The definition of EV

BetterSense · May 22, 2009

Wikipedia says that EV is defined as

log_2(N^2/t)

Or the base-two log of the quotient of the square of N and t. N is an f-number and

t is the exposure time (shutter speed) in seconds

The problem I'm having is that their description of t is ambiguous. It's unclear to me weather t=15 means 15 seconds or 1/15s. So, if t *really is* an amount of time, measured in seconds, I'm not sure if the real formula is

log_2(N^2*t)
or
log_2(N^2/t)

It could be either of the above formulas depending on how t is defined.

I further read that

EV 0 corresponds to an exposure time of 1 s and a relative aperture of f/1.0.

Oh, really? For all film speeds? There is a missing element here. Is EV defined for 100-speed film or something?

BradS · May 22, 2009

BetterSense said:
Oh, really? For all film speeds? There is a missing element here. Is EV defined for 100-speed film or something?

yes. That is one half of the equation. The shutter sped and aperture.
That EV must be equal to the EV of the film speed plus the light intensity.

S + L = A +T

where:
S is film speed
L is light
A is aperature
T is time (shutter speed)

I have to run but we can discuss further later.

Look for a reply from HelenB as she will have it sorted correctly.

Christopher Walrath · May 22, 2009

Wirelessly posted (BlackBerry9000/4.6.0.167 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/102 UP.Link/6.3.0.0.0)

And in anticipation of a good aforementioned friend, the EV recommended by a light meter stays the same. It is only when you input film speed that changes to the EV recommendations occur. An EV is a set combination of s/s and aperture. Those recommendations are different for different ISO's but EV0 will always be 1' at f/1.

Q.G. · May 22, 2009

Oh, really? For all film speeds? There is a missing element here. Is EV defined for 100-speed film or something?

Yes: really.

The thing to understand first and foremost: EVs are INDEPENDENT of film speeds.

The thing to understand first of all is that the EV system is shorthand for aperture-shutter speed combinations, and one that works thus that any combination that produces the same exposure gets the same EV designation (or vice versa).

For instance, f/11 at 1/60 will always be f/11 at 1/60, thus always EV 13.
Not matter what ISO film you would like to use, or how bright your scene is.

What that "same exposure" is, whether it is right/wanted or not for any given situation is neither here nor there.
So no ISO values, no LUX values in EVs.
ONLY f-stops and shutterspeeds.

Then, secondly, the formula.
It makes use of the nominator. So 1/60 -> 60.
So in my example: log(11^2 * 60) / log(2) = log(121 * 60) / log(2) = log(7260) / log(2) = 12.83, or close enough to 13.

The fault (should be 13 exactly) is due to the rounded nature of both shutterspeeds and f-numbers.
In a series in which shutterspeeds halve or double, 1/60 is not included. It would be 1/64.
And f/11 is part of a series progressing by a factor sqr(2). So f/11 would be f/11.3137085
Use those numbers, and you'll get 13 exactly.

BetterSense · May 22, 2009

For instance, f/11 at 1/60 will always be f/11 at 1/60, thus always EV 13.
Not matter what ISO film you would like to use, or how bright your scene is.

Why is it then, that I constantly see charts that say that a bright sunny day is EV 15, open shadows EV 12, etc?

Is it that one uses EV 15 as the recommended exposure *explicitly if he is using 100 speed film*, and must correct if he is using a different film speed?

In other words, if I was using 200 speed film, then on a bright sunny day, I would actually use EV17?

I had though that EV was an expression of actual brightness.

So in my example: log(11^2 * 60) / log(2) = log(121 * 60) / log(2) = log(7260) / log(2) = 12.83

So it would seem that in the original formula, t is a conventional shutter speed (in Hz), and not the shutter-open-time in seconds.

Q.G. · May 22, 2009

BetterSense said:
Why is it then, that I constantly see charts that say that a bright sunny day is EV 15, open shadows EV 12, etc?

Is it that one uses EV 15 as the recommended exposure *explicitly if he is using 100 speed film*, and must correct if he is using a different film speed?

In other words, if I was using 200 speed film, then on a bright sunny day, I would actually use EV17?

That's it.
Almost, because when you would use EV15 when using ISO 100 film, you would use EV16 when using ISO 200 film. Not EV17.
EVs and stops are of equal size. So a 1 stop step in ISO speed results in a 1 EV step, i.e. EV15 becomes EV16.
Provided that you want to get the same exposure on both films, of course.

