• 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!

On the ISO SOS and ISO REI metrics for sensitivity

Doug Kerr

Well-known member
We often see today in the specifications for digital cameras that the "sensitivity" is presented in terms of the metric(s) ISO SOS and/or ISO REI. What does that mean?

It is a long story, and I will start with some background.

Classical reflected light exposure metering

In what I call classical reflected light exposure metering (whether with a free-standing exposure meter or via an inbuilt automatic exposure system) the "instrument" measures the average scene luminance.

From that, plus the "sensitivity" of the film or digital imaging system, a photographic exposure (combination of shutter speed and aperture) is recommended or put into effect.

This is quite a leap of faith, given the variations in scene structure and the existence of various "exposure strategies" the photographer might want to follow. But it is simple to do, and for many years it was the basic for much exposure planning.

Now assume the following (we will assume a digital camera):

• The "calibration" of the automatic exposure system (we will work with that rather than a free-standing meter as the standards are more "singular") is per the applicable ISO standard.

• The ISO speed of the digital imaging system (an input to the equation of the automatic exposure system) is determined per the applicable ISO standard (using the "saturation" basis).

Now imagine that we shoot a scene with such a camera in which the average luminance is 18% of the peak (highlight) luminance (the infamous "typical scene".

Then, for a "metered" exposure, the result on the sensor will be that the photometric exposure for a highlight area will be 70.7% of the saturation photometric exposure ("1/2 stop below it"). The average photometric exposure will be 12.8% of the saturation photometric exposure.

A collateral implication is that if we take a metered exposure of a uniform-luminance scene (perhaps a frame-filling gray card), the photometric exposure anywhere on the sensor is 12.8% of the saturation photometric exposure.

You were expecting maybe 18%? Well, 12.8% is just about 18% of 70.7% of the saturation photometric exposure.​

The "1/2 stop" distance is often called the exposure metering "headroom", a term borrowed from a similar matter in sound recording. Its purpose is to avert overexposure in the event that, for the actual scene being shot, the average luminance is less than 18% of the peak (highlight) luminance. (It's not enough if the average luminance is below 12.8% of the peak luminance.)

Intelligent metering

Of course modern cameras use a much more sophisticated approach to automatic exposure control, which involves measuring the scene luminance at multiple points (maybe even at every pixel location), and concluding from that what the range of scene luminance is (at least over the regions that we are probably interested in).

Typically, the "calibration" of such a system is stated by consideration of the photometric exposure that would result from a uniform luminance scene. If we made that consistent with the "calibration" of a "traditional" automatic exposure system , one possible implication would be that in a hypothetical "typical" scene, the photometric exposure for highlights would be 1/2 stop below the saturation photometric exposure.

But this is a 1/2 stop waste of the "range" of the sensor system. No "headroom" is required, since the exposure control system deals with the actual (at least intelligently estimated) range of scene illuminance, not some simplistically-assumed range.

Getting back the "headroom"

Of course, we can readily reclaim the "headroom" by just making a 1/2 stop change in the calibration of the exposure control system. But there are some subtle downsides to that. One is that if we now use an external exposure meter (calibrated per the ISO standard for such) with the camera, for a "uniform luminance scene", the photographic exposure it will recommend will be different from that which the camera would adopt. This might make users think that the automatic exposure system in the camera was "miscalibrated".

So a crafty ploy is used. The "rating" of the various sensitivity settings in the camera, as ISO speed, is understated by 1/2 stop (compared to the ISO definition). This shifts the photographic exposure that would be set by the camera (or recommended by an external exposure meter) "hotter" by 1/2 stop, thus "burning the headroom".

Exposure index

Before I proceed with the story, I must speak of another term that will soon be part of the story.

When exposure metering first came into existence, there was of course no established numerical way to express the sensitivity of a certain film type. Thus, exposure meter manufacturers (such as Weston) devised there own premises and scales of sensitivity, and then did research to determine what the value of that metric was for available film types. These were listed in a pamphlet the user got with the exposure meter.

In many cases, this metric was called the exposure index of the film.

Eventually, standard schemes of rating film sensitivity were established, such as (in the US) by the American Standards Association. The metric was labeled "ASA film speed". "Speed" was a common colloquialism for film sensitivity, since, for a given scene luminance and a given aperture, the greater the sensitivity, the faster would be the needed shutter speed.

The ASA sensitivity system eventually was embraced by an international standard system, where the metric is called the "ISO speed" of the film.

Today, "exposure index" has taken on a new meaning: "what we tell the metering system is the ISO speed of the film or sensor". What does that mean?

Suppose we are shooting with film rated at ISO 200, but we recognize that the lighting situation is such that we would like to decrease the photographic exposure from what the meter will recommend by 1/2 stop (what today we would usually call -1/2 stop of exposure compensation - the formal term is exposure bias).

There is no "exposure compensation" setting on our meter, but we set the "ISO dial" to ISO 280. That will cause the meter to issue a 1/2 stop less photographic exposure recommendation that would be normal.

So the little dial marked "ISO" is where we set, not necessarily the ISO speed of the film, but rather the "exposure index" at which we want the exposure reckoned.

Back to the story

Before our little excursion into the significance of "exposure index", I mentioned that, in order to "burn the headroom" of the traditional exposure metering equation when we were using "intelligent" metering systems, we could understate the ISO sensitivity of the system that was fed to the inbuilt automatic exposure system (or set into an external light meter as the "exposure index"). But in fact that meant lack of compliance with the established ISO definition of ISO speed, at the least intellectually troublesome, and with some implications about understating the low-light potential of the camera.

So a few years ago, the ISO added to the standard for the "sensitivity" of digital camera imaging systems two new metrics (in addition to ISO speed):

• The ISO Standard Output Sensitivity (ISO SOS). For all practical purposes, this metric is determined the same way as the ISO speed using the "saturation" basis, but will have a value that is 70.4% of the ISO speed (almost exactly "1/2 stop lower).

If the ISO SOS value is fed to the automatic exposure system, or used as the exposure index into a free-standing exposure meter, then the "1/2 stop bump" in photographic exposure, needed to "burn the headroom", will occur.

• The ISO Recommended Exposure Index (ISO REI). This value can be chosen by the camera manufacturer, at its discretion, for use an an exposure index to attain an arbitrary bump in the photographic exposure compared to that based in the ISO speed which the manufacturer feels will be most useful for typical photographic tasks. There is no technical definition. It is to be chosen for the desired result.

If the camera manufacturer wants to adopt the exact "bump" in photometric exposure implied by the ISO SOS value, he can make that clear by stating the determined ISO SOS value and then stating the same number as the ISO REI value.

Thus, if for one position of the "ISO" control on the camera, the manufacturer says, in the specifications, that is is "ISO 400 (SOS/REI)", he is saying that:

• The ISO speed at that setting is about ISO 568.

• The automatic exposure system in this camera will choose the photographic exposure according to the industry standard exposure equation for ISO speed of ISO 568.

• It is recommended that, in using a free-standing exposure meter, you set the "ISO" setting (exposure index) to ISO 568.


Best regards,