Lighting archaeological objects for photography

A common practice in photographing archaeological objects is to place the object inside a light tent, where the light falls evenly from every side. For an archaeological object it is the wrong choice. Even light flattens relief, makes tool marks and wear harder to see, and leaves an object on a white ground looking like a cut-out. The PHOTARCH method gives its light a direction on purpose: one key light, from the upper left, falling toward the lower right, the same in every photograph. The reasons and the research behind them are under Principles & Setup; this guide is about doing it.

Where should the light come from?

Imagine the object at the centre of a clock face, seen from above. The key light stands at half past ten and 45° above the surface, and the shadow falls toward half past four.

Each part of that has a reason. The visual system assumes light from above and slightly to the left, and reads shape fastest when the light is there. Western painting has lit its subjects mainly from the left since the early fifteenth century, and archaeological illustration made top-left shading a drawing rule long before photography — so a photograph lit this way reads the way a drawing of the same object does. At 45°, a straight-sided object casts a shadow about as long as it is high, which gives an idea of its height: an estimate, not a measurement.

The leftward part is a convention rather than a law of vision, and the method says so. What makes it worth keeping is that every photograph made by the method uses the same one.

Why not even light?

The imaging standards for cultural heritage, FADGI and Metamorfoze among them, measure how evenly the light falls across the field, because they were written for flat originals: a page, a print, a painting, lit evenly from two sides, where an even field is the whole point. An object is not flat. Its form is only visible through the differences in light across it — the side that faces the lamp and the side that turns away from it.

A PHOTARCH photograph is therefore not meant to meet the standards' requirement for even light, and nothing in the method tries to. This is a deliberate departure, set out under Standards.

The same reason rules out two set-ups often recommended for small finds. A fill light as strong as the key light gives two shadows and cancels the shading the key light draws. A light tent, which surrounds the object with light from every side, takes away most of the shading: the picture looks clean, and the object looks flat and weightless. The aim is the opposite: that the step from its bright side to its dark side shows its form, and that the picture looks the way the object does in the hand.

Shadow is information

The shadow where the object meets the ground places it on that ground and shows how high it stands. Vision uses contact shadows to judge position and depth; remove the shadow and the object seems to float. Some institutional standards raise objects on glass precisely to get rid of shadows. PHOTARCH keeps them, and lets them fall on the white ground — see A white background.

Fill light: soften the shadow, don't remove it

Opposite the key light, place a fill light: a white card, or a second lamp clearly weaker than the key light. It lifts the shadow side enough to show detail in it while the shadow stays a shadow. Keep some light there: where the side turned away from the lamp gets no light at all it goes black, and black holds no information — no surface, no detail, nothing to measure. A card is often enough, and it cannot overpower the key light. A lamp used as fill light must stay clearly weaker than the key light: at the same strength it gives two shadows and flattens the object again.

The size of the key's modifier — the softbox, diffuser or beauty dish that shapes its light — decides how soft the shadow's edge is: the larger the light source compared with its distance, the softer the edge. Keep it the same within a series. A larger object needs the lamp further back to light it evenly, and then a larger modifier to keep its shadow as soft.

1. KEY LIGHT — 10:30 · 45° above
2. FILL LIGHT — opposite the key; lifts the shadow, never removes it
3. CAST SHADOW — as long as the object is high

Choosing the light source

• One LED panel is enough to follow the method, as long as it holds one kind of LED (see below).
• A high-CRI LED or a strobe with a modifier gives better colour and a cleaner result. The colour rendering index (CRI) says how faithfully a lamp shows colours compared with a reference light; a lamp with gaps in its spectrum can make two colours look alike that are not.
• One kind of light only. Switch off room lights and shade windows. Two sources of different colour cannot both be corrected by one white balance.
• The same relation to the object every time. From a small source, light weakens roughly with the square of the distance, so a lamp moved closer changes both the brightness and how quickly the light falls off across the object. The next section says what stays the same when the object is larger.

What is inside an LED panel?

An LED panel is many small lamps behind one diffuser, and they are not always alike. Some panels hold one kind of LED, all of the same colour. Others mix two whites, a warm one and a cool one, and set the colour of the light by turning one up and the other down. Still others mix red, green and blue, sometimes with whites, to make any colour at all.

Mixing has a cost. The light is then the sum of a few separate bands rather than one continuous spread, so its spectrum has peaks and gaps. Two colours on the object that differ only where the light has a gap come out alike, and no white balance tells them apart again — see Colour accuracy. Close to the object the separate LEDs also stand at slightly different angles to it, and the edge of the shadow can take a faint coloured fringe.

Choose a panel with one kind of LED, and a high CRI, rather than one that mixes tones.

Distance: the inverse square law

From a small lamp, light spreads over an area that grows with the square of the distance, so the light reaching a surface falls with the square of the distance: twice as far is a quarter of the light, two stops less.

