Uneven Development, Bromide Drag and Surge Marks

Streaks, mottle, sprocket-hole patterns, little clear circles, crescents, a sky that is not one tone: they look like six different faults and they are one mechanism, read at different scales and with different amounts of stirring. Learn the mechanism and the six become predictable, including which way each of them moves when you change the agitation.

The single mechanism behind the whole page

Development is a chemical reaction that runs at a surface, and it changes the solution immediately touching that surface.

Where the emulsion is dense — a sky, a highlight, a bright wall — a great many grains are developing at once. Two things follow, locally, in the thin film of solution against the gelatin:

1. Developing agent is consumed, so its concentration falls.
2. Bromide is released, because reducing silver bromide to silver puts the bromide ion into solution, and bromide is the developer’s own restrainer. Its concentration rises.

Both changes slow development. Where the emulsion is thin — a shadow — almost none of it happens, and the solution there stays fresh.

So a developing film has a map of solution composition lying over it that is the negative of the picture: depleted and inhibited over the dense parts, fresh over the thin parts. Kodak’s Z-133 states the whole purpose of agitation in one sentence against that background: it is “necessary to maintain uniform solution activity by removing exhausted solution from the emulsion surface and replacing it with fresh”, and it “must be uniform throughout the processing tank”.

Everything on this page is what happens when that removal is done badly — too little, too much, in one direction only, or through a hole.

⚗Why the loaded layer moves at all

The layer of solution over a dense area is not just chemically different. It is denser, because it carries dissolved bromide and oxidised developing agent that the fresh solution does not.

A denser fluid over a lighter one is unstable. It sinks — slowly, in a laminar sheet, following gravity rather than following the picture. Where it flows, it carries its own inhibition with it and retards development in whatever lies beneath.

That is the entire mechanism of bromide drag, and it explains the two things about drag that puzzle people: why the streak appears below a dense area rather than around it, and why “below” means below relative to how the film hung in the tank, not below in the photograph.

Bromide drag

What you see. Broad, soft-edged streaks of reduced density running away from a large dense area, in the direction gravity ran while the film was in the tank. Their width is millimetres to centimetres. They have no sharp boundary and no period.

Where you see it. Beside and below skies, snow, white walls, a bright window, a light background behind a portrait — anywhere the negative carries a large, continuous, dense region with something less dense adjacent to it.

Which way is “down”. This is the observation that identifies it, and it requires you to remember how the film sat. On a spiral in an inverted tank, the film is coiled and the vertical direction of the film changes as the tank turns; on a strip hung in a deep tank, down is along the strip; on a sheet in a hanger, down is along the sheet. Drag follows the tank, not the photograph. If a streak runs downwards in the picture on one frame and sideways on the next, it is following the picture and is not drag. If it runs the same way relative to the film on every frame regardless of what each frame shows, it is.

Which end of the agitation ladder it is on. Too little. Drag is the failure mode of reduced agitation, and it is the reason stand and semi-stand development have a reputation for it. Part XXVII’s agitation experiment puts the whole ladder on one axis and measures the profiles.

Surge marks

What you see. A regular series of density differences along the long edges of a 35 mm negative, repeating at the pitch of the perforations, and encroaching inwards from the edges. The emulsion beside each hole is developed further than the emulsion between them.

The mechanism. Agitation drives solution past the film. Where the film has holes in it, solution goes through the holes as well as across the face, and the emulsion beside a hole is therefore resupplied more vigorously than the emulsion between two holes. More resupply means more development means more density, in a pattern that has the film’s own geometry.

Which end of the ladder. Too much. Surge is what over-agitation looks like, and it is the mirror image of drag: drag is a flow that happens because you did not make one, surge is a flow you made too enthusiastically.

The diagnosis is the pitch, and you measure it rather than assume it. Nothing else in the processing chain repeats at a fixed interval set by the film. Lay a rule on the negative and compare the period of the marks with the period of the perforations on the same strip. If they match, the diagnosis is made.

⁇What is sourced here, and what is the course's own reasoning

The category is documented and the specific name is not. Kodak’s Z-133 troubleshooting table lists “streaks of non-uniform density” against “excessive or uneven developer agitation”, and “mottle — areas of non-uniform density” against “inadequate developer agitation”. Both entries are about unevenness and they point in opposite directions, which is the useful part and is a manufacturer’s statement.

