Push and Pull Processing: What Extended Development Can and Cannot Buy

Rate a 400-speed film at 1600, develop it half as long again, and you have done two things that are usually described as one. You changed the number on the meter, which changed the exposure the film received. You changed the development, which changed the slope of the curve that exposure was recorded on. Only the second of those is chemistry, and the first is the one that decides what is in the shadows.

This page is about what the second can and cannot repair. Its short answer is that extended development amplifies everything the exposure recorded and invents nothing it did not, so a push buys contrast and mid-tone separation, costs fog and grain, and buys a fraction of a stop of usable speed rather than the two stops written on the meter. The long answer is a decision you read off two curves you measured yourself in the development time series.

Why the speed point barely moves, and what that means for the shadows

Speed is defined by a point low on the curve, a stated density above base plus fog, and Part XIII’s criterion page sets out the construction the course uses. That point sits on the toe, and the toe is made of crystals that caught barely enough photons to form a latent-image speck at all.

Development is amplification, not creation. It reduces to metallic silver those crystals that already carry a speck, and it does so faster where the speck is larger. Part XIII’s density-data lesson puts the consequence in one sentence the whole of this page depends on: exposure places the shadows and development sets the slope. A crystal that received no photons has no speck, and no amount of time or activity will produce one. So a region of the negative that was two stops underexposed has recorded less information, and developing it longer amplifies the little it recorded along with the fog beside it.

That is why the toe moves so much less than the rest of the curve when development is extended, and it is why the speed criterion — which is anchored to base plus fog and therefore rises as fog rises — can move in the wrong direction if the fog floor climbs faster than the toe does.

One film, two development times: what moves and what does not

• Developed to a normal contrast index
• The same film developed much longer

Show the numbers behind this plot

Two characteristic curves for the same film, plotted as density against relative log exposure from 0 to 2.4. The lower curve, for normal development, starts at a base-plus-fog density of 0.10, creeps to 0.12 at log exposure 0.3 and 0.17 at 0.6, then climbs through 0.28, 0.45, 0.64, 0.83 and 1.00 to reach 1.14 at the far right. The upper curve, for the same film developed much longer, starts at a higher base plus fog of 0.16, reaches 0.19 at 0.3 and 0.27 at 0.6, then climbs far more steeply through 0.45, 0.73, 1.05, 1.36 and 1.62 to 1.80. Two horizontal guide lines mark the two fog floors at 0.10 and 0.16. The drawing's whole point is in the left-hand third: at the lowest exposures the two curves are separated by only about 0.06 to 0.10 of density, which is close to the fog difference alone, while at the right-hand end they are separated by 0.66. Extended development has multiplied everything the exposure recorded and lifted the floor beneath it, and it has done almost nothing at the exposures where a pushed negative's shadows actually land.

Series	Relative log exposure	Density
Developed to a normal contrast index	0.00	0.10
Developed to a normal contrast index	0.30	0.12
Developed to a normal contrast index	0.60	0.17
Developed to a normal contrast index	0.90	0.28
Developed to a normal contrast index	1.20	0.45
Developed to a normal contrast index	1.50	0.64
Developed to a normal contrast index	1.80	0.83
Developed to a normal contrast index	2.10	1.00
Developed to a normal contrast index	2.40	1.14
The same film developed much longer	0.00	0.16
The same film developed much longer	0.30	0.19
The same film developed much longer	0.60	0.27
The same film developed much longer	0.90	0.45
The same film developed much longer	1.20	0.73
The same film developed much longer	1.50	1.05
The same film developed much longer	1.80	1.36
The same film developed much longer	2.10	1.62
The same film developed much longer	2.40	1.80

Drawn to show the shape, not measured from any film: the two curves are plausible and illustrative, and your own pair from the development time series will differ. What transfers is the geometry — the separation between two development times grows with exposure, so the region that gains least is exactly the region a push has moved the shadows into. 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 the left-hand third of that drawing again. Between the two curves the gap at the lowest exposures is about the size of the fog difference, and the gap at the top is six or ten times larger. That is the whole argument in one picture: development pays out in proportion to what the exposure delivered, and a shadow that was underexposed by two stops has moved two stops to the left, into the region where the payout is smallest.

What does move, and by how much

Six things change when development is extended, and your own strips measure every one of them.

