Gray Card Basics

[Stephen Benskin]

Kodak gray card instructions will tell you to angle the card so as to point one third of the way between your camera and the main light source. The gray card should be in the same light as your subject.

The exposure equation is based on a plane that is at an approximately 41 degree angle from the sun and parallel to the camera. It also assumes a Lambertian surface which isn't physically possible. Anyway, the sun has an illuminance at a normal angle of around 10200 footcandles. A normal angle is a surface or meter pointed directed at the sun. At a 41 degree angle, the illuminance is at 7680 footcandles. The old axiom of having the sun over the shoulder at 10am or 2pm comes from this because a surface perpendicular to the ground (ie a standing person) approximates the 41 degree angle under those conditions.

 

[Lee L]

This: http://www.largeformatphotography.info/articles/conrad-meter-cal.pdf is the source that Hannemyr references for his meter calibration information.

Lee

 

[panastasia]

Well ...
Only if you are looking to do more than the thing is good for.
To take the varying and unknown reflective properties of whatever happens to be in front of your reflected light meter out of the equation, it provides all you need.


The fact, by the way, that the amount it reflects changes with its orientation is not a problem. It would be a problem if it did not.
It is no more than a 'reflection' of the fact that light falling on your subject changes with its orientation too.
So it's rather good that it does! (And that with no more advanced technology than a coloured bit of cardboard. No batteries. Imagine that ... ) Dare i suggest that if that causes confusion in our heads, it's because we don't really understand how light 'works' on subjects that are not flat?

I think light reflecting off a glass surface is one of the strangest phenomenons. I just learned from a Physicist that the light at 55 deg angle of reflectance is 100% polarized.

I know some enjoy playing with the math formulas and believe they can really nail it down and be precise in their methods, but nature doesn't allow us to know how everything works exactly. It's still fun, though.

 

[Stephen Benskin]

This: http://www.largeformatphotography.info/articles/conrad-meter-cal.pdf is the source that Hannemyr references for his meter calibration information.

Lee

I just wrote a lengthy post about Conrad's article and lost it before I could post it. So, here's another try without as much detail.

Conrad is mistaken on a few topics. First, his assumption for the checkerboard having the white squares equal 90% reflectance. He then calculates an average reflectance of 18 - 19%. He might just be using this as an example, but I've seen people attempting to shoehorn figures in order to achieve a predetermined result before. For theoretical exposure determination the highlight is assumed to be 100% reflectance. The statistically average scene luminance range is 2.20. The distribution of the highlight is about 0.95 above the mean log luminance or midpoint and the shadow is 1.25 below. Add a stop flare to the shadow and that brings the shadow up to 0.95 below the mean and you get 0.95 above the mean and 0.95 below. That's just under 3 1/3 stops and 3 1/3 stops is the difference between the midpoint exposure and the speed point (coincidence? Nope).

He also claims that K is only a metric conversion factor. That's incorrect. It's a light loss constant much like q is in the major exposure equation - q*L*T / A^2 + Hf = H.

He is also mistaken in claiming Hg or midpoint exposure as 11.38Hm. It's 10Hm. The midpoint uses a constant "P" which is equal to 8 lux. The standard says that it is more accurately 8.2, and I've found it works best with my calculations at 8.11 but 8 is fine for most situations. Consider the Sunny 16 rule and the basic exposure equation H = E x t. P would equal E and t would be the 1/ISO of the Sunny 16 rule. For a 125 speed film, that would equal 0.064 lux seconds for Hg. Speed point exposure or Hm would be 0.0064 lux seconds for a 125 speed film or 10x (1.0 logs) less. All this is described in Connelly's paper, Calibration Levels of Films and Exposure Devices, which I referenced in an earlier post.

He is on the right track with his explanation as to why Zone System practitioners generally obtain effective film speeds 1/2 that of the ISO rating. He doesn't go into much detain and never mentions the important influence of flare.

His statement, "ANSI/ISO 6-1993 specifies a higher contrast than usually is desired for practical photographic use," doesn't consider how the standards contrast parameters infers the Delta-X Criterion, which is a mathematical approximation of the fractional gradient method. According to Delta-X, film speeds change little with changes in processing. In fact, when rounded to the nearest third stop, a given film/developer combination will produce a single film speed for a development range between -1 to +2.

Conrad's problem is that he is mostly using the standards as his source and not the papers from which the standards are derived. The purpose of the standards is to describe how to do something. How to determine film speed or how to calibrate an exposure meter. They don't explain the reason behind it. Generally, if there are major changes proposed, a paper with all the reasoning and theory will be published around the time the committee is considering a new standard. Here again are three of the key papers explaining exposure and exposure meters.

Connelly, D, Calibration Levels of Films and Exposure Devices, Journal of Photographic Science, Vol 16, 1968.

