Part XXVII Overview: From Observation to Measurement

Part IX asked whether adding bromide lowers fog, and could answer only in the currency it had: a threshold step that moved some countable number of places along a wedge, and a masked patch that looked cleaner, or did not, than the one beside it. That is a direction, and a direction is worth having.

This part asks the next three questions. How far? How repeatably? And is the difference larger than the apparatus that reported it? Answering them changes what a practical page hands you. Until now the thing you carried out of a session was a negative, a print or a working instrument. From here it is a number, an uncertainty on that number, and a sentence written before the session saying what result would have shown the idea to be wrong.

What the second pass adds, and what it is careful not to repeat

Part IX ran three experiments with no instrument in the room. The activity series spent seven strips on three alkali levels and four bromide additions. The solvent series spent fourteen on sulfite from 10 to 100 g/L, on dilution, and on three agitation scripts. The developing-agent comparison spent eight more on six agent loads. Every one of those pages ends the same way: fix, wash, dry, label and sleeve every strip, because this part will read it and will not have you there to ask.

Reading a step wedge by eye buys three things honestly, and the experimental-design lesson is careful about which: the threshold step, which is a speed reading; the scale length in steps, which on a T2115 converts to a log exposure range at 0.15 a step; and base plus fog as a rank against the strips beside it. What it cannot buy is a gradient, because a gradient needs densities and the eye returns positions.

So the second pass is not the first pass repeated. Four of Part IX’s arms — pH, bromide, sulfite and dilution — come out of their sleeves and go under a densitometer aperture, one step at a time. Between them the two archived sessions hold fourteen cells for those four arms, plus the shared reference strip each session carried to tie itself to the others. Nothing is developed again except where an arm is missing or spoiled, and that top-up run is bounded to the strips it replaces.

One strip, read twice: what the instrument adds

1. 1Threshold step and scale length — positions on a wedge — a speed reading and a range, and nothing about slope
2. 2The masked patch — base plus fog, ranked against neighbours because the eye compares better than it measures
3. 3One aperture, twenty-one readings — each step read in turn, in a run order fixed in advance
4. 4Four quantities per strip — base plus fog, speed point, contrast index, maximum density — each with an uncertainty

The strip is the same object in both panels. What Part XXVII adds is not more film but a scale, and with it the right to say how much rather than which way.

Two of Part IX’s arms are deliberately not re-read as curve families. Its three agitation cells are in the archive and nothing stops you reading them, but the agitation experiment here runs its own strips, because its central measurement is a density profile along the strip and a continuous-tone negative, and the Part IX session was not exposed for either. Its developing-agent comparison is not re-read either: the closing assignment does that work again from scratch, and takes each developer to the same contrast index rather than the same time, which is a different experiment wearing the same name.

The three things you must already have

A sensitometer with a calibration record. Part XIV builds the instrument and, more importantly, the certificate: a relative log exposure uncertainty, the terms that went into it, and the rule used to combine them. That page’s worked budget adds a wedge term of 0.02, a uniformity term of 0.015, a drift term of 0.01 and a timing term of 0.002 to about 0.047 log H — roughly a third of a nominal wedge step — falling to about 0.029 when the wedge is a calibrated one. Those are that page’s figures for a worked example, not a specification your instrument has met; your own come from your own measurements. Notice which term dominates: a Stouffer T2115 is sold uncalibrated, and only the CC parts are supplied calibrated against a national standard, so the largest single term in most budgets is settled at the moment the wedge is bought, before anything is built.

A densitometer with a calibration record. Part XV ends in an uncertainty statement written per density band, because an instrument whose resolution changes across its range has no single figure. It also ends in two lines that this part leans on constantly: a working range, running from zero to the density at which the residual leaves ±0.02, and a geometry line reading modelled on ISO 5-2; conformance not claimed and not tested. For scale, X-Rite’s brochure for a commercial desktop transmission densitometer publishes a repeatability of ±0.01 D and a linearity of ±0.02 D from 0.0 to 5.0 D. That is what a metal-cased instrument claims about itself in a specification table. It is not a target your box has met, and this course makes no such claim for it.

