Steve
If point 4 on that graph has a log density of 0.10, then a point with 3 1/3 stops more exposure is far from an average meter reading (Zone V) if developed for an average gradient of (0.57 - 0.61). An average Zone V exposure creates around 0.72 log density. 1.0 log exposure less is a Zone I**, which averages around 0.19 log density, and that is, by the way, very close to the 0.17 speed-point density, which I use in my film testing. In other words, I like a speed point at point 4, but I doubt that it has a density of 0.10. It is likely more around 0.17 - 0.19.
Ralph, I want to address your post but I want to involve those at a less advanced level too, so you will have to forgive the background stuff that you already know. I’m also talking about photographic theory here using the standard model. I am attempting to have a discussion about how the system was intended to be designed and how that relates to the Zone System which is the predominant system in use. Your system, which works fine, only confuses the issue and shouldn’t be part of it. I’m having trouble figuring out where to begin because there are a number of basic concepts here I want to cover.
The graph came from a book by Jack Dunn and G.L. Wakefield titled Exposure Manual. Dunn was well known in the scientific community writing a number of scientific papers on exposure (with a focus on incident). He is the co-inventor of the S.E.I. exposure photometer and the modern Invercone for incident light meters. The graph itself represents the classic model representing the relationships of the different elements in exposure with the characteristic curve. It represents the system for which such things as film speeds and exposure meters were designed to. The curve itself is immaterial. I've also uploaded another graph showing the same relationship from a paper by Allen Stimson,
An Interpretation of Current Exposure Meter Technology, Photographic Science and Engineering, vol 6, no 1, Jan-Feb 1962. Stimson was later chairman on the ANSI subcommitee PH3-3 on Standards for Photoelectric Exposure Meters producing ANSI PH3.49-1971.
There isn’t an intrinsic connection between negative densities and print densities or between Zones and negative densities and print densities (except Zone I for negative densities). Take the mid-density points for 320TXP and 400TX for example. In order for the long toed TXP middle exposure density to equal the same density as TX middle exposure density, it would require an additional stop exposure. In other words, TX’s Zone V density is TXP’s Zone VI.
Meters also don’t see percentages, they read scene luminances, but if you take the standard model's statistically average scene with a highlight at 100% Reflectance, the amount of exposure the meter would place at the film plane would be produced at 12% Reflectance and not 18% Reflectance. So, if you are using the assumption that the meter reads Zone V then you are ½ stop off. The exposure meter reads at Zone IV ½.
The relationship between the meter reading and the ISO speed point is that average meter reading falls 1.0 log units above the b&w speed point. That’s a fact. That is how ISO speed works in conjunction with the meter. The exposure at the speed point is 0.8/ISO and the exposure at the meter reading is 8 * 1/ISO. That is locked and unchanging. When you set the ISO on your meter that determines the amount of exposure it wants to strike the film. So for an ISO of 125, the meter wants to place 0.064 meter candle seconds at the film plane. No matter what the lighting conditions are, that is what is going to strike the film. The ISO equation for determining the speed of the film is the exposure necessary for the film to produce a density at 0.10 over Fb+b divided into 0.80 or 0.80/Hm. You can determine the exposure necessary to produce the density for a given ISO by using 0.80/ISO. In our example above, the light meter wants to place 0.064 mcs for a ISO 125 at film plane at the point of the meter reading. Now at the speed point, it will need 0.80/125 or 0.0064 mcs or ten times less exposure of a 1.0 log difference. That’s the way the math works. I’d be happy to produce a list of references if anyone wishes.
So, doesn’t the Zone System testing recalibrate the meter? Not really. You might now have the meter make Zone V exposures instead of Zone IV ½, but since that part is not measured in relationship with the ISO or ZS testing, you can’t be sure what the resulting densities will be (ie my TX/TXP example). Also, since there is a set relationship between the shadows and the meter reading, you are disregarding the shadow placement (ie overexposing the shadows from where you initially tested for them to be). Let me just say for the thousandth time, there’s nothing wrong with additional exposure. It can help. I do it myself. It’s just not relevant to this discussion.
