked to do the most work. If the two strips differ most in their thin steps instead, look at the bromide column: a toe held back is a restraint signature, not a consumption one.

The three signatures, drawn so they can be told apart.

What the three arms should look like if they are different processes

• Sequential use: agent consumed, a steady fall that steepens
• Added bromide alone: contrast holds, the cost is elsewhere
• Aerial oxidation, no film through it: a knee, not a slope

Show the numbers behind this plot

Three traces of contrast index against loading, all starting together at 0.62 with nothing through the bath. The sequential-use trace falls steadily and slightly faster as it goes, reaching about 0.585 at four films per litre and about 0.495 at ten, a total loss of a little over 0.12. The added-bromide trace is almost flat, falling only from 0.62 to about 0.598 across the whole range, because bromide restrains the toe rather than flattening the straight line. The oxidised trace follows the other two closely at first, barely separating from them to about four films' equivalent, and then falls away sharply to about 0.33, forming a knee rather than a slope. The drawing's point is that the three processes are distinguishable by shape and not only by size: consumption is a gradual slope, restraint is nearly flat on this axis and shows itself on a different one, and oxidation is a plateau followed by a collapse. The oxidised trace's horizontal axis is elapsed storage rather than film, plotted on the same axis only so the three shapes can be compared.

Series	Loading, films per litre — or the estimated equivalent for the bromide arm	Contrast index
Sequential use: agent consumed, a steady fall that steepens	0.00	0.62
Sequential use: agent consumed, a steady fall that steepens	1.00	0.61
Sequential use: agent consumed, a steady fall that steepens	2.00	0.60
Sequential use: agent consumed, a steady fall that steepens	4.00	0.58
Sequential use: agent consumed, a steady fall that steepens	6.00	0.56
Sequential use: agent consumed, a steady fall that steepens	8.00	0.53
Sequential use: agent consumed, a steady fall that steepens	10.00	0.49
Added bromide alone: contrast holds, the cost is elsewhere	0.00	0.62
Added bromide alone: contrast holds, the cost is elsewhere	1.00	0.62
Added bromide alone: contrast holds, the cost is elsewhere	2.00	0.61
Added bromide alone: contrast holds, the cost is elsewhere	4.00	0.61
Added bromide alone: contrast holds, the cost is elsewhere	6.00	0.61
Added bromide alone: contrast holds, the cost is elsewhere	8.00	0.60
Added bromide alone: contrast holds, the cost is elsewhere	10.00	0.60
Aerial oxidation, no film through it: a knee, not a slope	0.00	0.62
Aerial oxidation, no film through it: a knee, not a slope	1.00	0.62
Aerial oxidation, no film through it: a knee, not a slope	2.00	0.61
Aerial oxidation, no film through it: a knee, not a slope	4.00	0.60
Aerial oxidation, no film through it: a knee, not a slope	6.00	0.56
Aerial oxidation, no film through it: a knee, not a slope	8.00	0.47
Aerial oxidation, no film through it: a knee, not a slope	10.00	0.33

Drawn to show three shapes that can be told apart, not measured from any developer. The numbers on both axes are illustrative, chosen so that they will not be mistaken for a result: the whole purpose of the session is to find out whether your own three arms separate at all, and by how much. The oxidised arm's points are placed against elapsed time and share the axis only for comparison. The curve is drawn to show the shape, not measured from a real material. Your own materials will differ, and measuring them is what the sensitometry part of the course is for.

Four things that would falsify the design rather than the chemistry. The two control strips disagreeing by more than the resolution sentence allows, which leaves the fortnight uncontrolled. No drift at all in the sequential arm by ten films per litre, which would mean the load carried far less silver than the estimate assumed. Nothing in the oxidised arm at day 14, which is a real possible result and is reported as one. And B-10 landing outside the bromide family Part IX already measured, which would put the fault in the stock, the syringe or the arithmetic before any conclusion about used developer could be drawn.

One observation that is not a fault. Your capacity will not equal the manufacturer’s. Kodak print 5 and 15 for one product; nobody’s bench matches a figure written for a different tank, a different agitation and a different film.

What is happening chemically

Three processes, one bottle, three different clocks

ILFORD’s process-control introduction names all three and the course adds nothing to the list:

1. The pH falls, making the developer more acidic.
2. The developing agents are used up.
3. Waste products, mainly bromide, are introduced into the developer.

And separately, and whether or not any film is processed: developers react with air, oxidation breaks some of the active ingredients down and reduces activity, and high temperatures speed that up.

Those are ILFORD’s statements. What follows is the course’s reading of them against the equation, marked as such because it is an inference rather than something ILFORD wrote.

The reaction as written consumes two hydroxide ions for every molecule of hydroquinone oxidised, which is one reason a working bath’s pH falls, and which puts the first two items on ILFORD’s list into one process rather than two: the alkali and the agent go down together in a fixed ratio, and it is the alkali going down that makes the remaining agent less active than its concentration suggests. A developer does not simply run out of agent; it runs out of the conditions under which the agent works. That is why the pH column is not decoration.