1. SMALL LAMP
2. AT 2 × DISTANCE — 4 × area, 1/4 of the light, −2 stops
3. AT 3 × DISTANCE — 9 × area, 1/9 of the light, −3.2 stops

In the set-up this has two consequences. The first is exposure: a lamp moved from 50 to 100 cm needs two stops more, which is why the distance belongs in the record of a sitting and stays the same through it. A larger object may need the lamp further back. It then goes back along the same line, up as well as out, so that it still stands at 45°, and behind a larger modifier, so that it looks as large from the object as before: what stays the same is the relation between the light and the object. The light is no longer the same, so it is a new sitting, and the colour card is photographed again. The second is the ground. Its near side is closer to a lamp at the upper left than its far side, so it is brighter. Across 30 cm of ground, from distance alone, the far side receives about 1.1 stops less light than the near side with the lamp at 50 cm, 0.6 at 100 cm and 0.3 at 200 cm: doubling the distance roughly halves the fall-off.

1. LAMP — 45° above the near edge; distances not to scale

The fall-off is part of a light with a direction, and the method does not try to remove it; the ground is taken to white in the published copies. Moving the lamp back evens the ground, but it costs light, and the shadow's edge gets harder, as the next section shows.

The size of the light and the edge of the shadow

How soft a shadow's edge is depends on how large the light looks from the object: the size of the source compared with its distance. Light from each part of a large source passes the object's edge at a slightly different angle, so the shadow does not end in a line but in a band, from full shadow to full light.

1. A SMALL SOURCE
2. A LARGE SOURCE — same distance

The width of that band is roughly the size of the light × the height of the edge ÷ the light's distance. A 60 cm softbox at 60 cm gives a band about 10 mm wide below an edge 10 mm high. The same softbox moved to 120 cm halves the band and makes the light harder. A larger diffuser also spreads the highlights on metal and glaze, as described below. Neither soft nor hard is the right answer: what matters is that it does not change within a series, where the same modifier at the same distance gives every picture the same shadow.

Metal, glass and glossy surfaces

Metal reflects the lamp rather than scattering it. A small, bare source makes a small, hard highlight that clips; a larger diffuser in front of the key spreads it into a broad, softer one. The white card opposite gives polished surfaces something to reflect on the shadow side. Check the highlight warning on the object itself, not only on the ground: a clipped area on the object holds nothing that can be recovered.

Glossy glaze, polished stone and lacquer behave like metal. A larger diffuser and the card opposite help them; moving the lamp does not, and breaks the series.

The shape of the modifier shows in every highlight, since a glossy surface reflects the light itself. Choose a round, octagonal or hexagonal one: its highlight reads like the sun's. A square or rectangular softbox puts a small window in every highlight. The exception is the strip box, long and narrow: on a vessel or another curved object, a bowl or a vase, its one long highlight runs along the edge and marks the curve, which defines the form.

1. KEY LIGHT — 10:30 · 45° above
2. CAMERA — plumb view
3. FLAT AND LEVEL — the light leaves at 45° the other way and misses the camera
4. CURVED TOWARD THE LAMP — the light goes straight up into the lens: a highlight

Glass lets the light through. On a white ground clear glass can almost disappear; black cards placed just outside the frame, at the sides, give its edges a dark line without moving the key. Coloured glass casts a coloured shadow, which is part of what the picture records.

Reflections, and polarised light

A glossy surface works like a mirror: it shows whatever is in front of it. On a polished object that is the lamp, as a highlight. The board is glossy too, and it can show the lamp, a window or a ceiling light as a pale patch on the ground, or the object itself mirrored beneath it. Seen from straight above, the lamp's own reflection leaves the board at 45° toward 4:30 and misses the camera. Light from the room does not, which is one more reason to work in a dark room.

Light can also be sorted by the way it vibrates. Ordinary light vibrates in every direction across its path. A polarising filter lets through only the part that vibrates in one direction and stops the rest: it is what takes the glare off water in polarised sunglasses. Light reflected from a shiny surface keeps the direction it had; light scattered from a matt surface loses it.

Cross-polarisation uses that. A polarising film goes over the lamp, and a second polariser on the lens is turned until it stands at right angles to the film. The reflections, still vibrating the lamp's way, are stopped; the light scattered from the surface, which carries its colour and its detail, passes. Glare from glaze, varnish and polished metal largely disappears, and colours look more saturated.

It has a price. It takes away much of what makes metal look like metal, it changes colour, and the filters cost light, so exposures are longer or apertures wider. It does not remove reflections of light that never passed the film, such as the room's. In the PHOTARCH method it is therefore a choice for a whole sitting, not a default: the colour card is photographed with the filters in place, and the choice is recorded with the pictures.

The camera moves, the light does not

Once set, the light stays where it is. Between views the camera moves round the object, which is turned only where a view needs it — over for the underside, a quarter for two of the diagonals — and the lamp is never moved. The same view of two objects is then lit the same way, in this studio and in any other that follows the method. Record the lamp, its modifier, its distance and the angle of the light with the pictures: no file can know them unless someone writes them down.

The PHOTARCH app's levels leave the light out, and say so: evenness is not a goal, because the object is lit from one side on purpose, and the direction of the light is the photographer's to record. Flat originals — paintings, documents, prints — are lit evenly from two sides instead, and the app does not measure how evenly.