More fuel on the fire: Are light meters set for 12% vs 18%?

[administrator]

This is from Wikipedia:

Before you delve in, allow me to offer my take on it:

1) Actual exposure meter calibration does not take account of reflectance.

2) Reflectance is implied by the K and C factors used to calibrate the meter.

3) The K and C values that reflect the real world better result in an implied reflectance of approximately 13%. Other K and C values are also used.

Before you react to this earth-shaking information, please read the following from wikipedia:

Exposure meter calibration
In most cases, an incident-light meter will cause a medium tone to be recorded as a medium tone, and a reflected-light meter will cause whatever is metered to be recorded as a medium tone. What constitutes a “medium tone” depends on meter calibration and several other factors, including film processing or digital image conversion.

Meter calibration establishes the relationship between subject lighting and recommended camera settings. The calibration of photographic light meters is covered by ISO 2720:1974.

[edit] Exposure equations
For reflected-light meters, camera settings are related to ISO speed and subject luminance by the reflected-light exposure equation:


where

N is the relative aperture (f-number)
t is the exposure time (“shutter speed”) in seconds
L is the average scene luminance
S is the ISO arithmetic speed
K is the reflected-light meter calibration constant
For incident-light meters, camera settings are related to ISO speed and subject illuminance by the incident-light exposure equation:


where

E is the illuminance
C is the incident-light meter calibration constant
[edit] Calibration constants
Determination of calibration constants has been largely subjective; ISO 2720:1974 states that

The constants K and C shall be chosen by statistical analysis of the results of a large number of tests carried out to determine the acceptability to a large number of observers, of a number of photographs, for which the exposure was known, obtained under various conditions of subject manner and over a range of luminances.

In practice, the variation of the calibration constants among manufacturers is considerably less than this statement might imply, and values have changed little since the early 1970s.

ISO 2720:1974 recommends a range for K of 10.6 to 13.4 with luminance in cd/m². Two values for K are in common use: 12.5 (Canon, Nikon, and Sekonic[1]) and 14 (Kenko[2] and Pentax); the difference between the two values is approximately 1/6 EV.

The earliest calibration standards were developed for use with wide-angle averaging reflected-light meters (Jones and Condit 1941). Although wide-angle average metering has largely given way to other metering sensitivity patterns (e.g., spot, center-weighted, and multi-segment), the values for K determined for wide-angle averaging meters have remained.

The incident-light calibration constant depends on the type of light receptor. Two receptor types are common: flat (cosine-responding) and hemispherical (cardioid-responding). With a flat receptor, ISO 2720:1974 recommends a range for C of 240 to 400 with illuminance in lux; a value of 250 is commonly used. A flat receptor typically is used for measurement of lighting ratios, for measurement of illuminance, and occasionally, for determining exposure for a flat subject.

For determining practical photographic exposure, a hemispherical receptor has proven more effective. Don Norwood, inventor of incident-light exposure meter with a hemispherical receptor, thought that a sphere was a reasonable representation of a photographic subject. According to his patent (Norwood 1938), the objective was

to provide an exposure meter which is substantially uniformly responsive to light incident upon the photographic subject from practically all directions which would result in the reflection of light to the camera or other photographic register.

and the meter provided for "measurement of the effective illumination obtaining at the position of the subject."

With a hemispherical receptor, ISO 2720:1974 recommends a range for C of 320 to 540 with illuminance in lux; in practice, values typically are between 320 (Minolta) and 340 (Sekonic). The relative responses of flat and hemispherical receptors depend upon the number and type of light sources; when each receptor is pointed at a small light source, a hemispherical receptor with C = 330 will indicate an exposure approximately 0.40 step greater than that indicated by a flat receptor with C = 250. With a slightly revised definition of illuminance, measurements with a hemispherical receptor indicate “effective scene illuminance.”

[edit] Calibrated reflectance
It is commonly stated that reflected-light meters are calibrated to an 18% reflectance,[3] but the calibration has nothing to do with reflectance, as should be evident from the exposure formulas. However, some notion of reflectance is implied by a comparison of incident- and reflected-light meter calibration.

Combining the reflected-light and incident-light exposure equations and rearranging gives


Reflectance R is defined as


A uniform perfect diffuser (i.e., one following Lambert's cosine law) of luminance L emits a flux density of πL; reflectance then is


Illuminance is measured with a flat receptor. It is straightforward to compare an incident-light measurement using a flat receptor with a reflected-light measurement of a uniformly illuminated flat surface of constant reflectance. Using values of 12.5 for K and 250 for C gives


With a K of 14, the reflectance would be 17.6%, close to that of a standard 18% neutral test card. In theory, an incident-light measurement should agree with a reflected-light measurement of a test card of suitable reflectance that is perpendicular to the direction to the meter. However, a test card seldom is a uniform diffuser, so incident- and reflected-light measurements might differ slightly.

In a typical scene, many elements are not flat and are at various orientations to the camera, so that for practical photography, a hemispherical receptor usually has proven more effective for determining exposure. Using values of 12.5 for K and 330 for C gives


With a slightly revised definition of reflectance, this result can be taken as indicating that the average scene reflectance is approximately 12%. A typical scene includes shaded areas as well as areas that receive direct illumination, and a wide-angle averaging reflected-light meter responds to these differences in illumination as well as differing reflectances of various scene elements. Average scene reflectance then would be


where “effective scene illuminance” is that measured by a meter with a hemispherical receptor.

ISO 2720:1974 calls for reflected-light calibration to be measured by aiming the receptor at a transilluminated diffuse surface, and for incident-light calibration to be measured by aiming the receptor at a point source in a darkened room. For a perfectly diffusing test card and perfectly diffusing flat receptor, the comparison between a reflected-light measurement and an incident-light measurement is valid for any position of the light source. However, the response of a hemispherical receptor to an off-axis light source is approximately that of a cardioid rather than a cosine, so the 12% “reflectance” determined for an incident-light meter with a hemispherical receptor is valid only when the light source is on the receptor axis.