Okay, the meter calibration point or meter reading point is 3 1/3 stops above the speed point which produces a density of 0.10. That’s a fact. The Zone System testing Zone I is found by stopping down 4 stops. Which means it is looking for Zone I 2/3 of a stop below where it would fall for a given film speed. A person would have to adjust their EI 2/3 of a stop in order to bring up the exposure enough to reach that point. That is why most people who do Zone System testing find their personalized film speeds to be ½ to 1 stop slower that the ISO film speed. And because the ISO standard is designed for the shadows to fall around 0.10 and those doing the ZS testing are rating their films at a lower EI than the film’s ISO, the Zone I exposure isn’t going to fall at 0.10 even though they think it will because they thought they tested it for it. Based on their own stated goals, they are overexposing.
This is why I kept making a distinction about EI and real film speed. I should have used the term Effective Film Speed (EFS). There are two types of EIs. One is however you set your camera which can include wanting to give it extra exposure, or wanting to “push for speed.” The other is the tested film speed derived outside the specifications of the ISO standard. Kodak did this with TMX and TMY when they first were introduced. The old ISO standard developer didn’t produce speeds reflective of real world use with those films, so Kodak developed them in D-76 or T-Max. Since they didn’t adhere to the standard, they couldn’t use the ISO prefix and had to use EI. I was attempting in the earlier posts, to make the point that ZS testing didn’t really produce an Effective Film Speed.
Prior to 1960, the Zone System was in closer agreement with the film standard because under that standard film speeds were a stop slower and that correlated with the type of speeds obtained with Zone System testing. In this regard, the Zone System testing method is out-of-date with the current ISO method. This is why, if you subscribe to the ZS method, you can’t make judgments about the legitimacy of it’s speed results over the ISO standards results. ZS practitioners are probably not aware of the difference because most don’t have the equipment to do a comparison test between the two methods and they are happy with the quality they are getting. Why shouldn’t they be? The old standard, which the ZS results reflect, produced excellent results. There’s nothing wrong with keeping the ZS method the same, but people should be aware of the differences and they should evaluate the two methods accordingly.
Some people might be wondering why the shadow exposure falls 4 stops under the meter reading while the speed point is only 3 1/3 stops below. Doesn’t the Zone System then compensate for what looks like the underexposure designed into the ISO standard? Well, the difference is actually 4 1/3 stops, and the reason why the shadow falls below the speed point is because, 1. the speed point doesn’t represent the minimum useable density point, 2. the speed point isn’t where the exposure is necessary supposed to fall, 3. FLARE. The average scene comes with at least one stop of camera flare. Flare adds at least one stop to the shadows effectively pushing them back up to around the speed point. So, the standard places the speed point one stop above where the shadows are going to fall knowing flare will raise them up. Without flare, film speeds would be a stop slower. Now, while the Zone System has almost the same meter to shadow relationship, its testing method doesn’t incorporate flare or the concept of flare factored into it.
Actually, since the average scene’s flare value using a 35mm camera is around 1 1/3 stops, which means the shadow exposure will fall 1/3 stop above the 0.10 speed point. In addition, the speed equation gives an additional 1/3 exposure (0.8/Hm and not 1.0/Hm). This is, of course, dependent upon the luminance range of the scene. The entire model is based on a 7 1/3 stop scene with average flare (see the Dunn graph and Connelly Graph). Under these normal conditions, the resulting exposure should come close to the values Ralph advocates. Given the testing discrepancies of the ZS testing should actually exceed those values.
Ralph, I’ve attached two three quadrant reproduction curves illustrating two exposure situations and the resulting mid-tone density on the print. Graph 1 is the standard model of exposure with one stop of flare. Graph 2 represents the results from ZS testing and I believe comes close to the results you support. There is little difference in the reflection density produced in the print for the mid-tone even though from a ZS perspective, one has the meter reading at Zone IV and the other at Zone V. (notice the reflection density for the meter calibration point - 0.92 = 12% Reflectance) The only real difference seen will come in the local contrast in the lower tones. But in order to properly indicate that difference, I would have to show a four quadrant curve. The degree to which that difference can be interpreted from a quality sense is subjective. There can also be inconsistencies in the subjective results based on tonal distribution of the shadows, level of flare, and curve shape. As fractional gradient theory shows, there aren't specific density points in exposure determination.
Here's a thought, since your numbers are derived for the purposes of increased shadow separation on the print (although they are close to many current results), why not make that point based on a gradient as it is the deciding factor? There are different gradients with different films for any point of specific density due to the shape of the toe. It seems the only consistent method would have to be based on gradient.
