Under versus Over (Exposure & Developing)

[Stephen Benskin]

How about something like "discussing BTZS" or "reviewing BTZS"?

Ron, could you include the rest of the article information (name, publication, date) in case some of us wish to look into this topic further? Thanks.

Steve

 

[Photo Engineer]

Color: Theory and Imaging Systems, 1973, C. SPSE, Chapter 5, "A review of old and new methods of evaluating the image structure of color films", page 80. Dr. Michael A. Kriss, Eastman Kodak Co.

Figures 8, 9 and 10 are relevant to this discussion and the entire field of color and B&W image structure are closely related.

PE

 

[df cardwell]

Not to get into a really technical N+ debate here, but I am confused.

Me too.

Here's a good PRAGMATIC experiment that will yield a gold mine of understanding with a minimum of work.

Using your plain old film and developer combination,
set your camera up and make a picture of your ideal or typical subject.
Use your normal films speed setting, and normal metering technique.

Make a picture !

Bracket 1 stop over, and 1 stop under. Repeat this sequence, correct, over, and under: use up the roll.

In the dark, cut the roll into 3 strips. Give the first strip normal development, the second 25% less, the third 25% more.

With the first strip, make the best print you can from the normal exposure, and from the over and under exposed negative. Does the normal exposure make the best print ? Does either of the other negatives make an interesting, or superior print ? Try different contrasts, and so on.

Repeat with the over and underdeveloped strips.

Make notes. Look at the prints a few days after the experiment. See what you like, what relationships have been established, and what leeway ( or latitude, or margin of error is inherent in each of the 9 negeatives you've printed.

WARNING: This may be exciting. Good. It takes an afternoon of pleasant work, and gets you to reliable - and informed - image making very quickly.

Re-read Gainer and PE for their comments about film & developer combinations. You can repeat this process when you switch film & developers.

As you get a feel, or understanding, of how the paper is affected by the exposure and development of the film, you may choose - or not - to use a densitometer to speed the process of investigation.

But it is no more accurate, nor informative, nor as much fun.

 

[gainer]

Patrick, it has been shown by a number of modern workers that the curve of the final print is the product of the film and paper curve. It is more complicated than that, in that it involves calculus and matrix algebra if you wish to compensate for spectral sensitivity of the paper and the tone of the developed silver, as well as correct for micro vs macro contrast.

The most significant contribution comes from the micro contrast curve, but the measurement of micro contrast must be done as a function of magnification, as the final result will be seen to vary as magnification changes.

There is an article by M. Kriss that covers this quite well. Figure 5-10 shows variations in contrast as a function of chemical adjacency effects from 1000 microns down to 10 microns.

PE

Matrix algebra and calculus were important in my work at NACA-NASA for analysis of flight and some wind tunnel test data, but I retired 24 years ago, have had encephalitis in the meantime, and would rather let someone else do that part of the job of making good prints. What usually happens, I think, is that we find a combination or combinations of film, developer and paper, not to mention technique, that work well for our purposes and cry a lot if any of them become unavailable. Perhaps that is what causes a lot of us to specialize in particular types of scenes. Maury Amsterdam, comedian and cellist, did a skit where he was playing a single note on his cello over and over. When the straight man asked why he didn't play different notes like other musicians, he said "Those other guys are looking for the right note. I've found it."

The part of this interchange that is pertinent here, IMOH, is that I'm seldom going to find straight line H&D curves in either film or paper, let alone both, and I'm not going to study integration by parts or information theory again. So, I spend time and money doing the trial and error thing, making crooked film and crooked paper go together to make pictures that I like and others like too. Sometimes.

 

[Photo Engineer]

Patrick;

I don't blame you.

PE

 

[mongo141]

Matrix algebra and calculus were important in my work at NACA-NASA for analysis of flight and some wind tunnel test data, but I retired 24 years ago, have had encephalitis in the meantime, and would rather let someone else do that part of the job of making good prints. What usually happens, I think, is that we find a combination or combinations of film, developer and paper, not to mention technique, that work well for our purposes and cry a lot if any of them become unavailable. Perhaps that is what causes a lot of us to specialize in particular types of scenes. Maury Amsterdam, comedian and cellist, did a skit where he was playing a single note on his cello over and over. When the straight man asked why he didn't play different notes like other musicians, he said "Those other guys are looking for the right note. I've found it."

The part of this interchange that is pertinent here, IMOH, is that I'm seldom going to find straight line H&D curves in either film or paper, let alone both, and I'm not going to study integration by parts or information theory again. So, I spend time and money doing the trial and error thing, making crooked film and crooked paper go together to make pictures that I like and others like too. Sometimes.

