Assignment: The Developer Comparison, Measured and Printed

Four developers, developed for the same number of minutes, produce four negatives that differ mostly in how far development got. That is not a comparison of developers; it is a comparison of development states, and it is what almost every developer comparison ever published actually shows. The fix is one sentence long and it costs a whole extra session: take each developer to the same contrast index, using each one’s own time series, and compare what is left over.

What is left over is worth having. Two developers at a matched contrast index still differ in where the speed point sits, in how the toe and shoulder are shaped, in fog, in grain, in acutance, in image colour, in what they cost, in how long they keep and in what they demand of the person handling them. This assignment measures the first four, assesses the next three with the method and the limits stated, reads the last three off documents, prints the results, and turns the whole thing into rows of the course’s film-developer comparison matrix, where every score carries a sentence of reasoning and a source.

What you produce

Four things, and the course will use all of them.

1. A lab report: four characteristic curves at a matched contrast index, a parameter table with uncertainties, and an explicit statement of which differences are real at your measurement resolution and which are not.
2. Four comparison-matrix rows, written to the formulary’s own rules, with a sentence of reasoning and a source behind every score and an empty cell wherever the evidence is missing — plus the limits paragraph that every comparison needs.
3. Four prints, one negative of one subject from each developer, on one paper at one grade, developed and dried identically, judged side by side.
4. A recommendation, one paragraph per developer, naming the use case it wins and the use case it loses, written so that another photographer could act on it.

Learning objectives

By the end of the assignment you will be able to:

• explain why matching development time across developers compares development states rather than developers, and design a comparison that avoids it;
• find each developer’s own time for a stated contrast index from a short bracket, using the manufacturer’s own acceptance rule;
• hold eleven variables still across four wet sessions and say which one you failed to hold;
• state what your instrument can and cannot measure about a stained negative, and why a spectral condition is part of a density rather than a detail of it;
• write an assessment that is not a measurement — grain, sharpness, image colour — with its method, its magnification, its light and its limits stated on the same line;
• score a comparison matrix so that a reader can disagree with a specific sentence rather than with a star, and leave a cell empty when the evidence is not there.

Prerequisites

This is the last practical page of the part and it leans on almost all of it. Designing an experiment that yields a number owns the method. The development time series owns the bracket-and-interpolate operation this assignment performs four times over. Four curve families, measured owns the uncertainty budget. From Part VIII, classical and alternative developing agents and staining and tanning developers own the chemistry of the four families, and Part IX’s developing-agent comparison is the semi-quantitative version of this page. From Part XV, what a density actually depends on is a hard prerequisite for the staining arm and is not optional. From Part XIX, test strips, base exposure and the black point and print development and session consistency own the print stage’s discipline, without which four prints are four accidents.

The four developers, and why these four

The comparison teaches something only if the four are far apart in mechanism, not merely different in name. One from each of four families:

#	Developer	Family	Why it is in the comparison
1	Kodak D-76	Metol–hydroquinone solvent	The reference against which everything else in this course is measured. 100 g/L of sulfite means real silver-solvent action; it is the developer Parts VIII, IX and XXVII all built on
2	Beutler	Dilute high-acutance	The opposite design: a working bath carrying about 0.83 g/L of metol and 4 g/L of sulfite, so little solvent action that the grain boundaries are left alone. The British Journal Annual files it under maximum sharpness at some loss of fine grain
3	Kodak XTOL	Ascorbate	The modern family, and the one whose composition you cannot see. Kodak publish two phrases — “ascorbic acid-based” and “no hydroquinone” — and nothing else in sixteen pages
4	PMK	Staining	A developer whose oxidised agent is part of the image. BERGGER describe a yellow-green tint surrounding each grain and filling the space between them, so that the density of a pyro negative is the conjunction of two densities, the silver’s and the stain’s

!One of the four is a bought product, and that changes what a row can say

D-76, Beutler and PMK are formulas: you weigh the ingredients, you know every one of them, and you can move one and watch what happens. XTOL is a product. Kodak publish a pH, a specific gravity, four make-up volumes, a capacity, a replenishment rate and eight tables of development times — and, of composition, two phrases in a features column. No patent in this course’s corpus names it, and no safety data sheet for it could be obtained at a citable address.

