Assignment: Characterising a Film and Developer with Your Own Instruments

Everything in this part has been building an instrument. This is the afternoon it does work.

You will expose nine strips with the sensitometer you built in Part XIV, process them in three batches that differ in exactly one thing, read every step of every strip on the densitometer you built in Part XV, and come out with a family of curves whose points carry error bars you can defend. From that family you will read a development time for a contrast you chose on purpose and an exposure index for the way you actually meter. Then you will turn the instrument’s other head on a sheet of paper and measure its base white and its maximum black, because six later parts quote those two numbers and nothing in this course has been able to measure them until today.

It is the first measurement in the course whose both axes are instruments rather than approximations, and the difference will be visible in the scatter.

What you are producing

Six things, and the last is the one later parts will actually read.

1. A family of characteristic curves: three curves, one per development time, plotted from your own densities against your own log exposures, with error bars on the points.
2. A contrast-index-against-time plot and a speed-against-time plot derived from it.
3. A development time for a stated target contrast, and the printing intention that chose the target.
4. A personal exposure index from your own speed point, with an uncertainty and the criterion named.
5. A reconciliation with the manufacturer’s published times and with the interim curve from Part XIII.
6. A calibration library entry: the film curves, and beside them one paper’s base white and maximum black in reflection density, with the processing and the drying interval recorded.

!Both instruments must be in certificate before you start

This assignment consumes film, chemistry and four hours. Do not begin it with an uncalibrated instrument, because every number it produces would inherit the fault and you would not know which.

The sensitometer needs its certificate: relative log exposure uncertainty, uniformity, warm-up rule, and whether its wedge is calibrated.

The densitometer needs its own: the working range, repeatability at two densities, the warm-up rule, the reflection repeatability figures and the tile serial.

Both certificates are attached to the submission. A curve without them is a drawing.

Safety, and what this session actually handles

Level A. The chemistry is the chemistry of every film-processing session in this course, at the concentrations those sessions use, and the rubric’s criteria for Level A are met: substances at most irritant or harmful in home quantities, nothing heated, no mains construction, and waste that is collected in labelled containers rather than discharged.

Hazards, and the controls that address them. Developer is alkaline and the developing agents are skin sensitisers; the control is nitrile gloves for every solution transfer, tongs rather than fingers in a dish, and the glove guidance for what the permeation data do and do not say. Fixer is a thiosulfate solution and its hazard is chiefly to the drain rather than to you; it is collected. The stop bath at working strength is a weak acid at the concentration Part X established. Splashes to the eye are the reason eye protection and an eyewash provision are assumed at every level of this course.

Ventilation is the ordinary control of an open window with solutions in dishes at room temperature; nothing here is heated and nothing evolves a vapour at these concentrations. Full hazard statements for each substance are on its encyclopaedia page and in the processing SOP, which is followed rather than restated here.

What is not a hazard. The instruments. Both run from a certified USB supply at 5 V and neither is opened during the session. The densitometer’s lamp is an indicator LED behind a diffuser inside a closed head.

One step deserves naming. Loading film onto reels happens in total darkness — ILFORD’s own instruction for HP5 Plus is to handle it in total darkness, and it applies to every panchromatic film. Darkness is a hazard control problem in itself: the bench is cleared and the chemistry is capped before the lights go out, and nothing sharp or spillable is in reach.

Part 1 — The plan, and the one variable

Write the plan before you buy anything, and write it as a plan rather than as an intention.

Choose one film and one developer, and stay with them. The course’s worked example throughout is HP5 Plus in ID-11, because the manufacturer publishes a table this page can quote and because both are widely available; FP4 Plus in ID-11 or Tri-X in D-76 work identically and the arithmetic is the same. Whatever you choose, the film is one emulsion batch and the developer is one mix.

Choose three development times that bracket the manufacturer’s recommendation. ILFORD’s table gives HP5 Plus in ID-11 at 7½ minutes at stock, 13 minutes at 1+1 and 20 minutes at 1+3, all at 20 °C in a spiral tank with intermittent agitation and a meter setting of EI 400.

✕Those three numbers are not three development times, and using them as your three runs would ruin the experiment

Look again at what varies between them. The dilution changes with the time. Stock, 1+1 and 1+3 are three different developers with three different sulfite concentrations, three different solvent actions and three different capacities, and running them as your three points would confound dilution with time in a way no amount of analysis can untangle afterwards.

One variable at a time means: pick one dilution, and bracket its recommended time. At 1+1 the recommendation is 13 minutes, so a defensible set is 9, 13 and 18 minutes — roughly a factor of two across the set, with the manufacturer’s figure in the middle.