BetterSense said:
I had though that EV was an expression of actual brightness.

It is not, no.

Only when you provide additional information, i.e. the brightness of the scene (!), the ISO value of the film used, and how you like to have the result to be, that it becomes really usefull for when you want to know how the image you produce will come out.

But the EV itself is just an easier to remember shorthand for aperture-shutterspeed combinations. Nothing more.

So it would seem that in the original formula, t is a conventional shutter speed (in Hz), and not the shutter-open-time in seconds.

If you like to consider shutterspeeds to be conventionally expressed in Hz: yes, exactly that.

Christopher Walrath · May 22, 2009

EV is an abbreaviation for Exposure Value. The amount of exposure. An EV is a reciprocal set of exposure setting combinations. Lower number is more exposure than a higher number. Each step indicates one stop of exposure difference from the other. Another way is through considering Film Speed and Additive Light Values (the S and the L as mentioned above in the SLAT rule). So you can have light and sensitivity working together with duration and intensity of exposure to determine your exposure value.

BetterSense · May 22, 2009

Cool. Learn something every day.

Christopher Walrath · May 22, 2009

It took me a while. I kept wondering what the EV's were with a different film speed. For like a couple of years. Some things take me a while to wrap my head around.

Stephen Benskin · May 23, 2009

In ANSI PH3.49 - 1971 Exposure values can refer to one of three different equations:

2^Ev = A^2 / T

reflected meter
2^Ev = B*S / K

incident meter
2^Ev = I*S/ C

A = f/number
T = shutter speed
B = luminance
K = constant
I = illuminance
S = film speed
C = Constant

For reflected light meters A^2 / T = B*S / K which is the reciprocal relationship between the scene and the camera settings. Since a meter reads the scene, it tends to use the B*S/K equation and then the meter's calculator uses the A^2/T equation. I have a Pentax Spot meter, and in its user manual it states that its Ev numbers to determine the scene are based on a 100 film speed.

In ISO 2720 - 1974 (R1994) Ev seems to only refer to 2^Ev = A^2 / t.

Q.G. · May 23, 2009

Christopher Walrath said:
Another way is through considering Film Speed and Additive Light Values (the S and the L as mentioned above in the SLAT rule). So you can have light and sensitivity working together with duration and intensity of exposure to determine your exposure value.

When trying to understand the Exposure Value (note the capitals) it is best not to say "to determine your exposure value".
We would better distinguish between the Exposure Value, and the value of the exposure.
(It's complicated enough, which is also shown by the quote of the two ANSI standards.)

Duration and exposure combined consitute the Exposure Value (note the capitals). And these two alone constitute the EV.

Combine this EV with the other two, amount of light present and the film speed, and you get the value of the exposure.

You can express the Exposure Value in Lux per second, producing its yet unknown value of exposure, and still have not calculated in the light present or the sensitivity of the film.
What you however do know already when you combine f-stops and shutterspeeds is that any other combination of having the same EV will produce an equal exposure in Lux per second. No matter that we still do not know how many Lux per second.

That is why the EV system was invented: a single number to describe a large number of possible combinations. Shorthand.
Useful even if you don't have a clue about how much light there is, or how much light your film needs.

It's only when you add in light and sensitivity, combine Luxs and seconds with EV, and processing parameters, that you produce something visible: density.

Doremus Scudder · May 24, 2009

OK, Q.G., you've confused me a bit here. Maybe you can clarify my misunderstanding.

First, my Pentax spot meter gives me a different number for different subject brightnesses. I assumed this was the Exposure Value and was dependent on the amount of light (Lux/sec) being "seen" by the meter. This was then translated into an aperture/shutter speed combination depending on film speed by means of the calculator dial. Certainly, the numbers my meter gives me are dependent on the amount of light reaching the meter's sensor; more light = higher numerical value; less light = lower numerical value.

If Exposure Value is ONLY defined by aperture/shutter speed combination, then f/11 at 1/60 in bright sunlight means a lot more light reaching the film than f/11 at 1/60 at night (or in total darkness, in which case the light reaching the film would be zero...), which would be different meter readings as far as I understand. My meter gives me a number based on the amount of light it sees, which does not correspond to what I understand your definition of Exposure Value to be.

Does this mean that the numbers I read in my meter's viewfinder are not "Exposure Value" numbers? Are there any Exposure Value numbers in this system at all?