That is not a reason to exclude it. It is the reason to include it, because most readers’ developers are bought and the honest handling of a bought product is a skill. But it does change what its matrix row may say: its behaviour cells can carry measurements, its composition cells cannot carry anything, and a “mechanism” sentence about it is an inference and must be marked as one. There is no open ascorbate film developer in this formulary to compare it against, and its own product page says so.

⚠The staining developer, the alternative route, and who should take it

PMK is classified Level B in this course and pyrogallol alone decides it. Its aggregated GHS classification carries harmful in contact with skin, harmful if inhaled and suspected of causing genetic defects at 100 per cent of notifiers, plus harmful if swallowed and harmful to aquatic life with long-lasting effects. It is a phenol and a fine dust. The Photographers’ Formulary sheet asks for gloves and a dust mask by name and says the pyro is weighed and added outside or under a ventilating hood; BERGGER’s sheet says it is toxic by inhalation, skin contact and ingestion. Catechol, the agent of the other common staining family, carries its own classification and is handled to the same standard.

The alternative route replaces developer 4 with the Rodinal-type p-aminophenol developer, used at 1+25 or 1+50, and the comparison stays worth making: it is still a fourth family — a single-agent, highly dilute, strongly alkaline concentrate — and it still contrasts with the other three on solvent action, dilution and keeping. What is lost is the stain, and with it the one arm that forces the spectral question in Stage 3. Say in the report which route you took.

Take the alternative route if you cannot meet the controls in full, and take it as a matter of course if you are pregnant, if you have any skin condition, or if you are working in a shared kitchen. Mixing the Rodinal-type concentrate is not a soft option either: it dissolves 215 g/L of sodium hydroxide, and concentrated alkali goes into water and never the reverse.

The design, and the one decision the whole thing turns on

Same time, or same contrast index: two comparisons that answer different questions

1. 1Same time — four development states compared, and every difference confounded with gradient
2. 2Same contrast index — gradient held by construction, so what differs is the developer
3. 3The price — one extra session — a short time bracket for each of the four before a single comparison strip is exposed

The contrast indices in the left panel are illustrative and not measured from any developer; they stand for the ordinary situation in which four baths of different activity, given one time, stop at four different places on their own curves. The right panel is what this assignment does instead.

Choose the matched contrast index before anything else, and choose it for your enlarger. Kodak’s published aims are 0.58 for printing with a diffusion enlarger and 0.43 for a condenser. The course’s reference pairing has a convenient anchor at the higher figure: Kodak’s own XTOL table gives ILFORD FP4 Plus at EI 125 a contrast-index aim of 0.58 in 8 minutes at 20 °C, so one of the four developers arrives with a published starting point for the exact target.

What is held, and what is allowed to move

Held, across all four developers and all four sessions. One film and one emulsion batch. One exposure session on the sensitometer, one lamp, one distance, one wedge. One temperature, 20 °C, logged at the start, middle and end of every cell. One agitation script, written in words. One volume per strip, above every maker’s stated minimum. One fixing and washing sequence, with the clearing time measured. One reader, one densitometer, one calibration routine. One paper, one grade, one print developer, one drying method. And a run order decided in advance and drawn from a hat, both for the wet cells and, drawn separately, for the reading.

Allowed to move: the developer, and its time. Nothing else. The time is not a second variable — it is the means by which the contrast index is held, and it is exactly what makes this a comparison of developers.

Two things that cannot be held and must be reported instead. The developers’ published exposure indices differ, and PMK’s is the outlier: BERGGER recommend EI 80 for FP4 Plus where ILFORD rate the film at ISO 125/22°. That is not a confound for the strips, which are exposed by a sensitometer through a wedge on a shared exposure axis, and it shows up as a shifted speed point, which is a result. It is a confound for the pictorial negatives, and the way to handle it is to expose those four films identically at the film’s own rating and let the difference appear where it belongs — in the print.

The second is the stain, and it needs its own stage.

