Hiding in Plain Sight

CPorter said:

Metering along the lens axis from the camera position, IMO, would be most representative.........the light has to travel that distance to the film. Therefore, the meter reading should also be from that distance, IMHO..

Click to expand...

I agree. When you meter at the camera, it is like you already factored in flare.

CPorter said:

With d-76 1:1, which is quite a straight-lined curve in my process, gamma can be easily determined, however, finding gamma, is not required to find development times.

Click to expand...

First of all Gamma refers to a specific methodology. I'm assuming you mean slope or gradient.

It’s of no consequence whether anyone does or doesn’t use contrast index to determine the contrast of their film. It is an analytical tool. Just another tool in the toolbox. It is one of many approaches. I believe that looking at a problem from only a single perspective won’t give you the whole picture. Come at anything from different directions and problems and answers that at first may not have been evident will reveal themselves.

Using some form of gradient is also helpful in communicating. A single number offers a lot of information. Just like placing ISO in front of the film speed. It's placement communicates to everyone that all the steps in the ISO Standard were followed. That way others can understand how the testing was performed, the implications of the results, and how to do the testing themselves to confirm the results. Just like they taught us in science class.

But don’t anybody kid themselves. Contrast determination is about film gradient. It's inherent in all the methods including the Zone System. It’s basically about input and output, and you can't have an output unless there's something put in. The Zone system uses a base density of 0.10 over Fb+f. The highlight is at 1.35 over Fb+f. The film’s density range is 1.25. The exposure range between the exposure that created the shadow density and the exposure that created the highlight density is 7 stops (2.10 logs). The input is 2.10 logs. The output is 1.25.

1.25 / 2.10 = 0.595

Without knowing the exposure range (input), it would be impossible to gauge the degree of processing. Try and describe how to do Zone System testing to someone without explaining the part about input.

The standard model in tone reproduction has an input of 2.20 logs minus some flare 0.30 to 0.40, and an output of 1.05 NDR.

1.05 / 1.80 = 0.583

The negative density range (output) of the tone reproduction model is different only because the range of the input is effectively smaller than the Zone System model. What this shows is that the film contrast is the same for both models and the different output values are only the result of using different input values.

Another way to look at is to think of the film density range and film contrast as two separate elements. The density range is a result of a given input range at a particular film contrast. The film contrast represents the degree of development. A smaller input luminance range will result in a small density range without a change in the film contrast. The film can have a certain contrast even though there is only a single point of density.

If you find the two points that make for a 1.25 negative density range with litho film will have under ½ stop of input. If it’s only about density range, than the results would look the same from a pictorial test, that required 7 stops to reach a NDR of 1.25, and a litho film, that required ~1/3 stop to produce a NDR of 1.25.

Without considering the input value as well as the output value, the picture on how the process works is incomplete. The real question should be which method has the best agreement with use over the broadest set of conditions. The first piece of evidence can be that Alan Ross has identified a problem with the a fixed density range method with greater than average luminance ranges.

CPorter said:

I simply cannot subscribe to the idea of there always being a 1 or 1 1/3 stop of flare present just because I have a lens on my camera..........flare is certainly a consideration, it is never a consistent thing, but I believe it has been my experience that the image contrast and subject contrast can, not always, be quite close (with quality multi-coated lenses), but is never perfectly matched either.

Click to expand...

We can't rely on anecdotal evidence. Look at all the way people saw the world before the Enlightenment. That's why there's the scientific method.

I've never asked anyone to accept anything I say based only on my word. I've included references, uploaded seminal papers, and presented facts and equations that illustrate how it all works. There is a preponderance of scientific evidence out there. You just have to take the time to look for it because it's hiding in plain sight.