• Contrast index rises, along a curve that flattens. On the six values Kodak publish in their sensitometry workbook, three minutes of extra development from five to eight buys 0.11 of contrast index and the last minute from twelve to thirteen buys 0.01.
• Mid-tone separation improves. This is the same fact stated in printing terms, and it is the real reason a pushed negative can look good: the tones that were recorded are spread further apart.
• Base plus fog rises, and unlike the others it rises without buying anything. Kodak’s own process-control tolerance for minimum density is asymmetric for this reason — an upper control limit and no lower one.
• Maximum density rises and then stops, when the crystals in the heaviest steps have all been developed and there is nothing left to reduce.
• Grain becomes more visible. Two Tier 1 documents say so directly. Kodak’s Tri-X sheet states that push processing brings a loss of shadow detail and an increase in graininess. ISO 6 makes the more general statement in its own clause on processing: speeds obtained using various procedures can differ significantly, other sensitometric and physical changes accompany the speed changes, and generally, processes which yield higher ISO speed will also increase the graininess of the negative and the final print. Part VIII’s grain lesson owns the mechanism.
• The usable exposure index rises by a fraction of the nominal push. This is the one that needs its own section, because it is where almost all of the confusion lives.

Three words that are not interchangeable

The course’s terminology table keeps three things apart, and a push involves all three:

Term	What it is
ISO speed	A measurement made to a standard, under stated conditions, for a film and process together
Exposure index	The number you set on the meter, which is a decision
A push	A change of exposure index plus a change of development, described as though it were a change of speed

!ISO 6 says a film has several speeds, and the makers say the box number is not one of them

The standard is stranger and more useful than its reputation. ISO 6 is a standard for film/process systems, and its own introduction says so: the speed of a film depends on the process used, a particular film may have several ISO speeds depending on the processes used, and — in the standard’s own words — this conflicts with the tradition of associating a specific speed value with a particular product. Its second edition, in 1993, removed the first edition’s requirement that speed be measured in a specified developer. Its only processing requirement is that the film be developed to a stated degree of contrast, so that speeds are quoted at a comparable degree of development. The course cites ISO 6 by number only and prints no threshold, formula or table from it anywhere.

The makers are equally explicit about what their own EI numbers are. ILFORD print, on both the HP5 Plus and the DELTA 3200 sheets, that the recommended exposure-index range “is based on a practical evaluation of film speed and is not based on foot speed, as is the ISO standard”. And ISO 6’s introduction explains why a box speed is usually the conservative end: manufacturers take advantage of a film’s overexposure latitude and give it a conservative speed value to protect users from underexposure.

So “ISO 400” is a measurement of one film-and-process pair, deliberately placed on the safe side. “EI 1600” is a decision you made. The two are not the same kind of statement, and a page that writes “I pushed the film to 1600 ISO” has said something that ISO 6 has no way of meaning.

The clearest published demonstration is a film whose name is a meter setting. ILFORD DELTA 3200 Professional has an ISO speed rating of 1000/31°, measured in ID-11 at 20 °C with intermittent agitation in a spiral tank, and ILFORD print that figure on the same sheet on which they say the film is designed to be exposed at EI 3200 and given extended development. The gap between the measured speed and the recommended setting is one and two-thirds stops, both numbers come from the manufacturer, and neither is wrong. One is a measurement under a standard; the other is a working recommendation.

Kodak’s push tables are contrast-index ladders, and they say so

The single most useful document for this page is Kodak’s XTOL data sheet, because its development tables carry two columns where most sheets carry one: an EI and, next to it, the contrast index that time is intended to reach. Read one film’s rows and the whole subject becomes arithmetic.

KODAK PROFESSIONAL T-MAX 100 in full-strength XTOL at 20 °C:

Meter setting	Contrast index aimed at	Development time	Relative to the box-speed row
EI 25/50	0.52	6¾ min	−10 % of the time, −0.04 of contrast index
EI 100/200	0.56	7½ min	the published normal
EI 200	0.62	8 min	+7 %
EI 400	0.72	9½ min	+27 %
EI 800	0.82	10½ min	+40 %

Five stops of meter setting, from EI 25 to EI 800, are covered by development times running from 6¾ to 10½ minutes — a factor of 1.56 — and by contrast indices from 0.52 to 0.82. Nothing in that table claims the film became more sensitive. What it says is that if you set the meter three stops above the box figure, Kodak’s recommendation is to develop to a contrast index of 0.82 rather than 0.56.

