Effective film speed (as in Lambrecht's method)

[Chuck_P]

I have the spreadsheet exactly as he published it back in 2010, along with the PDF of the relevant pages from Way Beyond Monochrome that describe how to use it. (He made both files available on the web back then.) I've also fixed a minor bug in the spreadsheet in the version I use.

I've recently tried to find those old Way Beyond Monochrome resources on the web, without luck. If none here can point to a new location for them, I can always make the copies I have available.

Thanks, I've not been able to find it either.

 

[Chuck_P]

Flare plugs right into the simple formula that you use to decide the film contrast aim to fit the target paper grade.

Say paper grade 2 has Log Exposure Range 1.05

You have a classically normal Subject Luminance Range of 7 stops, 2.1

Take flare 1 1/3 stops, 0.4

(2.1 - 0.4) / 1.05

0.62 is the aim to develop your film for normal.

Now suppose the scene has one stop less contrast, 6 stops, 1.8

(1.8 - 0.4) / 1.05

0.75 is the aim contrast to develop the film.

Thanks Bill for that info.

 

[Stephen Benskin]

Flare plugs right into the simple formula that you use to decide the film contrast aim to fit the target paper grade.

Say paper grade 2 has Log Exposure Range 1.05

You have a classically normal Subject Luminance Range of 7 stops, 2.1

Take flare 1 1/3 stops, 0.4

(2.1 - 0.4) / 1.05

0.62 is the aim to develop your film for normal.

Now suppose the scene has one stop less contrast, 6 stops, 1.8

(1.8 - 0.4) / 1.05

0.75 is the aim contrast to develop the film.

Hey Bill, I believe you accidentally reversed the equation. LER / (LSLR - Flare) = aim CI

For Doremus. Personal choices and preferences are an important part of the process. It's an artform after all. The subject is called tone reproduction theory and it is based on sensitometric determination of the type of results obtained in practice. Let's not forget Jones' first excellent print test. It doesn't necessarily account for each photographer's personal taste but it gets you close enough so all the photographer needs to do is tweak. I don't know of any Zone System photographer that doesn't use the ISO rating as a jumping off point in their testing.

"many of us (especially those that use ZS metering techniques) end up finding that an E.I. different than box speed works better." I've written extensively on this point. The different results are mostly attributable to Zone System testing being a different methodology than ISO testing with the results being approximately the same as the ASA speeds obtained before the 1960 changes. Technically the results from Zone System testing produce a false sense of precision, but fortunately exposure doesn't have to be precise and giving 1/2 to 1 stop more exposure is just allowing for a higher safety factor, or personal taste. I believe I wrote in my last post that testing for speed for the individual was rather pointless because of the level of precision necessary to produce real meaningful results. ISO testing has already created a starting point in which to begin shooting and evaluation of the results based on the individual's personal approach.

The attached paper is how I determine processing.

 

[Bill Burk]

Right, I have switched the two sides. I use an HP 12c calculator with reverse Polish notation and often use the button to invert the result when it looks wrong. I forgot to rewrite.

 

[tih]

Thanks, I've not been able to find [Ralph Lambrecht's WBM resources] either.

Here are copies of the spreadsheet and accompanying PDF of the relevant pages from WBM, as downloaded by me in 2013:

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.xlsx

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.pdf

There is a minor bug in the spreadsheet:

On the "Input Data" sheet, you can (optionally) change the "Relative Dmin/Dmax" values in cells K49 and L49. If you do, you won't get the expected effect. This is because the difference between these two values (the wanted range) should be used in a few places, but isn't; instead, the constant 1.2 is.

On the "Curve <N>" pages, there are formulas in cells K49 and L49 to calculate "N", and they are, respectively, "=(2.1-(1.2/K48))/0.3" and "=(2.1-(1.2/L48))/0.3". Those "1.2" constants shouldn't be; they should be the calculated difference between the absolute Dmin/Dmax in K43 and K44, so, "=(2.1-(($K44-$K43)/K48))/0.3" and "=(2.1-(($K44-$K43)/L48))/0.3".

On the "Summary" page, the constant 1.2 is again used in cells P33 to P37. Here, the range end points aren't readily available on the page, so replace "1.2" with "($'Input Data'.$L$49-$'Input Data'.$K$49)"

I've changed this in my working copy, but not in the one linked to above; that one is as distributed by Ralph Lambrecht.

