Part XXVIII Overview: Diagnosis as a Discipline

A negative that came back wrong is not a mystery. It is a record of everything that happened to it, written in silver, and most of that record is still legible. The streak has a direction. The fog is either in the rebate or it is not. The marks either repeat at the pitch of the sprocket holes or they do not. Each of those observations costs nothing and eliminates whole classes of cause.

This part is the diagnostic half of the course. Parts VIII, X, XI and XII taught why development, stopping, fixing and washing work. This part teaches what it looks like when one of them does not, and — more importantly — how to find out which one without changing four things at once and hoping.

Every defect is the visible end of an event

The premise of the whole part, and it is worth stating flatly because it is what makes diagnosis possible at all:

Every defect on this course’s list is the visible end of a chemical or physical event, and that event happened at one identifiable point between loading the camera and drying the print.

Nothing here is bad luck. A drying mark is dissolved solids concentrating as the last water evaporates. An air bell is a bubble of gas that sat against the emulsion while the developer worked everywhere else. Bromide drag is a real hydrodynamic flow of exhausted, bromide-rich solution off a dense area. In each case there was a moment at which the fault was created, and in each case something about the mark still points at that moment.

That premise has a corollary that is worth having early: a fault you cannot locate on the chain is a fault you have not finished diagnosing. “The film is bad” is not a diagnosis. “The film was stored above 30 °C for a summer and its fog floor has risen” is one, and it implies a different purchase next time.

The chain, drawn as a fault tree

The chain is longer than most people picture, because it starts before you own the film and it runs twice — once for the negative and once again, in a different chemistry, for the print.

The negative's chain, from the factory to the dry strip

Manufacture and distributionEmulsion, coating, packing. You inherit it; the batch number and the expiry date are the only handles you have on it.
Storage before useTemperature and humidity, in a shop and then in your house. Fog rises here and speed falls here, and neither is visible until the film is developed.
LoadingCassette felt, pressure plate, film path, a spiral forced in the dark. Scratches, cinch marks and crescents are made here.
ExposureMeter, shutter, aperture, filter factor, reciprocity on a long exposure, flare. Density lives here.
Latent-image keepingThe interval between the shutter and the developer. Long intervals and warm ones cost a little shadow density.
DevelopmentTime, temperature, dilution, agitation, capacity, contamination. Contrast lives here, and so do fog, mottle, drag and surge.
StopOr a rinse, or nothing. Carry-over into the fixer is decided here, and so are gas blisters where a carbonate bath meets a strong acid.
FixClearing is not fixing. Under-fixing, over-fixing, dichroic fog and the sludges begin here and show themselves months later.
WashWhat is left behind after the fixer has done its work, and how much of it leaves.
DryDrying marks, water spots, dust, and the last chance to touch a soft emulsion.
Storage after processingEnclosures, humidity, oxidants. Silver mirroring and sulfiding are storage faults as much as processing ones.

Then the whole chain again, in different chemistry and on a reflective support, for the print. A defect names a place on this diagram; the work of diagnosis is narrowing to one box.

Kodak’s Z-133 publication gives the professional laboratory’s version of the same list. Its Causes of an Out-of-Control Process are improper solution mixing, improper storage and keeping, solution contamination, incorrect temperature, incorrect time, improper agitation, improper replenishment, evaporation and equipment malfunction. Nine causes, and every one of them is a box on the diagram above. The home darkroom has no replenishment pumps, but it has the other eight in full.

What a strip of film says about itself

The single most useful thing about a piece of 35 mm film is that most of it was never in the picture.

A strip of 35 mm film, read as evidence rather than as anatomy

Rebate, above and below the frames — Never exposed in the camera. Its density is base plus fog and it is your process floor, free with every film.
Edge printing: film name, batch, frame numbers — Exposed at the factory and developed with your film. Faint or missing edge printing convicts the developer, never the camera.
Inter-frame gap — Unexposed by the lens but reachable by a light leak in the camera back. Fog in the gap but not the rebate is a strong pointer.
Leader — Exposed to full room light while loading. It should be at maximum density. If it is grey, the developer or the fixer is the fault.
Sprocket pitch — The ruler the film brings with it. A mark that repeats at this pitch was made through the perforations, in a tank.
Frame spacing — Even spacing says the transport worked. Overlapping or drifting frames are a camera fault, not a chemistry one.

Sheet film has the same evidence in a different arrangement: the notch code, the area under the holder's rebate, and any margin the holder masked. Roll film has no perforations, so the sprocket ruler is missing and the edge numbers do the same job.

