Doug Kerr
Well-known member
Most "traditional" hand-held photographic exposure meters for incident light metering have a "light collector" in the form of a hemisphere or thereabouts. The story behind this is very convoluted, but here is a synopsis.
In cinematography, say as of the 1940s, incident light exposure metering was commonly used to get consistent exposures. But one lighting situation often encountered was "key-fill" lighting, where one light (the "key light") illuminated the subject (typically an actor's head) from some angle to the side and another light (the "fill light") illuminated the subject from essentially the camera position. Simple incident light metering did not deal well with that.
It had been determined empirically that to get an exposure from that situation that seemed visually equivalent to that obtained with "head on" lighting (only ) under conventional incident light metering, one could take two measurements with an incident light meter, one with the meter's flat receptor aimed at the key light and another with the receptor aimed at the fill light. Those two measurements (in terms of the actual meter observation, in illuminance units) were then averaged, and the average fed into the exposure calculator (really a specialized circular slide rule) on the exposure meter to get the rerecorded topographic exposure. But this was of course relatively tedious compared to the metering of a shot with a more basic lighting setup.
Don Norwood, a former military photographer and wannabe cinematographer, envisioned some instrument that would, with one measurement, recommend the ideal photographic exposure for a key-fill lit subject.
He believed that such an instrument would be an incident light exposure meter whose receptor was a hemisphere, or alternatively, which had the usual flat receptor covered by a translucent hemispherical dome. "light collector".
His reasoning for this was, to my mind, faulty. But nonetheless, it turned out that prototype meters built this was in fact seemed to give, with one measurement, a photographic exposure recommendation that much of the time seemed to be essentially "ideal".
Pressed to explain the theoretical basis for this, Norwood did experiments that showed, as we moved a single light source, initially at the camera position, to greater and greater angles to the side, the photographic exposure needed to give what was visually adjudged to be the same exposure result increased in direct proportion to that angle.
Now a problem with this is that as the light source moved farther to the side, the "shading" of the subject's face changed, so it is not clear how the observers would have even judged as "equally exposed" images with the light coming from different directions. But I press on.
Norwood characterizes this by saying that the "effectivity" of the light source in exposing the subject declined as the light moved farther to the side, inversely proportional with the angle.
Thus, he said, if the sensitivity of the exposure meter's light detection system declined that same way with the angle of the arriving light from the "axis" of the meter, it would automatically take into account this varying "effectivity" of all the light sources that may have been involved. He went on to say that, happily, the theoretical response of a meter with a hemispherical light collector followed exactly that relationship.
Now, less happily, the theoretical response of a hemispherical collector meter with angle is not that relationship. But the two do not differ
"greatly"
Eventually, workers in this area concluded that if a hemispherical collector incident light meter would, with one measurement, reliably give the "best" photographic exposure in the key-fill lighting of an actor's head, it would also deal ideally with whatever lighting situation might exist for other scenes.
There was apparently some empirical confirmation of this over the years.
In any case, most incident light photographic exposure meters have a hemispherical light collector, or some other clever collector exhibits perhaps that same directivity pattern.
Doug
In cinematography, say as of the 1940s, incident light exposure metering was commonly used to get consistent exposures. But one lighting situation often encountered was "key-fill" lighting, where one light (the "key light") illuminated the subject (typically an actor's head) from some angle to the side and another light (the "fill light") illuminated the subject from essentially the camera position. Simple incident light metering did not deal well with that.
It had been determined empirically that to get an exposure from that situation that seemed visually equivalent to that obtained with "head on" lighting (only ) under conventional incident light metering, one could take two measurements with an incident light meter, one with the meter's flat receptor aimed at the key light and another with the receptor aimed at the fill light. Those two measurements (in terms of the actual meter observation, in illuminance units) were then averaged, and the average fed into the exposure calculator (really a specialized circular slide rule) on the exposure meter to get the rerecorded topographic exposure. But this was of course relatively tedious compared to the metering of a shot with a more basic lighting setup.
Don Norwood, a former military photographer and wannabe cinematographer, envisioned some instrument that would, with one measurement, recommend the ideal photographic exposure for a key-fill lit subject.
He believed that such an instrument would be an incident light exposure meter whose receptor was a hemisphere, or alternatively, which had the usual flat receptor covered by a translucent hemispherical dome. "light collector".
His reasoning for this was, to my mind, faulty. But nonetheless, it turned out that prototype meters built this was in fact seemed to give, with one measurement, a photographic exposure recommendation that much of the time seemed to be essentially "ideal".
Pressed to explain the theoretical basis for this, Norwood did experiments that showed, as we moved a single light source, initially at the camera position, to greater and greater angles to the side, the photographic exposure needed to give what was visually adjudged to be the same exposure result increased in direct proportion to that angle.
Now a problem with this is that as the light source moved farther to the side, the "shading" of the subject's face changed, so it is not clear how the observers would have even judged as "equally exposed" images with the light coming from different directions. But I press on.
Norwood characterizes this by saying that the "effectivity" of the light source in exposing the subject declined as the light moved farther to the side, inversely proportional with the angle.
Thus, he said, if the sensitivity of the exposure meter's light detection system declined that same way with the angle of the arriving light from the "axis" of the meter, it would automatically take into account this varying "effectivity" of all the light sources that may have been involved. He went on to say that, happily, the theoretical response of a meter with a hemispherical light collector followed exactly that relationship.
Now, less happily, the theoretical response of a hemispherical collector meter with angle is not that relationship. But the two do not differ
"greatly"
Eventually, workers in this area concluded that if a hemispherical collector incident light meter would, with one measurement, reliably give the "best" photographic exposure in the key-fill lighting of an actor's head, it would also deal ideally with whatever lighting situation might exist for other scenes.
There was apparently some empirical confirmation of this over the years.
In any case, most incident light photographic exposure meters have a hemispherical light collector, or some other clever collector exhibits perhaps that same directivity pattern.
Doug