Effective film speed (as in Lambrecht's method)

[Doc W]

I am a little confused as to the meaning of "effective film speed" in Lambrecht's method of film testing. He uses box speed in testing for development times. Once these have been established, he tests for EI by using the typical grey card method until he gets a density of 0.17 (Zone 1.5). This is his Normal EI.

I noticed that when I plug different EI's into his spreadsheet, it has a dramatic effect on "effective film speed" for the various develop times. Without arguing about whether he is correct (I REALLY REALLY don't want another zone system donnybrook), what does one do with this in the field? If I find a subject that requires N-1 development, do I meter the scene using the corresponding effective film speed? Or do I just shoot at the Normal EI? If so, what is the purpose of the other effective film speeds.

IIRC, Adams' method first determines the EI and then finds development times using that EI. Lambrecht establishes development time using box speed, the other way around.

I have an APUG friend who has been trying to explain all this to me but when it comes to math and curves, I am a little thick. I am hoping that another explanation may open up my understanding. It may just confuse me more, but then I will be no worse off.

Once again, please, no zone system wars.

 

[pentaxuser]

Sounds as if the key contribution here which might have the added benefit of stopping a zone war is one from Ralph himself. He's around quite a lot so here's hoping.

pentaxuser

 

[Doc W]

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

 

[aparat]

I hope I understood your question correctly. I apologize if I didn't.

That is a very interesting question. I assume you have the book Way Beyond Monochrome. My understanding is that the film test has two parts. First, you make five exposures of a step tablet on a light table (or via a suitable device) and plot the characteristic curves. You can expose the film at box speed, as it is not critical at this stage. The curves will allow you to derive important information about the effect of exposure and development on density. After you've analyzed the curves, the authors suggest you take the approach described on p. 223 in fig. 12. I attached a plot I made recently according to that method. You would then do another test, this time focusing on speed. You would expose a few frames at different ASA settings (It's similar to the Fred Picker approach described in Zone VI Workshop), and you read the values with a densitometer to see which frame gives you 0.17 (or whatever you choose) over BF. You would then use the graph I attached to locate your preferred G (e.g., 0.5) and mark the corresponding log exp value on the x-axis. This is going to be your EI. Then, you'd measure the relative distance to that point for each of the curves (see the white labels on my graph), where 0.1 log exp is about 1.3 stop.

Knowing your EI, you can then make a few related plots, to help you make exposure and development decisions. I attached an example of such a plot. Hope this helps.

 

[Doc W]

Thanks, aparat. That clears up a few things. But what about these effective film speeds? Adams first tested for EI, i.e., one film speed, which wouid be used for all exposures, whether for N-1, N+2, whatever. Lambrecht has a series of film speeds, each associated with a corresponding N development time. What do I do with these? What is their purpose?

 

[aparat]

Thanks, aparat. That clears up a few things. But what about these effective film speeds? Adams first tested for EI, i.e., one film speed, which wouid be used for all exposures, whether for N-1, N+2, whatever. Lambrecht has a series of film speeds, each associated with a corresponding N development time. What do I do with these? What is their purpose?

Yes, you are right. They took a very different approach. They studied the characteristic curves first, exposing at box speed. The problem with doing a film speed test first is that, ideally, you'd need a well-calibrated exposure system, or sensitometer, so you can comply with the ISO standard and compare your speeds to the box speed. But that's difficult for most photographers to achieve, so they suggested a different, more "fluid," approach. Yes, you have multiple film speeds (EFF), but that makes sense only in terms of selecting the most appropriate combination of factors for any given scene. These EFF speeds are an interpretation of the data. Film (and paper) can be characterized by multiple speeds, it's just that the ISO standard chooses to pick one specific speed point (0.1 over BF).

The idea is that once you determined your nominal film speed (for them, it is at 0.17 over BF), you can then reinterpret the curves. A lot of photographers are motivated by a specific Average Gradient (G) or Contrast Index (CI), and you can get that from the data very nicely. I am attaching another plot that shows you just one of the many possible interpretations of the data. For 8 minutes, the EFF is 80-, CI is about 0.6, etc., and from such a plot, you can derive a lot of useful information to meet the demands of your vision and the scene. So, if the Subject Brightness Range is about 6, and I want to develop for CI 0.6, I would set 80- into my light meter, expose, and process for 8 minutes. That's one example of how you can interpret the data. Different photographers would do it differently. But you have the raw data at your fingertips, so to speak.

 

[Adrian Bacon]

I primarily scan so I don’t do any of that and largely don’t care about the zone system.

I test development for a given contrast, usually 0.62 unless I have reason to go for something different, and once I have a development time for that contrast with a given developer, I determine the minimum exposure required to produce 5 full stops of exposure from middle gray to film base plus fog. That minimum exposure is the Maximum EI (eg EI 800) you can shoot at and still retain reasonably useful shadow detail. I then go the other way and determine the maximum amount of exposure you can go from 7 stops over middle gray until the film effectively stops getting denser, and the number of stops over the 7 stop mark is then subtracted from the maximum EI and that is the minimum EI (eg EI 50) you can safely shoot at. I then generally will pick a value roughly in the middle of those two values to use as my rated EI for that development time with a reasonably healthy margin of over/under exposure without having to modify development time. So for example, fomapan 100 with a given developer and time is EI 160 and I can expose 9 stops over middle gray, that would be 9-7=2 stops, so 160/4 (2 stops) = EI 40 as the minimum EI, the halfway mark between EI 160 and EI 40 is EI 80, with a nice healthy amount of shadow and highlight detail with a good amount of wiggle room on both sides.

