Negative testing - exposure

[Stephen Benskin]

Stephen,

I'll always defer to your expertise, but here's my thinking:

Making in-camera images with a conventional taking lens in "average" lighting and a scene with a SLR that registers as "normal" in Zone System parlance has to be influenced by flare from the optical system. In my case the flare inherent in my spot meter as well as the flare from the taking lens. Hopefully the amount of flare will be "average" as well. In this case, then, shooting conditions and testing conditions are at least similar, the variation coming into play with variations in luminance ranges and distributions in different scenes.

Then, making a "proper proof" with the enlarger instead of making a contact print will include the flare from the enlarging lens.

So, by including the whole tone-reproduction system from meter to taking lens to enlarging lens, I'm also approximating in my tests the flare that's going to happen when making actual photographs in the field. (I say approximating because I'm well-aware that the amount of actual flare will vary according to a number of factors like the distribution of luminances in the subject (not every subject has a "statistically average" luminance range), whether the lens is single or multi-coated or not coated at all, the IC of the lens, whether a lens shade is used, etc., etc.)

And, I've also included other variables like my own idiosyncrasies in metering, my personal visualizations of what tonality each Zone represents, etc.

In a nutshell, I make a test negative and develop and print it in exactly the same manner as I would for any photograph I would in actual practice. The difference being making the print a proper proof, i.e., ensuring that FB+fog density renders very, very close to maximum paper black under viewing light that I consider excellent gallery lighting. With this last, I'm also including my eye's ability to discern differences in dark tones under changing lighting conditions.

We mention the "first excellent print" a lot in discussions like this, but we rarely mention the actual range of tolerances in exposure and contrast gradient that will still yield excellent prints.

The "first excellent print" is predicated on the minimum exposure to get desired shadow values. There's a large window of exposure above that minimum that will still yield excellent prints until the overexposure actually begins to degrade the image by shouldering off the higher-density values or showing too much grain. While I see the value in working with the minimum exposure to get an excellent negative, e.g., being more often able to use optimum f-stops and shutter speeds, I also realize that overexposing a stop or even more, especially with large-format film, will still get the job done and provide a bit of insurance against underexposure errors in tricky situations.

Similarly with development. Just as the contrast controls available with VC papers and a good color head make it possible to make good prints from scenes with a wide spectrum of subject luminance ranges, it also makes it possible to make excellent prints from negatives that are a bit over- or under-developed. I.e., there's a range of possible developments that will still allow for excellent prints.

So, using my more empirical approach to determining personal E.I. and development schemes I can start by simply choosing an E.I. that "should" be close to ideal (in the case of Zone System metering practice, that's 2/3-stop slower than ISO speed) and a development that "should" be close to "normal," go out and make a few identical negatives of a "normal" scene, in which I have clear Zone III and Zone VIII values, develop a negative and print it as a proper proof, and see immediately what adjustments I have to make. Usually the adjustments consist of refining development times since the 2/3-stop slower than ISO speed usually gets me very close to an ideal personal E.I.

Since I have a couple of other identical negatives, I use those to refine the development time. If I need to adjust my E.I., I usually just make a note of that and use the new E.I. in the field, keep notes and refine from there. Doing the same test for expansions and contractions gives me my N+ and N- times.

So, yes, there's a bit of a "black box" approach here: I assume that there is flare at several stages of the system, but don't do anything to eliminate it during testing or compensate for it later. What matters is if the shadow tones I get in the print correspond to the placements I made metering and if the highlight tones fall where they were planned to.
I'd be interested in hearing your thoughts.

Best,

Doremus

Doremus,

I don't necessarily disagree with anything you've said. I believe we simply have a misunderstanding, and I apologize for my part in it. Let me see if I can clarify where this may have happened. What I'm referring to is testing conditions, not shooting conditions.

The majority of camera flare is caused by the subject. Two scenes with identical seven stop luminance ranges one made up mostly of white with a little black will have a higher degree of flare than one made up of mostly black with a little white. It's about range. Flare in your spot meter comes from a light area bleeding into the shadow area you are attempting to meter. What if you were only metering a solid toned subject that more than filled the frame?

Let's take the classic exposure progression example.