 

[RalphLambrecht]

(I was hoping that Ralph would be lurking...)

he is but You were asking for an explanation in somebody else's words and that is legit. I aleady said all I can about the subject,

 

[RalphLambrecht]

Here are copies of the spreadsheet and accompanying PDF of the relevant pages from WBM, as downloaded by me in 2013:

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.xlsx

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.pdf

There is a minor bug in the spreadsheet:

On the "Input Data" sheet, you can (optionally) change the "Relative Dmin/Dmax" values in cells K49 and L49. If you do, you won't get the expected effect. This is because the difference between these two values (the wanted range) should be used in a few places, but isn't; instead, the constant 1.2 is.

On the "Curve <N>" pages, there are formulas in cells K49 and L49 to calculate "N", and they are, respectively, "=(2.1-(1.2/K48))/0.3" and "=(2.1-(1.2/L48))/0.3". Those "1.2" constants shouldn't be; they should be the calculated difference between the abolute Dmin/Dmax in K43 and K44, so, "=(2.1-(($K44-$K43)/K48))/0.3" and "=(2.1-(($K44-$K43)/L48))/0.3".

On the "Summary" page, the constant 1.2 is again used in cells P33 to P37. Here, the range end points aren't readily available on the page, so replace "1.2" with "($'Input Data'.$L$49-$'Input Data'.$K$49)"

I've changed this in my working copy, but not in the one linked to above; that one is as distributed by Ralph Lambrecht.

Thanks for pointing that out. I'll have to check into that.

 

[Chuck_P]

Here are copies of the spreadsheet and accompanying PDF of the relevant pages from WBM, as downloaded by me in 2013:

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.xlsx

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.pdf

There is a minor bug in the spreadsheet:

On the "Input Data" sheet, you can (optionally) change the "Relative Dmin/Dmax" values in cells K49 and L49. If you do, you won't get the expected effect. This is because the difference between these two values (the wanted range) should be used in a few places, but isn't; instead, the constant 1.2 is.

On the "Curve <N>" pages, there are formulas in cells K49 and L49 to calculate "N", and they are, respectively, "=(2.1-(1.2/K48))/0.3" and "=(2.1-(1.2/L48))/0.3". Those "1.2" constants shouldn't be; they should be the calculated difference between the abolute Dmin/Dmax in K43 and K44, so, "=(2.1-(($K44-$K43)/K48))/0.3" and "=(2.1-(($K44-$K43)/L48))/0.3".

On the "Summary" page, the constant 1.2 is again used in cells P33 to P37. Here, the range end points aren't readily available on the page, so replace "1.2" with "($'Input Data'.$L$49-$'Input Data'.$K$49)"

I've changed this in my working copy, but not in the one linked to above; that one is as distributed by Ralph Lambrecht.

Thanks for that information, I look forward to checking into it.....and to Ralph's update if he will have one.

 

[Doremus Scudder]

Stephen and Bill,

Thanks for the replies. As you know, I don't have the equipment or the desire to do the type of testing that you both do. My testing approach is based on that expounded in The New Zone System Manual by White, Zakia and Lorenz. It is a visual approach and highly empirical as well as individual, relying on "proper proofs" and visual inspection. I get well-exposed and developed negatives that work well for my purposes.

I used to do all the ZS film-speed testing, but found, as you point out, that the ISO rating minus 2/3-stop for my ZS metering practices, works well as a starting point. Many of the film I use I end up rating at 1/3-stop slower than ISO, not really much of a difference and possibly simply developer-related.

I also agree that the ZS can lead to a false sense of precision. IM-HO, the main advantage of the ZS is its use as a tool for visualization. At least I can try to hit the window as far as exposure and development are concerned, that lets me make the print I envisioned when releasing the shutter.

Flare, however, seems to me to be such a variable factor in the equation that pinning it down to a particular value can only be a rough approximation. Using a well-coated lens in ideal lighting conditions with a hood and top-quality enlarging lenses introduces much less flare into the system than slapdash use of an uncoated lens without a hood and in situations with the sun or other brighter areas (even bright overcast) just out of the image area, etc. How do you deal with the variability of flare in your calculations? Simply find an average value and then let the chips fall where they may, or?

Best,

Doremus

 

[Sirius Glass]

Stephen and Bill,

Thanks for the replies. As you know, I don't have the equipment or the desire to do the type of testing that you both do. My testing approach is based on that expounded in The New Zone System Manual by White, Zakia and Lorenz. It is a visual approach and highly empirical as well as individual, relying on "proper proofs" and visual inspection. I get well-exposed and developed negatives that work well for my purposes.