Materials

Item	Quantity	Note
Test strips from one exposure batch	20	12 for the four brackets, 8 for the comparison run in duplicate. Two spares if the tin allows
Pictorial film, same emulsion batch	4 short loads of about 6 frames	Bulk-loaded from one cassette, exposed identically on one subject in one session
Transmission step wedge	1	The Stouffer T2115 or equivalent, the same one throughout
Card label slips and negative sleeves	24 slips, 12 pockets	The eight-field slip from the test-negative lab
Variable-contrast RC paper	1 packet of 25, 5 × 7 or 8 × 10	Four final prints plus test strips. One packet, one batch, one surface
Storage bottles, brown glass or HDPE	6	Two for PMK’s stocks, two for Beutler’s, one for D-76, one for XTOL. A named price gap
Measuring cylinders and a graduated syringe	4 cylinders, 1 × 10 mL syringe	The syringe is for PMK’s 1+2+100, which takes about 10 mL of A and 20 mL of B to a litre of working bath
Balance reading to 0.01 g	1	A named price gap in the course’s price file
Lab notebook sheets, curve-plotting sheets and formula version records	1 set + 20 + 4	One version record per developer mixed

Chemicals

Three of the four developers are mixed from solids, which is what makes this a Level B session. Every formula is on its own formulary page and none of them is reproduced here.

Chemical	Where it is used	Form
Metol	D-76 and Beutler’s solution A; PMK’s solution A	Solid, weighed
Hydroquinone	D-76	Solid, weighed
Sodium sulfite	D-76 at 100 g/L; Beutler’s A at 50 g/L	Solid, weighed
Borax	D-76	Solid, weighed
Sodium carbonate, anhydrous	Beutler’s solution B at 50 g/L	Solid, weighed
Pyrogallol	PMK’s solution A	Solid, weighed — the operation that sets the level
Sodium bisulfite	PMK’s solution A	Solid, weighed
Sodium metaborate, tetrahydrate	PMK’s solution B	Solid, weighed
p-Aminophenol, potassium metabisulfite, sodium hydroxide	The Rodinal-type concentrate, on the alternative route only	Solids, weighed
Rapid fixer, ammonium thiosulfate type	All four arms	Concentrate, diluted 1+4
Paper developer and wetting agent	The print stage and the final rinse	Concentrates

XTOL is bought as two sealed packets and mixed to the maker’s instructions. Nobody outside Kodak knows what is in them, this course does not guess, and a reader about to open them should get the current safety data sheet from the seller first.

Equipment

The sensitometer and densitometer from Parts XIV and XV with both certificates on the bench, and — for this page specifically — the densitometer’s blue channel fitted and calibrated, which the calibration page describes as the second source you were told to buy and fit later. Today is later, and Stage 3 explains why. A changing bag, one thermometer, a stopclock, a balance reading to 0.01 g with its calibration mass, an enlarger or a contact printing frame, a loupe of stated magnification and a light of stated colour temperature for judging the prints. None of it is consumed.

Estimated cost

£££ on the planner’s bands, and it is the dearest page in the part for three reasons that are all avoidable in part. The paper is the largest single item, because four prints need test strips before them. The raw chemicals are bought in pack sizes far larger than the assignment uses, so the first run carries the whole cost of a jar of pyrogallol and a kilogram of sulfite and every later run carries almost none of it. And the film is about three cassettes.

Estimated consumables cost

One run of the whole assignment, at the quantities the Materials and Chemicals sections above specify:

Consumed	This session	Sourced price	Cost this session
35 mm film, one emulsion batch	20 strips plus 4 short pictorial loads, about 3 cassettes	£6.37–£11.40 per 36-exposure roll	£19.11–£34.20
D-76, mixed from solids	1 L of stock	metol at £16.20 per 50 g, sodium sulfite at £13.68–£19.98 per kg, hydroquinone at £11.89 per 50 g, borax at £9.98 per 200 g	£3.30–£3.93 per litre
Beutler, mixed from solids	250 mL of A and 250 mL of B, making 3 L of working bath at 1+1+10	metol, sodium sulfite and sodium carbonate at £7.20 per 500 g	£1.16–£1.24 for the two stocks
XTOL	the smallest packet Kodak sell	£16.99 to £18.00 for the packet that makes 5 L of stock, at two retailers on 7 September 2026	£16.99 to £18.00
PMK, mixed from solids	one 250 mL batch of each stock	Pyrogallol at £44.99 per 50 g and sodium bisulfite at £40.00 per kg; sodium metaborate is still a named gap, so solution B cannot be costed	£22.70 for solution A’s pyrogallol and bisulfite; solution B, —
Rodinal-type concentrate, alternative route only	250 mL	None for the route this page mixes: the raw p-aminophenol hydrochloride and the potassium metabisulfite are unpriced and sodium hydroxide is a named gap. A bought p-aminophenol concentrate is priced, at £12.00 for 250 mL, and it is a different product made to a strength no maker publishes	—
Rapid fixer concentrate	400 mL, to make 2 L at 1+4	£21.05–£25.98 per 1 L of concentrate	£8.42–£10.39
Variable-contrast RC paper	about 16 sheets, 5 × 7	£16.06–£44.71 per 25 to 100 sheets, 5 × 7 in	£10.28–£28.61
Paper developer	1 L at 1+9	£10.52–£20.03 per 500 mL to 1 L of concentrate	£1.05–£4.01
Wetting agent	5 mL, for 1 L at 1+200	£28.70 per 1 L of concentrate	£0.14
Storage bottles, sleeves, card slips, squared paper	6 bottles, 12 pockets, 24 slips	None. Two named price gaps: storage bottles, and sleeving that passes the Photographic Activity Test	—