1,2,4,8,16,32,64,128

Add a stop of flare and you get. (side point: The effective luminance range has been reduced. So we are working not with a 7 stop range, but a 6 stop range at the film plane.)

2,3,5,9,17,33,65,129

Now what if you had a stop of flare but not the full range?

8, 16, 32, 64, 128

to

9,17,33,65,129

Flare has little affect and in this example I would say no noticeable affect. Let's take this further. How about if we have a subject with a single tonal value that falls in the middle of the scale. There is no shadow for any flare that may exist to affect. There is no bright tone to create the flare exposure. Even in an optical system, wouldn't you say this is effectively a flare free condition? Isn't this how Zone System testing is done?

That's all I'm saying. Plus the ramifications to the NDR.

 

[MattKing]

One of the things I've learned over the years is that when it comes to venerable things like the Zone System, there are many instances where the system works but, with respect to why and exactly how the system works, the explanation given by the creators of the system falls short.
Stephen is very much in to why and how.

 

[Doremus Scudder]

Doremus,

I don't necessarily disagree with anything you've said. I believe we simply have a misunderstanding, and I apologize for my part in it. Let me see if I can clarify where this may have happened. What I'm referring to is testing conditions, not shooting conditions.

The majority of camera flare is caused by the subject. Two scenes with identical seven stop luminance ranges one made up mostly of white with a little black will have a higher degree of flare than one made up of mostly black with a little white. It's about range. Flare in your spot meter comes from a light area bleeding into the shadow area you are attempting to meter. What if you were only metering a solid toned subject that more than filled the frame?

Let's take the classic exposure progression example.

1,2,4,8,16,32,64,128

Add a stop of flare and you get. (side point: The effective luminance range has been reduced. So we are working not with a 7 stop range, but a 6 stop range at the film plane.)

2,3,5,9,17,33,65,129

Now what if you had a stop of flare but not the full range?

8, 16, 32, 64, 128

to

9,17,33,65,129

Flare has little affect and in this example I would say no noticeable affect. Let's take this further. How about if we have a subject with a single tonal value that falls in the middle of the scale. There is no shadow for any flare that may exist to be affected. There is no bright tone to create the flare exposure. Even with an optical system, wouldn't you say this is effectively a flare free condition? Isn't this how Zone System testing is done?

That's all I'm saying. Plus the ramifications to the NDR.

"Isn't this how Zone System testing is done?"

Stephen,

No need to apologize for anything. I don't think there's really any misunderstanding either. I'm simply not trying to separate and understand the effects of all the variables in the process, but instead, arrive quickly at a practical, usable result.

What I'm advocating for testing is using a subject with a full range of tones (not using a grey card or whatever as a target) and, therefore, producing flare. If I (am lucky enough to) pick a subject with a "statistically average" range of luminances, then I also get a statistically average amount of flare.

Whatever this flare does to the range of densities on the negative, while interesting, is unimportant in my regime. All that matters is that I craft a negative that produces a print (flare in the enlarging system too...) that reproduces the tones as I desire to have them and how I envisioned them at the time of exposure using my (flarey Pentax digital) spot meter.

I understand all the charts and the science behind what you are saying. I'm just advocating a less-scientific approach to finding E.I. and development times using real-life subjects and accepting the flare at each stage of the process. My only effort to try and control flare is to try and meter carefully and not have bright areas in my meter's field of view and to try to choose a test subject that at least comes close to "statistically average" in terms of luminance range and distribution.

I'm not trying to determine the precise characteristics of the materials (emulsion speed, curves, etc.), rather just be able to have confidence that when I make a negative, I'm exposing adequately and able to develop it so that it presents me with a usable density range and contains the information I need to make the print I want.

Best,

Doremus

 

[Stephen Benskin]

Have a question regarding sensitometer - I found couple of those - they are dedicated to test X-ray films against 21 step wedge. But it is lit by green or blue LEDs. Is it suitable for our ortho or panchromatic B&W negs?

Personally, I'd recommend against it, but I don't have any hands on experience. @ic-racer can probably offer more on the topic.

 

[albada]

Yet, the LER for a grade 2 paper is based on approximately the same Zone I to Zone VIII range is 1.05.