How True, Amen!

 

[sanking]

Well, maybe we should start a new thread; we're now a rather long way from the original question.

Phil Davis is making things extremely complicated. One of the most basic matters: that curves come in different shapes, must have escaped his attention. As so many other zone-system proponents he seems focused on the idea that it is possible to make all films & developers to behave in the same way: N, N+1, N-1 and so on. But only a few can do it.

You appear to be asserting as fact about Davis, i.e. "that curves come in different shapes, must have escaped his attention" something I don't believe you will find stated in his writings. I have seen many film curve families produced by Davis, and some clearly show a difference in the shape of the curve with the same film but different developers. I doubt seriously that Davis would have missed something so obvious. And a friend of mine used for many years a developer that alters the typical shape of a Tmax-100 film curve, *specificlaly* because it was recommended to him by Phil Davis for that purpose.

Sandy

 

[Papa Tango]

Pass on the calculus...

I'm going to throw in my $0.02 just to add a different perspective.

The combinations one can get by varying exposure and varying development, for a given film and a given developer, are larger than they might first appear. Exposure and development seem to work together but in fact work differently and are not symmetrical. That is, if you expose and develop normally, then overexpose by 1/3 stop and underdevelop by 1/3 stop, you don't really end up in the same place like you'd think you would.

Think of it like this. Exposure for a given film determines where the image is going to be along the characteristic curve. This is particularly important at the toe of the curve. If you expose too little, you don't create a latent image since that data falls below the toe of the curve. If you expose more, you move the image up along the curve and pull shadow detail up over the toe. Thus the admonition "expose for the shadows."

Development determines maximum density and thus the slope of the characteristic curve (some call this the gamma, some call it the contrast index). Development doesn't move the image up or down the characteristic curve - it merely changes the slope of the curve. Thus the admonition "develop for the highlights." (Italics added for emphasis)

Bruce, that has to be one of the clearest & succinct explanations of what is occurring that I have ever read. The trouble with adequately understanding some of these concepts and applying them is many of us are "mere mortals" :confused: and are looking to achieve consistent results without applying scads of arcane technical details to the process. I do not see matrix algebra and calculus ever becoming part of my photographic repertoire. (apologies to those who do...)

Clear and concise understandings go a long way to develop technique. I have a Cosar densitometer, and have played with it to see different base factors and compare densities of a couple different films. I find this about as exciting and interesting as preparing my taxes. I have also gotten little out of the process, as in my "amateur" practice of photography a negative is either of sufficient (average) density to achieve a quality print, or it is not. Most of the time this can be confirmed easily on the light table without sensitometry. Hence, the question I have asked.

It is fairly obvious that a film will not behave the same in different developers. Nor will different films react the same for changes in exposure or development time. And neither do different size formats of film. So, it looks like the best option here is to choose 1 or 2 films in several formats. Bracket as DF Cardwell suggests, and document, document, document the results. Choose accordingly by paper characteristics (sigh, another set of tests).

I wish I could recall how I always got good negs and prints 30 years ago with PanX and grade 3 paper…

 

[Photo Engineer]

I would like to make it clear that I don't use matrix algebra at all in my work, nor do I advocated its use by anyone else here. I was merely pointing out that it is used in the DESIGN phase of film systems. And, I gave the reasons for that above.

For our use, making good prints is the final desired result. Using the developer and film combination that optimizes this result, or using 'what works for you' is my position and always has been.

The problem goes back to one of the early posts where someone got nearly identical characteristic curves with differrent results when printed. I was trying to explain the reasons why. Basically, it comes down to what I just said - "use what works", but if you continue to measure the macro H&D curve for prediction of final result, it just wont be a good predictor, as micro contrast is often a determining factor in the 'equation' if you pardon that use of the word here.

PE

 

[gainer]

You needn't worry. We know you are smart enough not to do that.

It has been said that Freudian psychoanalists can explain everything but predict nothing. Behaviorists OTH don't care how a patient got that way but just go about fixing things. That's an oversimplification, but it seems to me that you would be in favor of behavior modification. After all, you are PE, and engineering is more akin to behavior modification than to formation of theories of analysis.

There are probably a number of us APUGers who could analyze all the characteristics of a film and compute the characteristics of a paper that would be synergistic with it, but there would have to be a search for that paper or the formation of some algorithms for creating it. Behavior modification is more fun, and makes for better war stories.

Now please don't take me too seriously.