KODAK PROFESSIONAL T-MAX P3200 in the same developer takes the pattern to its limit, and the regularity is startling:

Kodak's own push ladder, plotted: from EI 1600 upward, each doubling of the meter setting buys exactly 0.10 of contrast index

• KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI

Show the numbers behind this plot

Seven published points from Kodak's XTOL data sheet for T-MAX P3200 film, plotted with the meter setting on a logarithmic horizontal axis and the intended contrast index on the vertical. The points run from EI 400 at a contrast index of 0.52, through EI 800 at 0.56, EI 1600 at 0.62, EI 3200 at 0.72, EI 6400 at 0.82 and EI 12500 at 0.92, to EI 25000 at 1.02. Because the horizontal axis is logarithmic, the equal spacing of the doublings makes the relationship visible as a straight line from EI 1600 upward: each doubling of the meter setting there is met with a rise of exactly 0.10 in the contrast index Kodak's published time is intended to reach. The two lowest steps are smaller, 0.04 from EI 400 to 800 and 0.06 from 800 to 1600, and the first of those is the step Kodak's own Tri-X sheet says needs no change of development at all. The corresponding development times, which are not plotted, run from 9.5 minutes at EI 400 to 19 minutes at EI 25000 - six doublings of meter setting bought with exactly a doubling of development time.

Series	Meter setting, as the logarithm of EI — 2.60 is EI 400 and 4.40 is EI 25000	Contrast index Kodak's time is intended to reach
KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI	2.60	0.52
KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI	2.90	0.56
KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI	3.20	0.62
KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI	3.50	0.72
KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI	3.81	0.82
KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI	4.10	0.92
KODAK T-MAX P3200 in full-strength XTOL: published EI, as log EI, against published CI	4.40	1.02

These are Kodak's published values, read off Table 1 of the XTOL data sheet J-109, revised February 2018. They are recommendations rather than measurements of speed: the contrast index column says what the development is aimed at, and the meter-setting column says what exposure the recommendation assumes. Six doublings of meter setting - a factor of 64 in exposure - are met with a doubling of development time and a rise of 0.50 in contrast index.

Say that last sentence slowly. A factor of sixty-four in exposure is answered with a factor of two in development time. If pushing worked the way the word implies, the two would have to move together. They do not, because the meter setting and the development are answering different questions: the meter setting decides where the subject lands on the exposure axis, and the development decides the slope it is recorded at.

Two more things in those tables are worth reading as evidence rather than as instructions.

The first stop is free, and Kodak say so. The TRI-X 400 row in XTOL is headed EI 400/800 — one row, one time, two meter settings — and the Tri-X film sheet gives the reason outright: because of these films’ exposure latitude you can underexpose by one stop and use normal processing times, and prints will show a slight loss in shadow detail. So the thing photographers call “a one-stop push” is, on the manufacturer’s own account, not a development change at all. It is latitude, and it is spent whether or not you know you are spending it.

The tables run out at the cold end. The 18 °C cell for TRI-X 400 at EI 3200 is blank, and the 18 °C cell for HP5 Plus at EI 3200 in XTOL is marked NR, not recommended, as determined by testing. A push is a demand for a large degree of development, and at a low temperature the time needed to reach it stops being practical before it stops being possible.

ILFORD publish the same ladder without the contrast column, and the difference matters

ILFORD’s HP5 Plus sheet carries a push table too, and it is instructive precisely because it lacks Kodak’s second column. At 20 °C in a spiral tank:

Developer	EI 400	EI 800	EI 1600	EI 3200
ILFOTEC DD-X 1+4	9 min	10	13	20
MICROPHEN stock	6½ min	8	11	16
ID-11 stock	7½ min	10½	14	—
ILFOTEC HC 1+15	3½ min	5	7½	11

Three things are visible in it even without a contrast index. The first stop is nearly free again: DD-X goes from 9 to 10 minutes for the first doubling, 11 per cent, and then from 13 to 20 for the last, 54 per cent. ID-11 stops at EI 1600 and prints no figure at 3200, which is ILFORD declining to recommend a general-purpose fine-grain developer for the biggest push — the same sheet sends you to DD-X or MICROPHEN for maximum film speed. And the four developers need wildly different times for the same job, from ILFOTEC HC’s 11 minutes to DD-X’s 20 at EI 3200, which is a fact about developer activity rather than about the film.

What the table cannot tell you is what contrast index any of those times reaches, and therefore whether the four rows are four routes to the same negative or four different negatives. Kodak’s table answers that question and ILFORD’s does not. When you set a published time beside your own curve, that is the missing quantity you are supplying.