I thought the standard LER rating of paper was based on Zone II to Zone VIII. Zone II is the first to show texture (in normal room lighting; I use EV 6), and Zone VIII is the last, so II-VIII is what AA called the "textural range". Furthermore, my own calibration of Ilford and Foma papers based on II-VIII gives results close to published specifications of those papers. My calibrations use contact prints of a calibrated Stouffer wedge. WBM also documents the ISO ranges, and I notice that 0.9*Dmax corresponds to Zone II. So my question is: How does Zone I relate to LER?

Mark

 

[kal800]

What about condenser enlarger as a light source? Will it work? Of course with conversion filter into daylight.

 

[Chuck_P]

How does Zone I relate to LER?

I'll take a stab at it and will know soon enough if I'm in the ballpark of an accurate statement............................It's the first "useful" negative density that can be put on paper to separate from the paper's Dmax. It's the speed point of the ZS method of establishing personal EI with the ISO rating that gives a negative density of 0.1 at Zone I. VIII (1.3) - I (0.1) = 1.2 NDR, but then I guess there is the subject flare issue.

 

[aparat]

You are asking very good questions, so keep at it. However, asking questions on a forum, such as Photrio, is likely to get a few advanced photographers involved, which may or may not be what you need. You're going to get bits and pieces of theory, bits and pieces of methodology, information about related issues, etc. Photrio can be an incredible resource, but, to an absolute beginner, it may seem chaotic and frustrating, at times. If you enjoy learning in this kind of environment, you can disregard what I say next.

My advice is to think of your questions as part of a framework. To answer the questions, one needs to pick a suitable framework in order for the answers to make sense. There have been a few such frameworks proposed since the 1930s. Pick one of them and learn it well before you move on to the next. Outside of a framework, most statements will lack the necessary explanatory adequacy. They will seem abstract, confusing, even contradictory.

A suitable framework consists of theory and methodology of not just photography, but other, related disciplines, including some mathematics, physics, and statistics. If you feel like you have a decent grasp of these areas, you can go straight to the early sensitometry literature, esp. the Journal of the Optical Society of America. Start in the 1930s and slowly work your way up to the 1960s. Most of the important progress happened in that period.

If you do not have experience in the areas I mentioned, start with a textbook specifically for photographers, such as The Negative (Adams), Way Beoynd Monochrome (Lambrecht and Woodhouse), Beyond the Zone System (Davis), Exposure Manual (Dunn and Wakefield), The Zone VI Workshop (Picker), or any of the others. But pick just one to start with. Otherwise, you'll end up frustrated and overwhelmed with detail.

 

[kal800]

You are asking very good questions, so keep at it. However, asking questions on a forum, such as Photrio, is likely to get a few advanced photographers involved, which may or may not be what you need. You're going to get bits and pieces of theory, bits and pieces of methodology, information about related issues, etc. Photrio can be an incredible resource, but, to an absolute beginner, it may seem chaotic and frustrating, at times. If you enjoy learning in this kind of environment, you can disregard what I say next.

Well, I've been photographing for last 20 years finishing quite nice photography school in the meantime. Since I was shooting only roll film I did not need to have it tested as I could not apply custom development for a roll containing multiple scenes. Now I need only advice on one step of the whole procedure which I mentioned at the very beginning.

I've got Way Beyond for a long time, being my photo bible, but I do not have BTZS - I'm looking to find one actually

 

[aparat]

Well, I've been photographing for last 20 years finishing quite nice photography school in the meantime. Since I was shooting only roll film I did not need to have it tested as I could not apply custom development for a roll containing multiple scenes. Now I need only advice on one step of the whole procedure which I mentioned at the very beginning.

I've got Way Beyond for a long time, being my photo bible, but I do not have BTZS - I'm looking to find one actually

Awesome! In that case, I recommend you get a copy of Fred Picker's Zone VI Workshop, book and/or video. He explains everything you need to test film and paper for a roll and/or sheet film photographer. His approach is practical, designed specifically for photographers. Also, since you already have a copy of Way Beyond Monochrome, you've got everything you need . It's excellent.

I have a spare copy of BTZS. I will be happy to send it to you.

 

[Stephen Benskin]

My bible was:
Photographic Materials and Processes
by Leslie; Compton, John; Current, Ira; Zakia, Richard Stroebel

Not the Basic version. Supposedly it was used as a textbook for the 1st year students at MIT. Used it so much the binding fell apart.