Two constraints on the bracket, both from the manufacturers. The short end must stay above five minutes, because ILFORD’s own temperature-compensation chart states plainly that development times below five minutes are not recommended owing to the risk of uneven development. And the long end should stay inside a factor of about two, because beyond that fog begins to move as much as contrast does and the plot you are making stops being about contrast.

Fix everything else, and write down what you fixed. Temperature 20 °C. Agitation: ILFORD’s script for a spiral tank is four inversions during the first ten seconds and four more during the first ten seconds of each further minute, and it is the script for all three runs. Water: the same source and the same temperature. Solutions: ILFORD ask that every bath be at the developer’s temperature or at least within 5 °C of it, so the stop and the fix are made up and brought to temperature before the first run rather than between runs.

One-shot, always. Kodak instruct that D-76 at 1+1 is diluted immediately before use and discarded after processing one batch of film, neither reused nor replenished. ILFORD publish no formula for ID-11 and this course does not assert that the two are the same product — what Part VIII established is that ID-11’s published pH of 8.60 to 8.70 puts it in the D-76 family’s working region — so take the instruction as applying to whichever diluted metol-hydroquinone-borax developer is in your tank.

That is not thrift, it is the experiment: a developer that has already processed a film is a different developer, because, as ILFORD note, the by-products released by each film act as a restrainer on the next one.

Nine strips, three per run. Three identically exposed strips per development time gives you a repeatability figure at every step for free, which is where the error bars on the family of curves come from. Plus, in every run, a control strip from your Part XIV control-strip stock and an unexposed strip of the same film for base-plus-fog. That is five pieces of film per run and fifteen in all.

Part 2 — Exposure: three runs, one instrument, one afternoon

Expose all fifteen strips in one session, before any processing. This matters more than it sounds: the sensitometer’s reproducibility across days is worse than its repeatability within a session, and exposing everything at once removes that term from the experiment entirely.

Obey your own warm-up rule, from the sensitometer’s certificate, and log the minutes since switch-on on every strip. Keep the cadence the certificate quotes, because strips made ten seconds apart and strips made two minutes apart are not the same measurement on a lamp that cools between exposures.

Wedge orientation is fixed and recorded. The step wedge goes into the exposure stage the same way up and the same way round for all fifteen strips, and the strip is marked — a notch at the clear end is the traditional answer and it survives processing. A strip whose orientation you cannot establish afterwards is a strip you must throw away, because step 3 and step 19 are not interchangeable.

One entry per strip in the exposure log: strip number, run it belongs to, requested interval, the firmware’s measured_us, the monitor mean and spread, minutes since switch-on, and anything unusual, written at the moment rather than afterwards.

Then put the exposed strips in the light-tight tin and process them the same day. Latent image fade is a real term and there is no reason to introduce it into an experiment about development.

Part 3 — Processing: making everything identical except one thing

Three runs, in whatever order you like — but write the order down, because if anything goes wrong it is usually the first run and knowing which that was matters.

Per run: mix the working developer from stock immediately before use; bring it to 20 °C; load the five strips of that run; process to the script; and note the temperature at the start, the middle and the end of development rather than only at the start.

!Why the temperature is recorded three times and to a tenth of a degree

Kodak’s process-control publication states that developer temperature variations greater than ±0.3 °C will affect process control and image quality. Your thermometer’s own uncertainty, from Part II’s commissioning session, is probably of that order — which is the point. A temperature you did not measure well enough to detect a 0.3 °C drift is a variable you have not actually controlled, and recording it three times is how you find out whether a tank standing in a cold room drifted through the run.

If it did, ILFORD publish the compensation: their chart converts a recommended time at 20 °C to the equivalent at another temperature, and its own worked example is 8 minutes at 20 °C becoming 5 minutes 30 seconds at 24 °C. Use it to understand a drift, not to excuse one: a corrected time is not the same experiment as a controlled one, and the notebook must say which you did.

Stop, fix, wash and dry identically in all three runs, and record the fixer’s age and the number of films through it. Dry the strips the same way, hanging in the same place, because a strip dried against a radiator is a strip with a different history.

The unexposed strip goes through every run with the others and is what base plus fog is measured on. It is not optional and it is not shared between runs: the whole of Part 7’s fog question depends on having one per run.

Part 4 — Reading: the densitometer session

Set aside an unhurried hour. Fifteen strips at twenty-one steps is 315 readings, plus repeats, and tiredness is a systematic error.