The course has not found a manufacturer publication that names the sprocket-hole surge mark specifically. The flow-through-the-perforations account above is presented as mechanism — it follows from the transport argument and from the geometry — rather than as a sourced identification, and the atlas entry says the same thing in the same words. If you find a Tier 1 publication that names it, that entry is where the correction belongs.

Practitioners are near-unanimous about the effect and the fix, and practitioner consensus is worth something; it is not worth a citation.

Mottle from too little agitation

What you see. A soft, irregular unevenness in areas that should be one tone — a patchiness with no direction, no edge and no period. Kodak’s table calls it exactly that: “Mottle — areas of non-uniform density”, against inadequate developer agitation.

Why it is the hardest of these to catch, and the most important thing on this page about testing. Mottle is invisible on a step wedge. A wedge is a series of small, uniform patches with sharp boundaries; a patchiness of a few millimetres inside a large even area has nowhere to show itself on one. It needs a real negative with a large area of continuous tone in it — a sky, a wall, an out-of-focus background — to become visible at all.

This has a practical consequence that catches out careful people. You can run a complete sensitometric programme, measure your contrast index to two decimal places, confirm that everything is in control, and still be producing negatives with mottled skies, because the instrument you chose cannot see the defect you have. Part XXVII’s evenness arm exists for exactly this reason and uses an evenly exposed area rather than a wedge.

The other route to it, which is not development at all. Kodak’s 1928 primer records a mottle produced in the fixing bath: when film in hangers is not agitated enough on first immersion, or is insufficiently rinsed between development and fixation, “development continues locally during the first few minutes of fixing and in these spots the image has greater density”. The primer adds that mottle also appears where the ends of a hanger protrude above the surface of the fixing bath. That is a plus-density mottle from local continued development, and its prevention is a rinse and a good initial agitation in the fixer — not more agitation in the developer.

Air bells, and why rapping the tank is a procedure rather than a superstition

What you see. Small, round, sharply bounded clear spots, usually few, usually towards one edge or grouped where the film was highest in the tank.

The mechanism. When the developer is poured in, air is entrained, and bubbles rise and lodge against the emulsion. Where a bubble sits, there is no developer; where there is no developer, there is no development; where there is no development, the fixer removes the halide and leaves clear film. The spot is round because the bubble was.

The fix is one physical action and both major manufacturers publish it. ILFORD’s first-film guide: “each time you invert the tank tap it on the bench to dislodge any air bubbles which may have formed on the film”. Kodak’s AJ-3 small-tank procedure opens with “tap tank on work surface to dislodge air bubbles” before the initial agitation, for invertible and non-invertible tanks alike.

Two manufacturers, independently, put the same physical rap in their published procedure. It is a step, not folklore.

Contact marks, crescents, and how to tell them from a scratch

Contact marks happen when the film touches itself, or touches the reel, so that solution cannot reach that patch. The result is a mark of reduced density — often a smooth curve or an arc where two turns of a badly loaded spiral pressed together — and it is a development fault by the same argument as an air bell: no solution, no development.

Crescents are different and come from force rather than from contact. Kodak’s Z-133 lists them under Pressure marks — plus-density areas (e.g. half-moons, crescents, etc.), against poor film handling, excessive tension and camera malfunction. Note the words plus-density: pressure on an emulsion before development makes the pressed area more developable, not less, so a crescent from forcing a film onto a reel prints as a dark mark on the negative and a light one on the print.

How to tell the three apart. They are three different things with three different signatures, and the discrimination is worth having because the fixes are unrelated:

	Density	Shape	Where	Made when
Contact mark	Less	Smooth arc or blotch, soft edge	Wherever two surfaces touched	In the tank
Crescent (pressure)	More	Sharp half-moon, a few millimetres	Where the film was bent or forced	Before development — loading, or in the camera
Scratch	Either, and it is a groove	A line, usually long and straight	Along the film path	Anywhere, and the side it is on says where

The crescent is a genuine hybrid — it is a physical event that produces a density change rather than a mark you can feel — which is why the manifest for this part places it in two pages at once. Its mechanical half, and the question of which side of the base it is on, belong to the emulsion under stress; its density half belongs here.