From Wikipedia: Anecdotal evidence

In science, anecdotal evidence has been defined as:

"information that is not based on facts or careful study"[4]
"reports or observations of usually unscientific observers"[5]
"casual observations or indications rather than rigorous or scientific analysis"[6]
"information passed along by word-of-mouth but not documented scientifically"


From Wikipedia: Scientific Method

In the 20th century, a hypothetico-deductive model[9] for scientific method was formulated (for a more formal discussion, see below):

1. Use your experience: Consider the problem and try to make sense of it. Look for previous explanations. If this is a new problem to you, then move to step 2.
2. Form a conjecture: When nothing else is yet known, try to state an explanation, to someone else, or to your notebook.
3. Deduce a prediction from that explanation: If you assume 2 is true, what consequences follow?
4. Test: Look for the opposite of each consequence in order to disprove 2. It is a logical error to seek 3 directly as proof of 2. This error is called affirming the consequent.[10]

A linearized, pragmatic scheme of the four points above is sometimes offered as a guideline for proceeding:[46]

Define a question
Gather information and resources (observe)
Form an explanatory hypothesis
Test the hypothesis by performing an experiment and collecting data in a reproducible manner
Analyze the data
Interpret the data and draw conclusions that serve as a starting point for new hypothesis
Publish results
Retest (frequently done by other scientists)

Stephen Benskin said:

But don’t anybody kid themselves. Contrast determination is about film gradient. It's inherent in all the methods including the Zone System.

Click to expand...

Yes, I am referring to the slope of the straight line of a curve when a straight line is actually present, that's all.

I follow you on all that, no problem-----CI is easily determined ZS useage or not, I only am stating a fact that CI in and of itself, is of little use to me in being fluid with ZS and of making good negatives and evaluating contrast.

Bill Burk said:

I think you read me wrong because I think we are in agreement.

Click to expand...

Thanks Bill, that was a clever way to get your message across better, and it worked. Did I read once that you use graded paper? I use VC paper, so Chuck probably won't be wanting to use Bill's paper!

CPorter said:

I follow you on all that, no problem-----CI is easily determined ZS useage or not, I only am stating a fact that CI in and of itself, is of little use to me in being fluid with ZS and of making good negatives and evaluating contrast.

Click to expand...

Just because you don't personally use some form of average gradient, doesn't mean it can't be used in a discussion or that any point made using some form of average gradient isn't valid because you don't use it.

Tell ya what. Explain to me using the example below how you go about determining your development. It's a family of curves with Fb+f zeroed out. I've even marked out a 1.20 NDR. Or upload an example of your own showing how you do it.

Stephen Benskin said:

Just because you don't personally use some form of average gradient, doesn't mean it can't be used in a discussion or that any point made using some form of average gradient isn't valid because you don't use it. That would just be insane.

Click to expand...

Oh, there's no need to be threatened by little old me.......................I'll tell you what, first, prove in my posts where I indicated that it can't be used in discussion because I pernsonnally don't use it and and that a point made regarding average gradient is not valid because I don't use it. It's not in post #101, #107 (this is where I acknowledge there is nothing wrong with CI or using it in the ZS or whatever), it's not in #108, and it's not in #113. You don't get something for nothing in this world . You do that and and admit how horribly wrong you are with these assertions and I'll provide you with an answer, gladly.

That's just an attempt to avoid answering the question. Bill and I have explained how gradient is an intrinsic part of contrast determination even if the gradient value isn't determined. This should be obvious to anyone taking the discussion seriously. Yet, you have maintained a position that it's somehow possible to determine contrast without it.

I'm simply asking you to support your statement, otherwise it's just another unsubstantiated anecdotal claim. Put your money where your mouth is, or maybe you're intentionally acting in bad faith, wasting everyone's time, and attempting to derail this thread. Don't they call that being a troll?

CPorter said:

Did I read once that you use graded paper? I use VC paper, so Chuck probably won't be wanting to use Bill's paper!

Click to expand...

Right... Bill has two choices: Grade 2 and Grade 3. I picked a paper, Galerie, over 30 years ago that is still available. How many people can say that?