Start with the four-reading preflight the break/fix page specifies: dark, zero, the calibration step, the opaque blank. The calibration step must read within 0.02 D of its certified value — the tolerance a commercial instrument specifies for exactly this check — or you recalibrate before a single strip goes on the stage. The blank must give the stray-light fraction your certificate quotes.

Obey the densitometer’s warm-up rule. Then, per strip:

1. Read all twenty-one steps with series(), in order, clear end first.
2. Read three steps again — one near the toe, one on the straight line, one near the shoulder — as separate readings with the strip lifted and replaced. Those three give you the replaced-sample repeatability for this session, on this film, rather than the certificate’s figure from another day.
3. Re-read the calibration step at the end of every third strip. A drift found at strip 6 costs you three strips; a drift found at strip 15 costs you the afternoon.

Read the unexposed strip of each run the same way and take its density as that run’s base-plus-fog. Read it at three places on the strip, not one, because fog is not always even.

Never transcribe a number. Capture the terminal to a file and work from the file. The single most common way an afternoon like this is lost is a digit typed wrongly at reading 200.

Part 5 — The family of curves, with error bars that mean something

Now the arithmetic. Three columns per strip and one plot.

The log exposure axis. Each step’s log exposure is the instrument’s log exposure at the film plane minus the wedge step’s own density, exactly as Part XIII established. If your wedge is calibrated, use the certificate; if it is not, use the nominal increment and mark the axis relative, because a nominal 0.15 is a manufacturing intention rather than a measured value.

The density axis. The mean of your three identically exposed strips at each step.

The error bars. At each step, the standard deviation of those three densities. Plot it, and then look at what it does across the curve — because a constant density spread is not a constant exposure spread.

σ(log H) = σ(D) ÷ G

Converting a density spread into a log exposure spread

σ(D) is the density spread at that step, G the local gradient of the curve there, and σ(log H) the equivalent uncertainty along the exposure axis. On the straight line, where G might be 0.6, a spread of 0.02 D is 0.033 log H — a fifth of a wedge step. Out on the toe where G is 0.1, the same 0.02 D is 0.20 log H, more than a whole step. This is the same conversion Part XIV’s calibration page insisted on, and it is why a speed figure read off the toe carries a much larger uncertainty than a contrast figure read off the straight line, from exactly the same data.

A family of curves: one film, one developer, three development times

• Short: the shortest of your three times
• Middle: the manufacturer's recommended time
• Long: the longest of your three times

Show the numbers behind this plot

Three characteristic curves of density against relative log exposure from 0 to 3.1, all rising from a common base of about 0.09 to 0.10 at the left. The shortest development time gives the lowest curve, leaving the base at about 0.6 log exposure and reaching 1.42 at 3.0. The middle time gives a curve that leaves the base at almost exactly the same place and reaches 1.93 at 3.0. The longest time gives the highest curve, leaving the base at the same place again and reaching 2.56 at 3.0. The three curves are nearly coincident through the toe and separate progressively through the straight line and the shoulder, so that the vertical gap between them grows steadily from left to right. A horizontal guide at 0.19, which is base plus fog plus 0.10, cuts all three curves within about a tenth of a log exposure unit of each other, which is the whole reason speed changes little while contrast changes greatly. Base plus fog itself rises slightly with development time, from 0.09 to 0.10.

Series	Relative log exposure	Diffuse density above base
Short: the shortest of your three times	0.00	0.09
Short: the shortest of your three times	0.30	0.10
Short: the shortest of your three times	0.60	0.14
Short: the shortest of your three times	0.90	0.26
Short: the shortest of your three times	1.20	0.44
Short: the shortest of your three times	1.50	0.63
Short: the shortest of your three times	1.80	0.82
Short: the shortest of your three times	2.10	1.00
Short: the shortest of your three times	2.40	1.17
Short: the shortest of your three times	2.70	1.31
Short: the shortest of your three times	3.00	1.42
Middle: the manufacturer's recommended time	0.00	0.09
Middle: the manufacturer's recommended time	0.30	0.11
Middle: the manufacturer's recommended time	0.60	0.17
Middle: the manufacturer's recommended time	0.90	0.34
Middle: the manufacturer's recommended time	1.20	0.59
Middle: the manufacturer's recommended time	1.50	0.86
Middle: the manufacturer's recommended time	1.80	1.12
Middle: the manufacturer's recommended time	2.10	1.37
Middle: the manufacturer's recommended time	2.40	1.60
Middle: the manufacturer's recommended time	2.70	1.79
Middle: the manufacturer's recommended time	3.00	1.93
Long: the longest of your three times	0.00	0.10
Long: the longest of your three times	0.30	0.13
Long: the longest of your three times	0.60	0.22
Long: the longest of your three times	0.90	0.45
Long: the longest of your three times	1.20	0.79
Long: the longest of your three times	1.50	1.15
Long: the longest of your three times	1.80	1.50
Long: the longest of your three times	2.10	1.83
Long: the longest of your three times	2.40	2.13
Long: the longest of your three times	2.70	2.38
Long: the longest of your three times	3.00	2.56

Drawn to show the shape the analysis expects, not measured from any film: the three curves fan out from a nearly common toe. Your own three curves will have this shape and different numbers, and if they do not fan from a common toe you have a variable you did not control. 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.