I used to do all the ZS film-speed testing, but found, as you point out, that the ISO rating minus 2/3-stop for my ZS metering practices, works well as a starting point. Many of the film I use I end up rating at 1/3-stop slower than ISO, not really much of a difference and possibly simply developer-related.

I also agree that the ZS can lead to a false sense of precision. IM-HO, the main advantage of the ZS is its use as a tool for visualization. At least I can try to hit the window as far as exposure and development are concerned, that lets me make the print I envisioned when releasing the shutter.

Flare, however, seems to me to be such a variable factor in the equation that pinning it down to a particular value can only be a rough approximation. Using a well-coated lens in ideal lighting conditions with a hood and top-quality enlarging lenses introduces much less flare into the system than slapdash use of an uncoated lens without a hood and in situations with the sun or other brighter areas (even bright overcast) just out of the image area, etc. How do you deal with the variability of flare in your calculations? Simply find an average value and then let the chips fall where they may, or?

Best,

Doremus

Maybe they should only use flareless cameras and flareless lenses.

 

[Stephen Benskin]

Doremus, it's an average. Just like normal processing is based on the statistically average luminance range of 2.20. There's an average level of flare for a scene having a general distribution of tones. Flare is variable and most of it is caused by the subject. Two scenes with the same luminance range can have very different levels of flare is one is predominantly black with some white and the other is mostly white with a little black. Paper and printing techniques can be utilized and there is also the idea that flare is acceptable and even expected that in certain cases. You don't expect back lit scenes to look the same as front lit. The lower local contrast is expected and part of the look and mood.

Average flare with modern lenses is about 1 stop for a luminance range of 2.20. 35mm lenses with more lens elements generally have slightly higher levels of flare. Based on their aim CIs, Kodak has apparently been having a little debate over the average degree of flare. What is considered normal development begins with a scene with a 2.20 log luminance range. A CI 0.56 has about a stop of flare to just over a stop depending on rounding. CI 0.58 is about 1 1/3 stops of flare. Under Dick Dickerson, Kodak used CI 0.58 for normal. Before they used CI 0.56. In the last few years, it looks like it's back to CI 0.56. The difference doesn't really mean much in practice, and the saving grace to photography is that it doesn't really matter because you can do almost anything and still have a printable negative under average conditions the majority of the time; however, most of us here probably want something slightly better. So the idea is to aim for the bullseye in order to have the best chance of hitting the target. As with any model, how good it's considered to be is how well it works in the greatest number of circumstances. With an emphasis on greatest number of circumstances. For example, Jones found the fixed density method for determining speed isn't as precise across the board as the fractional gradient method; however, most people get by just fine using the fixed density method. The difference under average conditions, depending on film type, is negligible, but it becomes more obvious when the conditions become more extreme.

I put together something that is kind of interesting. I took the normal distribution curve of luminance ranges, and asked if a negative was processed to Normal (CI 0.58) under the different luminance ranges, where would it fall on the various paper grades. In other words, if I just went out and shot and processed everything normal, what grade would theoretically produce an acceptable print. One standard deviation includes 68% of all data points or more than half of all encounters. Below they would fall in the middle of Grade 1 to Grade 3, excluding personal taste. This suggests most personal testing methods don't have to be too accurate while still producing a usable negative in most cases.

 

[Bill Burk]

he is but You were asking for an explanation in somebody else's words and that is legit. I aleady said all I can about the subject,

Maybe it feels that way but you know, as a teacher, students always need to hear it again.

 

[Stephen Benskin]

Stephen.....you state in the paper:

"A Contrast Index or Average Gradient of 0.58 is considered statistically normal. It was once considered to be 0.56 until the value of flare changed from 0.34 to 0.40. I believe the adjustment reflects a shift away from the use of larger format cameras to 35mm cameras with their higher proportion of lens elements creating the slightly higher average flare."

I use a 4x5 camera and lens system. Throughout all these discussions on flare, and concerning that, I have learned much from your insight, and I have a better appreciation for this ever so irritating concept that is called......."flare". Would it be reasonable, then, for large format users with modern lens coatings to assume that their potential flare value would be .34 or less.

Yes it would. Good catch.

 

[Doremus Scudder]

Thanks so much for your detailed response, Stephen. I'll keep aiming for that bull's eye and hoping I hit the target most of the time I'm doin' pretty well most of the time.

Doremus

 

[Chuck_P]

Hey Bill, I believe you accidentally reversed the equation. LER / (LSLR - Flare) = aim CI

Stephen, does this apply equally to an aim gamma? Thanks.