ILFORD’s own guide to a first film puts the reading rule in one sentence: the film edges — the rebates — should be clear, with legible frame numbers along the bottom, and a correctly exposed and processed negative should have a full range of tones, with some parts almost as clear as the rebates and other parts so dense you can only just read print through them.

Every clause in that sentence is a test.

Clear rebates say the fixer worked and the fog floor is where it should be.
Legible frame numbers say the developer worked, because those numbers were exposed at the factory and can only be brought up by development.
Some parts almost clear say the exposure reached the toe of the curve rather than sitting on it.
Some parts so dense you can just read print through them say development carried the highlights where they needed to go.

The three questions that resolve most cases

Before any chemistry, three observations. They are free, they are quick, and between them they eliminate more possibilities than anything else you can do.

Ask these before you think about causes

1. Is it on every frame, or on one?Every frame points at the film, the tank, the chemistry or the camera body. One frame points at that exposure — the light, the subject, a finger on the lens, a moment of shutter trouble.
2. Is it inside the image area, or does it reach beyond it?A defect that reaches the rebate and the inter-frame gap could not have come through the lens. That single fact excludes exposure, the shutter and the aperture in one stroke.
3. Does it follow a feature of the equipment, or a feature of the picture?A mark at the sprocket pitch follows the perforations. A streak below a bright sky follows the picture. The first is a tank fault; the second is a flow fault caused by what you photographed.

Write the answers down before naming a cause. The discipline is not the questions — anyone can ask them — it is recording what each answer eliminated, so that you cannot quietly un-eliminate it later when a more interesting explanation appears.

The second question deserves emphasis because it is the one people skip. Light that came through the lens cannot reach the rebate. The rebate is behind the frame mask; nothing focused by the lens ever touches it. So if the rebate is fogged, the cause is either something that happened to the whole film at once — chemical fog, age fog, heat, radiation, a light leak in the tank or the cassette — or something that happened after the film left the camera. This is why the fog page can dispose of half its candidates before running a single test.

The discipline that separates diagnosis from guessing

Three habits, and they are not optional if you want the second occurrence of a fault to take ten minutes rather than an evening.

Change one thing. A negative with three possible causes is settled by an experiment that isolates one of them, not by a new developer, a new time and a new agitation scheme all in the same session. If all three change and the fault goes away, you have learned nothing except that some combination of three things works, and you have not learned which.

Spend a sacrificial piece of material. A control strip, a clip test, a scrap of film dropped on the fixer, a sheet of paper cut into four. The whole of process control rests on putting something known through the process and reading the answer. Z-133’s instruction after every corrective action in its troubleshooting table is the same one, repeated fifty times: process another control strip to confirm that the change has returned the process to control before you resume normal work.

Keep the record that makes comparison possible. A bad session tells you almost nothing on its own. A bad session set beside a good session, with times, temperatures, dilutions, bath ages and agitation written down for both, usually tells you the answer in one line. The process-control chapter builds the formal version; the informal version is a notebook and takes thirty seconds a session.

The pages of this part

The material is grouped by mechanism, not by appearance, because two defects that look alike often have nothing in common and two that look different are frequently the same event at different scales.

Part28 of 28Level3 — AdvancedPages8Estimated time8.3 hoursHighest safety levelLevel A

3 of 8 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 / 8 lessons in this part completed

LessonReading the Negative as Evidence: Thin, Dense, Flat and Contrasty60 min
LessonFog and Where It Comes From Darkroom50 min
LessonUneven Development, Bromide Drag and Surge Marks45 min
LessonThe Emulsion Under Stress: Reticulation, Scratches, Pinholes and Drying Marks50 min
LessonFixer Faults, Washing Faults and the Stains They Leave50 min
LessonChemistry in the Bottle: Oxidation, Exhaustion, Precipitation and Contamination45 min
LessonPrint Faults: Weak Blacks, Mottle, Bronzing and the Rest Darkroom45 min
Break/fixBreak/Fix: The Film That Came Back WrongA Darkroom150 min

Six of those pages are mechanism pages. Density and contrast come first, because they are the faults of exposure and development read off the negative itself and they are the commonest thing anyone brings to a diagnosis. Fog is unwanted development or unwanted light. Flow marks are the developer’s motion over a surface. The emulsion under stress is one page rather than seven because reticulation, frilling, scratches, cinch marks, pinholes, drying marks and static discharge all share one physical premise: gelatin is a gel that swells when wet, is soft while it is swollen, shrinks as it dries and carries charge when it is dry. Fixing, washing and staining faults are all failures to remove silver or failures to remove what removing it left behind. And the last of the six is the chemistry in its bottles, which is where a surprising number of “development faults” were actually created, days before the film went into the tank.