Read four numbers off each run, and put each in the table below with its uncertainty.

Per run	How it is obtained	Where the convention is stated
Base plus fog, D₀	The unexposed strip of that run, three places averaged	Kodak define it as the density of the base plus the density of the fog in the emulsion, and prefer the term D-min
Contrast index, CI	The course’s own construction on the tabulated densities	Part XIII’s contrast-index convention
Speed point exposure, Hm	The exposure at density D₀ + 0.10, with the development condition checked	Part XIII’s speed criterion
Relative speed, S	0.80 ÷ Hm in lux-seconds, or the relative form if your exposure axis is relative	The same criterion

Then make the two derived plots. Contrast index against development time is the one you will use for the rest of your life with this film: Kodak’s workbook calls it the time-contrast-index curve and states its purpose exactly — to make it easy to find the development time for any desired contrast index. And speed against development time, which is the plot that shows you how little of it there is.

What the family gives you: contrast climbs steeply, speed hardly moves

• Contrast index against time
• Relative speed against time, normalised to the longest run

Show the numbers behind this plot

Two traces against development time from eight to nineteen minutes, plotted from the three-run bracket of nine, thirteen and eighteen minutes at one dilution. The first, contrast index, climbs steadily and nearly linearly from 0.47 at nine minutes through 0.61 at thirteen to 0.76 at eighteen, rising by about three fifths across the set. The second, relative speed normalised to its value at the longest time, rises only from 0.88 to 1.00 over the same range, a change of about a fifth of a stop where the contrast changed by three fifths. A horizontal guide at 0.58 shows how a target contrast index is converted into a development time by reading across to the contrast trace and down to the time axis, which lands at about twelve minutes. The two traces together are the assignment's central result: development is a contrast control and barely a speed control.

Series	Development time, minutes	Contrast index (left scale) and relative speed as a fraction of its longest-time value
Contrast index against time	9.00	0.47
Contrast index against time	13.00	0.61
Contrast index against time	18.00	0.76
Relative speed against time, normalised to the longest run	9.00	0.88
Relative speed against time, normalised to the longest run	13.00	0.96
Relative speed against time, normalised to the longest run	18.00	1.00

Drawn from the family above rather than measured; the contrast points are the three runs and the speed trace is illustrative. Your own version of this plot, with your own three points, is the deliverable — and it is the plot you will consult every time you develop this film. 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.

Part 6 — The two working numbers

The development time. Choose your target contrast index — Part 8 is where the choice is argued — and read across your contrast-index-against-time plot to the time axis. Interpolate linearly between your three points; with only three points a curve fit is over-claiming.

Then quote the time with a tolerance, and take the tolerance from your own data rather than from habit. Kodak’s process-control publication accepts a development time when it lands the contrast index inside a ±0.02 window, and that is the tolerance the course adopts as a defensible starting point. Convert it to minutes through the slope of your own plot: if contrast index rises by 0.03 per minute, ±0.02 of contrast index is ±0.7 minutes, and quoting the time to the nearest fifteen seconds is spurious.

The exposure index. Locate your speed point on the curve of the run at your chosen development time, using the course’s criterion, and compute the speed. Write it in the course’s form and no other:

“EI 320, measured under the criterion defined in Part XIII, which is modelled on ISO 6, on emulsion batch NNN in ID-11 1+1 at 20 °C for 12 minutes, read on instrument serial NNN under its certificate of [date].”

Never “the ISO speed is 320”, which is a claim the course does not make anywhere.

And the ±1 stop question. Take your chosen curve and mark where a subject shadow placed one stop under your exposure index would land, and one stop over. A stop is 0.30 in log exposure. Read the density at each. Then say, in one sentence, what each does to the shadow separation you would get in a print — and notice that the answers are not symmetrical, because the toe is curved and the straight line is not.

Part 7 — What is happening chemically

Three questions the family of curves answers, and one it only poses.

