Using a camera as a lux meter

A camera measures light from the object. In museum practice we wish to measure the light falling on the object, preferably in lux. For a given incident lux the light from the object in the direction of the camera can vary from about zero, for a dusty lump of coal, to a blinding glare from a shiny metal surface reflecting the light source.

The prospects for accurate lux measurement with a camera are apparently not so good. Even so, the camera is a useful tool. It will measure lux with adequate accuracy (I return later to the meaning of "adequate") if you point the camera at a white card and then use a few seconds with a pocket calculator or a table. It will measure the lux rather inaccurately if you point it at an object in a showcase, but if the alternative is several hours spent collecting permission to open the case, paralysing the alarm, struggling with the lock and calling for help when the glass door is too heavy for you to control, then maybe this article will be useful.

On the positive side, the calibration of the camera exposure meter is automatically checked every time you expose a film on a family outing and enjoy the perfect quality of the prints ( use colour transparencies for a more sensitive check).

It should therefore be possible for the conservator to dispense with the expensive specialised lux meter and wander round exhibitions with a Nikon slung round her neck, just like everyone else. It is also a discreet way of checking that loaned objects are being looked after properly.

The problem is: how can one deduce how much light is falling on an object from the light departing from it in the direction of the camera and how on earth does one derive the lux from f-number and exposure time for film of a given ISO speed?

The camera's photocell measures the luminance of the object, but it doesn't reveal this to you in the SI unit, which is candela per square metre. (See the chapter on lighting units if you want to be reminded of the definition of the candela and its relatives.)

The camera immediately translates the measured luminance into an aperture (f-number) and exposure time for your particular film, which you have to translate back again into cd/m².

The formula used for calibrating photographic exposure meters is:

Luminance in cd/sq m. = 12.4 x aperture²/ (exposure time in seconds x film speed in ISO units)

The value of the constant must be within the range 10.6 to 13.4, according to the standard. Text books usually quote the value 12.4. The logic behind this formula is quite complicated but fortunately not relevant to this article. We just accept it as a convention.

Now you can grab a sheet of perfectly mat, perfectly white card, focus your camera on it and use this formula to work out the luminance of the card.

Suppose we have ISO 100 film in the camera (a bad choice for museum photography as we shall soon see). We point it at a piece of white card and read: 1 second at f11.

Putting these numbers in the formula gives a luminance of 15 cd/m²

The lux illuminating the card is the luminance of the surface multiplied by pi. This mysterious appearance of a universal constant is of course to keep humanists in suspicious awe of scientists and mathematicians. For the moment we must push on, accepting the wisdom of those who formulate the basic units of quantitative existence.

The lux value (for white card only) is 15 x pi = 47 lux.

The lux values calculated from this formula tend to be low by about 30%. There are at least two reasons for this. The luminance value is low because most single lens reflex cameras, which are almost the only ones that still tell you what exposure they recommend, have through-the-lens light meters. The lens absorbs about 12% of the incident light before it reaches the meter. Other cameras with external light meters also make an adjustment for lens absorption. The second reason is that the white card does not scatter the light perfectly, having a diffuse reflectance between 80 and 90% (or maybe 110% if there is a fluorescent brightener in the card and uv in the light source!). If one puts in a very rough correction for these factors one ends up with a rather simple approximate formula for deriving lux falling on a diffusing white surface by pointing a camera at the surface:

Lux = 50 x fnumber²/ (exposure time in seconds x ISO film speed)

If you tend to use just one film type you can lump the speed and the constant together to further simplify the formula. You can then make a very simple table to give approximate lux values.

Up to this point the camera is no better than a lux meter, and more complicated to use than some. The value of the camera is that you can use it to get an approximate value for the illumination of an object in a showcase, or far out of reach. It is also good for rapid surveys of the likely damage done to furnishings in historic houses at varying distances from windows. The lux value will be accurate enough to allow you to decide if more precise, and time consuming, measurement is needed.

The main source of error is that you have to guess the reflectivity of the object and adjust the raw lux value upward to take account of the loss of light by absorption. Fortunately there are no further guest appearances by famous constants: if you estimate that the object is scattering only a fifth of the incident light towards you, then multiply the calculated lux for white paper by 5 to get the approximate lux hitting the object.

In photography, the typical everyday scene is assumed to be 18% reflective. The typical museum scene is probably considerably less reflective because of the accretion of dust and the liberal use of varnish. Fortunately, most museum scenes include white paper in the form of labels. An advanced camera with a spot meter, can measure just this limited area.

The reader can invent refinements to aid the estimation of the reflectivity of the specimen. One could, for example, shine a light that gives a known extra lux on the specimen and then calculate the reflectance from the difference in the luminance caused by this known lux (thanks to web-correspondent Alan Cottrell for this tip). Such refinements, however, spoil the essential simplicity of the method, which is designed to give an indication of the lux value as an aid to deciding if something more energetic needs to be done.

This method is particularly useful for surveying new exhibitions during the hectic stages of mounting and light setting. A camera allows very rapid assessment of the range of illumination in a showcase. It will show, much more quantitatively than the eye, the dark corners, the glare from the pale, too shiny background, the over-powerful lamp that must be changed rather than dimmed. It will do this without you having to reach in to hold a quivering lux meter 1 mm from a surface insured for millions. It is not necessary to translate to lux all the time: the luminance of the object is what the visitor senses and the relative luminances of objects and background are just as important as the lux hitting the object, if one considers the museum as a means of communication as much as a place for preservation.

If you do a lot of this kind of work it is worth investing in a spot meter. Minolta makes one which allows you to focus on quite small areas deep within a showcase. You can measure the reflectance of the back cloth of the case before objects are mounted. Then you can get quite an accurate value for the lux on an object by focussing on the nearby backcloth, or on the label. Fanatics for accuracy can put near, or on, each object a small patch of discretely coloured material of known reflectance as an internal standard for measurement. Note that the glass of museum display cases absorbs about 10% of the light before it reaches the camera.

The conscientious conservator will maybe accuse me of encouraging sloppy practice in publishing a method for very approximate measurement of lux values. I therefore conclude this article with a discussion of the extremely approximate value of the lux as the unit for evaluating the damage caused by light.

The lux as the museum unit for photochemical potency

Reference: International Standard ISO 2720: Photography - General purpose photographic exposure meters (photoelectric type) - Guide to product specification. First edition 1974.

The Minolta spot meter uses 13.9 as the constant in the luminance formula, which is outside the range allowed by the standard. I have tried comparing lux meter, spot meter and two cameras, pointed at ordinary typing paper. The values for the "constant" ranged from 14 to 17. A suitable value for general trouble shooting in museum lighting would be 16. This is unlikely to give too low a lux value. This is the value used in my approximate formula for measuring lux with a camera aimed at white paper.