The golden ratio in photography: where to put the subject.

Thirds, phi and the difference between them

The rule of thirds divides the frame into three equal strips each way and puts the subject on a line or a crossing: 33 percent in from an edge. The phi grid does the same with unequal strips — 38, 24, 38 — so the lines sit at the golden section, 38 percent in from each edge, and the crossings are a little closer to the centre than the thirds crossings. That is the whole difference: five percent of the frame width. It sounds like nothing and it changes the picture, because 38 percent is the point where the larger area and the smaller area relate as the whole relates to the larger. The eye reads that as settled; it reads thirds as slightly pushed out.

Neither grid is a law. Thirds is the better tool when the subject needs room to move into, when there is a strong horizon, or when the picture is about tension. Phi is the better tool for a single subject at rest — a portrait, a still life, a building, a shell — and for anything that will be looked at for a long time on a wall. Most of the pictures on this site are cut on the phi grid, and the calm people describe in them comes from that cut as much as from the subject.

The spiral, and why it confuses

The golden spiral overlay is a drawing of the construction — Fibonacci squares with a quarter-circle in each — laid over a viewfinder. It is meant to suggest that the eye enters the frame from the wide end and travels to the tight end, where the subject waits. That is true of some pictures, mostly ones with a curving line already in them: a road, a river bend, an arm. For everything else the spiral is decoration, and trying to make a face or a building sit inside it produces the stiff, over-composed look that people rightly distrust.

What the spiral does contain that is useful is its eye: the point it tightens toward. In a golden rectangle that point sits at the golden section both ways — 38 percent in from one side, 38 percent up from the bottom, or their mirror images. That is the same point the phi grid gives you at a crossing. So you can switch the spiral off, keep the grid, and lose nothing except the confusion.

Where the subject goes

For a single subject: put its visual centre — the eye in a portrait, the densest part of a still life — on a phi crossing, 38 percent in from the side it is facing away from, so that the larger space is in front of it. For a subject that faces the camera, either crossing works; choose the one that leaves the more interesting background in the larger area. For a symmetrical subject — a doorway, a face straight on — centre it and let the symmetry be the composition; the ratio then goes into the framing instead, by cropping so that the subject fills 62 percent of the height.

Horizons: put them on a phi line, not in the middle. Sky at 38 percent when the ground is the subject, at 62 percent when the sky is. The middle horizon is the one habit worth breaking first; it halves the picture and makes both halves compete.

Two subjects: the larger or nearer one at a phi crossing, the smaller on the opposite phi line, so that the two are separated by the golden section of the width. Three or more: choose one anchor at a crossing and let the rest fall where they will; the eye needs only one place to rest.

Cropping to the ratio

A golden rectangle is 1.618 to 1: wider than 3:2 (1.5) and narrower than 16:9 (1.78). Most editing programs let you set a custom crop of 1.618:1, and several offer a golden ratio overlay in the crop tool — in Lightroom it is under the crop guide overlay, cycled with the O key; in Capture One and Affinity it is an option in the crop settings. Crop first, then place the subject: the grid only means something once the frame is the right shape.

For prints, the golden rectangle lands close to a few standard sizes: 50 × 30, 65 × 40, 80 × 50 and 100 × 62 cm are all within a percent of it. A 3:2 file cropped to 1.618 loses about 7 percent of its width; a 4:3 file loses 17 percent, so compose with the crop in mind on a 4:3 camera.

When to ignore all of it

Any composition rule is a description of what tends to work, not a recipe. If the picture is better with the subject dead centre, or hard against an edge, it is better, and the grid is wrong. The honest use of the ratio is as a check after the fact: when a picture feels almost right but not quite, lay the phi grid over it and see whether the subject is sitting five percent off the section. Very often it is, and the crop that fixes it is small.

That is how I work on the pieces here. The construction is drawn first in the geometric works, so the section is built in; in the photographic and painterly ones, The Golden Aperture among them, the picture is made, then checked against the grid, then cut. Either way the proportion is there to be measured, which is the only claim about the golden ratio worth making.