A seventh page takes the print faults that have no negative equivalent, because a print fails in reflection density and in paper, which behave differently. The eighth is the clinic that closes the part: five cases with their session records, worked to a diagnosis, and two faults reproduced deliberately so their signatures are learned from a specimen you made yourself.

What this part cites and what it refuses to repeat

Every page here cites rather than restates. That is a deliberate constraint and it is worth knowing why.

The mechanism of fixing belongs to Part XI. The physics of washing belongs to Part XII. Development kinetics belong to Part VIII, and the shape of the characteristic curve to Part XIII. If this part explained them again it would drift from them, and a course that says two different things about the same chemistry in two places has a bug that the reader has to find.

So the subject of this part is narrower and sharper than “photographic chemistry gone wrong”. It is the signature and the diagnosis: what a fault looks like, what distinguishes it from the three faults it resembles, and what change prevents it.

The troubleshooting atlas, and which entries this part owns

The troubleshooting atlas is the course’s reference product for defects: one page per defect, in a fixed shape, so that a reader who has a mark on a negative can find it by appearance rather than by knowing its name already.

Every entry carries the same six sections, and the shape is the argument:

What an atlas entry contains, and why in that order

What you seeThe appearance, described precisely enough to be matched against your own material. This is the only section a reader can use before they know anything.
Likely causesIn the order they turn out to be the answer, not in the order they occur in the process.
The chemistry and physicsThe mechanism, from a source. This is what makes the entry teach rather than merely list.
Diagnostic questionsThe observations that separate one cause from another. An entry without a discriminator here is a description, not a diagnosis.
Corrective actionWhat to do with the material in front of you — including, often, that nothing can be done and why.
PreventionStated as a change to a procedure. "Be more careful" is not a prevention.

This part owns the negative and print defect entries. Each of the eight pages below names the entries it owns, and where an entry is shared with another part the ownership is stated rather than assumed:

Safelight fog is owned here as a defect entry, while Part XVI owns the test procedure.
Weak blacks are owned here as a defect entry, while Part XVIII owns the developer and paper chemistry and Part XIX owns the printing technique that fixes them.
Emulsion coating defects stay with Part V, because they belong to a material the student made rather than bought.

The atlas has no plates, and it says so on its own front page

The course holds no photographs of a drying mark or a dichroic veil, and its own rules forbid describing an image it does not have. The owner ruled on 5 September 2026 that both atlases publish as text and that plates are a later enhancement rather than a gate.

That constraint shapes this part. You will not find a page here describing a plate you cannot see. Every appearance is described in words precise enough to be matched against your own material, and every entry carries a confirming test in place of a picture — something you can actually do that separates this defect from the ones it looks like. The break/fix page goes further and has you produce two faults deliberately, because a specimen you made under your own safelight on your own paper is a better reference than any photograph of somebody else’s.

When to stop diagnosing

An honest part about failure has to say when the investigation costs more than the failure.

A roll of 35 mm film is a few pounds. An hour of your evening is worth more than that to most people, and a re-shoot is often cheaper than a diagnosis — unless the fault will recur, which changes the arithmetic entirely. The question is not “can this negative be saved?” but “will this happen again to the next thirty films?” A fault you can name and prevent is worth an evening. A fault that happened because you tripped over the tank is worth thirty seconds.

And some negatives are simply lost. If the shadows carry no density above base plus fog, the information was never recorded and no chemistry will invent it. If the gelatin has reticulated, it has set in its new shape. If a print has sulfided from residual thiosulfate over two years, the silver has become a different compound. The course says so plainly on each of those pages, because the alternative — implying that every fault has a rescue — wastes the reader’s chemicals and their afternoon.

What this part produces

Three things, and only one of them is knowledge.

A diagnostic habit. Evidence before hypothesis, one change at a time, and a written note of what each observation eliminated. It transfers directly to every other part of the course and to a good deal outside it.

A set of atlas entries, written to the course’s own format from your own evidence, which is the reference contribution the break/fix page asks for.

Two reference specimens you made on purpose: a sheet fogged by your own safelight at a stated distance and time, and a print fixed for deliberately too short a time and then kept. Both live in the darkroom rather than in a book, because a calibrated fault on your own materials is the only reference that answers the question you will actually ask, which is does my paper look like this?