Why contrast climbs with time. Development is autocatalytic and it starts at the specks the exposure created. A crystal carrying a large latent-image speck begins developing almost immediately; one carrying a marginal speck takes longer to start. Extra time therefore adds far more density where there were many developable crystals than where there were few, so the difference between a highly exposed region and a lightly exposed one grows. That is a slope, and it is why time is a contrast control.

Why speed hardly moves. Kodak’s own workbook says it in one sentence about the family of curves: the longer the development time, the steeper the slope, most of the change is in the straight line and the shoulder, and the toe remains basically the same. The speed criterion is a point on the toe — a density just 0.10 above base — and a region that barely changes gives a number that barely changes. Your own plot in Part 5 is the evidence, and the honest conclusion from it is that development is a contrast control and only marginally a speed control, which is the single most useful thing this assignment teaches.

What fog does, and where the course’s evidence stops. Kodak define base plus fog as the density of the base plus the density of the fog in the emulsion, and explain chemical fog as a few silver halide crystals developing without having been exposed at all. It follows from the same kinetics as everything above that a developer given longer will convert more of them — but the course has found no manufacturer statement quantifying the growth of fog with development time for any film and developer in its corpus, and it will not put a number to it from reasoning alone.

🧪So measure it: you already have the data

Your three unexposed strips, one per run, processed alongside the exposed ones, are three base-plus-fog values at three development times. Plot D₀ against time.

That plot is a small original measurement. It costs nothing extra — the strips were required anyway — and it answers, for your film and your developer, a question this course could not answer from any source it has read. Report the slope in density per minute, and say plainly that it is your measurement rather than a published figure.

Then use it. If fog grows by 0.01 per minute and you are considering a run at twice your normal time, you now know what it costs at the shadow end, where a rising floor eats the separation you were developing longer to get.

Part 8 — Art track: choosing a contrast index on purpose

A target contrast index is not a technical constant. It is a decision about a print, and this is where you make it and write it down.

The chain is short and every link is somebody’s choice. A negative’s density range is its contrast index times the subject’s log luminance range. A paper accepts a log exposure range fixed by its grade. A diffusion enlarger transfers the negative’s densities nearly as they are; a condenser enlarger, through the Callier effect, transfers them with more contrast. So the target contrast index is whatever makes your negative’s density range land inside your paper’s exposure range on your enlarger.

Write the printing intention as a sentence before you pick a number, in this form:

“Diffusion enlarger, ILFORD Multigrade RC at grade 2, subjects with an average outdoor range, and I want open shadows with separation in the first two zones above black.”

Then pick the number that serves it, read the time off your plot, and — this is the part that makes it an assignment rather than an opinion — state what you would change the target to if the intention changed. A condenser enlarger instead of a diffusion one wants a lower contrast index for the same paper. A habitually contrasty subject wants a lower one. A wish to print at grade 3 for local contrast wants a lower one again. Say which way each pushes and by roughly how much.

Your rule, in one line, is the deliverable of this part: “For this film in this developer, I develop to CI 0.58, which is N minutes at 20 °C at 1+1, and I rate the film at EI M.” That line goes on the box of film, in the notebook, and in the calibration library.

Part 9 — Reconciling with the manufacturer and with Part XIII

Two comparisons, and neither is a test of who is right.

Against the manufacturer. ILFORD publish 13 minutes for HP5 Plus in ID-11 at 1+1, at EI 400, and say what those times are aimed at in their own words: they produce negatives of average contrast suitable for printing in all enlargers, and they are intended as a guide which may be altered if a different result is needed. That is a statement of intent rather than a contrast index, and the manufacturer publishes no contrast index for it that this course has found.

So the comparison is not “is my 12 minutes right or is their 13”. It is: what contrast index does their recommended time give in my hands, and is it the one I want? Develop one extra strip at exactly the manufacturer’s time if your bracket did not include it, read its contrast index, and write the sentence down. Differences of a minute or two are entirely expected and the reasons are enumerable: your agitation is not their agitation, your thermometer is not their thermometer, your water is not their water, and your criterion for contrast index is the course’s rather than theirs.

Against Part XIII. You archived a step-wedge negative in Part XIII and read it with a lux meter, a spot meter and matched grey patches. Read it again now, and plot the interim densities against the instrument’s. The slope of that line is the calibration factor the interim method had all along; the scatter about it is what its precision really was. Both belong in the report, and the second one is the honest answer to the question “was building all this worth it?”