What those tables assume, and why a published push time is a starting point

Every figure quoted above carries a set of conditions in its own document, and a reader who takes the number without them is comparing two different processes.

• A tank, and a named agitation script. ILFORD’s spiral-tank times assume four inversions during the first ten seconds and four more during the first ten seconds of each further minute, and they instruct that continuous agitation calls for reducing those times by up to 15 per cent. Kodak’s small-tank times assume 5 to 7 inversion cycles in 5 seconds at 30-second intervals. Fifteen per cent of a push time is two or three minutes.
• A temperature, held. Every row is a column heading. Kodak’s own process-monitoring publication states that a developer temperature varying by more than 0.3 °C affects process control and image quality, and the temperature coefficients the time series assembled run from about 5 to 11 per cent of the development time per degree depending on which published table you compute them from.
• A fresh, unreused bath, and for a push both makers say so explicitly.
• A film batch, and a meter the maker has not seen. The tables assume the exposure landed where the meter setting says it did. A shutter running slow or a meter reading a stop out moves the negative and not the curve, and no development time corrects either.
• A contrast index as the actual target. This is the one that makes the rest legible: Kodak print their aim in the table, and the times exist to reach it. Which means the honest way to use a published push time is not to copy it but to read its contrast index and find your own time for that contrast index on your own curve.

Both makers say the same thing in their own words. ILFORD’s development times are “intended as a guide” which “may need adjusting to suit individual processing systems and working practices”; Kodak’s are “starting-point recommendations”, and for push processing specifically the Tri-X sheet’s instruction is to expose a test roll to determine the film speed that gives the best results. A published push time is the manufacturer’s answer to a question you have not yet asked in your own darkroom, and this part of the course exists to let you ask it.

Reading the decision off your own data

Everything above is somebody else’s film in somebody else’s tank. Your own two plots turn it into a decision. Both come out of the development time series, and neither can be borrowed.

One: the contrast-index-against-time curve. Choose the contrast index you want and read the time under it. Kodak’s own aims are the sensible anchors — 0.58 for printing with a diffusion enlarger and 0.43 for a condenser — and their published push ladder shows what a two-stop push is worth in the same currency: about +0.16 to +0.26 of contrast index above the normal aim, depending on the film. Read your own time for that target, with the tolerance the curve’s local slope gives you.

Two: the speed point against contrast index. The time series asks you to plot this precisely because contrast index is what you choose and the speed point is what it costs. Convert the speed point at your chosen contrast index into an exposure index under the course’s criterion, and you have the meter setting your own data supports — as opposed to the one on the meter.

🧮The stopping rule, computed rather than felt

There is a point past which extra development is a bad bargain, and it is calculable from two curves you already have.

Plot contrast index against time and base plus fog against time on the same horizontal axis. At each rung, take the local slopes: dCI/dt and d(B+F)/dt. The ratio is what a minute of development is worth:

R = (dCI/dt) ÷ (d(B+F)/dt)

What a minute of extra development buys, per unit of fog it costs

R falls as development is extended, for two reasons at once: the contrast curve flattens while the fog curve does not. Choose a value of R in advance and stop there. On Kodak’s six published contrast indices, dCI/dt is 0.04 per minute between five and six minutes and 0.01 per minute between twelve and thirteen — a factor of four — so the last rung of a long push buys a quarter of what the first one did, at a fog cost that has not fallen at all.

State R with its units and its date, and put it on the processing table beside the times. It is the only part of a push decision that is genuinely yours rather than the manufacturer’s, because it depends on how much fog your printing can absorb.

Three cautions on interpreting your own answer. The speed point’s uncertainty is larger than the contrast index’s, because a density uncertainty becomes an exposure uncertainty by division by the local gradient and the toe’s gradient is small — the time series works that arithmetic. A contrast index measured on a strip whose base plus fog has moved has had its construction’s zero moved with it, so read fog first and quote it beside the result. And your number is an exposure index under the course’s criterion, in the course’s form; it is never “the ISO speed”, which is a claim this course does not make anywhere.

Pull processing: the half nobody photographs

A pull is the same two moves in the other direction: a meter setting below the box figure and a shorter development. It is used when the subject’s luminance range is long — a lit interior with a window in it, a landscape with snow and shadow — and a normally developed negative would deliver a density range no paper grade can hold.

