Doug Kerr
Well-known member
The APEX system (Additive System of Photographic Exposure) provides for logarithmic expression of various quantities involved in the area of photographic exposure determination. Its original intent was to allow a photographer to "work the standard exposure equation" (perhaps working from arbitrary values of scene luminance for various situations) using only addition and subtraction.
The widespread availability of exposure meters (with their own little circular slide rules for solving the exposure equation) and the emergence of in-camera exposure meters and then fully-automatic exposure control systems made the system of little value to most photographers, and it never came into widespread use.
Still, the logarithmic notation is very intellectually attractive, and APEX notation is today often used in scientific papers and in standards relating to the matter of exposure. (And, for example, various quantities are encoded into Exif metadata in APEX form.)
But only one APEX value (the logarithmic expressions of quantities are all called "values" as a cue to their nature) survives in visibility to most photographers: exposure value (Ev). This actually describes photographic exposure, the joint effect on exposure of the shutter speed and the aperture (as an f/number).
There are many cases in which we would like to express a scene luminance in logarithmic terms, for which purpose we have the APEX value brightness value (Bv). This could be a convenient way to describe, for example, the lowest scene luminance that can be measured by a camera's automatic exposure control system.
But sadly, when camera manufacturers began to do this, they did not express that luminance in terms of Bv (likely concerned that few photographers has ever heard of Bv). Rather, they expressed it in "Ev". But Ev is not a measure of luminance. What can that mean?
Well, when a luminance is expressed as "Ev 5", that means "the illuminance for which the 'standard exposure equation', for an ISO speed of ISO 100, would recommend a photographic exposure (shutter speed and f-number) that would constitute Ev 5.
Ugh.
*************
But that's not really what I am here to speak of (I just like to bitch about it whenever I get the opportunity). But the background I gave is relevant to my real point here.
There is another distasteful curiosity built into the APEX system itself.
By "exposure equation" I mean an equation that takes the (measured average) scene luminance and ISO speed as inputs and delivers a recommended photographic exposure.
Well, how could there be a "correct" version of that equation, considering:
• Different distributions of luminance across the scene can dramatically change the "exposure result" of the image for any given measured average scene luminance and an ISO speed.
• Different photographers have different "strategies" as to how they want the exposure result to come out.
So of course there is no unique "correct" or even "ideal" version of the exposure equation.
This was recognized many years ago by the ISO standard for the "calibration" of free-standing exposure meters (which among other things, implies a certain exposure equation).
As a result, the standard essentially defines the exposure equation to include an arbitrary "calibration constant", and a wide range of values of that is "allowed". Thus the exposure meter manufacturer was free, over a fairly wide range, to implement the exposure equation it felt would be most beneficial to the users of its exposure meters.
Now, APEX emerges. Since its objective was to allow a user to "work" the exposure equation simply, we could hardly have it include a term for the calibration constant (and what value would the photographer use). So a fixed exposure equation was defined, thus:
Now Ev (exposure value) indicates the joint effect of Av and Tv thus:
But this was done in the context of the existing doctrine under which the exposure meter manufacturer could choose the exposure equation over a fair range of "calibrations". Doesn't the establishment of an exposure equation with no variable calibration factor (as just shown) foreclose that?
No.
Because the definition of the scale of Bv (what luminance, in cd/m^2, is "Bv 0") is not fixed, but involves - the exposure meter calibration constant (for the exposure meter involved).
Oh great!
How will this work out if we are using Bv in a scientific paper, where no exposure meter is involved (much less some particular one)?
Not well at all.
So, tables of the Bv equivalent of various luminance values are published that say nothing about the meter calibration constants, but in fact are predicated on some arbitrary choice of one (often "near the middle of the allowed range").
And the scientists (often unaware of the situation) work with that as if that were a stable definition of Bv (like the length of a millimeter).
Best regards,
Doug
The widespread availability of exposure meters (with their own little circular slide rules for solving the exposure equation) and the emergence of in-camera exposure meters and then fully-automatic exposure control systems made the system of little value to most photographers, and it never came into widespread use.