Part 10 — The reflection deliverable: base white and maximum black

Six later parts quote a paper’s maximum black or its base white, and until this attachment existed every one of those numbers was a judgement by eye. Today you measure two of them, and you fix the procedure by which they are measured, because a number read under one set of conditions is not comparable with one read under another.

!The course's own procedure, stated here because a later part depends on it

The course could find no manufacturer statement fixing how a paper’s maximum black should be measured or how long after drying it should be read. Part XIII established that no sourced magnitude for silver-gelatin dry-down exists in this course’s corpus at all. So this page fixes a convention by fiat, states that it is a convention, and asks every entry to record it — because a stated arbitrary rule that everyone follows produces comparable numbers, and an unstated one does not.

The Pure Silver paper-anchor convention.

1. One sheet, two areas. A base white area given no exposure at all, and a maximum black area given at least four times the exposure that first produced a visually maximum black — because a black that is still deepening is not a maximum.
2. Processed to the manufacturer’s own recommendation, at their dilution, their temperature and their time. For ILFORD Multigrade RC in Multigrade developer that is 1+9 at 20 °C for one minute, with the manufacturer’s warning attached: prints developed for shorter times may be underdeveloped and lacking in contrast and density. Stop bath used, which ILFORD strongly recommend; a non-hardening fixer, because a hardening one reduces washing efficiency.
3. Washed and dried fully, flat, in air, by the same method you use for prints you keep.
4. Read 24 hours after the sheet is dry to the touch, at room temperature. The interval is the convention: dry-down is an optical change as the gelatin collapses, its magnitude is unsourced, and a fixed interval makes two entries comparable even while the absolute value stays unknown.
5. Read at five places on each area and record the mean and the spread, because paper is less uniform than film and one reading is an anecdote.
6. Read against the tile and the trap, freshly anchored, on the same backing you fixed in the optical head build.

Then record the entry, and record enough of it that somebody in three years can tell whether it is still comparable with a new one:

Field	Why it is in the entry
Paper: make, product, surface, weight	Glossy and pearl of the same paper are different measurements
Batch or box code, and purchase date	Emulsions change; a batch is the unit that does not
Developer, dilution, temperature, time	The manufacturer’s recommendation, and whether you followed it
Stop and fixer, and the fixer’s age	Fixing affects base white more than it affects black
Wash and drying method	Air-dried, heat-dried and glazed papers are not comparable
Drying interval before reading	The convention above, and any departure from it
Base white, mean and spread of five	The top anchor of every print curve in later parts
Maximum black, mean and spread of five	The number six later parts quote
The difference between them	The paper’s reflection density range, which is what actually decides a print
Instrument, mode, tile serial, certificate date	Because the scale is relative to that tile
Reader and date	So an outlier can be asked about

And write the scale honestly. Your reflection densities are relative to your tile. Write “Dmax 2.08 relative to tile serial NNNN” every time, because the entry is going into a library that later parts will read, and the day somebody compares it with a published figure is the day the omission costs something.

Alternative route

This assignment needs a darkroom, or something that stands in for one, and the requirement is specific: total darkness for loading film onto a reel, and safelit or dark working for the paper sheet in Part 10. Everything else — exposing, reading, plotting, deciding — happens in an ordinary lit room.

For the film, there is a complete alternative and the course has used it since Part II. A changing bag replaces a darkroom entirely for loading, and a daylight developing tank does the rest at a kitchen sink with the lights on. Nothing in Parts 1 to 9 needs a darkroom at any point once the film is on the reel. If you have no changing bag, a windowless room at night with a towel along the door has loaded a great deal of film, and the test is the one Part II gave: five minutes with your eyes adapted and no light visible anywhere.

For the paper sheet of Part 10, there is a partial alternative and one honest limit. The exposure itself can be made by contact under any small lamp with the paper on a bench in a room lit only by an amber safelight — no enlarger is required, because the sheet is exposed to uniform light rather than to an image. What has no alternative is the paper being in the dark until it is fixed. A reader with no safelight and no dark space can still complete Parts 1 to 9 in full, and should record the calibration library entry as outstanding rather than estimating it. An estimated Dmax entered into the library is worse than a missing one, because later parts cannot tell the two apart.

If you have no sensitometer, the whole assignment can be run with the enlarger exposure series of Part XIII instead, at a stated cost: the exposure axis becomes relative rather than absolute, so the contrast index survives intact and the exposure index becomes a relative figure that can be compared only with your own other measurements. Say so in the report; do not quietly present it as the same thing.

What to submit

A single document, and the two certificates.