The tray, the sheet and the print

Everything above was a tank. A dish is a different hydrodynamic problem and fails differently.

Shuffling sheets. Sheet film developed in a stack, shuffled from bottom to top, gets its agitation in the shuffle. The sheet at the bottom is pressed against the one above it, solution is squeezed out from between them, and the result is uneven development wherever contact was maintained. It is a contact mark with a large area.

Too little solution. ILFORD’s paper-developer sheet asks that the volume made up be “enough to fill a dish/tray to a depth of about half full”, and their first-film guide’s worked example is 300 mL of working solution for one 35 mm film — 60 mL of concentrate and 240 mL of water. A tank with the film half out of the liquid, or a dish with a centimetre of developer in it, cannot deliver uniform resupply and will not.

A cold spot. A dish standing on a cold stone worktop, or half over the edge of a bench in a draught, develops a temperature gradient. Nine per cent of the development time per degree Celsius is the working figure this course takes from ILFORD’s compensation chart, so a two-degree gradient across a tray is close to a twenty per cent difference in effective development from one side to the other.

The large sheet whose developer runs out under the black areas. This is the print equivalent of bromide drag and it is real, but it is worth quantifying rather than asserting, because the numbers say when it matters. ILFORD publish the capacity of one litre of MULTIGRADE developer at 1+9 as 100 sheets of 8 × 10 inch RC paper or 50 of fibre. One 16 × 20 inch print is four 8 × 10s, so a litre holds a dozen of them at most on fibre — and the capacity figure is for a whole sheet’s average, while the developer under a large area of deep black is doing a great deal more than average work.

The fault this produces is not usually a failure to develop; it is a slow drift from the middle of a big print outwards, or a patchiness in an even dark tone. The prevention is volume and agitation: a dish deep enough, and continuous, changing-direction rocking rather than a rhythm.

✕A rhythm is not agitation

The single commonest cause of directional unevenness, in a tank and in a dish alike, is agitation that is perfectly regular and always the same.

Rock a dish front-to-back at a steady rate and the solution sets up a standing pattern; the same regions get swept every time and the same regions never do. Invert a tank the same way, at the same speed, with the same hand, and the flow inside repeats.

Kodak’s own procedure builds the break into the motion — “extend your arm and twist your wrist 180°” — which turns the tank about a different axis each time. ILFORD’s dish instruction is to rock the tray, and the useful addition, which both sheets imply rather than state, is to change which edge you lift. Randomising the flow is the point. Regularity is what you want in the timing, not in the direction.

Rotary processing, and why more even in one direction can be less even in another

A rotary processor rolls a drum containing a small volume of solution. It has real advantages: the film is continuously and vigorously resupplied, so mottle and drag are largely designed out, and the small volume makes one-shot working affordable.

It has its own artefacts, and they come from the same place. Rotation is a flow with a direction, and a constant one: solution runs the length of the drum and back, so what surge is to a spiral, the ends of the drum are to a roller. Continuous vigorous agitation also changes the development rate, which is why ILFORD’s sheet asks for spiral-tank times to be reduced by up to 15 per cent in a rotary processor without a pre-rinse, and by the same amount for continuous agitation in a dish.

That sheet also carries a warning worth pulling out, because the practice it warns against is widely recommended: “A pre-rinse is not recommended as it can lead to uneven processing.” The reasoning is not given, and the course will not invent one; but the instruction is a manufacturer’s, on the technical sheet for the film, and a worker whose rotary negatives are uneven should try removing the pre-rinse before trying anything else.

One more manufacturer statement belongs here because it is a limit rather than a preference: ILFORD’s temperature compensation chart carries, in capitals, “development times below 5 minutes are not recommended due to the risk of uneven development”. Short times are the general enemy of evenness, whatever the equipment: the pour, the first agitation and the drain all become a large fraction of the total.