Stephen Benskin said:

That's just an attempt to avoid answering the question. Bill and I have explained how gradient is an intrinsic part of contrast determination even if the gradient value isn't determined. This should be obvious to anyone taking the discussion seriously. Yet, you have maintained a position that it's somehow possible to determine contrast without it.

I'm simply asking you to support your statement, otherwise it's just another unsubstantiated anecdotal claim. Put your money where your mouth is, or maybe you're intentionally acting in bad faith, wasting everyone's time, and attempting to derail this thread. Don't they call that being a troll?

Click to expand...

I see your ego and your arrogance won't let you go where I want you to go, therefore I won't go where you want me to go. I have your answer, but I don't feel compelled to prove anything to you, nobody should, as this will do nothing but feed your ego, so believe what you want to believe. If the consensus in this thread believes that I am a troll, I'll leave it, in the mean time, get on with it.

I take the discussion seriously, nothing I've said implies a denial of the importance of gradient in determining contrast (those are your words). The only position that I've maintained is that determining CI is not integral to using the ZS, anyone who chooses can determine this for themselves. Please note where I have maintained that it's possible to determine contrast without gradient.

Chuck, I believe you want to make a strong statement about the simplicity of the Zone System. I think a position such as that is easy to defend.

Steve, I believe you want to generously share technical details "from the other side" and discuss things the Zone System glosses over.

It's not wrong to say "this doesn't pertain to me," but sometimes it causes a squelching effect in a technical thread. I got the feeling that was what happened here. A friend of mine used to have a rule: If you don't like something, don't just tell me to stop it. Suggest something you like in its place.

I like to say the Zone System is so easy to get your head around and easy to get immediate results from.

But I am interested in delving in the details that were overlooked to see if they can be "folded into" Zone System practice, and I want to have you both involved in the discussion.

I was going to draw a curve family to include flare.

Then I froze in my tracks when I saw how the spotmeter seems to include flare.

Steve,

I think you have some references that break down what I am seeing, maybe it is distance effect.


To get moving again, I am going to act like I don't care about what the spotmeter sees.

I will plan to factor that back in later, I think this may cancel out.

---

For this exercise, I assume flare is related to subject brightness range.

I admit if there was a direct or predictable relationship we'd have the numbers, but for now I will assume it is a tendency.

A starting point might be to make a table like this:

8 stops SBR? (somewhere above 0.4)
7 stops SBR = 0.4 flare.
6 stops SBR? (somewhere between 0.0 and 0.4)
...
1 stop SBR = 0.0 flare.
---
Are there "flare for subject range" numbers out there to fill in the blanks? Otherwise I will make a few assumptions and draw the curves...
---

For those who are still here and still interested, I found a previous film curve example of Chuck’s. The top curve is what was originally uploaded. In the bottom curve the parts that I've worked are in red.



It should be obvious to anyone who is taking the discussion seriously that gradient is an integral part of contrast determination even if the value isn't determined. But if you determine it, you'd know that if you measured the curve at a different point, say at 1.80, you’d still produce the same gradient value which means the same contrast. Gradient is also useful in predicting density ranges for different situations.

Let’s say Chuck wants to use a condenser enlarger. According to The Negative, the density range for a condenser enlarger is 1.15. At what contrast would Chuck produce a negative density range of 1.15 for his conditions? 1.15 / 2.1 = 0.55.

Gradient also reduces the need to testing. You do a family of curves, determined the gradient for each, and you’ve determined you processing for silver printing on a diffusion enlarger, but then you want to shoot some negatives for platinum printing, you don’t have to test the film again. If you know the LER of the platinum paper, you just recalculate. My program has a gradient calculator. Just enter the set of conditions, and it gives the gradients for -3 to +3 as well as the processing times. To find the processing times, I use a CI / Time curve. Phil Davis lays this out pretty well in Beyond the Zone System.

Do you need to know the gradient in order to shoot. Of course not. With the shear preponderance of subjects near the statistical average combined with the varying personal preferences of the viewer, you really don’t need to know much of anything to produce a workable negative at least 60-70 percent of the time.

