Edges: The First Decision

A painter faces a white sheet and decides what to put on it. A photographer faces a world already overflowing and decides what to cut away.

That sounds like a difference in phrasing, but it changes the sentence running through your head when you lift the camera. Beginners think I want to photograph that thing — so the thing goes in the middle, the shutter goes off, and the other 90% of the frame arrives by accident. The more useful sentence is a different one:

What am I deciding to leave out?

This chapter is an answer to that question. It needs no equipment and no travel. Turn ninety degrees, take one step forward, drop the camera five centimetres, and all four edges have changed.

Why cutting into something is not a mistake

The belief worth dismantling first is that a subject touched by the edge means a botched picture.

Look at your own frames and the instinct is to step back and “get it all in”. Then open any serious book of photographs and notice how few of the pictures in it are whole. That is not carelessness. It is that your visual system handles interrupted objects by a rule quite unlike the one you assume.

Shown an object that is occluded or cut off, the brain completes it for you — perceptual psychology calls this amodal completion. A cat half hidden behind a sofa is not experienced as half a cat; you simply see a whole cat, and the completion is involuntary and free. To the brain, the border of a photograph is one enormous occluder: the world outside the frame is hidden, not abolished, and the viewer’s perception carries it onward.

Three consequences follow, and between them they account for which cuts look considered and which look accidental.

• Completion needs enough evidence. Take a big bite out of something and what remains is still plenty to identify it and to say where it continues. Completion fires instantly, and what the viewer feels is a world running on past the border.
• Cutting too little never triggers completion at all. Clip a small corner and the brain does not read an occluded whole; it reads an object with a piece missing. That gap becomes the most conspicuous thing in the picture. This — not taste — is why a timid cut looks like a slip.
• Cutting at a joint leaves the direction ambiguous. Wrists, ankles, knees and necks are exactly where a shape changes direction, so they offer the fewest cues about where the rest of the body goes. Cut mid-calf, mid-thigh, mid-forearm — anywhere the form simply continues — and completion costs the viewer nothing.

Which lets the three kinds of cut be restated, this time with the mechanism attached:

• Clean (the subject sits whole inside the frame with air around it). No completion is triggered. The picture is quiet, specimen-like, catalogue-like. That is often exactly right for still life and architecture.
• Decisive (the edge takes a substantial bite). Completion runs at full speed. The picture gains tension and the viewer knows the world continues.
• Hesitant (a small corner just caught). Completion never starts, and the notch reads as damage.

One rule covers it: either give it room or cut it hard. The middle ground is the only wrong answer.

The four edges are the strongest lines you have

A second mechanism runs alongside completion and matters just as much.

Among the first things the visual system does is find edges — large populations of cells in the visual cortex respond to light–dark boundaries at particular orientations, and every later stage of shape recognition is built on that. Now consider the border of a photograph: four perfectly straight lines, the longest in the picture, with image on one side and blank paper or black screen on the other. Nothing inside the frame competes with them for contrast.

So anything near an edge holds the eye, deserving or not. A bright patch of sky in a corner, an unnoticed passer-by at the margin, a head grazing the top — each of them keeps pulling the viewer back out of the picture.

Two everyday experiences follow. The first is the photograph that feels off without your being able to say why; nine times out of ten something in a corner is competing with the subject. The second is that a subject pressed against an edge feels uneasy — it is held by the border and has nowhere to retreat.

No photograph can demonstrate this, because any picture good enough to publish has already solved it. A procedural rule serves better: after you take a frame, look at the four corners before you look at the subject. In the other order you will not see them.

The cut is the content

The Naschmarkt frame is the extreme case of cutting decisively: not a single object in it is whole, and it is not remotely chaotic. Had the photographer stepped back five paces and included the stall entire, with its neighbours either side, the picture would have turned into a record. You would know it was a market stall. You would not feel that you were standing in it.

Which is why “step back and get it all in” is usually the wrong advice. Stepping back does not merely shrink things; it switches completion off and shows the viewer out of the scene.

The limiting case: the subject is not in the picture

This one rewards a long look. It makes what was left out the subject: the person is entirely outside the frame and only their evidence is inside it. The hexagonal cracking of the salt gives that shadow a near-infinite, uniform ground with nothing to compete against.

Note how it differs from the familiar accident of letting your own shadow into a picture. That version arrives at an angle from a corner and the photographer never noticed. This one enters at the centre of the bottom edge, points into the depth of the frame, occupies the whole central axis, and has been given clear space on every side. Same element; one is an error and the other is the subject, and the only difference is whether it was arranged.

There is a small exercise in that. Next time the sun is low, look at how long your own shadow is before you decide whether it belongs in the frame. Including it is fine and excluding it is fine. Not seeing it is not.

Empty space is not waste

Beginners read this as waste and want to move in. But emptiness is not absence — it is scale. How much room you give the surroundings is how you tell the viewer whether this subject is tiny or towering.

The working test is one question: is this a photograph of that object, or of the place that object is in? If it is the place, the subject should be small. If you cannot answer, the picture does not yet know what it is about, and that is the real problem to solve.

Empty space does one more thing that is rarely mentioned: it gives completion somewhere to happen. The room around a subject is where the viewer’s eye imagines the world beyond the border. A frame packed to the corners is forceful, but it grants no such latitude.

Six symptoms and their causes

The format throughout is: the picture shows X → the usual cause is Y → change Z.

