Actually, the facts are in the standard and in the spec sheets of the exposure meters, as I already mentioned. Since you have all the standards, check ISO 2720:1974 and look for the calibration constants 'K' for reflection meters and 'C' for incident meters. You will find that the standard suggests a fairly large range for both, 10.6-13.4 and 240-400, respectively. As you may know Sekonic is using a K-factor of 12.5, but Minolta and Pentax are using 14.
As you can see, the standard has a wide range and major manufacturers don't even stick to it. You may call this an outdated standard (1974) but keep in mind that many currently used exposure meters are from that time or were designed with this standard being up-to-date.
Wikepedia suggests an approximation to correlate K to reflectance, which suggests that a K of 10 is around 12%, 12.5 around 16% and a K of 14 is around 18%. That's interesting, because AA was using a Pentax spotmeter! He was also referring to the Kodak Gray Card as an example of Zone V, which has 18% reflectance.
Yes, I can see why you might think that; however, there are a couple of things to consider. ANSI/ISO 2720 -1974 (R1994) has two ranges for K – K1 = 10.6 to 13.4 and K2 = 13.3 to 16.9. As for the manufacturers using the higher values of K previously before they were included in the standard, you are right. Standards only come under consideration for updating every five years, and often it takes all the parties some time to agree on any changes. One of the subcommittees Jones headed took almost ten years to agree that color was a psychophysical phenomenon and not a psychological one. As with the ISO b&w speed standard of 1960, Nelson observes in his intro to his paper
Safety Factors in Camera Exposures, Photographic Science and Engineering, vol 4, n. 3, Jan-Feb 1960,
“Many photographers give less camera exposure for black-and-white films than is indicated by exposure meters used with the American Standard exposure index of the film.” So, it’s not uncommon for the standards to lag behind.
One thing I learned from having all the exposure meter standards is that I am able to see all the changes made over the years. As the actual camera exposure equation has never changed (except for a slight tweak to adjust for a more accurate light loss factor but the target exposure never has changed), the changes to the exposure meter standards have more to do with making that exposure possible. One of the biggest changes happened in the early 1960s when they changed the color temperature of the light source. This change eliminated the need for EI ratings for tungsten and daylight use. The change was mostly required because of the preferred type of photo cells was changing from one that was overly blue sensitive. Schudder, Nelson and Stimson talk about in their paper
Re-evaluation of Factors Affecting Manual of Automatic Control of Camera Exposure, Journal of the SMPTE, vol 77, 1968, how the long equation used in finding K can be reduced to a number of variables because most are of the factors are constant. These variables include only r, which is the actual spectral response ratio of light detector, and t, which is the actual lens transmittance. You can also find this point summed up in the 1971 standard’s appendix C5. This indicates the biggest influence to the value of K comes from the spectral sensitivity of the photo cell. Jeff Conrad from your post seems to agree with this point.
It’s also important to remember it is not the purpose of the various standards to explain the reasoning or theory behind them. Sometimes they include an appendix which is not part of the standard that explains certain aspects, but the standards themselves only address the issue they are designed for. General Purpose Photographic Exposure Meters (Photoelectric Type) – Guide to Product Specification, will not explain why there is a range for K, or how the exposure meter relates to film speed and camera exposure. Fortunately at least in the past, when major changes are being made to a standard, someone associated with its development will write a paper explaining the reasons. I find these papers essential help toward evaluating the standards.
For those wondering what K is, it is a light loss constant. As the light travels through the lens, a certain amount of the light is absorbed or reflected. In order to then have an accurate indication of exposure, you need to apply this constant to the basic exposure equation. With the camera exposure equation, that value is “q”. With exposure meter calibration, it is “K”. But also with the light meter, the value of K involves adjustments when the meter doesn’t actually agree with the test target. This can happen for several reasons, but photo cell sensitivity is a big one. The exposure meter calibration equation is:
A^2 / T = (B * S) / K
Where
A = f/stop
T = Shutter speed
B = field luminance, in footlamberts
S = Film speed
K = light loss constant
You can reduce the equation down to A^2 = B in order to find the theoretical value of B without the light loss. The concept of Sunny 16, I believe, comes from or is reflected in the standard models of exposure and exposure meter calibration. Here A = 16 so the theoretical value of B = 256. The standard has B = 297 which makes the value of K = 1.16. One way to look at K is that during the calibration process, if the value of B doesn’t agree with the standard’s value, you adjust the K factor so that it does. This way, you should always obtain the same exposure at the film plane.
This is how that works. The camera image equation is.
(q * Lg / A^2) * t +Hf = Hg
Where
q = constant = 0.65
L = Mean Scene luminance, in candelas per square meter
A = f/stop
T = shutter speed
Hf = flare
Hg = exposure in mcs at the film plane
For our purposes, we don’t need t or Hf. The equation with all the variables in it is:
0.65 * (297 * 10.76) / 16^2 = 8.11
If you recall from an earlier post the constant for midpoint exposure is P = 8. You can calculate the exposure at the film plane for the midpoint for any film speed by multiplying 8 into the shutter speed – for example 8 * 1/125 = 0.064mcs. B&W film speed is calculated using the equation 0.8/Hm. That is 10x less exposure than the midpoint or 0.8/0.0064 = 125. That’s how it is all supposed to link together. As further evidence, there is an equation involving all three constants - P/q = K.