1. The plan, as written before the session, including the three times and the reason for the bracket.
2. The exposure log for all fifteen strips.
3. The processing record per run: developer batch and its formula version, dilution, three temperatures, agitation script as words, fixer age.
4. The reading log, as the raw file, plus the preflight results at the start and every third strip.
5. The family of curves, with error bars, on the curve-plotting sheets.
6. The contrast-index-against-time and speed-against-time plots, with your three points marked.
7. The base-plus-fog against time plot, which is your own measurement of the fog question.
8. The two working numbers, each with an uncertainty and each written in the course’s form.
9. The printing intention, as a sentence, and the one-line rule that follows from it.
10. The two reconciliations: against the manufacturer’s time, and against the Part XIII interim curve with its slope and its scatter.
11. The calibration library entry for one paper, complete on every field above.
12. Both instrument certificates, attached, with their dates.

Filing, and the calibration library

The data goes out as comma-separated values, in the shape the firmware already writes plus four analysis columns: run, strip, step, log_H, density, sd, n. Keep the raw counts file beside it. A density can always be recomputed from counts, a zero and a slope; counts cannot be recovered from a density.

📓Two things the course does not yet have, recorded here rather than glossed

There is no curve tool. The course’s own loader for this data has not been built and no schema for it is published anywhere. The column set above is what a tool would read, and it is stated here so that files made today will still load when one exists. Until then the spreadsheet is the tool.

The boundary with the later experimental part is not settled. This page owns the protocol for characterising one film in one developer with the course’s own instruments, and it is written once, here, because this is the first part where both instruments exist. Part IX’s developer laboratory owns the comparison of developers by the interim reading method and stands as written. The later part on experimental photographic chemistry is unwritten, and when it is written it should cite this protocol rather than restate it. That is a decision recorded here so that the same protocol is not published twice with two sets of conventions.

The calibration library is the durable product of this part. Physically it is a section of the calibration records — the worksheet set records honestly that it does not yet carry a densitometer sheet, so use the shape of the balance and thermometer sheets, which is reading, reference, deviation, action. Into it go: both instrument certificates; this assignment’s film curves; the paper entry from Part 10; and, from here on, one line per session recording the calibration step’s reading and the date.

That last habit is the one that pays. Three years of a single number, read at the start of every session, answers questions about your instrument and your wedge that no datasheet in this course’s corpus could.

What this assignment puts into the library, and which later parts draw on it

1. 1Sensitometer certificateRelative and absolute log exposure uncertainty, uniformity, warm-up rule, wedge serial. Every exposure figure the course quotes hereafter cites it.
2. 2Densitometer certificateWorking range, repeatability at two densities, stray-light ceiling, reflection repeatability, tile serial. Every density figure hereafter cites it.
3. 3The film family of curvesThree curves, a contrast-index-against-time plot, a speed-against-time plot, and a base-plus-fog-against-time plot, all with your uncertainties.
4. 4Your working ruleOne line: development time for a stated contrast index at a stated dilution and temperature, and the exposure index that goes with it.
5. 5The paper anchor entryBase white and maximum black in reflection density, with paper, batch, processing, drying interval, instrument mode and date — read against your own tile and declared relative to it.
6. 6One line per session, for yearsThe calibration step's reading and the date. Cheap, dull, and the only thing that will ever tell you whether your instrument or your wedge has moved.

The library is not a filing exercise. It is the reason a number quoted in a later part can be traced to an instrument, a date and an uncertainty rather than to a memory.

Recap

Nine strips, three development times, one variable. The family of curves fans out from a nearly common toe, and that shape is the whole result: development steers contrast strongly and speed hardly at all, because the speed point sits on the part of the curve that barely moves — which is what Kodak’s own workbook says about a family of curves, and which you have now seen in your own data.

From the family come two working numbers and one sentence. The development time comes from a target contrast index you chose for a printing intention you wrote down, quoted with a tolerance you derived from your own slope rather than inherited. The exposure index comes from your own speed point under the course’s own criterion, written in the course’s own form and never as an ISO speed. The sentence is the rule you will actually use.

Three things this assignment measured that the course could not look up: the growth of base plus fog with development time; what the Part XIII interim reading method was really worth, as a slope and a scatter; and a paper’s base white and maximum black, under a convention this page had to fix because nobody publishes one.

And the fourth thing, which is not a number. Every point on those curves now has an error bar that traces back through two certificates to a calibrated wedge and a stated criterion. That is the difference between a measurement and an opinion, and it took two parts and an afternoon to buy.

Check your understanding

Question 1. A reader proposes using the manufacturer stock, 1+1 and 1+3 times as their three development times. What is wrong with the plan?