The arithmetic is the same chain in reverse. If the subject spans nine stops, that is 9 × 0.301 = 2.71 in log luminance, and at a contrast index of 0.56 the negative’s density range would be about 1.52. ILFORD’s instruction for MULTIGRADE RC DELUXE is to multiply the effective negative density range by 100 and take the nearest published range figure; their softest, filter 00, is 160, so 152 is inside the paper’s reach but only just, with nothing in hand. Develop the same subject to a contrast index of 0.43 and the negative range is about 1.17, which lands near the range figure of 110 and gives you two grades of room to print into.

What a pull costs is shadow density. Shorter development lowers the whole curve, the toe included, so the darkest tones sit closer to base plus fog and closer to the point where they cannot be separated from it. That is why a pull pairs with generous exposure: you give the film more light to put the shadows higher up the curve, and then develop less to keep the highlights from running away. The two moves work together, which is exactly what makes a pull safer than a push — a push asks less of the exposure and more of the development, and a pull asks the reverse.

Two published pull tables in this corpus are worth reading, and they are less dramatic than most readers expect.

Kodak’s XTOL table publishes settings below box speed. T-MAX 100 at EI 25/50 is aimed at a contrast index of 0.52 in 6¾ minutes, against 0.56 in 7½ at EI 100/200. Two stops of pull cost forty-five seconds of development. Almost all of the pull is in the exposure; the development change is a trim.

ILFORD’s rescue table for HP5 Plus is not a pull table at all, and the way it is not is instructive. For film “inadvertently exposed at settings below EI 250/25”, they give PERCEPTOL stock at 9 minutes at both EI 50 and EI 100, and 11 minutes at EI 200 — a longer time for the setting that is closer to box speed — with the note that the quality of negatives processed this way will not be as high as conventionally processed ones. The answer to three stops of overexposure is not a drastic cut in time. It is a change of developer, to a fine-grain solvent formulation ILFORD describe as being for use when very fine grain is required and a decrease in film speed is not important. Overexposure has already happened; what remains is to keep the highlights printable and the grain small, and a different developer does that better than a stopwatch.

And a pull is also a contrast decision with a printing name. Kodak’s T-MAX developer sheet gives the instruction in exactly those terms: to adjust contrast for printing with a condenser enlarger rather than a diffusion one, reduce the development time by 20 to 30 per cent. Nothing about the meter changes. That is a pull performed for the enlarger, and it is the commonest pull anybody actually does.

The whole chain: a negative meets a paper, not a verdict

Two routes to the same print, and what each one spends

1. 1The subject — a short luminance range at a light level that dictates the shutter speed
2. 2Route one — the push — meter high, develop long; buys the shutter speed, spends shadows, fog and grain
3. 3Route two — expose and print harder — meter at box speed, develop normally, print on a hard grade; spends shutter speed or depth of field and nothing chemical
4. 4The print — both routes can reach a full-scale print; the difference is what was spent and what was never recorded

The two routes are not equivalent and the drawing does not pretend they are: route two records shadow detail that route one never captured, and route one exists because route two is sometimes not available at any shutter speed the picture can survive.

The arithmetic behind that drawing is Part XIII’s, and this page only applies it:

Negative density range ≈ CI × subject log-luminance range

Negative density range

A dim interior is usually a short subject — five stops, or 1.51 in log luminance. At a contrast index of 0.56 that gives a negative density range of 0.85, which by ILFORD’s instruction becomes a range figure of 85 and lands near filter 2 on MULTIGRADE RC DELUXE. Push the same subject to a contrast index of 0.82 and the range becomes 1.24, near the range figure of 130, which is filter 0 — the second softest the paper has.

That result surprises most readers, and it is the point of the chain. A pushed negative does not need a harder grade; it usually needs a softer one, because the contrast has already been built into the film. Which means the two routes can arrive at prints of similar contrast, and the difference between them is not grade but what was recorded and what it cost: route two has shadow detail that route one never captured, and route one has fog and grain that route two never made.

✕Joining a contrast index to a paper grade is this course's inference, not a manufacturer's instruction

ILFORD publish range figures for their papers and Kodak publish contrast-index aims for their films. Neither publishes the join. Putting the two together is this course’s own arithmetic on two manufacturers’ documents that were not written to be read together, and Part XIII says so where the join is first made.

There is also a measurement problem in it worth knowing about. ILFORD’s range figure is for the effective density range of the negative as projected on the enlarger baseboard, which already includes your enlarger’s flare and its Callier effect. Your densitometer measures something else. Treat the grade the chain predicts as a starting filter for a test strip, not as an answer.