But who doesn’t prefer to have more information? Who doesn’t prefer additional insight? If the tool is available, why not use it when it can be useful? Understanding gradient made it possible to analyze two seemingly disparate methodologies and determine they are not dissimilar at all. They just measure the film curve in different places. The apparent difference between the negative density range of the Zone System’s 1.25 and the ISO paper grade 2 LER value of 1.05 is explained simply by using gradient.

Bill Burk said:

Are there "flare for subject range" numbers out there to fill in the blanks? Otherwise I will make a few assumptions and draw the curves...
---

Click to expand...

The rule of thumb is 1/3 stop change per stop luminance range. I believe it is only intended to be applicable for a limited +_ range of say 2 to 3 stops.

I've posted this before. It's a CI / LSLR graph of a number of different flare models.



I believe the reason Kodak used a fixed flare model for their chart was due to the unpredictable nature of flare. Next time you are in the testing mood, take two targets, one with a small black square in the center surround by white, and the other with a small white center surrounded by black. Meter a gray card and make exposures of both targets. Check the density range. They have the same luminance range, but will they have the same density range?

Another way of looking at Kodak's decision is that there are a lot of different areas of photography where errors tend to cancel each other out. Allen Stimson in Measuring and Judging Photographic Exposure of Color Film for Automatic Exposure Controlled Cameras writes about exposure, "The many tolerances and uncertainties in exposure control systems add up to an alarming possible error. However, many of the errors are usually opposite in direction and, therefore, are mutually compensating."

It could also be a question of the need for precision because there's the whole psychological aspect.

Stephen Benskin said:

The rule of thumb is 1/3 stop change per stop luminance range. I believe it is only intended to be applicable for a limited +_ range of say 2 to 3 stops.

I've posted this before. It's a CI / LSLR graph of a number of different flare models.

Click to expand...

Great! Let me see if I can work backwards from "practical" flare and my family of curves. I'll label them as N+ / N / N- for ease of visualization.

Bill Burk said:

Great! Let me see if I can work backwards from "practical" flare and my family of curves. I'll label them as N+ / N / N- for ease of visualization.

Click to expand...

The variable flare model reflects the 1/3 stop factor. The practical flare model uses the average between the fixed and the variable flare model. It's designed to include one of the exceptions to the LER / NDR relationship. From Jones, "For the soft papers, the density scales of the negatives should in most cases exceed the sensitometric exposure range of the paper, whereas, for the hard papers, the density range of the negatives should in most cases be less than the sensitometrically determined log exposure range of the paper."

The variable flare model agrees best with reality. I believe the practical will work best with use.

Stephen Benskin said:

The variable flare model agrees best with reality. I believe the practical will work best with use.

Click to expand...

I guess it's one thing to know what flare does, and another thing to know what you really need in the darkroom. I'll go with what you need in the darkroom.

Stephen Benskin said:

For those who are still here and still interested, I found a previous film curve example of Chuck’s.

Click to expand...

Here is the curve family that I ultimately arrived at for d-76 1:1 and TMX, absent any flare factor of course. What I found interesting with the test for EI is that it actually turned out to be 100 for me as the curve reached the Zone VIII target , just about right on the money. When +2 dev is planned, I reduce the exposure by 1/3 stop due to increased fog density and when -2 dev is planned, I increase the exposure by 1/3 stop due to reduced fog density. The curves don't show enough deviation in fog density at +1 and -1 to make a similar exposure adjustment. To date, these times have worked beautifully for me, of course, all I have to offer is my own anecdotal evidence of my own analysis and I'm not able to offer any more than that.

The "N" development comparison curves just represent, for me anyway, marked behavioral differences between d-76 and hc-110 with TMX in each of the three divisions of the curve (toe, straightline, shoulder). Knowing the CI values of each curve would be informative, I guess, but no decision would hinge on it. I evaluate differences in CI, figuratively speaking, in each major division of the curve when compared to another, in this way, the entirety of the curve is given full attention. Perhaps this explains my somewhat simplified point of view better, IDK.

