Museum lighting

Why are the paintings a museum values most often hung in its dimmest rooms?

Walk through a large museum and you meet something that looks like a mistake. The sculpture court is flooded with daylight, the gallery of oil paintings is comfortably lit — and then you reach a room where the most precious things in the building, the drawings and manuscripts and early photographs, sit behind glass in what feels like dusk, and you must stand still a while before you can see them properly.

If light were simply the medium through which we see, this would be exactly backwards: you would light the treasures brightest of all. So light must be doing something besides revealing. What is it costing?

Start with what light physically is when it lands on a painting. Not information — energy. A photon absorbed by a pigment molecule delivers its energy into that molecule, and if that is enough to break a chemical bond, the molecule is changed. That is the whole mechanism: photochemical degradation. Dyes fade, paper yellows and embrittles, textile fibres lose strength, varnish discolours. Shorter wavelengths carry more energy per photon, which is why ultraviolet is disproportionately destructive, blue next, and the long red end least.

Now the property that turns this from a nuisance into a governing constraint: the change is cumulative and does not reverse. A faded watercolour does not recover in the dark; it merely stops getting worse. So every object carries a finite quantity of viewing, spent a little at a time by everyone who looks — and the museum's real decision is not how bright the room should be, but how to divide that quantity between the people alive now and the people who are not.

Can that be made quantitative? Approximately, and the approximation is what conservation lighting rests on. Damage is treated as depending on the dose — illuminance multiplied by exposure time, counted in lux-hours — so fifty lux for eight hours costs the same as a hundred lux for four. It is a working rule rather than a law, and departures from it are documented for some materials, but it is good enough to plan with, and it hands the curator a budget spendable either as brightness or as hours.

That reframes the dim room entirely. It is not an aesthetic preference and not caution for its own sake; it is a choice about how to spend an allowance. Want an object on the wall permanently, and you must show it dim. Want it bright, and you must show it briefly and rest it in the dark for years — which is why the most fragile things appear in rotations rather than the permanent hang. The rotation is the budget being spent in the other currency.

Materials differ enormously, which is why one building holds three lighting worlds. Standard practice sorts them into sensitivity classes. The highly sensitive — works on paper, watercolours, textiles, dyed leather, many photographic processes — are shown at around fifty lux. The moderately sensitive — oil and tempera paintings, undyed leather, bone, ivory, wood — at around two hundred. Stone, ceramic, glass and most metals have essentially no photochemical limit. The glass roof over the sculpture court and the gloom of the print room are one policy applied to different chemistry.

One question remains: how is fifty lux enough to see by, when an office runs several hundred and an overcast day runs thousands? Because the eye adapts, and adaptation takes minutes. A dark-adapted eye resolves fine detail at fifty lux surprisingly well — which is why those galleries are approached through progressively dimmer spaces, and why a bright doorway, or a phone screen, quietly ruins the room for everyone in it. What degrades faster than acuity at low levels is colour discrimination, and that is the genuine loss the visitor is asked to accept.

Three refinements, each changing how the trade should be read.

The first is that the published lux figures are conventions, not thresholds. There is no level below which nothing happens; there is only a slower rate. Fifty lux is a negotiated compromise between the rate of loss and the point at which a visitor can no longer see — distilled from research on fading, but chosen rather than discovered. Treating it as a safety line is the commonest misreading of the practice.

The second is that light-emitting diodes have moved the frontier without abolishing it. A well-chosen LED renders colour faithfully while emitting essentially no ultraviolet and very little heat, so the cheapest and most gratuitous portion of the damage can simply be removed. Visible light still fades things, so the budget remains; but the same illuminance now buys more viewing, and lamps that come up only when someone enters mean the unattended hours cost nothing.

The third is easy to miss while admiring the engineering. Underneath the lux-hour arithmetic sits a value judgement in technical clothes: how long is this object supposed to last, and how do we weigh a visitor today against one in three hundred years? The science gives the exchange rate between brightness and lifetime. It cannot say which side of the trade to take, and museums answer differently — a study collection consulted by twenty scholars a year is not managed like a painting a million people come to see.

• Why humidity and temperature swings damage some objects faster than light ever will, and how the two budgets interact.
• How the same reasoning applies outside museums — to books on a sunlit shelf, to a carpet by a window.

Every term the collection defines is gathered in the glossary.