 

[ic-racer]

My bible was:
Photographic Materials and Processes
by Leslie; Compton, John; Current, Ira; Zakia, Richard Stroebel

Not the Basic version. Supposedly it was used as a textbook for the 1st year students at MIT. Used it so much the binding fell apart.

My book that is falling apart is Zakia and Todd's book "Photographic Sensitometry" I got in graduate school in the early 1980s. It never went into as much detail about flare as I see in your postings here. For example it shows a four-quadrant tone reproduction family of curves without explaining how they were created.

 

[Mr Bill]

Supposedly it was used as a textbook for the 1st year students at MIT.

You probably meant RIT, where the authors were from.

 

[StrangestStranger]

My bible was:
Photographic Materials and Processes
by Leslie; Compton, John; Current, Ira; Zakia, Richard Stroebel

Not the Basic version. Supposedly it was used as a textbook for the 1st year students at MIT. Used it so much the binding fell apart.

Actually, RIT. Stroebel was a guest professor on occasion while I was there. It's been updated and there is a much more recent version available than the one that I used for the class.

 

[Stephen Benskin]

You probably meant RIT, where the authors were from.

I did. I did mean RIT. Wrote the post quickly on lunch break.

 

[Stephen Benskin]

I believe I have an example that clearly illustrates the cause of the discrepancy between the Zone System and the tone reproduction NDR aims for a Normal negative. Kal800 began this thread by asked about three testing methods. I've emphasized over the years that testing done incorrectly not only leads to the wrong conclusions, it also leads to a false sense of accuracy / confidence. The same is true for interpreting the testing data which this example also addresses. I started making a series of posts on Methodology and Curve Interpretation last Spring but got bored and stopped before finishing up the introductory stuff and getting to the interesting part.

A film's characteristic curve is a graphic depiction of how a given film / developer combination responds to exposure. Exposure is generally done by contacting a step tablet with the film to be tested. This eliminates many potential variables that can obscure the results. Including eliminating flare, making for a flare free test. In order to properly interpret the film curve as it applies to shooting conditions, the elements associated with taking a photograph need to be considered and included.

Below are examples of two ranges of exposure applied to the same film curve: 1.80 log-H and 2.10 log-H. The exposure starts at 0.10 over Fb+f and extends 7 stops (Δ2.10 log-H) to the right. In this example, the resulting negative density is ~1.35 for a NDR of 1.25. While a film curve does not include flare, a photographed a scene with a camera and it's optical system does. Camera flare reduces the effective exposure range at the film plane. A scene with a normal Luminance range has about 1 to 1 1/3 stops flare. One stop of flare will reduce a 7 stop (log 2.10) scene to a 6 stop (log 1.80) scene at the film plane. While the example doesn't accurately represent the way flare affects exposure, it does show the influence. The 7 stop Luminance range becomes a Δ1.80 log-H range. The resulting density would then fall at ~ log 1.15 for a NDR of 1.05. This corresponds to the LER for a grade 2 paper using a diffusion enlarger. All the elements are the identical between the two examples: same scene, same camera, same film and development. The only difference is the interpretation of the data.

 

[aparat]

Thank you @Stephen Benskin for a very clear demo of how flare affects exposure range at the film plane (also described as "log image luminance"). Camera flare also effects the resulting negative's D-log-E curve in three major ways: (1) it increases the length of the toe, (2) it adds curvature to the lower part of the curve, and (3) it reduces the shadow contrast. The highlights are much less affected. I put together a simple demo (from actual data) of how flare affects the characteristic curve. Here, I apply the concept of flare density (0.02), rather than flare factor (log subject luminance ratio to log image luminance ratio), but it does not change the overall result. In the sensitometry literature both flare concepts have been used. Whether we incorporate flare density in film test analysis is subject to debate. I've seen it done both ways.

 

[Bill Burk]

What about condenser enlarger as a light source? Will it work? Of course with conversion filter into daylight.

I think you're asking about using an enlarger as a sensitometer, by placing a step wedge in contact with film being tested. That will work.

You don't have to think of the fact it's a condenser enlarger until later when you decide what contrast of negatives you want to aim for.