I'm obviously not a technical minded as you and Bill because my road to the end result, seems just my own, within the confines of this thread that is, but I do take it seriously and find it interesting. If you are dieing to know the development times for the curves you posted last night, I arrived at:

At neg density 1.3, based off of development times from between 4 and 8 minutes and using Zone VIII as my target high value:

+2 dev, Zone VI (at log exp 1.5) = time interpolated at 7:30
+1 dev, Zone VII (at log exp 1.8) = time interpolated at 6:45


Normal dev, Zone VIII (at log exp 2.1) = ~6:00, the curve is within the range of 1.25 to 1.35


-1 dev, Zone IX (at log exp 2.4) = ~5:00, the curve is within the range of 1.25 to 1.35
-2 dev, Zone X (at log exp 2.7) = ~4:00, the curve is withing the range of 1.25 to 1.35

If I knew the ISO, I would break down my evaluation of the EI's for those four curves.

Chuck,

Those curves are beautiful and I think I see where flare is hiding.

As I get ready to draw my TMY-2 curves with flare included, I imagined your curve of N+1 with Zone I a little bit denser. To visualize only, what if Zone I really came out on film at 0.3 instead of 0.1.

You wind up with a camera negative that you developed with an aim of 1.2 NDR, but because of flare it comes out closer to 1.0 NDR.

And how do you deal with that? You don't need to. Your model anticipates flare by telling you to choose an artificially high NDR target. Higher than you really want. On your enlarger you want 1.0 NDR just like me.

Now since I picked a real negative that was really 1.0 NDR - I have to make sure to estimate flare before I decide development time because my charts don't have flare included and my aim NDR doesn't include adjustment for flare.

There may be more to it than that, but that's what I think I see right now.

Bill Burk said:

You wind up with a camera negative that you developed with an aim of 1.2 NDR, but because of flare it comes out closer to 1.0 NDR.

And how do you deal with that? You don't need to. Your model anticipates flare by telling you to choose an artificially high NDR target. Higher than you really want. On your enlarger you want 1.0 NDR just like me.

Click to expand...

This is what I've been saying for years! It's about the way the film curve is interpreted. The film curve is created under no flare test conditions, but there's flare in actual shooting conditions. In order to have agreement between the two, a flare value needs to be applied to the no flare film curve during the interpretation process (speed and contrast determination). The other way is to create a false aim value, such as a higher NDR to compensate for flare which is to be used only for the determination of contrast. Both approaches work, but one way reflects use and the other is a "fix" that can cause confusion if the concept of the "fix" isn't readily known. This is the difference that is observed between the Zone System NDR of 1.25 and the Grade 2 LER aim of 1.05. The difference is an illusion created by two different approaches but resulting in the same NDR even if they aren't aware of it. This is one reason why I like to say people aren't getting what they think they're getting.

The use of some form of average gradient in determining film contrast eliminates this potential confusion, in much the same way it was used to explain it.

Chuck are you carrying through specific subjects/zones measured in the scene or are your zones defined at the negative?

What I'm getting at is that if measured at the negative then camera/scene based flare wouldn't be represented, right?

markbarendt said:

Chuck are you carrying through specific subjects/zones measured in the scene or are your zones defined at the negative?

What I'm getting at is that if measured at the negative then camera/scene based flare wouldn't be represented, right?

Click to expand...

Look alone the bottom of the x-axis. The log-H range is 2.10 (7 stops). That's approximately the normal scene luminance range, not the in camera illuminance range at the film plane (exposure range). The important point is that the combination of the NDR and the log-H range still equates to a normal contrast negative, but that the 1.20 or traditional ZS 1.25 is a projection or construct in order to compensate for interpreting real world conditions on something produced in non real world conditions. This method can be used instead of adding a separate camera image curve or adjusting the log-H range of measurement to reflect the exposure range.