That geometry line has a consequence this part lives with on every page. The Part XV head has an opal diffuser below the film and a large photodiode a millimetre above it, which makes it diffuse on both sides — neither of the two conditions ISO 5-2 defines — and a green LED with a bare silicon photodiode, which is a spectral product but not a specified one. So your densities are internally consistent and comparable with each other, which is all a curve family needs. They are not the standard diffuse visual densities a manufacturer plots its published curves in, and Part XV’s geometry lesson is where that difference is worked out. Compare your strips with each other freely; compare them with a datasheet only with that caveat written next to the comparison.

The habit of processing identically. Kodak’s process-monitoring publication states that a developer temperature varying by more than 0.3 °C affects process control and image quality, and ILFORD’s compensation chart shows why: on its 8-minute row, 20 °C becomes 9 minutes 45 seconds at 18 °C and 6 minutes 30 seconds at 22 °C, from which the course takes about 9 per cent of development time per degree near 20 °C. A temperature you did not hold is a development time you did not run. ILFORD’s own process-control sheet recommends a good-quality liquid-in-glass thermometer for checking a built-in sensor, on the grounds that very little can go wrong with it — advice this part takes literally, because one thermometer that is precise and wrong will move every curve in the same direction and look like chemistry.

!The sentence that separates a finding from an impression

Write the resolution decision down before the first reading: the smallest difference this apparatus can distinguish, taken from the budget rather than from hope.

Then compare two claims. “This developer is more active” cannot be shown to be wrong; no measurement contradicts it, which is exactly what makes it worthless as a result. “At 11 minutes at 20 °C this bath gives a contrast index 0.08 higher than the control, against a combined uncertainty of 0.03 on the difference” can be shown to be wrong by anybody with a strip and an instrument — and that is what makes it a result. The brief’s seventh rule asks that fact and interpretation be kept apart; the practical form of it here is that a finding must be capable of being wrong, and a difference smaller than your budget is not a finding at all. Precision and accuracy are also different words, and the difference matters: a repeatable instrument reading against a wrong reference is precisely wrong every time.

What this part hands to the rest of the course

Four products, and each of the later parts and the capstone consumes at least one.

• A contrast index against development time curve for at least one film in one developer, with a three-temperature arm alongside it, which becomes your personal processing table.
• A personal exposure index, measured under the course’s speed criterion rather than adopted from the box.
• A measured capacity figure with the tolerance that defines it written beside it, because capacity is a decision about how much drift you will accept and not a property of the bottle.
• Comparison-matrix rows for the film developer matrix, each score carrying a sentence of reasoning and a source, which is what the brief means by no unexplained stars.

What a time–contrast index curve looks like: Kodak's numbers, for a film it declines to name

• Kodak H-740 answer key, six published values
• The two readings the workbook takes back off its own curve

Show the numbers behind this plot

Six published points rise from a contrast index of 0.51 at five minutes through 0.55 at six, 0.62 at eight, 0.67 at ten and 0.72 at twelve to 0.73 at thirteen minutes. The rise is steep at the left and almost flat at the right: the three minutes from five to eight buy 0.11 of contrast index, while the last minute from twelve to thirteen buys 0.01. A second, dotted series marks the two values the workbook reads back off the curve it asks the student to construct, 0.58 at seven minutes and 0.70 at eleven, which is the operation the plot exists for: a target contrast is chosen on the vertical axis and a development time is read down from it. The curve's flattening at the top is the honest content of the plot, because it shows that beyond a certain time development stops buying contrast at a useful rate.

Series	Development time, minutes	Contrast index
Kodak H-740 answer key, six published values	5.00	0.51
Kodak H-740 answer key, six published values	6.00	0.55
Kodak H-740 answer key, six published values	8.00	0.62
Kodak H-740 answer key, six published values	10.00	0.67
Kodak H-740 answer key, six published values	12.00	0.72
Kodak H-740 answer key, six published values	13.00	0.73
The two readings the workbook takes back off its own curve	7.00	0.58
The two readings the workbook takes back off its own curve	11.00	0.70

These are Kodak's published values for the worked example in its sensitometry workbook, which names neither the film nor the developer and calls them XYZ and A. The shape is the lesson; the numbers describe no material you can buy. Your own six points, for your film in your developer at your temperature with your agitation, will not land on this line, and the whole of the development-time experiment is about producing them.

The workbook also answers, in one line, why this part holds so much still: the four factors affecting contrast index are time, temperature, agitation and developer. Every experiment here moves one of those four, in one arm at a time, and holds the other three.