The priced rows come to £83.15 to £123.22, at the retail ranges read on 5 and 7 September 2026 and recorded in the laboratory planner. That is a floor, not a total: two of the 11 rows carry no dated price at all, and PMK’s solution B carries none either, because sodium metaborate is still a named gap. A priced entry is a dated range to plan against, never a quotation.

Two things about that floor are worth saying aloud, because they go straight into the matrix’s cost column. Three of the four developers can now be costed and the fourth cannot. D-76 works out at £3.30 to £3.93 per litre of stock, or £1.65 to £1.97 per litre at 1+1; Beutler’s working bath at 1+1+10 is the cheapest developer in the course at about £0.39 to £0.41 per litre, because ten of its twelve parts are water; XTOL is £3.40 to £3.60 a litre of stock, or half that worked at 1+1, which makes it the cheapest of the packaged developers the file carries and still dearer than Beutler mixed from solids. PMK is costed by halves — £22.70 of pyrogallol and bisulfite in solution A, nothing for solution B — and the Rodinal-type concentrate cannot be priced at all on the route this page mixes. The matrix’s rule is that a cell which cannot be sourced says so rather than guessing. And the first run pays for the packets. A 50 g jar of metol makes twenty-five litres of D-76; the cost of this assignment is not the cost of the jar.

Waste streams

Four, kept apart and labelled per the container SOP.

1. Spent film developers, alkaline, about 4 L across the four arms.
2. Spent PMK, and its rinse water, kept separate from stream 1 because pyrogallol’s classification includes harmful to aquatic life with long-lasting effects; the PMK page’s own instruction is that its solutions and rinse water are collected rather than drained.
3. Spent fixer and the first rinse after it, acidic and silver-bearing, to the silver stream — and this session’s is unusually rich, because both film and paper pass through it.
4. Spent paper developer, which is a different bath from the film developers and is collected as its own stream.

Streams 1 and 3 are never combined, because mixing an alkaline sulfite solution with an acidic silver-bearing one spoils the recovery.

Alternative route

Two facilities are declared and each has its own route.

Mains. The enlarger is the only mains item, and the print stage is the only part of the assignment that needs it. With Part XVI’s contact printing frame and a bare lamp on a timer, the same four negatives can be printed at contact size on one paper at one grade, developed and dried identically, and judged side by side — which is the whole of what the assignment asks the prints to settle.

A darkroom. The measured half needs a changing bag and nothing else: loading is done in the bag, development in a closed tank, and reading in room light. The print half needs a room that can be darkened, and Part XVI’s reversible blackout is the route. With no darkroom at all, the measured half stands unchanged and produces its matrix rows; what is lost is the art half — which print is better, and why — and that judgement cannot be recovered from the numbers. Say so in the report rather than quietly omitting it. A comparison that names what it could not test is still evidence; one that hides the gap is not.

And the third route, which is about the chemistry rather than the facility, is the substitution of the Rodinal-type developer for PMK, set out in the safety callout above. It is a full route, not a reduced one, and the assignment is complete without a staining developer in it.

Stage 1 — Four time brackets, about 90 minutes

Each developer needs its own time for the matched contrast index, and the fastest honest way to find one is the method Kodak use themselves.