 

[Stephen Benskin]

Stephen, does this apply equally to an aim gamma? Thanks.

As Normal for gamma is in the low 0.70s and normal for CI is in the high 0.50s, it probably wouldn't be a good fit. Gamma wouldn't be my first choice as it only applies to the straight-line portion. Almost any average gradient method would be more applicable.

 

[Chuck_P]

As Normal for gamma is in the low 0.70s and normal for CI is in the high 0.50s, it probably wouldn't be a good fit. Gamma wouldn't be my first choice as it only applies to the straight-line portion. Almost any average gradient method would be more applicable.

Thank you

 

[Leszek]

Poniżej znajdują się kopie arkusza kalkulacyjnego i towarzyszącego mu pliku PDF odpowiednich stron z WBM, pobrane przeze mnie w 2013 r.:

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.xlsx

https://hamartun.priv.no/tih/photo/FilmTestEvaluation.pdf

W arkuszu kalkulacyjnym znajduje się niewielki błąd:

W arkuszu „Dane wejściowe” możesz (opcjonalnie) zmienić wartości „Względne Dmin/Dmax” w komórkach K49 i L49. W przeciwnym razie nie uzyskasz oczekiwanego efektu. Dzieje się tak, ponieważ różnica między tymi dwiema wartościami (pożądany zakres) powinna być użyta w kilku miejscach, ale tak się nie dzieje; zamiast niej używana jest stała 1,2.

Na stronach „Krzywa <N>” w komórkach K49 i L49 znajdują się formuły do obliczania „N”, odpowiednio: „=(2,1-(1,2/K48))/0,3” i „=(2,1-(1,2/L48))/0,3”. Stałe „1,2” nie powinny występować; powinny być obliczoną różnicą między bezwzględnymi wartościami Dmin/Dmax w K43 i K44, czyli „=(2,1-(($K44-$K43)/K48))/0,3” i „=(2,1-(($K44-$K43)/L48))/0,3”.

Na stronie „Podsumowanie” stała 1,2 jest ponownie używana w komórkach od P33 do P37. W tym przypadku punkty końcowe zakresu nie są łatwo dostępne na stronie, dlatego należy zastąpić „1,2” wartością „($'Dane wejściowe'.$L$49-$'Dane wejściowe'.$K$49)”.

Zmieniłem to w swojej kopii roboczej, ale nie w tej, do której link znajduje się powyżej; tę rozprowadza Ralph Lambrecht.

Can you insert a file corrected in the old version of the .xls spreadsheet?

 

[RalphLambrecht]

(I was hoping that Ralph would be lurking...)

lurking and thinking...

 

[spookyphoto]

Being a retired photo analyst I might suggest you take a look at the book "Exposure Manual" by Dunn & Wakefield. It's out of print but can still be found if you Google it. The list of references in it is worth the price of the book. It will list the many ways film speed has been calculated over the years and will explain the relation of film speed and gamma. The Jone's point (fractional gradient point) derived by calculating delta X is where you will find the answer to most of your questions. Sensitometric history is really quite interesting.

 

[Alex Benjamin]

Being a retired photo analyst I might suggest you take a look at the book "Exposure Manual" by Dunn & Wakefield. It's out of print but can still be found if you Google it. The list of references in it is worth the price of the book. It will list the many ways film speed has been calculated over the years and will explain the relation of film speed and gamma. The Jone's point (fractional gradient point) derived by calculating delta X is where you will find the answer to most of your questions. Sensitmetric history is really quite interesting.

Good suggestion, but most of these questions were asked eight years ago...

 

[spookyphoto]

Good suggestion, but most of these questions were asked three years ago...

I was answering these same questions 40 years ago and even that was 40+ years late

 

[Alex Benjamin]

I was answering these same questions 40 years ago and even that was 40+ years late

Well, we all reach that age where, as with film, our own speed becomes kind of less and less effective...

 

[MattKing]

Well, we all reach that age where, as with film, our own speed becomes kind of less and less effective...

And fog becomes more and more prominent! .....

 

[RalphLambrecht]

after taking more test shots than actual pictures,I arrived at the same conclusion as Doremus and I think what AA proposed but as an Engineer, I feel more comfortable with my conclusions with the data backing hem up; different EIs for different developmentsmake sense because, development affects foot speed.

Doremus Scudder said:
It's not true that Adams used the same E.I. for all development schemes. I believe he advocated establishing a separate E.I. for each development scheme.
exactly. That is the whole point of having several EIs.