• Nothing; those are the times the manufacturer publishes
• The dilution changes with the time, so development time and developer composition are confounded and cannot be separated afterwards
• The times are too far apart to interpolate between
• Stock developer cannot be used one-shot

Show the answer and why

Answer: The dilution changes with the time, so development time and developer composition are confounded and cannot be separated afterwards

Those three numbers are one time for each of three different developers. Stock, 1+1 and 1+3 differ in sulfite concentration, solvent action and capacity, so a curve that changed between them would be answering two questions at once and no analysis can untangle which. One variable at a time means picking one dilution and bracketing its own recommended time, which for HP5 Plus in ID-11 at 1+1 gives something like nine, thirteen and eighteen minutes. Two constraints bound the bracket: the short end stays above five minutes, which ILFORD state plainly, and the long end stays inside about a factor of two so that fog does not begin to move as much as contrast does.

Question 2. Why does development time change contrast strongly and speed hardly at all?

• Because speed is fixed by the manufacturer and cannot be changed
• Because the speed criterion is a point on the toe, and extra development changes the straight line and the shoulder far more than the toe
• Because the speed point is measured before development
• Because contrast and speed are unrelated quantities

Show the answer and why

Answer: Because the speed criterion is a point on the toe, and extra development changes the straight line and the shoulder far more than the toe

Kodak own workbook says it in a sentence about a family of curves: the longer the development, the steeper the slope, most of the change is in the straight line and the shoulder, and the toe remains basically the same. The speed criterion locates a density only 0.10 above base, which is on the toe, so a region that barely moves gives a number that barely moves. Development is therefore a contrast control and only marginally a speed control, and the practical consequence is that a printer who wants more shadow detail must give more exposure rather than more development.

Question 3. Three identically exposed strips differ by 0.024 in density at a step where your own curve has a local gradient of 0.12. What is that worth on the exposure axis, and what does it tell you?

• 0.0029 log H, which is negligible
• 0.20 log H, more than a whole wedge step, and it tells you the figure came from the toe rather than from the instrument
• 0.024 log H, because density and log exposure are the same units
• It cannot be converted without the absolute exposure scale

Show the answer and why

Answer: 0.20 log H, more than a whole wedge step, and it tells you the figure came from the toe rather than from the instrument

Dividing by the local gradient gives 0.024 divided by 0.12, which is 0.20, against a nominal wedge step of 0.15. The same 0.024 on the straight line at a gradient of 0.6 would be 0.040, a quarter of a step. Nothing about the instrument changed between those two numbers: what changed is the slope of the film curve where the reading was taken. That is why a repeatability figure has to be quoted with the density it was measured at, and why a speed figure read off the toe carries a much larger uncertainty than a contrast figure read off the straight line, from exactly the same data. Option 1 multiplies where it should divide.

Question 4. You measure a paper maximum black as 2.08 on your reflection head. How should the entry read in the calibration library?

• Dmax 2.08
• Dmax 2.08, ISO 5-4
• Dmax 2.08 relative to tile serial 0041, with the paper, batch, processing, drying interval, instrument mode and date
• Dmax 2.08 plus or minus the transmission repeatability from the certificate

Show the answer and why

Answer: Dmax 2.08 relative to tile serial 0041, with the paper, batch, processing, drying interval, instrument mode and date

A reflection density is a ratio against a reference white, and if the reference is an uncertified tile then the number is relative to that tile and to nothing else. Naming the tile makes the entry honest and, just as importantly, makes it comparable with your own later entries. The rest of the fields are what let somebody in three years decide whether the entry is still comparable, since paper, batch, developer, drying method and drying interval all move the number. Option 4 borrows the wrong repeatability figure: the reflection mode has three of its own, and the largest is the one that applies after the head has been refitted.

Question 5. The course could find no source quantifying how base plus fog grows with development time. What does the assignment do about it?

• It quotes a figure from a photographic textbook instead
• It omits the question, since no source supports it
• It measures it: one unexposed strip goes through each of the three runs, giving three values of base plus fog at three times
• It assumes fog is constant, since the effect is small

Show the answer and why

Answer: It measures it: one unexposed strip goes through each of the three runs, giving three values of base plus fog at three times

The three unexposed strips are required anyway, one per run, because the speed criterion is a density above measured base plus fog rather than above zero. Plotting those three values against development time therefore costs nothing extra and answers, for your film and your developer, a question the course could not answer from any source it has read. Report the slope in density per minute as your own measurement rather than as a published figure, and then use it: if fog grows by 0.01 per minute, doubling your development time has a cost at the shadow end that a rising floor will take back from you.

Check my answersClear