Developer choice, and the honest size of the gain

The market for push processing is a market in developers, and the claims are worth reading closely, because almost none of them is what a reader assumes.

Developer	What the maker claims about speed	Is it a claimed gain?
ILFORD MICROPHEN	“gives an effective increase in film speed. A speed increase of up to half a stop can be achieved with most films but with faster films such as HP5 Plus, Delta 400 Professional and Delta 3200 Professional it is more”	Yes, and it is the only quantified one in this corpus
ILFORD ILFOTEC DD-X	“gives full film speed”; “highly recommended when fast films need to be push processed”	No — a claim of not losing speed
Kodak XTOL	“full emulsion speed”; “excellent emulsion speed with normal and push processing”; at 1:1, “slightly greater film speed… and slightly more grain”	No — the same claim, plus an unquantified comparative
Kodak T-MAX Developer	“enhanced shadow detail in normally processed and push-processed films”; higher image quality “than current push-processing developers”	No — a claim about shadow detail and comparative quality
ILFORD PERCEPTOL	for use “when very fine grain negatives are required and a decrease in film speed is not important”	A published loss, stated as such

One quantified claim, and it is half a stop. Set that beside the arithmetic of a push: two stops is a factor of four in exposure, and half a stop is a factor of 1.41. Even taking ILFORD’s figure at face value and allowing that they say it is more on the fastest films, a speed-increasing developer supplies something on the order of a quarter of what a two-stop push asks of it. The rest of the two stops is underexposure, recorded on a steeper curve.

No manufacturer in this corpus publishes a measured speed gain at a fixed density criterion for a pushed process. They publish times, they publish the contrast index those times reach, and they publish adjectives. That is not a criticism — a maker’s job is to give a working recommendation — but it means the number a photographer most wants is one the literature does not contain, and the only way to have it is to measure it under a stated criterion on your own strips.

As for how a speed-increasing developer does what it does, the honest answer is that the course can name the family and not the formula. MICROPHEN is a bought product whose composition ILFORD do not publish, and ILFORD attribute its grain behaviour to its low alkalinity. XTOL is described by Kodak in two phrases — “ascorbic acid-based” and “no hydroquinone” — and nothing else in sixteen pages names a substance in the packets, which is why its formulary entry is a product page rather than a formula. Part VIII’s superadditivity lesson is where the mechanism a phenidone-ascorbate pair is believed to work by is set out, with the same care about what is established and what is inferred.

!A pushed film exhausts a developer faster, and two makers say so

Extended development in a reused bath is the one combination both makers rule out in writing. ILFORD say, on the powder-developer sheet and again on the DD-X sheet, that they do not recommend push processing using reused developers. Kodak’s T-MAX developer sheet says the capacity is lower when the developer is used for push processing, and that such a solution is discarded after one batch rather than reused.

The reason is in the exhaustion experiment: a longer development develops more silver per film, which consumes more agent and releases more bromide, and a push is asking the bath for its maximum activity at exactly the moment it has least. Use fresh developer for a push. It is the cheapest insurance in the darkroom.

The picture, the look, and the decision that is not chemical

Every number above serves one question, and it is not a chemical one: what did the picture need?

A push is bought with a shutter speed, an aperture and a depth of field. A hand-held frame at 1/60 in a dim room, a moving subject that must not blur, a lens that must stay at f/4 because f/2 will not hold the face — those are the reasons, and they are photographic. If the picture would survive a tripod and a half-second, route two exists and costs nothing chemically. If it would not, route one is the price of having the picture at all, and a grainy negative with empty shadows is worth infinitely more than a clean one you did not take.

And there is a second reason, which the measurements cannot adjudicate. A pushed negative has a look: hard edges, shadows that go to black without detail, visibly coarser grain, and mid-tones separated more than the scene separated them. Half a century of reportage looks like that, and for some subjects it is the right look rather than the compromise. The honest way to hold both is to say which you are doing. “I pushed because the light gave me nothing else” and “I pushed because I want the shadows empty and the grain visible” are different decisions that produce the same negative, and only one of them should be repeated when the light is better.

The measurements are what let you choose deliberately. Without them, the look is what happened; with them, it is what you asked for.

The record to keep

Four lines in the notebook, and they are what make the next decision better than this one.