The diagnostic questions, in order

From a mark on the film to a named flow fault

1. 11. Is the mark periodic, and does the period match a piece of equipment?Measure it against the perforations on the same strip. A match is a surge mark and the diagnosis stops here. No period, go on.
2. 22. Does it have a direction, and does that direction match how the film sat in the tank?Consistent relative to the film across frames that show different things: bromide drag. Consistent relative to the picture: it is following the subject, so look at flare, the lens or the light instead.
3. 33. Is it worse next to dense areas or next to thin ones?Reduced density beside and below dense areas is drag. Increased density in a general patchiness with no relation to the picture is inadequate agitation, or the fixing-bath mottle of the 1928 primer.
4. 44. Is it round, small and clear?Air bells. Count them and note where they sat: they collect where the film was highest in the tank.
5. 55. Is it a sharp half-moon of extra density?A pressure mark, made before development by force on the film — loading or the camera — and not a development fault at all.
6. 66. Every film, or one session?Every film points at your standard scheme, your tank or your volume. One session points at what was different that day: a rushed pour, a cold bench, a tank filled short.

The fixes, and what each one costs

Every fix on this page changes something else at the same time, and a page that offered them as free would be lying.

Standardise the agitation and write it down as inversions and intervals. This is the first fix and the cheapest, and the reason to write it down is that the manufacturers do not agree, so “normal agitation” is not a specification. ILFORD ask for four inversions in the first ten seconds and four again in the first ten seconds of each further minute. Kodak’s AJ-3 asks for five to seven inversion cycles in five seconds, repeated at thirty-second intervals. Kodak’s is roughly twice as much agitation as ILFORD’s. Both are correct for the times their own tables give — which means the time and the agitation are one specification and not two. Adopting Kodak’s agitation with ILFORD’s time over-develops; adopting ILFORD’s agitation with Kodak’s time under-develops. Cost: if you change the scheme, your development time changes and your established contrast index is no longer yours.

Increase the solution volume. More liquid means more reservoir and a smaller change in composition over the run. Cost: more chemistry per film, and on a reused bath a different capacity arithmetic.

Change the dilution. A more dilute bath exaggerates local exhaustion — which is the whole point of compensating development, and the whole problem when you did not want it. Going more concentrated reduces drag and mottle. Cost: less compensation, more contrast in the highlights, and a different time.

Change the tank or the reel. A tank whose geometry sets up a resonance with your inversion rhythm will keep doing it. A reel that lets film touch itself will keep doing that too. Cost: money, and a new set of habits to establish.

Rap the tank. Free, and it is in both manufacturers’ published procedures.

Fill the dish properly and rock it irregularly. Free.

The boundary with the version of this you want

The same mechanism, at a scale a thousand times smaller, is one of the most prized effects in photography.

Where a dense region abuts a thin one, the bromide released on the dense side diffuses sideways, a few tens of micrometres, into the thin side — and fresh agent diffuses the other way. Development runs slightly faster just inside the dense edge and slightly slower just inside the thin edge, producing a density overshoot and undershoot at the boundary. That is the adjacency effect, and it makes edges look crisper than the lens delivered. Photographers pay for it.

Drag is the same chemistry at the same time, and the only difference is how far the loaded solution travelled before it met solution that was being replaced. Tens of micrometres and it is an edge effect. Centimetres and it is a defect. Part VIII owns the wanted version and Part XXVII measured both — including the instrumental point that a 2 mm densitometer aperture averages the edge effect away entirely and resolves the drag streak directly, so the instrument that can see one cannot see the other.

That relationship is the reason this page and the acutance page must be read together rather than separately. You cannot reduce agitation to gain edge effects without moving towards drag, because they are one dial. What varies is where on the dial your subject, your dilution and your format put you.

Atlas entries this page owns

Seven entries in the troubleshooting atlas belong here: bromide drag, surge marks, air bells, mottled or uneven development, reel and contact marks, uneven print development and insufficient solution volume.

Uneven print development is here rather than on the print page because its mechanism is identical to the film case: it is the same local exhaustion in a dish instead of a tank. The print faults that have no negative equivalent — weak blacks, bronzing, dry-down — are the print page’s. Cinch and pressure marks are owned by the emulsion under stress and cross-linked from here, because their cause is force rather than flow.