Still, the logarithmic notation is very intellectually attractive, and APEX notation is today often used in scientific papers and in standards relating to the matter of exposure. (And, for example, various quantities are encoded into Exif metadata in APEX form.)
But only one APEX value (the logarithmic expressions of quantities are all called "values" as a cue to their nature) survives in visibility to most photographers: exposure value (Ev). This actually describes photographic exposure, the joint effect on exposure of the shutter speed and the aperture (as an f/number).
Yes, it ignores lens transmission and bellows factor.
There are many cases in which we would like to express a scene luminance in logarithmic terms, for which purpose we have the APEX value brightness value (Bv). This could be a convenient way to describe, for example, the lowest scene luminance that can be measured by a camera's automatic exposure control system.
But sadly, when camera manufacturers began to do this, they did not express that luminance in terms of Bv (likely concerned that few photographers has ever heard of Bv). Rather, they expressed it in "Ev". But Ev is not a measure of luminance. What can that mean?
Well, when a luminance is expressed as "Ev 5", that means "the illuminance for which the 'standard exposure equation', for an ISO speed of ISO 100, would recommend a photographic exposure (shutter speed and f-number) that would constitute Ev 5.
Ugh.
*************
But that's not really what I am here to speak of (I just like to bitch about it whenever I get the opportunity). But the background I gave is relevant to my real point here.
There is another distasteful curiosity built into the APEX system itself.
By "exposure equation" I mean an equation that takes the (measured average) scene luminance and ISO speed as inputs and delivers a recommended photographic exposure.
Well, how could there be a "correct" version of that equation, considering:
• Different distributions of luminance across the scene can dramatically change the "exposure result" of the image for any given measured average scene luminance and an ISO speed.
• Different photographers have different "strategies" as to how they want the exposure result to come out.
So of course there is no unique "correct" or even "ideal" version of the exposure equation.
This was recognized many years ago by the ISO standard for the "calibration" of free-standing exposure meters (which among other things, implies a certain exposure equation).
As a result, the standard essentially defines the exposure equation to include an arbitrary "calibration constant", and a wide range of values of that is "allowed". Thus the exposure meter manufacturer was free, over a fairly wide range, to implement the exposure equation it felt would be most beneficial to the users of its exposure meters.
"Yes, I bought the Ajax Expo III. It gives me much nicer-looking pictures than I got with my Brandon Expowizard."
Now, APEX emerges. Since its objective was to allow a user to "work" the exposure equation simply, we could hardly have it include a term for the calibration constant (and what value would the photographer use). So a fixed exposure equation was defined, thus:
Av + Tv = Bv + Sv
where Av is the aperture value (describes the f-number), Tv is the time value (describes the exposure time, or "shutter speed"), Bv is the brightness value (describes the measured average scene luminance), and Sv is the speed value (describes the ISO speed of the film).Now Ev (exposure value) indicates the joint effect of Av and Tv thus:
Ev = Av + Tv
and so we can write the exposure equation this way:Ev = Bv + Sv
But this was done in the context of the existing doctrine under which the exposure meter manufacturer could choose the exposure equation over a fair range of "calibrations". Doesn't the establishment of an exposure equation with no variable calibration factor (as just shown) foreclose that?
No.
Because the definition of the scale of Bv (what luminance, in cd/m^2, is "Bv 0") is not fixed, but involves - the exposure meter calibration constant (for the exposure meter involved).
Oh great!
How will this work out if we are using Bv in a scientific paper, where no exposure meter is involved (much less some particular one)?
Not well at all.
So, tables of the Bv equivalent of various luminance values are published that say nothing about the meter calibration constants, but in fact are predicated on some arbitrary choice of one (often "near the middle of the allowed range").
And the scientists (often unaware of the situation) work with that as if that were a stable definition of Bv (like the length of a millimeter).
Best regards,
Doug