1. The exposure index you set, and why — the light level, the shutter speed the subject demanded, the aperture you would not give up.
2. The development time you actually ran, at a logged temperature, on a named agitation script, in a bath whose age and version code you wrote down.
3. The contrast index and base plus fog you measured, with the uncertainty from your own instrument records, and the exposure index your own speed point supports under the course’s criterion.
4. What the print needed: the filter you actually used, whether the shadows held anything, and whether you would make the same trade again.

The fourth line is the one most people skip and the only one that closes the loop, because a push is judged in a print rather than on a plot. Add these to the personal processing table the time series produced, as extra rows for the higher contrast indices, and the next low-light evening starts from data instead of from a rule of thumb.

Recap

A push is two moves and only one of them is chemistry. Exposure places the shadows and development sets the slope, so extended development amplifies everything the exposure recorded, invents nothing it did not, and raises the fog floor while it works. What rises is contrast index, mid-tone separation, maximum density up to a limit, fog and grain; what barely moves is the speed point, because it lives on the toe, where the payout from extra development is smallest.

The manufacturers agree, and their own tables show it: Kodak index their push times by the contrast index those times reach, and six doublings of meter setting on T-MAX P3200 are met with a doubling of development and half a unit of contrast index. The one quantified speed-gain claim in this corpus is ILFORD’s half a stop for MICROPHEN, against the four-fold exposure change a two-stop push assumes. And before any of that, the first stop is latitude rather than development, on Kodak’s own account.

Read the decision off your own two curves, stop where the fog stops being worth the contrast, and remember that the negative meets a paper grade rather than a verdict — because the same print is often available from a correct exposure and a harder filter, at no cost in fog or grain, whenever the picture can survive the shutter speed that requires.

Check your understanding

Question 1. Kodak’s XTOL data sheet lists T-MAX P3200 at EI 400 with a contrast-index aim of 0.52 and 9½ minutes, and at EI 25000 with an aim of 1.02 and 19 minutes. What does that pair of rows establish?

• That a factor of about 64 in meter setting is met with a factor of 2 in development time and a rise of 0.50 in contrast index, so the recommendation is a contrast decision rather than a claim that the film became more sensitive
• That the film’s speed increases by six stops when it is developed for twice as long
• That contrast index and exposure index are proportional to one another
• That 19 minutes is the longest useful development time for any film

Show the answer and why

Answer: That a factor of about 64 in meter setting is met with a factor of 2 in development time and a rise of 0.50 in contrast index, so the recommendation is a contrast decision rather than a claim that the film became more sensitive

The two columns are the point of the table. Kodak index each published time by the contrast index it is intended to reach, which makes the recommendation legible as what it is: a slope to develop to, given an assumption about where the exposure landed. Six doublings of meter setting is a factor of 64 in the light the film received, and it is answered with a doubling of the time and 0.50 of contrast index. If a push made a film more sensitive in proportion to the meter setting, those two factors would have to track each other, and they are nowhere near doing so. The film’s sensitivity did not change; what changed is the slope on which whatever light arrived was recorded.

Question 2. ILFORD print two numbers on the DELTA 3200 Professional sheet: an ISO speed rating of 1000/31° and a recommended meter setting of EI 3200. Which statements are true?

• Both come from the manufacturer and neither is wrong, because they are different kinds of statement
• The ISO figure is a measurement made under stated conditions — ID-11 at 20 °C with intermittent agitation in a spiral tank — while the EI is a working recommendation
• ILFORD state on the same sheet that the recommended EI range is a practical evaluation of film speed and is not based on foot speed as the ISO standard is
• The film is misnamed, since a film called 3200 should have an ISO speed of 3200

Show the answer and why

Answer: Both come from the manufacturer and neither is wrong, because they are different kinds of statement, The ISO figure is a measurement made under stated conditions — ID-11 at 20 °C with intermittent agitation in a spiral tank — while the EI is a working recommendation, ILFORD state on the same sheet that the recommended EI range is a practical evaluation of film speed and is not based on foot speed as the ISO standard is

This is the cleanest published demonstration of the distinction the terminology table insists on, because both numbers come from the same page of the same document and differ by one and two-thirds stops. ISO 6 is a standard for film and process systems, and its own introduction says a particular film may have several ISO speeds depending on the processes used, which conflicts with the tradition of associating one speed value with one product. An exposure index is a decision about the meter; an ISO speed is a measurement under a standard. The fourth option treats the product name as a speed claim, which is exactly the confusion the sheet’s own wording is written to prevent.