1. The subject is jammed against one edge and the frame feels tense. → The border has hold of it and it has nowhere to retreat. → Give it more room on that side, or go the other way and cut a third of it off, so it reads as decisive instead of falling out.
2. A person is cut at the wrist, ankle or neck and looks amputated. → The cut landed where the form changes direction, leaving no cue for completion. → Move it to mid-calf, mid-thigh or mid-forearm, where the shape simply carries on.
3. A small corner of something is clipped and it nags at you. → Too little was cut, so completion never fired and the notch reads as damage. → Step in and cut properly, or step back and include it whole. Do not stop in between.
4. A bright patch of sky or a reflection in a corner keeps taking your eye. → Edge attraction and high luminance working together. → Move half a pace to put it outside the frame, or drop your viewpoint so a building covers it.
5. A portrait with a slab of empty space above the head and the chin nearly touching the bottom. → Focused on the centre of the viewfinder and fired without re-composing. → Take the headroom out; better to clip a little hair than to leave the face floating.
6. At home it feels like something is missing, yet everything is in the frame. → You photographed cautiously, leaving every object its full margin, so nothing in the picture is more important than anything else. → Photograph the same scene again fitting half as much in.

Symptoms three and six are two faces of the same illness: hesitation. The cheapest cure on location is not to think harder but to take two frames — one cautious, one hard — and compare them later. It does not take long to notice that you always choose the hard one.

Aspect ratio: a rhythm you inherit

Aspect ratio is not a line on a specification sheet. It changes how you arrange things.

Ratio	Feel	Suits
3:2	Directional, leaning dynamic; the long side takes leading lines easily	Street, landscape, most situations
4:3	Steadier and more solid, close to a comfortable field of view	Portraits, still life
1:1	The most stable and most closed; pushes you to centre or to strict symmetry	Single subjects, pattern
16:9	Emphatically horizontal; subjects drown in their surroundings easily	Environmental storytelling, panoramic landscape

What makes a square peculiar is that it has no long side. A 3:2 frame has a direction to spread along — leading lines, foreground and background, a row of things. In 1:1 neither axis has the advantage, so the picture collapses toward its centre. That is why square photographs are so often either strictly symmetrical or a single centred subject. It is not a preference. It is the shape insisting.

The X-T5 and X-E4 both let you set 1:1 or 16:9 before the exposure, and the viewfinder shows the cropped frame directly. That is far more useful than cropping afterwards, because cropping can only choose among things you already photographed, whereas setting the ratio on location changes where you walk and where you stand.

Where these numbers came from

Worth knowing: almost none of the ratios in that table were invented for aesthetic reasons.

16:9 is a compromise calculated on paper. When engineers were choosing a single widescreen television standard in the early 1980s, Kerns Powers of the SMPTE laid out the existing formats — Academy 4:3, widescreen 1.85:1, anamorphic 2.35:1 — scaled to equal area, and found that they all fitted inside, and all contained, a rectangle of about 1.78:1. The geometric mean of the extremes lands on the same figure. So 16:9 is not a shape anyone found beautiful; it is the rectangle that wasted the least when converting between everything else. Every phone you have held since has inherited that arithmetic.

Square arrived because turning the camera was awkward, and returned because a database was. Twin-lens reflexes like the Rolleiflex are used at waist level, where rotating the body for an upright frame is genuinely unpleasant, and the 6×6 negative removed the need. Kodak’s Instamatic gave consumers a square for a different reason — a simple cartridge and a simple print size. Then, decades later, Instagram launched square because a uniform grid was easier to build and to browse, and abandoned the restriction in 2015. Three squares, three engineering motives, no aesthetic argument anywhere.

Printing full-frame is a position, not a default. Cartier-Bresson held that composition had to be finished at the instant of exposure, and had prints made showing the ragged black edge of the negative — a declaration that nothing had been cropped, that the four borders you see are the four he saw. Ansel Adams took the opposite view and cropped freely in the darkroom, since for him the print was where the photograph was performed.

Both positions produced great photographs, so this is not a question with a right answer. What it does show is that where the border gets decided is itself something you may choose.

Decide now, or crop at home

Both are legitimate. But if you are starting out, spend a while deliberately practising the first.

The reason is that cropping rescues a composition without rescuing your eye. Once you know you can crop at home, you stop looking hard at the edges on location, and in time you become someone skilled with a mouse rather than someone who sees frames.

There is a technical reason too. Cropping discards pixels, and what you crop away is usually the few steps you should have walked. Moving closer changes perspective, background and the relationship between near and far all at once. Cropping only enlarges the same photograph.

The decision table

Run it before the shutter. The first four rows are about edges, the last two about shape.

Your instinct says	It usually means	Do this
“I should step back and fit it all in”	You are avoiding a cut, which switches completion off	Step forward instead and cut harder
“That crop looks odd somehow”	Too little cut, or cut at a joint	Move the cut: take more, or shift to mid-limb
“There’s something in the corner”	The border is competing with your subject	Half a pace to put it outside, or lower your viewpoint
“Surely the subject is too small”	First settle whether the subject is the object or the place	If the place, leave it; if the object, walk closer rather than crop
“This scene is symmetrical / has one subject”	The long side has no work to do	Set 1:1 on the camera and look again
“This scene is really about the surroundings”	You need horizontal spread	Set 16:9, then check the subject does not vanish