Nelson, C.N., Safety Factors in Camera Exposures, Photographic Science and Engineering, vol. 4, no. 1, Jan-Feb 1960.

Stimson, Allen, An Interpretation of Current Exposure Meter Technology, Photographic Science and Engineering, Vol. 6, No. 1, Jan-Feb 1962.

 

[Q.G.]

I don't like maths, so i'll just ask:
Is this a matter of trying to figure out mathematically what "middle" grey would be, while we have always been told that meters are set for the typical average scene these thingies are pointed at (i.e. no mathematics involved, except for finding a statistical mean)?

 

[Lee L]

Thanks Stephen. I expect that reading your sources would require a trip to a major research library or an interlibrary loan for most of us.

Lee

 

[Photo Engineer]

Well ...
Only if you are looking to do more than the thing is good for.
To take the varying and unknown reflective properties of whatever happens to be in front of your reflected light meter out of the equation, it provides all you need.


The fact, by the way, that the amount it reflects changes with its orientation is not a problem. It would be a problem if it did not.
It is no more than a 'reflection' of the fact that light falling on your subject changes with its orientation too.
So it's rather good that it does! (And that with no more advanced technology than a coloured bit of cardboard. No batteries. Imagine that ... ) Dare i suggest that if that causes confusion in our heads, it's because we don't really understand how light 'works' on subjects that are not flat?

Well, the real problem is that our eyes are linear and adjust their scale based on light level. The eye is also, by definition, linear with a response slope of 1.0. Film is not adjustable, and it has a non-linear response slope.

So, having more points helps usually as you can look at the exposure and the response both with a step scale. Kodak also makes a chart similar to the MacBeth checker but has a full page 18% gray card on the facing page.

PE

 

[Q.G.]

I see that your use for a card is as a test subject. Not to take a reading off.

Use it for the latter, and there is no real problem: it doesn't matter at all what our eyes do or don't, etc.

The cards are great to take the unknown reflective property of the bits in your scene out of the equation. You could do that with greater ease using an incident light meter. But the card works fine.

 

[Photo Engineer]

I use it for both!

PE

 

[Q.G.]

Then you'll know that the OP should not be confused by mentioning problems that do not exist when the card is used to take a reading off.

 

[Photo Engineer]

Then you'll know that the OP should not be confused by mentioning problems that do not exist when the card is used to take a reading off.

Good point.

I was directing my answers to some of the later posts which seemed to want to use the card to "derive" the zone system by various aperture sizes and etc. You can do all of this with a card and a checker.

We used the checker because there is an 18% gray in it, you can get your speed, and you can see white, black and contrast.

PE

 

[Q.G.]

I use(d) a Q 13 for that, of course*.

(*Eastman may have had an unnatural liking for the letter "K". I must admit i am quite partial to the letter "Q". )

 

[Stephen Benskin]

Thanks Stephen. I expect that reading your sources would require a trip to a major research library or an interlibrary loan for most of us.

Lee

Any major university should do. I got most of research in the stacks at the science library at UCLA. Although, I noticed that they were moving some of the specific photographic journals out of the stacks to their regional library which has shorter hours and where you can't browse. It appears they are yet another casualty of the digital age.

Those papers and the ones by Loyd Jones are well worth the effort to track down.

 

[Stephen Benskin]

I don't like maths, so i'll just ask:
Is this a matter of trying to figure out mathematically what "middle" grey would be, while we have always been told that meters are set for the typical average scene these thingies are pointed at (i.e. no mathematics involved, except for finding a statistical mean)?

Meters use math. There's a correlation between film speed and exposure placement. The meter reads luminance and not reflectance. It wants to place that luminance at a specific point on the film's curve ( 8/ISO). This point was determined based on analysis the physical world and through psychophysical testing. The only accurate way to describe it all is through math...and graphs.

 

[Stephen Benskin]

For a simple example of how K works as a light loss factor, take the exposure meter calibration equation.

A^2/T = B*S/K

Take away K for the moment and assume the Sunny 16 rule of f/16 at 1/ISO. This would cancel out T and S leaving only.

A^2 = B

16^2 = 256 so B = 256 footlamberts

This is how in a perfect mathematical world the reciprocal aspects of exposure would work, but you need to take into consideration how the lens isn't a perfect transmitter of light. There is some light loss as the light travels through the many lens elements. Therefore, in order to have the same amount of light passing through the real world lens which causes some of the light to be lost, you need to have the aim value be higher than the theoretical value. By calculating the many light loss variables, it was determined that that number is 1.16 times more than the mathematical model. So K = 1.16 and B = 297 footlamberts.

 

[Q.G.]

Meters use math. There's a correlation between film speed and exposure placement. The meter reads luminance and not reflectance. It wants to place that luminance at a specific point on the film's curve ( 8/ISO). This point was determined based on analysis the physical world and through psychophysical testing. The only accurate way to describe it all is through math...and graphs.