Why bromide is not the same as consumption, and does not act where you think

Bromide restrains by raising the barrier to development at crystals that have only a marginal latent-image speck. That is a threshold effect, so it shows on the toe, which is where the speed point lives; the straight line, made of crystals that have plenty of speck, is comparatively unbothered. Kodak’s 1928 primer puts the whole thing in one phrase: the restraining action of accumulated bromide and iodide is analogous to cutting down the exposure. Cut the exposure and you lose shadows, not slope.

Consumption is the opposite shape. Running short of agent starves the whole reaction, and it starves it worst where the reaction has most to do, which is the top of the scale. So the two mechanisms move different ends of the same curve, and that is the whole reason the third arm exists.

One subtlety makes the third arm an approximation even if the concentration were exactly right. Sheppard and Mees noticed it in 1907, in a kinetics experiment on ferrous oxalate: their most dilute developer’s rate constant drifted downward, which they attributed to the rapid accumulation of bromide and other reaction products which do not diffuse rapidly, and so influence the velocity by remaining in the reaction layer. The bromide that matters most is not the bromide in the bottle but the bromide in the millimetre of liquid against the emulsion, and that depends on agitation. Adding bromide to the bulk raises both; developing a film raises the local one first. If B-10 and U-10 disagree, this is one place to look, and the agitation experiment is where the boundary layer was measured.

What the sulfite is doing, and why the oxidised arm has a knee

The oxidised arm’s shape is not a slow decline because the preservative is not a slow decline. Sulfite is sacrificial: it reacts with dissolved oxygen on its own account, and it also regenerates the developing agent from its oxidation product at its own expense.

2 SO₃²⁻ + O₂ → 2 SO₄²⁻

Sacrificial oxidation: the preservative spending itself on oxygen, written as the ions rather than the sodium salts

C₆H₄O₂ + SO₃²⁻ + H₂O → C₆H₆O₂ + SO₄²⁻

Regeneration: quinone reduced back to hydroquinone, and another sulfite gone

D-76 carries 100 g/L of sulfite against 2 g of metol and 5 g of hydroquinone, an enormous reserve, and while it lasts the agents are protected and activity barely moves. What comes out the other end is sulfate, which is not a preservative and does nothing. So the expected shape is a plateau followed by a fall, and the sulfate accumulating meanwhile is why a bottle that has been standing does not return to normal when you cap it. Part VIII’s sulfite lesson works the mechanism out; this page borrows the consequence.

Two visible things follow and both go in the data table. The colour is hydroquinone’s report, not metol’s — Part VIII’s oxidation experiment established that a metol solution loses activity without darkening nearly as much, so a pale bath is undiagnosed rather than good. And fog is the oxidised arm’s signature quantity, which is why base plus fog is read from every strip’s own masked patch.

What is consumed, what accumulates, and what merely changes

One litre of developer, and the four quantities that move as film passes through it

1. 1Developing agent, consumed — falls in proportion to silver developed; neither maker publishes the rate, and this session infers it from the curve rather than measuring it
2. 2Sulfite, barely touched in a closed bottle — 100 g/L against 7 g of agents — the reserve is why the sequential arm is a slope and the oxidised arm is a knee
3. 3Bromide, accumulating — the course's own estimate from Kodak's published silver content and film area — 0.15 g/L of KBr-equivalent per film
4. 4Alkalinity, falling — two hydroxides per molecule of hydroquinone oxidised; the one quantity here you measure directly, with a pH meter, at every cell

The bars show directions and relative magnitudes, not measured values: only the bromide row carries numbers, and those are this course's arithmetic on Kodak's silver-recovery documents rather than a measurement of any developer. The two dashed lines are the manufacturers' own published capacities, drawn to show that they fall at very different places on the same axis.

Data to record

Copy this into the notebook before the first cell.

Programme header. Dates of all three sittings; who; room temperature at each; the bath set point; every instrument with its checked offset; the sensitometer and densitometer certificate dates; the pH meter’s calibration slope and date.

The mixing block. The formula version of the 2.5 L mixing, its date, the solids’ batch numbers if you have them, its measured pH at a stated temperature, and the four portion volumes as poured. One version code covers all four arms, because they are one mixing.

Exposure batch block. Film, format, emulsion batch number; the sensitometer’s lamp, warm-up and cadence; the wedge part number and whether it is calibrated; the interval from exposure to development.

Load block. The load film’s type; its flashing exposure in full; the density a clip developed to; and the area of each of the ten portions, in square inches, with the running total in films per litre beside it. The tally is data. ILFORD warn that miscounting is one of the real risks of reuse, and on this page a miscount is not an inconvenience — it is the x-axis.