With the Pentax value of K, I found something questionable in the manual. It says the meter’s K value is 1.4 (in cd/ft^2) or 14 (in cd/m^2). That would make the conversion value 10, but should it be 10.76? So, one of the two values must be wrong.
Ralph, as your post didn’t include the math for how you obtained the equivalent percentages for K, I have taken the trouble to do them myself.
A value for K of 1.16 converts to 12.5 (1.16 * 10.76), but to do the percentages you need to use 1.16 * pi. According to K / C, where C = 30 footcandles (C is the calibration constant for incident exposure meters), that would equal 12%. You can also check it by converting to footcandles and dividing it with the approximate highlight value, 297 * pi / 7680 = 0.12. That would make it 12% for 12.5 and 13.6% for a K of 14. 16.9 would only equal 16.4% and 10.6 would equal 10%. Note: One of the things I hate about dealing with all the different equations is that each one seems to use different systems of measurement and having to remember the proper conversion factors.
Let’s look at the range. From 10.6 to 16.9, the value of B would be from 252 to 402 footlamberts or a 2/3 stop range. Let’s looks at the difference in footlamberts between 14 and 12.5. For the 12.5 and 14 (if it really is that) range that most manufacturers use, the range is from 297 to 332 footlamberts or 1/6 stop. I don’t think that is enough to worry about.
10.6 = .985 = 252 footlamberts
12.5 = 1.16 = 297 footlamberts
14 = 1.30 = 332 footlamberts
16.9 = 1.57 = 402 footlamberts
Here’s the other thing to remember, even though different values of K might read different luminance values as their target, they all produce the same illuminance value on the film plane for a given ISO. In other words, they place the exposure at the same place on the curve. The differences you might see with the suggested exposure setting with different meters has little to do with K, but more to do with the color temperature of the subject they are reading and the spectral sensitivity of the particular meter’s photo cell. If you check the three quadrant reproduction curves in post #97, you will see that even though the meter reading is at 12% (log 0.92), it still produces a print reflection density of close to 0.74 (18%). So, in a way, the meter does read Zone V and produces 18%, but it doesn’t see 18%. Of course, since there is no fixed relationship between negative densities and print densities, people can simple make adjustments in printing. That is also part of exposure theory, but it also is a different subject.
No matter what the value of K is, it still doesn't change the fact that the film speed / meter exposure relationship is 1.0 log-H units difference.
I noticed the 1974 standard doesn’t have much of a forward, so just to point out that the standards aren’t just some intellectual endeavor that ignores real world participants. Here is a list of the organizations included in the deliberations for the 1994 revised version.
Association of Audio-Visual Technicians
Canadian Standards Association
Leitz USA
National Association of Photographic Manufacturers
Optical Society of America
Philips Lighting Company
Photographic Society of America
Professional Photographers of America
Society for Imaging Science and Technology
Society of Photo-Technologists
US Defense Logistics Agency
US Department of the Army
US Department of the Navy
US General Services Administration
Xerox Corporation
Since AA used a Pentax spotmeter, it's not a misconception
Apart from the meters not seeing percentages and how K = 14 is 13.6% (I showed the math), there is also the question as to the edition of the book and if that information was ever updated. He always had 18% from the first edition of the book (and the older meter's calibration was actually closer to 18%). There’s a question whether he updated that aspect in the later books or simply carry over the earlier value. I still maintain that is what happened with the pre and post 1960 standards and Zone System speed testing. Then there's just the probability that Adams was wrong (He was wrong about the relationship between the meter and the speed point). There’s no way of knowing for sure.
This is pure speculation, but I also think Adams was heavily influenced by the Munsell scale. Middle gray (Step 5) was 18% and there were 10 steps (I believe Munsell’s company later changed middle gray to 19 point something percent to reflect the CIE findings). Think about it. The average scene is 7 1/3 stops. Each Zone is supposed to represent one stop. There are 10 print Zones. It never really equated. The occasions where I've seen attempts at justifying it remind me of how young Earthers try to justify the Earth being only 10 thousand years old. Again, a different subject.
but yes, it would be beneficial to understand better how exposure meters are working and what they are telling us.
One of the reasons why I harp on some of these ideas is because there are many people who assume something they have previously read, like Adams, is the authority and everything contradicting it must be mistaken. Chuck's response to discovering the 3 1/3 stop relationship between film speed and the meter's exposure point is an example of this. He assumed Adams was correct and consequently rejecting the real authority and standard model, as defined in the graph, as wrong. It is the same for ZS speed claims against the "box speed."