1. Start from a published time where one exists. For ILFORD FP4 Plus at 20 °C: D-76 stock is 8 minutes on ILFORD’s own sheet; XTOL full strength is 8 minutes on Kodak’s, aimed explicitly at a contrast index of 0.58; PMK at 1+2+100 is 10 minutes at 21 °C for FP4 Plus at EI 80; and the Rodinal-type route starts from ILFORD’s 9 minutes at 1+25. Beutler has no manufacturer’s time at all — Windisch gives 7 to 10 minutes at 18 °C for low-speed films and the British Journal Annual gives 8 to 15 at 20 °C for a slightly stronger working dilution — so its bracket is wider and starts from those two ranges rather than from a number.
2. Run three strips per developer, at the published time and at roughly ±30 per cent of it. Kodak’s own bracket for this job is arithmetic — 4, 6, 8, 10 and 12 minutes — because it is hunting for one point rather than drawing a curve, and this is the same job.
3. Read the three, plot contrast index against time, and interpolate the time for your aim. Kodak’s acceptance rule is that a strip landing within ±0.02 of the aim is recorded and used; otherwise fine-tune from the closest one and run another.
4. Write down the tolerance on each time, from the local slope of that developer’s own three points. Four developers will have four different slopes, which means the same ±0.02 in contrast index is worth a different number of seconds in each bath — and that is the first real finding of the assignment.

✕Three strips is a bracket, not a time series

Three points fix an interpolation and nothing else. They do not establish the shape of any developer’s contrast-index curve, they cannot show where it flattens, and they must not be quoted as one. Say “interpolated from three strips at 6, 8 and 11 minutes” every time the number appears, and keep it out of any sentence that begins “this developer’s contrast index curve…”.

Stage 2 — The comparison run, about 60 minutes

5. Two strips per developer at its interpolated time, in the drawn run order, at 20 °C, on the written agitation script, in fresh one-shot developer. Two rather than one, because a difference between two developers means nothing until you know the spread between two strips from the same developer.
6. One pictorial load per developer, on the same reel run or immediately after, from the four identically exposed short loads.
7. Respect each developer’s own minimum volume. They are not the same. Kodak require at least 100 mL of full-strength XTOL for each roll’s worth of an ordinary dilution; the Photographers’ Formulary give PMK a floor of 300 mL of working solution per 80 square inches of film, and BERGGER cap it from the other end at 1,000 cm² of film per litre. A dilute developer starved of volume is a compensating developer you did not intend to make.
8. Fix, wash and dry identically, fixing for twice the clearing time, and give PMK’s negatives its own after-bath if you are following the Formulary’s sequence. Hang everything in the same still air, and do not squeegee.

Stage 3 — Reading, and the problem the stain creates

9. Run the densitometer’s warm-up and daily check, read all twenty-one steps and the masked patch of every strip in a separately drawn reading order, and read three strips twice with the sample lifted and replaced between readings. That replaced-sample spread is your Type A term.
10. Compute, for every strip: base plus fog, the speed point under the course’s criterion, the contrast index under the course’s convention, and maximum density — each with its uncertainty.

!A stained negative has two densities, and your instrument reads one of them

Part XV’s geometry lesson establishes the principle: a density value is defined only when both its geometric and its spectral conditions are given, which is what ISO 5-3 exists to specify. On a neutral silver image the spectral condition barely matters. On a stained one it decides the answer.

BERGGER put the physical fact plainly: a yellow-green tint surrounds each silver grain and fills the space between them and becomes an intrinsic part of the image, so the density of a pyro negative is the conjunction of two densities, that of the silver and that of the coloration. A yellow-green stain absorbs blue light strongly and green light far less. Your densitometer’s ordinary transmission channel is a green LED at about 515 to 545 nm. Ordinary silver-gelatin enlarging paper is blue-sensitive, so the density that decides the print is essentially a blue reading.

So the green channel under-reports what the paper will see, and by an amount that grows with the stain. The instruction that follows is simple and is the reason Part XV told you to buy a second LED: read every strip on both channels, and read the stained ones twice over. Then report three numbers for the staining developer — the green reading, the blue reading, and their difference, which is the stain measured — and say in the matrix row which channel each score came from.

Two honesties go with it. The course’s densitometer is modelled on the standards and conformance is neither claimed nor tested, so a blue channel is an approximation to a printing density rather than one. And a contrast index computed from green readings on a stained negative and compared with one computed from green readings on an unstained negative is a comparison of two different quantities wearing one name. Where you cannot fit a second channel, say that the staining developer’s numbers are not comparable with the other three’s and leave the affected matrix cells empty with that reason written in. An empty cell with a reason is a result; a filled one is a mistake with a number on it.