TIA

Doremus Scudder
www.DoremusScudder.com

Q.G. · May 24, 2009

Doremus Scudder said:
If Exposure Value is ONLY defined by aperture/shutter speed combination, then f/11 at 1/60 in bright sunlight means a lot more light reaching the film than f/11 at 1/60 at night (or in total darkness, in which case the light reaching the film would be zero...), which would be different meter readings as far as I understand.

That's exactly right.

If f/11 at 1/60 would give proper exposure in bright sunlight, it obviously would not at night.
So though the EV number indicates the same aperture - shutterspeed combination, it obviously is not the one to use at night.
And accordingly, a meter will indeed not indicate the same EV in both situations.

Doremus Scudder said:
My meter gives me a number based on the amount of light it sees, which does not correspond to what I understand your definition of Exposure Value to be.

It does.
The confusion possibly is between what a particular EV stands for on the one hand, and what particular EV to set to get the proper exposure for a given situation on the other hand.

What your meter does is suggest what EV to use, depending on how much light it sees. It does the latter of the two mentioned above.
The number, i.e. the EV value suggested for proper exposure, should be higher when there is more light, lower when there is less light.

As mentioned before, EV numbers are shorthand for f-stop - shutterspeed combinations.
What f-stop - shutterspeed combinations are described by, say, EV16 is a fixed thing. What f-stop - shutterspeed combination to use, what EV to use for any given situation depends on film speed and light level.

Doremus Scudder said:
Does this mean that the numbers I read in my meter's viewfinder are not "Exposure Value" numbers? Are there any Exposure Value numbers in this system at all?

Some meters will employ similar single numbers to transfer a reading to a more elaborate scale. And some do not adhere to the 'standard' EV system in doing so, so they would not be proper EV numbers. Even when they put an 'EV' next to the scale.

As far as i can tell from the manual, the EV numbers Pentax use are the same as the ones described here. (They show an example of EV12 matching up to - among other combinations - f/8 to 1/60. And you can use the formula mentioned earlier to see that that is correct, a 'proper' EV value.).
But i don't know the Pentax well enough to be sure.

All Gossen meters, to name another brand, do use the standard EV system.

Lee L · May 24, 2009

Here are a couple of articles that might help.

http://camerapedia.org/wiki/Light_and_Exposure_Values_(LV_&_EV)
http://www.earthboundlight.com/phototips/exposure-stops-here.html

Lee

Stephen Benskin · May 24, 2009

It's hard, if not impossible, to be absolutely positive when interpreting the standards intentions. The latest standard just has 2^Ev = A^2/t, but this could mean that for the purposes of the standard, which is as a "Guide to Product Specification", that is all that is required. Standards rarely include theory. Older standards and scientific papers have two additional equations that equal 2Ev. This may not mean that the term Ev can be used with them, but it does show the relationship between them. It also means you can determine the luminance of a subject by knowing its Ev.

In Allen Stimson's paper "An Interpretation of Current Exposure Meter Technology," for the Journal of Photographic Science and Engineering, he has an equation:

S*B / K = 2^Ev = A^2 / t

So, perhaps, the official Ev is the one that represents the reciprocal relationship between the f/number of shutter speed, but it doesn't preclude it being used to calculate luminance values.

I've attached a page from the 1971 standard.

Q.G. · May 24, 2009

I think the entire confusion rises because of the EVs being "exposure" values.
They are. And as such naturally connected to anything else that has to do with exposure.

There is one important bit that is not in the formula, and that is our expectation that we do use calculations to get a certain result.

As a shorthand for f-number - shutterspeed combinations, there is nothing in them that is linked to a desired or expected results.
Just like f-numbers are just rather round-about ways of saying how big the (effective) hole in a lens is, without any reference to how big we need or like it to be, EV numbers are just a measure for 'how big' groups of f-number - shutterspeed combinations are.

Would of course all be pointless if we didn't relate them to other things, putting them to good use to arrive at the desired result

And that results is not just an exposure of our films, but the 'proper' exposure of the film, given a few other parameters as well.
And given that the EVs and those other parameters are linked by that desired or expected result, the proper exposure of our films, you can of course work back from there and, for instance, use EVs to calculate luminance values.

But only (!) given that assumption that the EV number itself is indeed the one that gives a 'proper' exposure.
EV numbers by themselves do not (!) contain any information at all about illuminance values, nor film sensitivities.