Question 3. A negative is pushed two stops. Which of these does the print show first, and why?

• The empty shadows, because no development creates density where no photons were recorded, whereas the raised fog and the coarser grain both act on tones that are still there
• The coarser grain, because graininess is the largest of the three changes
• The raised base plus fog, because it affects every part of the negative equally
• The improved mid-tone separation, because it is the largest change in density

Show the answer and why

Answer: The empty shadows, because no development creates density where no photons were recorded, whereas the raised fog and the coarser grain both act on tones that are still there

Fog and grain are degradations of information the negative holds; empty shadows are the absence of information it never held. A raised fog floor lifts the whole print slightly and can be printed through; coarser grain is a texture laid over tones that are still separated. But a shadow that received too few photons to form latent-image specks has no developable crystals, and extended development amplifies nothing there — which is why the two curves in this page’s first plot are separated by little more than their fog difference at the lowest exposures. The mid-tone separation does improve, and it is what makes a pushed negative look good in the mid-tones while the shadows are gone.

Question 4. ILFORD publish, for HP5 Plus inadvertently exposed below EI 250, a PERCEPTOL stock time of 9 minutes at both EI 50 and EI 100 and 11 minutes at EI 200. What does that table teach about pull processing?

• That the answer to serious overexposure is largely a change of developer rather than a drastic cut in development time
• That the published times rise rather than fall as the meter setting approaches box speed, which is the direction the contrast-index chain predicts
• That ILFORD regard the result as a rescue rather than a technique, since they note the quality will not be as high as conventionally processed negatives
• That overexposed film should always be developed for less than half the normal time

Show the answer and why

Answer: That the answer to serious overexposure is largely a change of developer rather than a drastic cut in development time, That the published times rise rather than fall as the meter setting approaches box speed, which is the direction the contrast-index chain predicts, That ILFORD regard the result as a rescue rather than a technique, since they note the quality will not be as high as conventionally processed negatives

ILFORD head the table "accidental exposure only" and name PERCEPTOL, a fine-grain solvent developer they describe as being for use when very fine grain is wanted and a decrease in film speed is not important. That is the whole strategy: the overexposure has already happened and cannot be undone, so what remains is to keep the highlights printable and the grain small, and a different developer does that better than a stopwatch. The rising times across the row are the ordinary contrast relationship, not an oddity: a setting closer to box speed needs more development to reach a printable gradient. The fourth option is the folk rule the table refutes.

Question 5. You are photographing a dim interior of about five stops of luminance range and you must hold 1/60 second at f/4. Using the chain in this lesson, which statements are sound?

• Pushing to a contrast index near 0.82 gives a negative density range near 1.24, which on ILFORD’s range figures for MULTIGRADE RC DELUXE lands near filter 0 — a softer grade than the unpushed negative would need
• If a tripod and a longer exposure would serve the picture, exposing at box speed and printing on a hard filter reaches a similar print at no cost in fog or grain
• The join between a film contrast index and a paper range figure is this course’s inference from two manufacturers’ documents, and the predicted grade is a starting filter for a test strip rather than an answer
• Pushing is the correct choice here because a five-stop subject always needs a higher contrast index

Show the answer and why

Answer: Pushing to a contrast index near 0.82 gives a negative density range near 1.24, which on ILFORD’s range figures for MULTIGRADE RC DELUXE lands near filter 0 — a softer grade than the unpushed negative would need, If a tripod and a longer exposure would serve the picture, exposing at box speed and printing on a hard filter reaches a similar print at no cost in fog or grain, The join between a film contrast index and a paper range figure is this course’s inference from two manufacturers’ documents, and the predicted grade is a starting filter for a test strip rather than an answer

The arithmetic is negative density range equals contrast index times subject log-luminance range: five stops is 1.51, so 0.56 gives 0.85 and 0.82 gives 1.24, which by ILFORD’s instruction become range figures of 85 and 124 and point at filters 2 and 0 respectively. The pushed negative wants the softer grade, because the contrast has been put into the film rather than the paper. The second option is the real alternative and the honest comparison: it records shadow detail the push never captured, and its cost is the shutter speed. The third is the caveat the course attaches wherever this join is made, since ILFORD’s figure is for the range as projected on the baseboard, complete with enlarger flare and the Callier effect. The fourth confuses a subject property with a decision: a short-range subject can be given more contrast in the negative or in the paper, and the choice between them is what this whole page is about.

Check my answersClear