 

[Stephen Benskin]

@aparat, that approach was very popular when curves were predominately hand drawn. Combining exposure with the film curve is less labor intensive than having to calculate and draw a separate camera image curve and align it to the film curve; however, in my opinion it has the potential to be conceptually misleading unless you’re aware it is a construct. It doesn’t depict how flare actually works. This is one of the problems I had with the Phil Davis plotting program.

I’ve included an example from a paper by Jack Dunn that I believe clearly shows how the plotting method works. Curve A is the film’s characteristic curve. Curves B – D include progressively increasing degrees of camera flare. To depict flare’s effect, each flare curve locates the density on the film curve that corresponds to the difference in log-H for the flare factor. For example, curve B has a flare factor of 2, or one stop, or a Δ 0.30 log-H. The density at that point to the right on the curve is then projected back to the position of the shadow exposure becoming the new shadow density for a flare factor of 2. The flare curve is then drawn. In the example, a flare factor of 4 uses the density value from Δ 0.60 log-H units to the right and the flare factor of 8 uses the density value from Δ 0.90 log-H units to the right. Also note how the "Range as 'seen' by the camera" is calculated from the point where the dotted lines intersect with the film curve and not from the exposure at relative log 0.

While this is method is quick to plot and straight forward to evaluate, it’s doesn’t represent how flare works. Dunn’s description of “effective printing characteristics” is a good explanation of why the approach is used. It represents the apparent changes caused by flare on the film’s characteristic curve as seen in the printing process, but not actual changes to the film curve. Camera flare does not change the film’s characteristic curve. It only apparently does. The actual cause comes from the camera image’s relationship to the original subject and where it falls on the film curve, and this is where misconceptions can occur with this plotting approach. Don't get me wrong, this does not invalidate the method. It remains an effective evaluation method.

A film’s characteristic curve is a graphic depiction of how the film responds to varying degrees of exposure when the film is processed under certain conditions. It is basically fixed and unchanging under those conditions. Any value of exposure will always produce the same amount of density. Any change in exposure only shifts its position along the X-axis to the corresponding density for that exposure. One stop of camera flare will double the shadow exposure, which moves its position 0.30 log-H units to the right of its original position. The change in density comes from the exposure falling on a different part of the film curve. It doesn't change the curve itself.

How one, two, and three stops of flare actually work on the characteristic curve. Interesting fact. Without flare, normal shadow exposure will fall approximately one stop below the 0.10 fixed density speed point to where the fractional gradient speed point is. Which means that even if the degree of flare is less than average, the exposure for an average scene will still produce a good negative.

Camera flare affects the shadows proportionately more than the mid-tones and highlights. It compresses the shadow’s illuminance range which can appear to have the same resulting prints as a film with different toe characteristics.

 

[Stephen Benskin]

Thank you @Stephen Benskin for a very clear demo of how flare affects exposure range at the film plane (also described as "log image luminance").

Actually, I was defining a resolution to the apparent discrepancy between the aim NDR between Zone System testing and that indicated by tone reproduction. I did state that the example did not accurate represent how flare effects exposure. The example was used to illustrate the effective log-H range and resulting NDR.

The top table is from Photographic Materials and Processes. The bottom set of data is from The Negative.

What I was illustrating is the discrepancy is due to a misconception interpreting the data, ie flare vs non-flare. The problem comes from the assumption that since Zone System testing is done with a camera, it contains flare. My example shows applying the same luminance range from Zone System testing to a flare free test (film curve) will result in the same aim NDR from Zone System testing which indicates Zone System testing in a camera is essentially flare free. Since the OP indicated he wants to do more advanced negative testing, I thought it important to emphasis good testing methodology and to be alert to possible testing errors and misconceptions.

There was a company that made machines that would create negatives from digital files - LVT. A calibration negative would be generated to test the digital output with the film development. After calibration, they would create the negative from the digital file. Unfortunately, the resulting negatives always were too flat. I don't know if they ever resolved it, but the problem came from a misinterpretation on how to apply the aim NDR of 1.05. Instead of matching it to the paper's LER range they though it applied to the full digital range of 0 - 255 and was supposed to fit on the full log exposure range of the paper from paper D-min to D-max.