Four developers at one contrast index: what is left to compare once gradient is held

• A metol–hydroquinone solvent developer
• A dilute high-acutance developer
• An ascorbate developer
• A staining developer, read on the green channel

Show the numbers behind this plot

Four characteristic curves of density against relative log exposure from 0 to 2.4, drawn so that all four have closely similar gradients through their straight-line portions, because all four have been developed to the same contrast index. What separates them is everything else. The dilute high-acutance curve has the lowest base plus fog at 0.10 and the longest, flattest toe, not lifting appreciably until log exposure 0.6 and reaching only 1.15 at the top. The metol-hydroquinone solvent curve starts at 0.14 and reaches 1.24. The ascorbate curve starts at 0.12, lifts off its toe earliest of the four and reaches the highest maximum density of 1.27. The staining curve, read on a green channel, starts highest of all at 0.18 because the stain contributes density even where there is no image, climbs earliest through the toe, and then flattens soonest at the top, reaching only 1.14. The four therefore differ in three visible ways at matched gradient: the height of the fog floor, which spans 0.08 across the four; the exposure at which the toe lifts, which decides where the speed point falls; and the shoulder, where two of the four have begun to compress while two have not.

Series	Relative log exposure	Density
A metol–hydroquinone solvent developer	0.00	0.14
A metol–hydroquinone solvent developer	0.30	0.16
A metol–hydroquinone solvent developer	0.60	0.22
A metol–hydroquinone solvent developer	0.90	0.36
A metol–hydroquinone solvent developer	1.20	0.56
A metol–hydroquinone solvent developer	1.50	0.77
A metol–hydroquinone solvent developer	1.80	0.96
A metol–hydroquinone solvent developer	2.10	1.12
A metol–hydroquinone solvent developer	2.40	1.24
A dilute high-acutance developer	0.00	0.10
A dilute high-acutance developer	0.30	0.11
A dilute high-acutance developer	0.60	0.15
A dilute high-acutance developer	0.90	0.30
A dilute high-acutance developer	1.20	0.53
A dilute high-acutance developer	1.50	0.76
A dilute high-acutance developer	1.80	0.95
A dilute high-acutance developer	2.10	1.08
A dilute high-acutance developer	2.40	1.15
An ascorbate developer	0.00	0.12
An ascorbate developer	0.30	0.15
An ascorbate developer	0.60	0.23
An ascorbate developer	0.90	0.39
An ascorbate developer	1.20	0.59
An ascorbate developer	1.50	0.79
An ascorbate developer	1.80	0.98
An ascorbate developer	2.10	1.14
An ascorbate developer	2.40	1.27
A staining developer, read on the green channel	0.00	0.18
A staining developer, read on the green channel	0.30	0.21
A staining developer, read on the green channel	0.60	0.28
A staining developer, read on the green channel	0.90	0.43
A staining developer, read on the green channel	1.20	0.62
A staining developer, read on the green channel	1.50	0.81
A staining developer, read on the green channel	1.80	0.97
A staining developer, read on the green channel	2.10	1.08
A staining developer, read on the green channel	2.40	1.14

Drawn to show what a matched-contrast-index comparison makes visible, not measured from any developer: the four shapes are illustrative and your own four will differ, possibly in the opposite direction. Note also that the staining curve is labelled with the channel it was read on, because on a stained negative that label is part of the measurement rather than a footnote to it. 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.

Stage 4 — The prints

The measurements settle what is on the film. The prints settle what it is worth, and the brief’s assignment path asks for both.

11. Choose one negative from each developer — the same frame of the same subject — and print all four on one paper, at one grade, developed identically and dried identically. One packet, one batch, one surface. Follow Part XIX’s session discipline, which exists for exactly this: four prints made across two hours with a drifting developer are four different papers.
12. Establish the base exposure and the black point once, per the test-strip page, and then hold the grade. If one negative needs a different exposure to reach the same maximum black, record the difference in seconds — it is a measurement of the four negatives’ printing densities, and for the stained one it is the most honest measurement of the stain you will make all evening.
13. Predict the grade before you print. The negative density range is the contrast index times the subject’s log-luminance range, and ILFORD’s instruction for MULTIGRADE RC DELUXE is to multiply the effective range by 100 and take the nearest published range figure — 160, 130, 110, 90, 70, 60 and 50 for filters 00 to 5. Since all four negatives were developed to one contrast index and photographed on one subject, they should all predict the same grade, and any that does not is telling you something about its shoulder.
14. Dry all four the same way and judge them dry. A wet print is a different print, and this is a comparison.