Lightproof · May 24, 2009

The trick is, as posted above, that the EV is equatable by several terms. Thus, we can use the whole thing to compress a lot of information in a single number.
The light meter puts out an EV, taking into account film speed and the amount of available light. When transferred to a camera capable of EV-coupling, the single number is decompressed to the maximum number of solutions. This maximum is confined by the total range of aperture and time scale.

If there is no EV coupling on the camera, the information can be decompressed by looking at a EV table.

Ian David · May 24, 2009

Q.G. said:
As far as i can tell from the manual, the EV numbers Pentax use are the same as the ones described here. (They show an example of EV12 matching up to - among other combinations - f/8 to 1/60. And you can use the formula mentioned earlier to see that that is correct, a 'proper' EV value.).
But i don't know the Pentax well enough to be sure.

Yes, the Pentax uses the normal EV numbers. The digital readout always assumes ISO 100.

Stephen Benskin · May 24, 2009

iandavid said:
Yes, the Pentax uses the normal EV numbers. The digital readout always assumes ISO 100.

The Pentax users manual contains an error. It has K = 14 (cd/m^2) or 1.4 (cd/f^2) suggesting a conversion factor of 10 except that it should be 10.76. K should then equal 1.3 and not 1.4. So, if you plan to play with the equations for a Pentax digital spot meter use K = 1.3.

BetterSense · May 24, 2009

Yes, the Pentax uses the normal EV numbers. The digital readout always assumes ISO 100

This is probably where my confusion stems from. I have used several times over, even had one taped to my camera, exposure charts that say that daylight is EV15. The assumption I did not know about was that it was EV15 if you are running 100 speed film. I also didn't know that EV was defined with the equation I posted...I thought it was just a scale of luminance values, separated by one stop. Since EV0 is "pretty dark" it made sense to me. So even though I was using this exposure chart and it was working, I was doing it wrong. This didn't get me into any trouble at all because I knew sunlight (15) was "Sunny 16" and so I auto-compensated for film speed...if the chart said EV 13, I just added two stops of exposure to whatever Sunny 16 would give me.

Now that I know what EV really means, I'm not really going to change anything. I've been shooting with meterless cameras more and more, and getting more comfortable with guessing brightness values (and printing poorly exposed negatives). I have since lost my exposure chart; does anyone know of a source of another one? Basically I know that daylight is sunny 16, hazy daylight needs +1 stop, overcast +2 stops, daylight shadows a good +3 stops, but for indoors or night photography the seat of my pants isn't very helpful.

Ian David · May 24, 2009

BetterSense said:
I also didn't know that EV was defined with the equation I posted...

For all practical, as opposed to academic or curiosity, purposes, you don't need to know anything about the equation in order to use EVs in your photography.

BetterSense said:
I have since lost my exposure chart; does anyone know of a source of another one? Basically I know that daylight is sunny 16, hazy daylight needs +1 stop, overcast +2 stops, daylight shadows a good +3 stops, but for indoors or night photography the seat of my pants isn't very helpful.

Somebody else here recently posted a link to the following chart, which looks like the sort of thing you are talking about:
http://www.fredparker.com/ultexp1.htm#Light Intensity Chart

Doremus Scudder · May 25, 2009

Q.G. and Lee (and others too),

Thanks for the explanation and the links to the articles. As I understand now, EV only links to luminance values when a light meter tells you that a certain EV is necessary to "properly" expose ISO 100 film. What the meter is doing is determining the brightness, and "converting" it into a suggestion for a series of aperture/shutter speed combinations for ISO 100. The number itself represents the series of combinations, but, knowing the parameters the meter uses, light values can be deduced from the number by using equations or tables. Most of us, however, simply transfer the number, as if it were a luminance value, to the calculating ring of the meter to determine exposure. (Do correct me if I'm wrong...)

Best,

Doremus Scudder
www.DoremusScudder.com

Q.G. · May 25, 2009

You're absolutely right.

(An additional use of EV readings: some of us transfer the EV numbers to our leaf shutter lenses that have EV-scales on them.
The aperture and shutterspeed scales on these same lenses, plus a mechanism that links both, then allows us to select the aperture or shutterspeed we like to use, while when doing so the shutterspeed or aperture changes accordingly, keeping the EV constant.
Very useful.)

The definition of EV

Privacy & Transparency

Privacy & Transparency