The limits, stated at the start rather than in the conclusion

One film. One water supply. One worker, with one pair of hands and one habit of inversion. One thermometer, one balance, one wedge, one lamp, and a room whose temperature drifts through the evening. Everything this part measures is measured through that apparatus.

What generalises is the method, and usually the direction that Part VIII derives from mechanism: raise the pH and activity rises, add bromide and the toe is restrained, dilute and the upper scale flattens. What does not generalise is the number. A contrast index you measured at a stated time is a fact about your film batch, your bath, your thermometer and your agitation on that evening, and quoting it as a fact about the developer is the error this part exists to train out of you.

That is also why the course prints its own measurements as its own. Where a figure comes from a manufacturer, the page says the sheet states. Where it comes from a session at this bench, the page says the course measured, names the conditions and gives the uncertainty. Where a claim is common in the literature and cannot be traced to a source meeting the standard this course sets itself, the page says that too, and leaves the question open rather than closing it with a plausible sentence. ILFORD prints a pH range for fresh ID-11 stock of 8.60 to 8.70 measured under controlled laboratory conditions and then advises users to make their own control measurements from their own solutions — which is the same discipline, coming from the people who made the developer.

How the strips are spent

The part is planned so that one film’s worth of work is not wasted, and so that the archive does most of the heavy lifting.

Page	What it reads	What it develops	What it produces
Designing an experiment that yields a number	Nothing	Nothing	An experiment plan and an uncertainty budget, written before any strip is spent
Four curve families, measured	The archived Part IX pH, bromide, sulfite and dilution cells	Only a bounded top-up run, where an arm is missing or spoiled	Four curve families and the uncertainty budget that governs the whole part
The development time and temperature series	Its own strips	Six times at one temperature, three temperatures at one time	The contrast-index-against-time curve, the temperature coefficient and the personal processing table
Agitation, stand and semi-stand	Its own strips and its own negatives	Three agitation regimes, on strips and on a continuous-tone subject	An evenness measurement and a measured account of what stand development does
Exhaustion, capacity and what a used developer does	Its own strips	One strip per film processed, across three ageing treatments	A capacity figure with the tolerance that defines it
Push and pull processing	The development-time series data	Nothing	Push and pull decisions read off your own curve rather than a rule of thumb
The developer comparison	Its own strips, negatives and prints	Four developers, each to a matched contrast index	Comparison-matrix rows, a lab report and four prints

Two consequences for buying film. First, a 36-exposure cassette yields ten or eleven strips of 135 mm at the exposure geometry Part IX specified, so the new development here is counted in cassettes rather than frames, and the comparison assignment alone needs a short time bracket for each of its four developers before a single comparison strip is exposed. Second, buy from one emulsion batch, and if you can, buy it at the same time as Part IX’s. A top-up strip from a different batch is a different material, and the difference will sit inside a curve family looking exactly like chemistry.

What it costs

Time. The seven pages total 1,030 minutes of time on task, of which 750 are the four experiments. That is not a weekend; it is a season of evenings, and the sequencing above exists so that none of them is spent twice.

Money. ££, on the planner’s bands. The recurring cost is film and the four developers’ raw chemicals; the fixer, stop bath and wetting agent are the ordinary ones. Three items this part needs are not priced in the course’s price file, and the planner shows their band alone: a pH meter with buffer standards, small storage bottles for the exhaustion experiment’s ageing arm, and sleeving that passes the Photographic Activity Test for the strips you are going to keep. Any subtotal you compute for this part from the price file is therefore a floor rather than a total, and the experiment pages say so in their own consumables tables.

⚠What raises the classification here, and what does not

The measuring half of this part is instrument work on dry, fixed, washed strips. That half involves no chemical handling at all, and the hazards in front of you are the low-voltage ones of the densitometer and its lamp.

What raises the level is mixing developers from raw chemicals. Weighing metol, hydroquinone, phenidone, pyrogallol or catechol means handling fine powders that must not be inhaled; metol, hydroquinone and phenidone are skin sensitisers; and adjusting pH means handling alkali in concentrate. Under the course’s rubric that is Level B — Level A’s gloves, eye protection, ventilation and dedicated labelled vessels, plus chemical splash goggles and an apron whenever concentrate is open, weighing over a tray with a particulate mask where the safety data sheet requires one, and an eyewash and spill kit within reach. Concentrated alkali goes into water, never the reverse. The curve-families page is Level B only for its bounded top-up run: if no top-up is needed, that session is instrument work and the Level B controls stay in the cupboard.