One row per cell, from the table drawn in Preparation: run order, intended and actual time, developer temperature at start and end, pH, colour against a stated reference, volume remaining in the bottle, and then base plus fog, speed point, contrast index and maximum density after the reading.

The oxidised arm’s own block. Headspace as a fraction, whether the cap was on or off, where the bottle stood, the room’s temperature range over the fortnight, and whether the spot was dark.

Clearing time, measured three times: at the start, at film 7 and at the end. A fixer that exhausted during the session is a confound for maximum density on every strip fixed after it did.

Deviations. A cell run at 20.6 °C. A bottle left uncapped overnight by mistake. A load portion mismeasured. Write them at the time, because a deviation remembered afterwards is a deviation reconstructed to suit the result.

Analysis

1. The two controls, and what the fortnight cost

Before anything else, compare C-start with C-end. They are the same solution from the same full, capped bottle, developed a fortnight apart, and every difference between them belongs to the film, the instruments, the room and you — not to any treatment.

Add that difference to your resolution sentence. If C-end differs from C-start by more than the Type A spread the time series measured on its three control strips, then the programme’s own drift is larger than its instrument’s, and every difference reported below must clear the larger figure. Say which of the two you used and print both.

2. The drift curves, one quantity at a time

Plot four separate small plots, not one crowded one: contrast index, speed point, base plus fog and maximum density, each against loading, with all three arms on each. Four plots is the right number because the whole hypothesis is that different arms move different quantities, and a single plot with four quantities on it hides exactly that.

Then answer four questions in writing, each with a number and its uncertainty:

• Which arm lowers contrast index most, per unit of nominal loading?
• Which arm moves the speed point most, relative to how much contrast index it cost? This ratio is the discriminator between restraint and consumption, and it is more robust than either quantity alone because a common error in strip exposure moves both.
• Which arm raises base plus fog? Kodak’s own control-chart tolerance for D-min has no lower limit — fog that falls is never a fault — and that asymmetry is a hint about which direction matters.
• Does maximum density move at all, and when? It should be the last thing to go and the surest sign that agent, rather than restraint, is the limit.

3. Testing the manufacturer’s compensation, which is a claim and not an instruction

This is the step the session was designed around. For each sequential-arm strip, take the measured contrast index onto your own contrast-index-against-time curve and read off the time at 20 °C in fresh developer that would have produced it. That is the cell’s equivalent time. The extra time the bath now needs, as a percentage of N, is:

extra time = (t_equivalent ÷ t_control) − 1, at each loading

The compensation your own data asks for

Then tabulate your figure against the maker’s ladder — 10 per cent per film for ILFORD, 15 per cent after every four rolls per gallon for Kodak — and say which of three things you found:

What you found	What it means	What to do with it
Your compensation is close to the published ladder	The ladder describes your bath as well as theirs, which is a stronger result than it sounds	Adopt it, and say at what loading you stopped checking
Your compensation is smaller than the ladder	Your load films carried less silver than the ladder assumes, or your agitation is gentler, or your volume per film is larger	The load’s flashing density is the first place to look, and the second is the volume you actually poured
Your compensation is larger than the ladder	The bath is running down faster than published — check the pH column, the age of the stock and the headspace of the U bottle before concluding anything about the developer	If the pH has fallen further than the equation accounts for, you may be measuring a mixing error rather than exhaustion

Two cautions. Equivalent time is not the whole story: the compensation restores contrast index, and if the speed point has moved too, the extended time gives a negative of the right slope in the wrong place — check both columns. And the ladder is arithmetic on a curve that flattens: ILFORD’s N+90 % at the tenth film is 15 minutes on an 8-minute time, and a percentage of a long time buys much less contrast per minute than it did at the start.

4. Is a bromide addition a used developer? The comparison the third arm exists for

Set B-4 against U-4 and B-10 against U-10, four quantities each, with uncertainties. Then answer, in one sentence with numbers in it, the question the whole arm was built to answer: at equal estimated loading, does adding bromide to fresh developer reproduce what use does?

Three outcomes and what each licenses you to say:

• They agree within the resolution. Then, at these loadings and on this film, the used bath behaves as though bromide were the only thing that had changed. Say it that carefully: it does not follow that bromide is the only thing that has changed, only that nothing else has changed enough to measure yet.
• The bromide arm restrains more than use does. The estimate is probably too high, which would not be surprising given what it was built on. Recompute it against your measurement and report the implied silver content.
• Use costs more contrast than bromide does. This is what the consumption mechanism predicts: bromide alone barely touches the straight line, so a used bath that has lost slope has lost something bromide does not explain. Kodak’s 1928 remedy for a bath that needs seasoning — add about 5 per cent of old developer — would be an absurd instruction if bromide were the whole of it, since a weighed addition would be simpler, cheaper and exactly repeatable.

Whatever you find, put the estimate’s five stated assumptions beside the conclusion. A result that rests on an estimated axis has to carry the estimate with it.