The assessments that are not measurements

Three of the properties this comparison is for cannot be measured on this bench, and the honest way to handle them is to assess them with the method stated on the same line as the assessment.

Grain. What you can do is rank four negatives by eye at a stated magnification, under a stated light, on a stated area of the negative — the same step of the same strip on all four. What you cannot do is measure granularity, which is a statistical quantity requiring a microdensitometer with a specified scanning aperture, and which this course has no instrument for. So write the assessment as “ranked by eye at 20× on the step nearest a density of 0.6, under a 5000 K lamp, by one observer”, and never as a granularity figure. Note also that the matrix’s grain axis asks a different question — where the maker places the developer in its own grain vocabulary — so your ranking belongs in the lab report and in the row’s reasoning sentence rather than in the score itself.

Sharpness. Edge appearance at a stated magnification is a ranking of the same kind. If the earlier acutance work established a scanner edge trace you can repeat, use it and say what scanner, what resolution and what edge; if it did not, say that the assessment is visual and leave it visual. Beutler’s own claim, as Windisch reports it, is a resolution of up to 141 lines per millimetre — a figure about a target and a lens as much as about a developer, and not something four negatives of a street can confirm or refute.

Image colour and stain. Describe it against a stated reference under a stated light: the unstained control strip beside it, on a light box of stated colour temperature, with the words chosen from a fixed small vocabulary rather than invented per strip. BERGGER’s own words for PMK’s failure mode are useful here — uneven colouring varying from olive green to yellow where agitation is insufficient — which is also a reminder that an image-colour observation can be a fault report rather than a property.

Building the matrix rows

The course’s film-developer matrix already exists, with its scales defined once and its axes each carrying their own question. Reuse them; do not invent a scheme on this page. The scales you will score against are published there: Coarse, Moderate, Fine, Extra fine, Minimum for grain; Low, Moderate, High for sharpness; Published loss / No loss claimed for the speed effect; Soft, Normal, High for contrast; Hours, Days, Weeks, Months, Indefinite for shelf life; One-shot, Reusable, Replenished, Either; the four cost bands; Simple, Moderate, Involved, Specialist; the four safety levels; and the five waste routes.

What a matrix cell has to carry before it is allowed to be a score

1. 1A point on a declared scalechosen from the scale published in the matrix, never invented for this row
2. 2A sentence of reasoning, at least eighty characterssaying why this point and not the one beside it — the schema refuses a score without one
3. 3At least one sourcea document with its section, or the course price file for a cost cell, or this page for a measurement the assignment made
4. 4The claim type, where the argument is the course’smarked inferred when the score was reasoned out of something a source states rather than read off it
5. 5Otherwise: no score, and a reasonan empty cell carries a why saying what is missing — and five of the sixteen axes are empty on purpose

The rule in one sentence: a score is a score, its reasoning and its source, or it is not a score. The schema refuses it, the component refuses to render it, and the validator refuses it a third time.

Four practical points about scoring your four rows.

A measurement you made is a source, and it is the best kind. Where a cell records something this assignment measured, cite this page, name the conditions — film, batch, temperature, agitation, matched contrast index, instrument certificate dates — and give the uncertainty. That is what turns a star into a score.

Silence is not a score. The matrix’s own speed axis says it in its missing-cell note: an unstated speed effect is unknown, not zero. If a maker publishes no speed claim, the cell stays empty with that reason, even if your own measurement suggests a direction — because the axis asks what the maker publishes, and a different question needs a different axis rather than a different answer.

Mark your inferences. The matrix has a field for it. If you scored acutance from the sulfite load rather than from a measurement, that is an inference and the row must say so.

Where a cell disagrees with a page, the page is right and the cell is a bug. These four rows restate what your lab report establishes, with a link back to it. If you find yourself writing a cell that your own report does not support, the fix is in the cell.

The recommendation, and the limits paragraph

One paragraph per developer, each naming the use case it wins and the use case it loses, written so that a reader could act on it. A recommendation that only praises is not a recommendation. Anchor each one to something you measured or something a maker published, and say which: “D-76 at 1+1, for a subject of ordinary range printed on a diffusion head, because it reached the aim in a time whose tolerance is the widest of the four and because it is the only one of the four with a published replenishment scheme” is actionable; “D-76 is a good all-rounder” is not.