What you need first

This part is a synthesis and it leans on nearly everything before it. You do not need all of it to start, but you will not get much from the pages below without these:

The characteristic curve and contrast index, because half of this part is stated in that vocabulary.
Development kinetics and restrainers and antifoggants.
How fixer works and fixer capacity and exhaustion.
The physics of washing and permanence and image deterioration.
Gelatin, the photographic binder, for the page on the emulsion under stress.
For the print pages, maximum black, base white and dry-down and the print that will not go black.

Alternative route

Seven of the eight pages need no darkroom at all. They are read with a negative, a light box or a window, a loupe and a notebook, and the diagnosis is done on material you already have. If you have been sending film out for processing, the negatives that came back are exactly the evidence these pages teach you to read — and reading them is, if anything, more useful when you did not control the process, because it tells you what to ask the laboratory.

The one page that needs a darkroom is the break/fix clinic, and specifically the part of it that produces the two deliberate specimens: a safelight-fogged sheet and a deliberately under-fixed print. Without a darkroom, work the five cases and the sixth blind case in full — they are paper exercises and lose nothing — and substitute for the specimens as follows.

For safelight fog, the substitution is a room test rather than a safelight test. Take one sheet of photographic paper in a changing bag, cut it in two, keep one half in the bag, and expose the other on a covered surface in the room where you would print, under whatever light you have, for a stated time. Develop both later, or have them developed. You will have measured your room rather than your safelight, which is the test you can run and a genuinely useful one.
For the under-fixed print, there is no substitute that produces a specimen, and the course says so rather than inventing one. What can be done without a darkroom is the reading half: find a print, in a family album or a junk shop, that has yellowed at the edges or in the highlights, and work the fixing-and-washing page’s location-based diagnosis on it. An old, badly processed print is the specimen somebody else made for you.

Everything else in this part — the atlas entries, the fault tree, the evidence lists, the three questions — is unaffected.

Recap

A defect is a record of an event, and the event has a place on a chain that runs from the film’s manufacture to the dry print and then runs again for the print. The rebate, the edge printing, the inter-frame gap, the leader and the sprocket pitch are evidence the film brings with it about its own history, and they are free. Three questions — every frame or one, inside the image area or beyond it, following the equipment or following the picture — settle most cases before any test is run. What separates diagnosis from guessing is changing one thing, spending a sacrificial piece of material, and writing down what each observation eliminated.

Sources for this page

3 cited · checked 2026-09-07

01Monitoring and Troubleshooting KODAK Black-and-White Film Processes, publication Z-133E, bound with How to Process and Print Black-and-White Film, publication AJ-3Eastman Kodak Company, 2005§ Z-133E — Causes of an Out-of-Control Process, which lists improper solution mixing, improper solution storage and keeping, solution contamination, incorrect processing temperature, incorrect processing time, improper agitation, improper solution replenishment, evaporation and equipment malfunction; Diagnosing and Troubleshooting Processing Problems, with its instruction to verify the out-of-control condition before acting on it by checking the densitometer and processing a second control strip; and Troubleshooting from the Appearance of Processed Film, the table that pairs an appearance with a possible cause and a corrective action and ends every entry with the instruction to process another control strip to confirm that the change returned the process to control125px.com/docs/techpubs/kodak/z-133-2003_03b.pdftier 1, primary2026-09-07
02Processing your first black and white film, information leafletHARMAN technology Limited (ILFORD Photo), 2003§ Step 17, Examining the negative — the film edges (rebates) should be clear, with legible frame numbers along the bottom, and a correctly exposed and processed negative should have a full range of tones with some parts almost clear like the rebates and other parts so dense you can only just read print through them; step 10, Agitation, and the instruction to tap the tank on the bench at each inversion to dislodge air bubblesilfordphoto.com/wp/wp-content/uploads/2017/04/Processing-your-first-black-and-white-film.pdftier 1, primary2026-09-07
03How to Process and Print Black-and-White Film (KODAK Publication No. AJ-3, Technical Data / Black-and-White Film)Kodak Alaris Inc., 2016§ Small-Tank Agitation Procedures — the instruction to tap the tank on the work surface to dislodge air bubbles before the initial agitation, five to seven inversion cycles in five seconds for an invertible tank and four or five reel rotations for a non-invertible one, repeated at 30-second intervals; and the processing table's note that agitation is very important for even developmentbusiness.kodakmoments.com/sites/default/files/files/resources/AJ-3.pdftier 1, primary2026-09-07

Formulas, hazard statements, historical dates and process descriptions on this page were checked against the sources above on the date shown. Safety data changes: obtain the current safety data sheet for the product you actually buy before you open it.