5. Turning drift into capacity, which means choosing a tolerance first

Capacity is not a property of the developer. It is the loading at which the drift crosses a line you drew, and if you draw the line somewhere else you get a different capacity from the same data. That is not a weakness in the method; it is what the word means, and it is why two Kodak documents can print 5 and 15.

One drift curve, three defensible tolerances, three different capacities

• A drift curve of the shape the sequential arm is expected to give

Show the numbers behind this plot

A single falling curve of contrast-index deviation from the fresh control against loading in films per litre. It starts at zero with nothing through the bath and falls away with increasing steepness, reaching minus 0.015 at two films, minus 0.035 at four, minus 0.060 at six, minus 0.090 at eight and minus 0.125 at ten. Three horizontal guide lines cut it, all of them Kodak's own published numbers for black-and-white film process control. The highest, at minus 0.02, is the acceptance window Kodak use when choosing a development time, and the curve crosses it at about 2.4 films per litre. The middle line, at minus 0.07, is Kodak's lower action limit, and the curve crosses it at about 6.6 films per litre. The lowest, at minus 0.12, is Kodak's lower control limit, beyond which Kodak say results will be unsatisfactory for shadow density or contrast, and the curve crosses it at about 9.7 films per litre. One curve therefore yields three capacities differing by a factor of four, and which one is right depends entirely on which line was drawn before the data was taken.

Series	Loading, films per litre	Contrast index, as a deviation from the fresh control
A drift curve of the shape the sequential arm is expected to give	0.00	0.00
A drift curve of the shape the sequential arm is expected to give	1.00	-0.01
A drift curve of the shape the sequential arm is expected to give	2.00	-0.01
A drift curve of the shape the sequential arm is expected to give	4.00	-0.04
A drift curve of the shape the sequential arm is expected to give	6.00	-0.06
A drift curve of the shape the sequential arm is expected to give	8.00	-0.09
A drift curve of the shape the sequential arm is expected to give	10.00	-0.13

The curve is drawn to show the construction and is not measured from any developer. The three horizontal lines are not drawn: they are Kodak's published tolerances for their own black-and-white film process, the ±0.02 acceptance window used to pick a development time and the lower action and control limits from the plotting form. One curve, three lines, three capacities — 2.4, 6.6 and 9.7 films per litre — and the only thing that decides between them is the tolerance you committed to before you had the data. The curve is drawn to show the shape, not measured from a real material. Your own materials will differ, and measuring them is what the sensitometry part of the course is for.

Work it in that order and write the sentence out:

For this film in this developer at this temperature with this agitation, developed to a fixed time with no compensation, the capacity is ____ films per litre at a tolerance of ____ in contrast index, with an uncertainty of ____ on the crossing point taken from the local slope of the drift curve.

The last clause is the one most often left out: the crossing point’s uncertainty is your contrast-index resolution divided by the slope of the drift curve there, exactly as the time series turned a contrast uncertainty into a time uncertainty. Where the drift curve is shallow, the capacity is poorly determined even though every contrast index on it is well measured.

Three published anchors, and none of them is a default. Kodak’s acceptance window for choosing a development time is ±0.02 in contrast index; their plotting form gives action limits of +0.07 to +0.20 and −0.07 to −0.12 around an aim of 0.58, with control limits beyond; and your own resolution sentence sets the floor, because a tolerance tighter than your measurement cannot be enforced.

!Kodak's limits are asymmetric, and the asymmetry is the lesson

Look again at the published tolerance table. Contrast index may run to +0.20 above aim before the process is stopped but only −0.12 below. Speed may run to +22 above aim but only −17 below. And D-min — base plus fog — has an upper control limit of +0.03 and no lower limit at all.

None of that is arbitrary. A negative that is too contrasty can be printed on a softer grade; a negative that is too flat has lost separation that no paper restores. A negative with too much shadow density prints; one with too little has nothing in the shadows to print. And fog that falls is never a fault, which is why the column has no floor.

A tolerance encodes what the consequence of being wrong is, in which direction. When you choose your own, choose it in the direction that costs you a print, and say what that direction is. A symmetric tolerance chosen because symmetry looks tidy is a tolerance nobody thought about.

6. Replenishment, explained and not run

The industrial answer is not compensation but replenishment: instead of letting the bath drift and lengthening the time, add a solution that puts back what use took out, so the published time keeps working. Kodak’s rate for D-76 is 22.2 to 29.6 mL of Replenisher D-76R per 135-36 roll or 8 × 10 sheet, and it raises the capacity from 16 sheets per gallon to 120 — a factor of seven and a half — with no increase in development time at all.

Three things about it are worth having even though this page runs none of it.

A replenisher is a different formula, not more of the same one. ILFORD’s September 2012 withdrawal notice for their ID-11 replenisher shows it without disclosing anything: the improvised substitute takes two packs, dissolves both part A sachets and only one part B, and discards the other. Doubling one half against the other is what a replenisher is — the consumed component at higher concentration, the rest unchanged, and no accumulated by-products. Part VIII works it through; this page adds only that it is why “just add more developer” does not work.