The closing assignment is the highest-risk page in the part, because it mixes four developers in one session and may include a pyrogallol or catechol staining formulation. Both are readily absorbed through skin and both are toxic; that page carries an alternative route replacing the staining developer with a non-staining one, and it is the right route for anyone pregnant, anyone with a skin condition, and anyone working in a shared kitchen.

The exhaustion experiment adds one hazard of its own that has nothing to do with toxicity: it keeps a part-full open bottle of deliberately oxidised developer for a stated period. Label it, and store it where it cannot be mistaken for working solution.

Spent developer is alkaline and is collected separately from the acidic, silver-bearing fixer, which goes to silver recovery; the two are never combined. The disposal page explains why the route is jurisdictional. Check your local regulations; they govern.

The notebook, and the code that makes a result a result

Every experiment in this part produces numbers that will be read again months later, by you, in an argument with yourself. Three records make that possible, and all three already exist in the course.

The lab notebook sheets carry the session: what you set out to test, what you actually weighed, what you saw, and the one change you will make next. The curve-plotting sheet carries one strip — the conditions that made it, the twenty-one step readings with the reading method’s own resolution beside them, and the four figures read off the plot with their constructions named. And the formula version record carries the bottle.

That last one is the discipline this part cannot work without. The developer variants generated here — a base with the sulfite moved, a bath trimmed to a measured pH, a dilution the manufacturer does not publish — are not formulary entries and are not presented as published formulas. They are versioned in the course’s own scheme, STEM-INITIALS-SEQUENCE, by the versioning SOP: D76-EB-001 is a claim about parentage, an identifiable pair of hands and a batch nobody else mixed. A curve tied to “D-76 with a bit more bromide” is a curve that cannot be checked, repeated or argued with, and a result that cannot be checked is not a result. Write the code on the label before the bottle is filled, and on every strip, plot and print the batch produces.

Alternative route

Six of the seven pages need no darkroom. A changing bag, a daylight tank, a tap, a bench and the two instruments will run the whole measured programme: the curve families are read on dry archived strips in room light, and the time-and-temperature, agitation and exhaustion experiments load in the bag and process in the light, exactly as Part IX’s sessions did.

The exception is the print half of the closing assignment, which needs a room that can be darkened and, ideally, an enlarger. Two routes exist. With Part XVI’s reversible blackout and its contact printing frame, the same negative can be printed from each developer 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. With no darkroom at all, the measured half of the assignment stands unchanged and produces its matrix rows; what is lost is the art half, the judgement of which print is better and the argument for 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, and one that hides the gap is not.

What you need first

From Part IX, all of it, and in particular the archive — this part is only as good as the labelling on those sleeves — together with the laboratory report, which is where the distinction between a position and a density was first argued. From Part XIII, the characteristic curve, gamma, contrast index and average gradient and film speed and exposure index, which own the three constructions every page here reads off a curve. From Part XIV and Part XV, the two calibration pages and their certificates, plus what a density actually depends on, because a number compared across two geometries is two different numbers. From Part II, measurement and uncertainty, which owns the rule this part combines terms by.

If you have done Part XV’s characterisation assignment, you have already made a three-point version of this part’s central curve. The development-time series here is that assignment taken seriously: six points instead of three, a temperature arm beside it, and an uncertainty on every one of them.

The pages of this part

Part27 of 28Level4 — SpecialistPages7Estimated time17.0 hoursHighest safety levelLevel B

✱4 of 7 pages in this part need a darkroom, a UV source or mains-powered equipment, marked below. Each says what can be improvised and, where one exists, gives an alternative route.

0 / 7 lessons in this part completed

• LessonDesigning an Experiment That Yields a Number45 min
• ExperimentExperiment: Four Curve Families, MeasuredB240 min
• ExperimentExperiment: The Development Time and Temperature SeriesA✱ Darkroom180 min
• ExperimentExperiment: Agitation, Stand and Semi-Stand DevelopmentA✱ Darkroom180 min
• ExperimentExperiment: Exhaustion, Capacity and What a Used Developer Actually DoesA✱ Darkroom150 min
• LessonPush and Pull Processing: What Extended Development Can and Cannot Buy45 min
• AssignmentAssignment: The Developer Comparison, Measured and PrintedB✱ Darkroom · Mains180 min