The following is an article that originally appeared in Modern Photography magazine, July 1980. This would still work for DSLRs.
How To: Incident Light For Your SLR
Through-lens metering is certainly adequate for the vast majority of picture situations. But if you've ever tried shooting portraits of someone wearing a brilliant white or dark black dress with the main light source anywhere but front, especially when using pure white or black backgrounds, you must have realized the disadvantages of metering reflected light from the camera position.
A spot meter is often helpful when zeroing in on a subject's face, but unless the face has the same reflective value as 18 percent gray, you will have to make an additional calculation to compensate for correct skin tone. If the skin is very light or very dark, you just don't accept the meter's reading. This accounts for all the complicated dials and charts (and cults) associated with reflected-light metering.
An 18 percent gray card does solve most of the problems of reflected light metering, but it is awkward to carry, and holding it at the unshadowed exact angle can be painfully tricky. The angle is critical if a considerable portion of light comes from the sides. After all, the flat card hardly resembles the human face, or most other three-dimensional subjects for that matter. After years of hassling with the problem, I had come to the conclusion that an incident light meter was what I needed, and decided to invest $85 to $100.
But it suddenly occurred to me that I might be able to get around that expense by making my own spherical dome to fit my SLR. The theory is quite simple: Just slip the dome over the lens, go to subject position, point the camera toward taking position and take a reading of light falling on the subject. Amazingly, the dome actually works! Obviously, this is why most incident meters use a spherical shape resembling the average three-dimensional subject, especially that part of the human face which requires adequate lighting.
To make the dome I cut a piece out of a styrofoam cup, warmed it over the stove and pressed it into a hemisphere with a metal lime squeezer (warmed up to the same degree), then trimmed it to fit snugly into the threaded recess of my lens. Gluing this to a snap-on filter holder would make a really professional-appearing attachment. Regardless of the light source, my readings match the reflected light reading from the 18 percent gray card.
Versatile Approach
Incident-light metering is practically ideal as long as one can approach the subject or establish that, even at camera position, the light measured is the same as that falling on the subject. Also, if you want some areas of your picture to appear lighter or darker than they actually are, a little experience at compensation is the fastest and easiest tool. Or, you can spend a few hundred on a fancy reflected light meter with a spot attachment and enough dials, tables and instructions to fairly stifle any desire to take pictures.
Incident-light gathering attachments for SLRs are not new. But if any of these are still available on the market, I've not been able to find them. Anyway, they would certainly cost more than a piece of styrofoam cup!
While I suspect that most professional reflected-light proponents prefer manual or, at most, semi-automatic operation, the simplicity of incident-light metering blends perfectly with the automatic mode. Once you've taken a reading and made the necessary compensation, if any, you can shoot to vour heart's content without further attention to dials, scales or memory locks, since the meter in your automatic camera continues to monitor the slightest light change caused by voltage fluctuation, moving clouds, or the sun.
Also, the styrofoam dome attachment is amazingly convenient to use outdoors. Instead of carrying an expensive additional instrument or clumsy gray card, or having to approach your central subject, perhaps wading over a muddy field or stream or scrambling through a thorn patch, to photograph a confusing scene containing highly contrasting areas, one can just slip on the dome, face away from the subject towards camera position and take a reading. Then adjust, turn and shoot. Almost invariably you'll get a reading matching a complicated averaging of spot readings or one taken with the tricky gray card. And, in case you want to show preference to any light or dark areas, it's easy to compensate from the incident reading.
Great For Low Light
Incident-light metering is also useful in extremely weak light situations where neither the camera's reflected-light sensitivity nor that of the styrofoam dome is sufficient to register a usable reading. In this case you simply choose a thinner, more translucent material for constructing the dome. I use two layers from an ordinary plastic milk bottle instead of styrofoam. This transmits approximately four times as much light as the styrofoam, thus producing a reading of two full stops less than a reflected light reading from an 18 percent gray card. We then apply the 4X factor to the initial reading to determine proper exposure. For example, if the milk carton dome indicates 1/15 sec. at a given aperture, use ¼ sec. instead. However, since working in such low light conditions will often result in exposures longer than ¼ sec, we must also include the appropriate reciprocity factor. Suppose we are using Plus-X film and arrive at a working exposure of 2 sec. Then we would apply the reciprocity factor of 2X and actually expose 4 sec. Should the indicated exposure be 10 sec., the reciprocity factor becomes 4X, resulting in a full 40-sec. exposure. The factors for color film are somewhat smaller but definitely more critical.
A word of caution: Should you be tempted to use a material of even greater light transmission, remember that translucence is mandatory. If any image of the light source penetrates the light-absorbing dome, it could easily fall on an area of your metered frame which is considerably more or less sensitive than the overall center-weighted pattern designed for your camera. This would plainly result in a biased reading. The object of the dome is to measure the intensity of light falling upon the subject, regardless of the shape, size or proximity of the source.
—Kenneth D. O'Brien (Venice, Calif.)