Replenishment cannot be run open-loop. ILFORD say the optimum rate is found with a process control system; Kodak say to monitor with control strips and adjust in steps. The rate depends on your throughput, your carry-out and your tank, so the only way to know it is to plot a control strip’s speed, contrast and D-min against date and watch — which is what the control-strip SOP is for, and why a replenished process without a chart runs on faith.

Throughput cuts the other way from intuition. ILFORD’s withdrawal notice says that as the volume of film through the developer reduces, the replenishment rate should be increased, to compensate for oxidation. A busy tank is replenished for consumption; an idle one for air. Kodak instruct a low-utilisation D-76 system to be discarded after a month however few rolls have gone through it.

Which is why one-shot working is right for almost every home darkroom. Not because reuse is bad chemistry, but because a replenished process is a controlled one, and control costs a control strip, a densitometer session and a chart every week. ILFORD recommend one-shot where image quality, reliability and consistency matter more than economy, and both makers say reuse is the more inconsistent route. A fresh litre costs less than an evening spent finding out why a negative is thin.

7. The seasoned-developer question, answered from the sources and from your data

The folk claim is that a developer improves once it has had a film or two through it. It is a testable claim, and this session tests it: if it were true, some measured quantity would be better in an early sequential cell than in the fresh control, and better in a stated direction.

Before you read your own data, read what the makers actually publish, because it is more interesting than either side of the argument.

• ILFORD state flatly that there will always be a difference in performance between a tank of fresh developer and a tank of used or seasoned developer. A difference, not an improvement.
• Kodak’s process-monitoring publication gives the direction: some developer and replenisher systems are formulated so that a fresh solution gives a higher speed than a highly seasoned, properly replenished one, so a declining speed trend as a fresh tank seasons is normal until it levels off. They name D-76 with Replenisher D-76R as one such system, and their footnote to the speed aim says the same in passing. Kodak’s D-76 sheet adds the contrast half: the modified T-MAX replenisher exists to help avoid a small speed loss and an increase in contrast as seasoning occurs.
• The industry’s answer to seasoning is not to season but to fake it. A starter makes a fresh bath behave from the first film as a seasoned tank does. ILFORD’s ILFOTEC DD STARTER goes in at 1+250, four millilitres per litre, and their sheet quantifies its worth: without it, the published times must be reduced by 20 per cent and then raised as the tank seasons. Kodak’s XTOL sheet gives the same effect at a different size, and Kodak’s monitoring note closes the loop — systems mixed with an appropriate starter do not show the declining-speed trend.

Read together, those documents agree. Seasoning is real, it is a loss of speed and a gain in contrast, it is worth 10 to 20 per cent of a development time, and the industry abolishes it with a starter rather than seeking it. No source in this course’s corpus states that a seasoned developer is better in any measured respect.

That is not the same as saying the folk claim is false, and the page does not say so. It says the claim is unsupported in the direction it is usually made, and that the one historical remedy resembling it — Kodak’s 1928 advice to add about 5 per cent of old developer — is offered among developer troubles and aerial fog as a remedy for fog rather than as a claim of superiority. Your U-1 and U-2 strips are the local test: if either beats C-start in a quantity you named in advance, by more than your resolution, write it up with its conditions rather than as a vindication.

8. The other exhaustion this session reveals for free

Two other baths were quietly working all evening, and both have published capacities that this session comes close to.

Bath	Published capacity, films of 135-36 per litre	Source
ILFORD ILFOSTOP at 1+19	15	ILFORD powder-developer sheet
ILFORD ILFOSTOP PRO at 1+19	22	ILFOTEC DD sheet
ILFORD RAPID FIXER and HYPAM at 1+4	24	ILFORD powder-developer sheet and ILFOTEC DD sheet
Kodak Rapid Fixer	32 per litre, 120 per US gallon	Kodak E-103CF chemicals table
Kodak Fixer, working strength	26 per litre, 100 per US gallon	Kodak E-103CF chemicals table
Kodak Indicator Stop Bath	discard when the colour changes to purplish-blue	Kodak E-103CF chemicals table

This programme puts about eleven films’ worth of area through one litre of fixer, against ILFORD’s rated 24 per litre — inside capacity, with roughly half of it spent. Had the load been twice as large, or the fixer half a litre, the last strips would have been under-fixed in a way that lowers maximum density and looks exactly like a developer that has run out. That is why the clearing time is measured three times and why every strip is traceable to a fixing bath. Part XI’s fixer capacity lesson and its clearing-time experiment own that measurement, and this page cross-refers rather than repeating it; the same goes for Part X’s stop-bath capacity experiment.

What cannot be concluded from this session

Not a capacity for the developer. It is a capacity for your film, your loading, your agitation, your volume, your temperature and your tolerance, and the tolerance is a choice.

