• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname


    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Emulating incident light metering

Doug Kerr

Well-known member
Any complete discussion of exposure metering requires discussion of the "exposure strategy"; that is, how do we want the various scene elements, having different luminances, to be mapped onto the "photometric exposure range" of our camera.

Photometric exposure is the physical phenomenon to which our sensor, for the most part, responds. At any point on the sensor, the phtometric exposure is the product of the illuminance on that point and the exposure time (the time the illuminance persists).​

Two strategies are widely recognized:

A. "Expose to the right". Here we aspire to have the "brightest" scene element receive a phtometric exposure that is at (or almost at) "saturation", the highest phtometric exposure to which the sensor can properly respond.

It is sometimes said, in discussing a "downside" of this strategy, that "It results in a photo of a gray cat on an ash heap looking like a photo of a white cat on a snowdrift".​

B. (I won't give this a name yet.) Here we aspire to have each scene element receive a phtometric exposure that, as a fraction of the saturation phtometric exposure, is equal to the reflectance of the element.
That is, if a scene element has a reflectance of 56%, then it would get a phtometric exposure of 56% of the saturation phtometric exposure.​

It turns out that, for all practical purposes, this strategy is the same as the underlying premise of the famous "Zone System" for planning photographic exposure.

It can be said of this strategy that "It results in a photo of a gray cat on an ash heap looking like a photo of a gray cat on an ash heap."​

No simplistic exposure metering system can follow strategy A. Classical "reflected light" metering (as practiced by classical free-standing exposure meters does not practice either of those strategies.

While in our "cat analogy" mood, it can be said of classical reflected light exposure metering that "It results in a photo of a black cat on a coal pile, or a gray cat on an ash heap, or a white cat on a snowdrift, looking like a photo of a gray cat on an ash heap."​

Modern "intelligent" reflected light exposure metering systems (those integrated into modern cameras) improve on the classical model, more consistently practicing an exposure strategy that can't be simply described. And in any case, my real point here is not about reflected light metering.

Rather, my point here revolves around incident light exposure metering. It turns out that its claim to fame is that it allows us to reliably practice exposure strategy B.

It works by measuring the illuminance of the incident light upon the scene, and from that (and its knowledge of the ISO speed of the camera) makes a recommendation as to photographic exposure (the combination of aperture and shutter speed). This type of metering has for many years been beloved to the field of cinematography, but is also widely used in still photography.

Many exposure meters have an "incident light" mode, so they can be used to practice this mete4ring mode when we feel it will lead to the result we seek (typically, under "strategy B").

But there is a handy alternative. We can place a test target with a "diffuse" surface and a certain reflectance in the scene and then have our camera's integrated exposure metering system observe that test target and from that choose and set a photographic exposure.

But what should that "certain reflectance" of the test target be? For many years, it has been common to use a test target with a nominal reflectance of 18%. The historical reason for that is beyond the scope of this note.

Is that reflectance "appropriate". Well, to assess that, we must make certain assumptions.

We will start with this set of the most prominent assumptions:

• The "calibration" of the camera's integral exposure metering system is consistent with the ISO standard for such systems.

That standard of course is predicated on a "classical" metering scheme, which operates upon the measured average luminance of the entire "frame". Of course, modern cameras use more sophisticated forms is reflected-light metering, which work on numerous luminance measurements made across the frame. But typically these are calibrated so that in the case of a "scene" of uniform luminance (typically a frame-filling test target), the chosen photographic exposure would be consistent with that from a "classical" exposure metering system observing the same scene.​

• The reckoning of the ISO sensitivity of the camera (a parameter fed to the metering algorithm as the "exposure index") is consistent with the ISO definition for ISO speed.

• We have in mind to fulfill exposure strategy B.

Then, it turns out that, theoretically, we need to use a target ("gray card") with a reflectance of 12.8%.

But of course, the common "gray card" has a nominal reflectance of 18%. So basically, we need to meter on such a gray card and then use a photographic exposure i/2 stop "hotter" than the metering system, in its normal state of mind, would enact.

We can perhaps do that by setting an exposure compensation of +1/2 Ev unit.

In fact, during some eras, the instructions accompanying a famous "gray card", made by Eastman Kodak, suggested a 1/2-stop "bump" in the exposure over that indicated by an exposure meter.

But today a new wrinkle has come into the picture. Very commonly, the sensitivity of the camera sensor is reckoned in terms not of ISO speed but rather in terms of ISO Standard Output Sensitivity (ISO SOS), which basically is "1/2 stop" less than the ISO speed.

The result (in fact, the object) of this is to make a reflected light metered exposure be 1/2 stop "hotter" than under the classical doctrine. The concept is to "eat the headroom" that is included in the classical doctrine to cover cases where the distribution of scene brightness would lead to overexposure of the highlights. (The premise is that the more "sophisticated" metering procedures take care of that matter directly, and "don't need no stinkin' headroom" to avert overexposure.)

But the basis of our emulated incident light metering did not include any concept of headroom. So this change in how the exposure index is reckoned in modern cameras can screw up the whole story.

Accordingly now, again holding to our trail of assumptions, when emulating incident light metering by using reflected light metering on a "gray card", using a gray card with a nominal reflectance of 18%, and using "normal" exposure metering practice (no "bump"), is a very reasonable place to start.

Amazing how reality can catch up to folklore!

Best regards,


Asher Kelman

OPF Owner/Editor-in-Chief
Explain why the analogy of "hotter" when using a 12% reflectance card is apt. I would have thought that reflecting less it would be "colder"!

You are not, after all, referring to color temperature but just the flux of light in a certain time interval.


Doug Kerr

Well-known member
Hi, Asher,

Explain why the analogy of "hotter" when using a 12% reflectance card is apt. I would have thought that reflecting less it would be "colder"!

You are not, after all, referring to color temperature but just the flux of light in a certain time interval.


Recall that the scenario here is:

1. We have the camera regard the test card and conclude what photographic exposure it would use to photograph the card (which is what the camera thinks we are getting ready to do).

2. We then use that same photographic exposure for the actual shot of the scene.

If we reduce the reflectance of the card, then it will be less bright (under the illumination on the scene, whatever that is) and this, in step 1, will cause the camera to conclude that it will need a greater photographic exposure to properly expose a shot of the card.

Then, in step 2, we use this greater photographic exposure to take the actual shot, leading to a "hotter" phtometric exposure.

Best regards,