Phil Davis discusses this concept, in brief, in BTZS 3rd ed, Analyzing Film Curves, p 95.

The question has never been about quality in the final results. The discussion has always been about understanding what's going on in the process.

Have you ever seen a film curve that shows the effects of flare like below?



This is a similar concept to ZS contrast determination in that it too is a construct. It shows how flare effectively changes the film density. Putting it all on one graph eliminates the need for a camera image curve, but it doesn't illustrate how flare acts upon the film curve in reality. For that, a two quadrant curve is required.

markbarendt said:

Chuck are you carrying through specific subjects/zones measured in the scene or are your zones defined at the negative?

What I'm getting at is that if measured at the negative then camera/scene based flare wouldn't be represented, right?

Click to expand...

Mark, I have to admit I'm not quite sure what you are asking me here. The curve is created by an in-camera exposure of a Stouffer 21-step tablet, the densities read and plotted. Perhaps my mind is foggy this morning, sorry. Perhaps rephrase the question.

Chuck I think you answered the question, but here's a rephrased version.

The zones you are using are defined by the step wedge in the scene, correct?

CPorter said:

But, a ZS practitioner (if he or she considers themselves strict about it) does not really consider CI or any other measurement of contrast that connects points on the curve with a straight line. It's the comparison of the entire curves of at least two films or one film with different developers that is considered most valuable.

Click to expand...

Chuck, I was confused at first by this response until I realized it was paraphrased from The Negative. Let me post the whole section on Gamma and contrast from page 88. I’ve bolded the part that concerns Adam’s opinions.

“Gamma and Contrast

The film’s contrast can be measured by determining the slope of the straight-line portion of the curve. The slope is the ratio of the density change to the exposure change in the straight-line region, and is given the name gamma. Gamma, then, equals change in density divided by change in log E in the straight-line region. A film with a higher value for gamma will have more contrast than a film with lower gamma; it should be clear, for example, that a one-zone exposure change produces a greater density interval with a higher gamma film than with a lower gamma film.


There are some limits to the usefulness of gamma as a measure of negative contrast, however. For one thing, not all films have a long region that closely approximates a straight line. Even when a true straight-line region is present, gamma gives no measure of the nature or extent of the toe region of the film, which is of great importance in photography. Several different contrast measures have been developed over the years with the intention of supplanting gamma as a contrast indicator. One such index is called mean gradient (written G), and another is the Contrast Index (CI) system used by Eastman Kodak. Both approaches define limits on the film curve to be connected by a straight line, and the slope of this line is then measured. I find such systems confusing and uncertain in applied photography, and I still consider gamma the most useful figure. No index, including gamma, is nearly as informative as comparing the entire curves of two films, or of one film under different processing conditions. In addition, many contemporary films have exceptionally long straight-line sections which, in my opinion, definitely change the concepts of contrast measurement.”

First I have to admit it’s hard to make a proper evaluation when you only spend two paragraphs on a subject that could fill chapters, but…

Adam’s is misrepresenting the purpose of the gradient methods when he attempts to associate the determination of contrast with comparing film characteristics. The different versions of the average gradient methods are no more about comparing the characteristics of film curves than is the Zone System's method for contrast determination. I believe this is just an attempt to quickly rationalize away concepts he doesn't want to deal with nor understands well enough to properly evaluate. I'm sure he also didn't want put in too much effort making a case for a method other than his own. It’s hard to sell books if you can’t make people believe they should use your methodology.

I believe I am correct in concluding he didn’t have a firm grasp on the different forms of contrast determination because 1). he admitted that he found the systems confusing, and 2). he considers Gamma to be the most useful "figure" even though the original unmodified version of Gamma had already been thoroughly challenged by that time.

Personally, I don’t think there’s enough included in those two paragraphs to make any conclusions on the viability of any methodology.