Not the bromide concentration a used bath reaches. The third arm’s axis is an estimate built on a radiographic film’s published silver content and a hospital’s exposure assumption. It is a good enough estimate to make the arm worth running and not good enough to quote as a concentration.

Not a measurement of agent consumption. Nothing on this bench measures how much metol or hydroquinone is left. The curve reports the consequence of consumption, mixed with everything else that changed.

Not a settled answer on seasoning, from two strips at low loading in one developer on one evening — though it is a real test of a claim that usually gets none.

And not a rule that stays true. Change the film, the flashing density, the tank, the agitation or the volume per film and the drift curve moves. Re-measure with one control strip when any of those changes, rather than trusting a figure you wrote a year ago.

Troubleshooting

What you see	Likely cause	What to do
C-start and C-end differ by more than the resolution sentence allows	The control bottle was not full, was opened repeatedly, or the fortnight moved the film, the lamp or the instrument	Widen the resolution sentence to the measured spread and say so. Check the volume column first; a control that oxidised is the commonest cause and the easiest to prevent next time
The sequential arm shows almost no drift by ten films per litre	The load carried far less silver than assumed — a thin flashing exposure is the usual reason — or the volume per film was much larger than the tally implies	Develop a clip of the load film and read its density. Report the loading in square inches as well as films per litre, because the square inches are what you actually know
The oxidised arm has moved nothing at day 14	The sulfite reserve is still ahead of the oxygen, which at 100 g/L is entirely possible	Report it as the result it is, with headspace, temperature and light stated. Extend to a month if you want the knee, and say that the reported figure is for a longer period than the design specified

| B-10 does not sit inside the bromide family Part IX measured | The bromide stock has changed concentration by evaporation, the syringe was misread, or the arithmetic is wrong | Fix this before drawing any conclusion about used developer. Make up a fresh 10 % w/v stock, weigh it, and re-run the cell. A third arm that disagrees with the family it is supposed to join is not evidence about anything | | Maximum density falls on every strip after film 7, in all arms | Suspect the fixer before the developer, and the clearing times will say | If the last clearing time is much longer than the first, re-fix the affected strips in fresh fixer, re-read them, and record that they were re-fixed |

| The drift curve is not monotonic — one cell sits above its neighbours | Almost always a process deviation rather than chemistry: a missed agitation, a short cell, a mismeasured load portion, or strips read before they were bone dry | Check the deviations column before excluding anything, and apply the exclusion rule you wrote in advance rather than one invented now |

Clean-up

Empty each cylinder into the developer waste container as its cell finishes rather than at the end; cylinders of similar-looking brown liquid standing about is how one gets confused with another. Rinse each before refilling.

The U bottle is emptied only after the last strip is read, not on the night, because a strip that has to be repeated needs the bath it was developed in. Cap it, label it with the loading it reached, and keep it until the reading session is finished. The O bottle goes to waste after O-14 and its label goes with it into the bin, so that it cannot become an unlabelled container six months later.

Wash the trays, the tank, the reels and the thermometer. Rinse the pH electrode and return it to its storage solution — an electrode left dry is an electrode to recalibrate or replace.

Storage

The strips, once bone dry, go into sleeves with their card slips and into the Part IX archive, kept together as one programme. Keep the control pair findable: the developer comparison will want a fresh-developer reference for this film. Store cool, dry and dark, in sleeving that passes the Photographic Activity Test where you can get it.

Any developer stock that remains goes back capped and full, dated, with its version code legible, because the storage tables on this page are about exactly that. A part-full bottle is a different material, and you have now measured how different.

The capacity figure itself goes on the personal processing table the time series produced, in its own row, with the tolerance beside it. A capacity without its tolerance is not a number anybody, including you, can use.

Disposal considerations

Spent developer, about 2.5 L across the programme, is alkaline and holds sulfite, sulfate, borate, metol, hydroquinone, their oxidation products and the bromide released by about eleven films’ worth of area. Sulfite is an oxygen scavenger and exerts an oxygen demand on whatever receives it; the aminophenol and the phenol are what a water authority is likely to care about. The oxidised arm is chemically the same stream, further along.

Spent fixer and the first rinse after it are acidic and carry dissolved silver as thiosulfate complexes. This session’s fixer carries more silver than most, which makes it the one stream in the programme with recoverable value; it goes to the silver route, and the two streams are never combined, because mixing them compromises that recovery.

Where each of those streams may go is a question of jurisdiction rather than of chemistry, and the disposal page sets out why this course describes the chemistry and the general practice and issues no universal instruction. Check your local regulations before the first bottle is emptied, because they govern and they differ from one municipality to the next.