Well, i always learned that the calibration was not a mathematical exercise aimed towards a mathematical 'middle' value, but the result of analyzing pictures people took. (Much like that other thing people like to calculate to death: depth of field.)
Or rather, from the masses of pictures, a typical, 'normal' scene emerged.

The aim of meter calibration was (and is, isn't it?) to make it ensure the largest amount of 'well exposed' images with the least effort.
Science, yes. But not so much that of the maths and graphs type. Merely statistics.

 

[Stephen Benskin]

Well, i always learned that the calibration was not a mathematical exercise aimed towards a mathematical 'middle' value, but the result of analyzing pictures people took.
Or rather, from the masses of pictures, a typical, 'normal' scene emerged.

The aim of meter calibration was (and is, isn't it?) to make it ensure the largest amount of 'well exposed' images with the least effort.
Science, yes. But not so much that of the maths and graphs type. Merely statistics.

Yes, that's part of it and the most important part too. That's the psychophysical part I mentioned, and that part can be found in the papers by Jones which I also mentioned. It's a combination of the two, but even the psychophysical part can be described through math. And all of this it isn't a mathematical exercise but an mathematical explanation that reflects real world conditions.

There's the physical world with it's range of illuminances and luminance ranges. There's what we think makes a good photograph and how to define it, and then there's how to mesh the two. Film speed was originally determined through the psychophysical test known as the first excellent print test. In order for the results from the images people judging to make any sense, they had to have the technical data connected with it which includes scene luminance ranges and the placement of the scene on the curve and the resulting densities.

The first excellent print test ultimately gives you a number that can be plugged into the exposure equation (film speed). This equation takes into consideration the luminance of a subject, camera settings, influence of the lens, and veiling flare. The placement of the exposure on the characteristic curve is based on the statistical average scenes luminance range of 2.2 or 7 1/3 stops. The midtone is used to peg exposure and the highlights and shadows are assumed to fall within an acceptable range on the curve.

Everything is based on how to obtain a quality image within the largest range of conditions, but it also has to be quantified.

 

[Q.G.]

Sure.
But there is a huge difference between doing the many complicated thingies that were mentioned, and taking the mean of a typical scene's contrast range.

 

[Stephen Benskin]

Sure.
But there is a huge difference between doing the many complicated thingies that were mentioned, and taking the mean of a typical scene's contrast range.

The thingies aren't supposed to be done when shooting. They are the theory and reasoning behind it all in order to make it possible for everyone to just point and shoot.

 

[panastasia]

It doesn't work when a dramatic scene is constantly changing. You'll end up with only stories about the keepers that got away - like the fishermen have. And what about those wonderful happy accidents that surprise?

Retard (definition):

-verb (used with object)
1. to make slow; delay the development or progress of (an action, process, etc.) hinder or impede.

-verb (used without object)
2. to be delayed.

A quote by my EX: "JUST TAKE THE PICTURE FOR GOD SAKES!!!"

 

[Q.G.]

The thingies aren't supposed to be done when shooting. They are the theory and reasoning behind it all in order to make it possible for everyone to just point and shoot.

I'm not talking about a "when shooting" thingy.
I'm talking about how the middle grey value once (almost 100 years ago now) emerged.
Find the typical scene, determine it's range, and take a mean of that, and there it is: the "middle grey".

All that math is pertinent, and important, yes. But not quite how the value was produced.

 

[Stephen Benskin]

All that math is pertinent, and important, yes. But not quite how the value was produced.

And how was that? Are you talking about Munsell?

 

[Q.G.]

Ah, names ... Difficult.

Not Munsell, no.
Rather people like Hurter and Driffield, L.A. Jones, Mees' bunch at Kodak, and such.

 

[Stephen Benskin]

Ah, names ... Difficult.

Not Munsell, no.
Rather people like Hurter and Driffield, L.A. Jones, Mees' bunch at Kodak, and such.

I actually referenced Jones, and all the information I listed comes from the photographic scientific community. I'm sorry, I don't understand the point you are attempting to make.

 

[Chuck_P]

Newbie time!

I just bought an 18% gray card after my brother suggested to add it to my arsenal of photo supplies. I was wondering what situations would it be favorable to use a gray card to meter instead of metering off of the actual subject? Thanks.

The main thing you should remember about the use of a gray card is that it cannot compensate for shadow values in the scene, which is why it is best used when all reflective surfaces in the scene are equally illuminated. With the gray card, you are "placing" the exposure on a known middle value and then letting all other reflectances or luminance values "fall" on the gray scale relative to the luminance of the gray card. But the card can fail you totally in some shadow areas----and it may even fail you in other shadow areas that seem to have developed with some density but perhaps not enough to suit you. The shadows, as you know, are not illuminated equally.