Zone System

The Zone System is a numbering scheme for tones. When a critique asks for Zone III-IV separation, it wants the darkest shadows that still show texture (III) to hold a visible step apart from the fuller shadows just above them (IV), so a dark area keeps its form instead of merging into one flat block. Everything below is how to see that step, light it, and rescue it. Adams and Fred Archer codified the scale around 1940 so a photographer could name, before the shutter, exactly how light or dark any part of the scene would print. Eleven steps, Roman numerals 0 to X, each one stop apart. Zone 0 is solid black with nothing in it. Zone V is middle grey, the 18% grey card, and the tone a reflected meter renders whatever it reads as: point one at anything and it will offer to place that thing on Zone V. Zone X is paper white, the tone of a light source or a specular glint. Between them the scale has a texture band and two rims. Zone I is the first tone above black and Zone II the first with a hint of texture; Zone III is the darkest value that still shows surface clearly, the shadow you can read fabric or bark in; Zone IV sits one stop above III, an open shadow with air still in it: dark foliage in open light, or the shaded side of a form in sunlight. Zones VI and VII are the light tones with full texture (light stone, snow in raking light), Zone VIII the last white with texture, Zone IX white with a trace of tone. Skin has no fixed rung: Adams put dark skin at Zone V, average light skin at VI and very light skin at VII, and all of them hold full texture. Adams called II through VIII the textural range, because that is where surfaces still describe themselves.

Two things make the scale useful rather than academic. First, it is a scale of exposure, not only of print value: each zone is one stop, so placing a tone is a metering act. Meter the shadow you care about, close down two stops from what the meter suggests, and you have placed it on Zone III. Second, it separates the tones a picture is anchored by from the tones it is made of. A frame needs a black and a white somewhere, but it lives in the steps between them, and the steps that fail most quietly are the dark ones: III and IV, the difference between a shadow that describes a form and a shadow that erases one.

On a digital file the numbers transfer and the chemistry does not. Adams controlled contrast by developing each negative longer or shorter and choosing a paper grade; you control it with a tone curve, reversibly, and you can do it to one part of the frame without doing it to the rest, which is a better tool for the same job. What you lose is film's shoulder: a sensor clips hard at the top, so the priority inverts. Film exposed for the shadows and developed for the highlights; digital exposes for the highlights and lifts the shadows in post, and the floor under those shadows is not paper black but the sensor's noise. On a neutral rendering the landmarks sit roughly where a Lightroom histogram would put them: Zone V near the middle, Zone III and IV where the Shadows slider does most of its work, Zone I and II down where Blacks does. Treat those as landmarks, not thresholds. The controls overlap heavily and adapt to the image, so on a contrasty frame Blacks reaches up into III, and a profile that carries contrast pushes the whole map lower. Slider names here are Lightroom's: Camera Raw, Luminar and DxO use the same two words, Capture One calls them Shadow and Black, and darktable does this job in EV rather than in named bands, with the tone equalizer, which is the closest thing any editor has to a zone control.

A 6 has a full histogram and a dead shadow end. The black point is set, a true white exists, the range reaches both walls, and every global measure says the tonality is correct. Look into the dark side of the subject and there are two values in it: black, and not quite black. The shadow side of the face, the hair, the coat and the shadow the figure casts all sit within a whisker of each other, so the region reads as a single shape with a lit edge, a cut-out, what a critique will call a graphic block. Form is gone because form is a gradient, and a gradient needs steps to be made of. Zone III and Zone IV have been folded into one.

An 8 has a ladder in its shadows. The darkest textured value (III) is a clear step above the near-black (II), and the fuller shadow (IV) is a clear step above that, so the dark side of a form still turns, the fabric still reads as fabric, and the shadow under the jaw can be told from the shadow of the jaw. Nothing in the histogram announces this: a frame with merged shadows and a frame with a shadow ladder can share the same black point, the same white point and the same overall spread. The judge sees it in one place only, inside the dark regions, at the size the picture will be viewed.

Zone III-IV separation is the phrase a critique uses for exactly that step, and it is among the zone pairs a critique flags most, because the low tones are the ones contrast pushes on first. Steepen the curve for punch, deepen the blacks for mood, convert to monochrome with a heavy hand, and the low tones are the first to slide together, which is why a striking, contrasty frame so often scores a 6 while looking like an 8 at thumbnail. The opposite failure exists and scores the same: shadows lifted until Zone III sits at Zone V, the frame floats, and the dark form has no anchor left to be separated from. The 8 keeps a real black in the frame and builds its steps above it.

At the scene, find the darkest thing you need to read. Not the darkest thing in the frame: the darkest thing whose surface matters to the picture, the shadow side of the face, the bark, the black coat with a weave, the interior of the doorway. That is your Zone III candidate, and everything darker than it can fall where it likes. Then find the shadows that give the subject its shape: the turn of a cheek away from the light, the underside of a leaf, the fold in the fabric. Those are Zone IV, and the question this lesson turns on is whether they sit a stop lighter than the Zone III surface or on top of it. Squint. If the dark side of the subject collapses into one flat shape when you do, the separation is not in the light and no exposure will put it there; you need fill, a different angle, or a different hour.

If your camera has a spot meter, use it as a zone ruler. Meter the Zone III surface, then the Zone IV shadow: readings one stop apart mean the step exists; readings half a stop apart mean an S-curve will fold them together. Zebras and blinkies tell you where the highlights die and say nothing about the shadows, so read the left wall of the histogram directly: a hard spike against it means tones you wanted at III have already been pushed to 0.