Recap

One mechanism: developer is consumed and bromide released fastest where density is highest, so a map of depleted, inhibited solution lies over the film in the shape of the negative of the picture. That loaded layer is denser than the bulk and sinks, and what the agitation does or fails to do about it is written on the film. Too little and you get drag — soft streaks of reduced density running away from dense areas in the direction gravity ran in the tank — and mottle, which a step wedge cannot show you. Too much and you get surge marks at the pitch of the perforations. Air bells are bubbles and are prevented by the rap both manufacturers publish. Crescents are plus-density pressure marks made before development and are not a flow fault at all. Every fix changes the development time as well, because agitation and time are one specification; the manufacturers’ own schemes differ by a factor of about two, which is exactly why yours has to be written down.

Check your understanding

Question 1. A 35 mm negative has a regular pattern of density along both long edges, encroaching inwards, at the same spacing as the perforations. Name the defect, the cause and the single change that prevents it.

• Bromide drag, from too little agitation; agitate more
• Surge marks, from solution driven through the perforations by excessive or uneven agitation; agitate less vigorously, and change the axis of the inversion so the flow does not repeat
• Air bells, from bubbles at the film edges; rap the tank
• Reticulation, from a temperature shock; match the bath temperatures

Show the answer and why

Answer: Surge marks, from solution driven through the perforations by excessive or uneven agitation; agitate less vigorously, and change the axis of the inversion so the flow does not repeat

The pitch is the diagnosis: nothing else in the process repeats at an interval set by the film's own geometry, which is why measuring it against the perforations on the same strip settles the case. Note the direction of the fix. Drag and surge are the two ends of one lever, so an agitation change that cures one moves you towards the other; the aim is the middle of the range with an irregular flow direction, not simply "less".

Question 2. A photographer halves the number of inversions per minute. Which two defects become more likely, and which becomes less likely?

• Drag and mottle become more likely; surge marks become less likely
• Surge marks and air bells become more likely; drag becomes less likely
• Reticulation and frilling become more likely; mottle becomes less likely
• Nothing changes, because agitation only affects speed

Show the answer and why

Answer: Drag and mottle become more likely; surge marks become less likely

Kodak's troubleshooting table places streaks of non-uniform density against excessive or uneven agitation and mottle against inadequate agitation, so the two ends of the ladder fail in opposite ways. Reduced agitation moves you towards the inadequate end: drag, because the loaded layer is left to sink undisturbed, and mottle, because resupply becomes patchy. There is a third consequence that the question does not ask about and that matters in practice: the development rate falls too, so the contrast index drops unless the time is extended.

Question 3. Why can a full sensitometric programme on step wedges show a process perfectly in control while the negatives still have mottled skies?

• Because step wedges are exposed differently from negatives
• Because a wedge is a series of small uniform patches with sharp boundaries, and a patchiness of a few millimetres inside a large even area has nowhere to appear on one — the defect needs a large area of continuous tone to become visible at all
• Because a densitometer cannot read low densities accurately
• Because mottle only affects the highlights, which a wedge does not include

Show the answer and why

Answer: Because a wedge is a series of small uniform patches with sharp boundaries, and a patchiness of a few millimetres inside a large even area has nowhere to appear on one — the defect needs a large area of continuous tone to become visible at all

This is a measurement-design point rather than a chemistry one, and it generalises: an instrument only reports the defects its sampling geometry can see. The wedge answers questions about the transfer function — speed, contrast, fog — extremely well and is blind to spatial unevenness inside a tone. Part XXVII's evenness arm therefore uses an evenly exposed area rather than a wedge, and reads a profile across it.

Question 4. ILFORD ask for four inversions in the first ten seconds of each minute; Kodak's AJ-3 asks for five to seven inversion cycles in five seconds every thirty seconds. Roughly twice the agitation. Which is right?

• Kodak, because more agitation is always more even
• ILFORD, because gentler agitation avoids surge marks
• Both, for their own development tables — agitation and time are one specification and not two, so mixing one maker's agitation with the other maker's time changes the effective development
• Neither; the correct scheme is continuous agitation

Show the answer and why

Answer: Both, for their own development tables — agitation and time are one specification and not two, so mixing one maker's agitation with the other maker's time changes the effective development

A published development time is the time that gives the intended contrast under the agitation published beside it. More agitation means faster resupply, faster development and higher contrast at the same time on the clock — which is exactly why ILFORD tell you to cut times by up to 15 per cent for continuous agitation or a rotary processor. The practical consequence is that "normal agitation" is not a specification, and your own scheme belongs in your processing table alongside the time and the temperature.

Check my answersClear