Then the paragraph every comparison needs and most omit.

Limits. One film, one emulsion batch, one subject, one paper, one worker, one water supply, one thermometer, one wedge, one evening. Four developers each represented by a single mixing. Every difference reported above is a difference at one matched contrast index, and nothing here establishes how any of them behaves at another. The grain and sharpness assessments are rankings by one observer at a stated magnification, not measurements. The staining developer’s densities were read on a channel that is an approximation to a printing density and conformance to any standard is neither claimed nor tested. What would have to be repeated before any of this generalises: a second film of a different emulsion type, a second subject of a different luminance range, a second matched contrast index, a second mixing of each developer, and a second observer for the visual assessments.

Every clause of that paragraph names something a later worker could go and do. That is what makes it a limits paragraph rather than an apology.

The art half, which is the point of all the measurement

Four prints on a table, all of one subject, all at one grade, all developed and dried the same. Now the question the numbers cannot answer: which is the best print, and why?

Answer it in writing, in terms somebody could disagree with. Not “the second has better tonality” but “the second holds separation in the brick shadow where the fourth has closed it, and its highlight on the window frame keeps texture where the first has bleached it”. Name the areas. Name what you are looking at. A judgement stated that precisely can be argued with, and one that can be argued with is worth having.

Then do the harder half: separate the developer’s contribution from the printing decisions. You chose the exposure, the grade, the dodging and the burning. If print three needed four seconds more to reach the same black, that is a fact about the negative; if you also dodged its foreground, that is a fact about you. Write down what you did to each print, and then say what remains after you have accounted for it. Most of the difference between four prints made by one person in one evening is that person, and a comparison that does not say so is measuring the printer.

And choose which negative to print in the first place with the same honesty. A subject with a short luminance range flatters a developer that builds contrast; one with a long range flatters a compensating one. Say which you chose and what it favours, because that choice is part of the result.

The record to keep

The assignment is finished when six things exist and are filed together.

1. The session records. Dates, room temperature, bath temperature logged at start, middle and end of every cell, the agitation script in words, the run orders as drawn, the deviations as they happened.
2. Four formula version records, one per developer mixed, with the STEM-INITIALS-SEQUENCE code, the mixing date, the measured pH where you took one, and the pack or batch of every solid. A curve tied to “the pyro I mixed in the spring” is a curve nobody can check.
3. Twenty curve-plotting sheets, one per strip, each carrying the conditions that made it, the twenty-one step readings with the reading method’s own resolution beside them, the channel each reading was taken on, and the four figures read off the plot with their constructions named.
4. The parameter table, four developers by four measured quantities, every entry with its uncertainty, and beneath it the sentence that makes it a result: the smallest difference this session may report is ____ in contrast index, ____ log H in the speed point and ____ D in base plus fog.
5. The four prints, dry, labelled on the back with the developer, its version code, the negative, the grade, the exposure and any local work, and stored flat with the negatives they came from.
6. The four matrix rows and the limits paragraph, plus the four recommendation paragraphs.

Two of those six are what the rest of the course consumes. The matrix rows go into the film-developer matrix, where they will be read by someone choosing a developer who has never met you. And the prints go into the archive with their negatives, because the capstone will ask you to look at them again — and a print you can re-read against a number you wrote down is the only kind of evidence a photographer accumulates.

Recap

The design is one decision: match the contrast index, not the time, and pay for it with a bracket session. What that buys is a comparison in which the gradient is held by construction, so that speed, toe shape, fog, maximum density, grain, sharpness and colour become properties of the developer instead of symptoms of how far it got.

Everything after that is bookkeeping of a demanding kind. Hold eleven things still and name the one you failed to hold. Read a stained negative on a channel that matches the paper, or say that you could not. Write every assessment with its method on the same line. Score a matrix only where a source or your own measurement supports it, and leave the rest empty with the reason showing. Then print the four negatives identically, argue for one of them in words that can be disagreed with, and separate what the developer did from what you did.

The four rows you hand in will outlive the evening. That is why every one of them has to carry its reasoning and its source, and why the most valuable cells in the matrix are the ones that say, in a sentence, exactly what nobody yet knows.