At the screen, run the same census. Zoom to the dark side of the subject at 100% and count the values you can see. Two (black, not-black) is the block; three or more, each doing a job, is form. Read the low end of the histogram for a mass jammed against the wall, and if your editor shows a tone readout, hover the shadow you care about. On a neutral rendering, in the gamma-encoded percentages every raw editor reports, Zone III sits roughly in the low twenties and Zone IV in the low-to-mid thirties. Those figures are false by about five times in a linear readout, and on a strongly colored surface you get three numbers rather than one, so read a neutral patch nearby instead. Then convert a copy to monochrome, because color hides merged shadows: a dark blue jacket and a dark green hedge look separate in color and land on the same zone the moment the hue is gone.

Separation in the shadows is bought with light first and exposure second. Fill is the honest tool: a reflector, a white wall, a bounced flash two or three stops under the key lifts the shadow side from Zone II to Zone III or IV without touching the highlights, which is precisely the move no slider can make cleanly afterwards. Direction matters as much as amount. Front light flattens the whole subject into one value; light from the side lays a gradient across the form and writes the III-to-IV step into the file before any editing exists. Wait for the hour when the sun rakes, or move so it does.

Then place the shadow. A reflected meter renders whatever it reads as Zone V, so metering the darkest surface you need texture in and closing down two stops puts it on Zone III; closing three puts it on Zone II, where only the first hint of texture survives. That is the film rule, and on a digital body it is one action: spot metering, the spot on that surface, exposure compensation at minus two, or in manual the needle driven to minus two. It comes with a caveat that changes the order of operations: check the highlights first. If the key highlight survives at the exposure that places the shadow on III, shoot it. If it does not, protect the highlight, because a clipped highlight is gone and a dark shadow is only quiet, and plan to lift the shadows at the desk. A modern full-frame sensor at base ISO holds something like five or six stops below middle grey that still look clean when they are lifted, and a smaller sensor or a higher ISO holds fewer; the failure is not that the data is absent but that it is noisy, and noise is what fills the gap between Zone II and Zone III when you try to open it later. So keep the ISO at or near base, use the tripod when the scene is still, and where the range beats the sensor, bracket rather than crush.

Skin has no single zone, so place it the same way. Adams' table names a complexion at four rungs, dark skin at Zone V, average light skin at VI, very light skin at VII and highlights at VIII, and it names a fifth without one: Zone IV, the recommended shadow value for a portrait in sunlight, whatever the face. That unqualified rung is the useful one here, because it makes the arithmetic complexion-dependent. A Zone VI face can fall two stops into its shadow side and still land on IV. A Zone V face has only one stop to spend before that shadow drops below IV and starts merging with III. So the step this whole page is about is structurally tighter on darker skin, and it is bought with fill rather than with exposure. Meter the face in front of you rather than reaching for a number, then meter its shadow side: the gap between those two readings is your lighting ratio.

Shooting for monochrome, remember that color contrast will not save you. Two surfaces of different hue and equal brightness sit on the same zone in the conversion, and the only capture-side levers are a filter and the light itself: a yellow or orange filter darkens a blue sky against cloud the way Adams did it on film, and a green filter lifts foliage off the shadows; a red filter darkens the sky hardest and takes green foliage toward black, buying sky drama at the cost of exactly the III to IV step this lesson is about. On a digital raw that filter's job moves to the channel mixer at the desk, so the capture-side lever left is the light, and a change of angle can turn a hedge from the same value as a jacket into a step above it.

The shadow end of the tone curve is where zones are separated or merged, and one idea governs every move: slope. Where the curve runs steeper than the diagonal, adjacent tones spread apart; where it runs flatter, they crush together. Zone III-IV separation is a request for slope through the low tones, not for brightness. Lifting Shadows globally moves III and IV up together, and that buys some readability, because the same one-stop step occupies more of the scale higher up. What it does not buy is slope. The slider lifts the darkest tones hardest and eases off as it climbs, so past a moderate setting it flattens the very step you were trying to open, and the shadow goes grey without turning. The move that works is a curve point placed around the Zone III value and a second one a stop above it, the lower held and the upper eased up, so the segment between them steepens while the black point stays where it is. Put a third point near the midtone to contain it: a curve's span is fixed, so whatever you steepen between III and IV is paid for by the segment just above. Blacks and Shadows are two controls for two jobs here: Blacks sets where Zone 0 and I sit, Shadows moves the mass around III and IV, and the mistake in most contrasty frames is that Blacks was pulled so far that III went with it.

Local beats global. A luminance range mask running from about Zone II to Zone IV, with its falloff dying out by Zone V, lets you add contrast inside the shadows without dragging the mids with it, and it reads the rendering you have now, so set it after the global tones are settled, and a brush over the one region that turned into a block (the shadow side of the face, the interior of the doorway) is the dodge-and-burn lesson applied to a single step. For monochrome, the channel mixer, called B&W Mix in Lightroom and Camera Raw, is the tool the curve cannot replace: two surfaces that share luminance but not hue are separated by weighting their colors differently before the conversion, the darkroom filter done at the desk, and the difference between a mono frame with steps and one with blocks is very often that mixer alone.

Now the honesty clause. Post can only separate tones the file recorded as different. Shadows crushed in an 8-bit JPEG hold a handful of values, and steepening a curve through them bands and posterizes: the gradient steps instead of rolling. Shadows captured so far under that they sit inside the sensor's noise hold a value plus noise, and steepening amplifies the noise faster than any detail. Neither one is a step you can recover. When the 100% view shows no distinct values to spread, the separation is capture-limited and the frame needs fill or a re-shoot, not a fourth curve. Check at 100% to find out whether the values exist, then judge at the size the frame will be seen: 1:1 on a high-resolution file magnifies the noise that is hiding the step. The opposite over-correction is just as real: pull Zone III up to Zone V to be safe and the frame floats, its anchor gone. Keep a true black in the picture and build the steps above it.