Questions

1. Kodak print a capacity of 5 rolls per litre for XTOL in one document and 15 in another. Give the strongest reading of each figure that makes both of them true statements, then say what a reader who has only one of the two documents cannot know.
2. Your sequential arm shows a contrast-index drift of −0.04 at four films per litre, and your resolution sentence allows differences of 0.03. State your capacity at Kodak’s ±0.02 window, at their −0.07 action limit, and at your own resolution, and say which one you would write on your processing table and why.
3. The bromide estimate assumes that half the silver in the film is developed. Recompute the estimated bromide at ten films per litre for a photographer who shoots almost entirely at night, and say what that does to the third arm’s interpretation.
4. Your O-14 strip is dark brown in the bottle but gives a contrast index within 0.02 of the fresh control. What have you learned, what have you not learned, and which of the four data columns would settle the difference?
5. A colleague argues that since ILFORD publish a compensation ladder, measuring the drift is a waste of ten films. Give their strongest argument, then give the strongest answer to it that uses only figures published on this page.

Further experiments

Run the bromide arm as a proper ladder rather than two cells. Four cells at 0.15, 0.59, 1.47 and 2.21 g/L of KBr-equivalent — one film, four films, ten films and fifteen films of estimated loading — turn the comparison from two points into a curve that can be laid on top of the sequential arm and subtracted from it. What is left over after subtraction is everything that is not bromide, which is the quantity nobody publishes.

Replace the estimate with a measurement. The bromide released is proportional to the silver developed, and the silver not developed goes into the fixer. A residual-silver estimator of the kind Part XI uses, applied to a fixer that has taken a known area of your own film, gives the undeveloped half; the developed half is what remains. That is a route to the silver content of your film, and it would replace this page’s largest assumption with a number.

Turn the control cell into a control chart. Every session in this part develops at least one strip at a control condition. Plot their contrast index, speed and base plus fog against date, with action lines from your own certificate, exactly as ILFORD’s process-control introduction and Kodak’s plotting form describe — the same three quantities, LD, HD−LD and Dmin, that a commercial laboratory watches. After a year of it you will have the one thing this session cannot give you: a measurement of how much of your own variation is the process and how much is you.

Test the starter claim. ILFORD publish a number that is unusually easy to check: without ILFOTEC DD STARTER, reduce the given times by 20 per cent. That is a testable prediction about how much more active a fresh bath is than a seasoned one, in a developer that is designed to be replenished. Two strips in fresh working-strength ILFOTEC DD, one at the published time and one at 80 per cent of it, against a strip in a bath that has had ten films through it, is a three-cell experiment that puts a manufacturer’s own quantified seasoning claim on a density scale.

Check your understanding

Question 1. ILFORD list three things that processing a film does to a developer. Which of them does adding potassium bromide to a fresh bath reproduce?

• Only the third — waste products, mainly bromide, introduced into the bath — leaving the agent consumption and the pH change unreproduced
• All three, which is why a bromide addition is the standard simulation of a used developer
• The second and third, since bromide also consumes developing agent
• None of them, because the bromide in a used developer is chemically different from the bromide in the jar

Show the answer and why

Answer: Only the third — waste products, mainly bromide, introduced into the bath — leaving the agent consumption and the pH change unreproduced

ILFORD name three effects: the pH falls, the developing agents are used up, and waste products — mainly bromide — are introduced. A bromide addition reproduces the third exactly and neither of the other two, and that is precisely what makes the third arm informative rather than redundant. If the used bath and the bromide-only bath behave the same at equal estimated loading, then at those loadings nothing else has changed enough to measure. If they differ, the difference is the size of the agent consumption and the pH change together. The last option is wrong for a chemical reason worth keeping: a bromide ion released from silver bromide and a bromide ion from a jar of potassium bromide are the same ion. What differs is where it is — Sheppard and Mees pointed out in 1907 that reaction products accumulate in the layer of liquid against the emulsion and do not diffuse away quickly, so development raises the local concentration before it raises the bulk one.

Question 2. Kodak publish a capacity of 5 rolls per litre for XTOL in their chemicals comparison sheet and approximately 15 rolls per litre in the XTOL data sheet itself. What is the best account of the disagreement?

• They answer different questions: the 15 is explicitly with a compounding time compensation reaching about 32 per cent extra development, and the 5 sits in a table of tray and tank working with no compensation attached
• One of the two documents contains a misprint that Kodak has not corrected
• The 5 is per litre and the 15 is per gallon, and the units have been confused
• The figures are for different dilutions of the product

Show the answer and why

Answer: They answer different questions: the 15 is explicitly with a compounding time compensation reaching about 32 per cent extra development, and the 5 sits in a table of tray and tank working with no compensation attached

Both documents say per litre and both are about full-strength XTOL, so units and dilution are not the explanation. What differs is the working practice each figure assumes. J-109 states its figure as approximately 15 rolls per litre with time compensation and prints the ladder: rolls 1 to 5 at the normal time, 6 to 10 at plus 15 per cent, 11 to 15 at a further plus 15 per cent, which compounds to about a third more development by the end. E-103CF prints 19 per gallon and 5 per litre in a table of keeping properties and useful capacities across tray and tank working, with no compensation column at all. A bath worked to a fixed time and a bath worked to a rising time have different capacities. Neither figure is wrong; what is missing from both is the sentence saying which question is being answered, which is exactly why this page insists that a capacity is quoted with its working practice and its tolerance.