Below is from Appendix 2: Film Test Data in The Negative. I've added the red lines. Zone I to Zone VIII, NDR 1.25 or the same results from my example. It's interesting to note the book states the testing in this section eliminates the K-factor in the calculation of the EIs. Another example of a misconception.

 

[aparat]

@aparat, that approach was very popular when curves were predominately hand drawn. Combining exposure with the film curve is less labor intensive than having to calculate and draw a separate camera image curve and align it to the film curve; however, in my opinion it has the potential to be conceptually misleading unless you’re aware it is a construct. It doesn’t depict how flare actually works. This is one of the problems I had with the Phil Davis plotting program.

I don't mean to hijack this thread, so I will reply briefly. Perhaps a new thread is needed to discuss this further?

Yes, but the same can be said of most sensitometry science (at least with regard to B&W film and paper), which has its roots in an era when plots were made with the French curve and calculations made with the slide rule. Still, both the flare ratio and flare density concepts can be turned into a modern statistical / computational model.

It can be argued that the flare density model has the advantage of being more parsimonious in both computing and showing the effect of flare on the characteristic curve, especially to "practical" photographers, though I do agree, in principle, that in a comprehensive tone analysis it is important to consider the formation of the camera image as a distinct process, one that, obligatorily, includes the contribution of flare. Also, if flare is considered in terms of a "flare factor," as proposed by Jones and Condit (1941), it is necessary to know, a priori, the object luminance scale before the resulting object-to-image density relationships can be computed. As I said before, both notions of flare have their pros and cons, but both models, fundamentally, do consider the formation and contribution of flare in the photographic process and both can be applied successfully. To me, this is not a binary.

"Actually, I was defining a resolution to the apparent discrepancy between the AIM NDR between Zone System testing and that indicated by tone reproduction. I did state that the example did not accurate represent how flare effects exposure. The example was used to illustrate the effective log-H range and resulting NDR."

Sorry! I though that was self-evident. I merely meant that the basis for the discrepancy was in the concept of flare related to the object-to-image luminance ratio. Your demo was very clear and persuasive.

 

[kal800]

Got a further question. I checked Stouffer wedge with transmission densitometer and the results were slightly different than referential, so this means that i calibrated the wedge. The question - what I can do with those numbers? During testing I will expose that wedge on the film and then measure densities of different steps. If for instance step 5 should have D0.60 and my result is D0.58, then what should I do with those D0.02 of difference when I will measure the same step on my developed negative?

 

[Stephen Benskin]

Got a further question. I checked Stouffer wedge with transmission densitometer and the results were slightly different than referential, so this means that i calibrated the wedge. The question - what I can do with those numbers? During testing I will expose that wedge on the film and then measure densities of different steps. If for instance step 5 should have D0.60 and my result is D0.58, then what should I do with those D0.02 of difference when I will measure the same step on my developed negative?

Kodak used to have a calibrated step tablet for two or three times the normal price which was one of their normal step tablets that they read on a very precise densitometer. What you have is a question of calibration and tolerance. Most densitometers are lucky to be within +- 0.01. Unless you want to calculate film speed and have a calibrated sensitometer, I would say close enough. You can either use what the steps should be or how your densitometer reads them.

 

[Bill Burk]

What you do with those numbers… use them instead of marking x-axis coordinates with the nominal 0.15 spacing. This makes your curves smoother because they are more accurate.

 

[ic-racer]

Have a question regarding sensitometer - I found couple of those - they are dedicated to test X-ray films against 21 step wedge. But it is lit by green or blue LEDs. Is it suitable for our ortho or panchromatic B&W negs?

Those are great for B&W process control and testing development. The color of the exposure did not influence the film's response to development time chanages when I compared 4 sensitometers with different lamps. Likely no development parameter would be altered significantly by the exposure color.

Also comparative film testing with the caveat that both films would be know to have similar sensitivities to the lamp color. For example FP4 was about 45% slower than Trix, irrespective of sensitometer exposure time or lamp color.

Since those tests I have obtained a sensitometer with a electroluminescent light source, but have no reason to think it would produce results any different than the other 4.

Wejex = 1 second incandescent exposure
EG&G = 1/100 xenon flash exposure
Green and Blue LED = about 0.5 second exposure