Question 3. On Kodak’s plotting form for black-and-white film, the contrast-index action limits are +0.07 to +0.20 above aim but only −0.07 to −0.12 below, and D-min has an upper control limit with no lower limit at all. What do these asymmetries encode?

• That a negative which is too contrasty can be printed on a softer grade, while one that is too flat has lost separation no paper restores
• That fog which falls below aim is never a fault, so the column needs no floor
• That the consequence of an error is different in each direction, so a tolerance has to be chosen in the direction that costs you a print
• That the densitometer is more accurate at high densities than at low ones

Show the answer and why

Answer: That a negative which is too contrasty can be printed on a softer grade, while one that is too flat has lost separation no paper restores, That fog which falls below aim is never a fault, so the column needs no floor, That the consequence of an error is different in each direction, so a tolerance has to be chosen in the direction that costs you a print

A tolerance is a statement about consequences, not about measurement. Kodak allow nearly twice as much drift upward in contrast index as downward because the two failures are not equally recoverable: excess contrast is a printing problem with a printing solution, and lost separation is gone. The same logic gives D-min an upper control limit and no lower one, since a base-plus-fog reading below aim costs nothing. The fourth option confuses tolerance with resolution: instrument accuracy across the density range is a separate question, addressed by the densitometer certificate written per density band, and it sets the floor below which no tolerance is enforceable rather than the shape of the tolerance itself.

Question 4. Four Tier 1 documents bear on the claim that a seasoned developer is better than a fresh one. Which statements are supported by them?

• Kodak state that some developer and replenisher systems, D-76 with D-76R among them, are formulated so that a fresh solution gives a higher speed than a fully seasoned one
• ILFORD state that there will always be a difference in performance between a fresh tank and a seasoned one
• ILFORD quantify the seasoning effect at about 20 per cent of development time for ILFOTEC DD used without its starter
• Kodak state that a seasoned developer produces finer grain than a fresh one

Show the answer and why

Answer: Kodak state that some developer and replenisher systems, D-76 with D-76R among them, are formulated so that a fresh solution gives a higher speed than a fully seasoned one, ILFORD state that there will always be a difference in performance between a fresh tank and a seasoned one, ILFORD quantify the seasoning effect at about 20 per cent of development time for ILFOTEC DD used without its starter

The three supported statements are printed in Kodak Z-133, ILFORD’s process-control introduction and the ILFOTEC DD sheet respectively, and together they describe seasoning as a real, sizeable, mostly unwanted change: a loss of speed and a gain in contrast, worth 10 to 20 per cent of a development time, which the industry answers by adding a starter so that a fresh tank behaves as a seasoned one from the first film. Kodak’s D-76 sheet says the same thing from the other side, describing the modified T-MAX replenisher as helping to avoid a small speed loss and an increase in contrast as seasoning occurs. No document in this corpus makes any grain claim about seasoning, and the course does not invent one. Note what this does and does not settle: the folk claim is unsupported in the direction it is usually made, which is not the same as disproved.

Question 5. Your drift curve crosses −0.02 in contrast index at 2.4 films per litre and −0.07 at 6.6, and the local slope at the second crossing is 0.013 in contrast index per film. Your resolution sentence allows differences of 0.03. What follows?

• The capacity at the −0.07 tolerance carries an uncertainty of about ±2.3 films per litre, because the crossing point’s uncertainty is the contrast-index resolution divided by the local slope
• The capacity at the −0.07 tolerance is 6.6 films per litre with an uncertainty of ±0.03, the resolution itself
• The capacity at the −0.02 tolerance should be preferred, since a tighter tolerance always gives a more precise answer
• Neither capacity can be quoted, because the resolution of 0.03 is larger than the −0.02 tolerance

Show the answer and why

Answer: The capacity at the −0.07 tolerance carries an uncertainty of about ±2.3 films per litre, because the crossing point’s uncertainty is the contrast-index resolution divided by the local slope

A crossing point is read along the horizontal axis, so a vertical uncertainty has to be divided by the slope to become one: 0.03 ÷ 0.013 is about 2.3 films per litre. That is a large band on a capacity of 6.6, and it is honest — the drift curve is shallow there, so the loading at which it crosses a line is poorly determined even though every individual contrast index is well measured. The second option makes the classic mistake of quoting a vertical uncertainty on a horizontal quantity. The third is wrong twice over: a tighter tolerance gives a smaller capacity, not a more precise one, and the slope is shallower near the start, so the ±0.02 crossing is in fact the worse determined of the two. The fourth is too pessimistic but points at something real — a tolerance tighter than your resolution cannot be enforced, so the ±0.02 capacity should be quoted with that stated rather than dropped.

Check my answersClear