Not surprisingly, most of the films I’ve done this with, I’ve ended up pretty close to the box speed if not right at the box speed. You can of course develop to a lower or higher contrast and have different effective film speeds, but for me, this is the simplest and is also pretty informative as to what the film will tolerate if you’re not trying to fit it on paper via analog means.

 

[Doremus Scudder]

Thanks, aparat. That clears up a few things. But what about these effective film speeds? Adams first tested for EI, i.e., one film speed, which wouid be used for all exposures, whether for N-1, N+2, whatever. Lambrecht has a series of film speeds, each associated with a corresponding N development time. What do I do with these? What is their purpose?

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.
In any case, N+ development increases effective speed a bit and N- reduces it. As a long time ZS practitioner, I've arrived at a point where I have 5-6 different E.I. for a given film, depending on the development I plan to use. These have been tested and tweaked with field experience.

However, in practice one can often ignore the speed increase for N+ exposures, since additional density supports the shadow details and can be printed through. And many just "support the shadows" for N- development (i.e., give more exposure by "guestimate" rather than testing), which can work just fine as long as one errs on the side of too much exposure. An extra 1/2-1 stop won't hurt much.

FWIW, after lots of extensive testing, I've found that my E.I. for N comes in consistently at 1/3-2/3 stop slower than box speed. For new films now, I simply start testing development schemes at 2/3 slower than box speed and refine that with field tests. For all of the films I currently use, I have an E.I. of 1/3-stop slower than box speed. For N+ developments I increase speed by 1/3 stop for each number (I rarely develop more than N+1 these days) and for N- developments I now use SLIMTs for everything, but add an extra 1/3 stop anyway, even though the SLIMT procedure retains film speed well.

Best,

Doremus

 

[RalphLambrecht]

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.
In any case, N+ development increases effective speed a bit and N- reduces it. As a long time ZS practitioner, I've arrived at a point where I have 5-6 different E.I. for a given film, depending on the development I plan to use. These have been tested and tweaked with field experience.

However, in practice one can often ignore the speed increase for N+ exposures, since additional density supports the shadow details and can be printed through. And many just "support the shadows" for N- development (i.e., give more exposure by "guestimate" rather than testing), which can work just fine as long as one errs on the side of too much exposure. An extra 1/2-1 stop won't hurt much.

FWIW, after lots of extensive testing, I've found that my E.I. for N comes in consistently at 1/3-2/3 stop slower than box speed. For new films now, I simply start testing development schemes at 2/3 slower than box speed and refine that with field tests. For all of the films I currently use, I have an E.I. of 1/3-stop slower than box speed. For N+ developments I increase speed by 1/3 stop for each number (I rarely develop more than N+1 these days) and for N- developments I now use SLIMTs for everything, but add an extra 1/3 stop anyway, even though the SLIMT procedure retains film speed well.

Best,

Doremus

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.

 

[aparat]

As michael_r pointed out, there are potential shortcomings related to using a fixed EI, particularly within the context of estimating exposure based on a fixed, low density point, such as 0.1 over BF. There is, however, a potential advantage to using a more fluid interpretation of the data, i.e., allowing the EFS values to vary (as the authors of WBM point out). For example, if you typically expose for shadows (e.g., Zone III), you can analyze the data by "shifting" the speed point to Zone III, which might give you a more accurate exposure, or at the very least, make the "placement" process more intuitive.

My personal view is that there are many possible interpretations of the raw film (or paper) data. Nothing is written in stone, and each photographer can interpret the data in a way that suits their process. Hence, as the OP (Doc_W) very nicely pointed out, there is no need for Zone System wars by allowing for multiple interpretations of the raw data, which, ultimately, serve the photographer's artistic vision.

 

[Doc W]

For those of you doing zone system testing, how do you produce the first negatives? Here is what I do for the first set of negatives.

With my spot meter set at box speed, I measure the light from my light table (which is at a colour temperature close to that of daylight) until it is stable. It always settles on EV 13. I open up five stops from that reading. I load a film holder with both a negative and the step wedge. I focus on infinity and shoot five negatives and develop them at five different times to supply data to Ralph's spreadsheet.

I do it this way because the light table is a constant and consistent light source. Taping the step to a window does not work for me because there is no window in my house that will give me consistent light. Furthermore, to get close enough so that the step wedge will be large enough on the negative would probably require bellows extension. Ralph says to use either an "average reading" or "a spot meter for the medium gray bars." I am not sure what he means in either case. Ralph?

What do you folks do to get this initial data?

Getting some control over contrast has been driving me crazy for ages.

 

[Doremus Scudder]

One shortcoming of this type of EI test (fixed density criterion) is that it overestimates the change in speed with changes in gradient.

Michael,
Of course, this is correct and becomes a bit of an issue when one does all the Zone System calibrations using a densitometer and some arbitrary FB+Fog value as a target. I simply make proper proofs and compare my low-value placements with what really happens. If there is an important difference, I'll tweak the E.I. for a particular development time to compensate. Not as precise as it could be done with a lab full of equipment, but it gets me the connection I need between metering and low-value rendering. FWIW, my usual adjustment between N and N-1 or N+1 is 1/3 stop; not very much.

Best,

Doremus

 

[aparat]

What do you folks do to get this initial data?

Getting some control over contrast has been driving me crazy for ages.

I use the method described in the BTZS book, and in some of the old BTZS newsletters, i.e., using an enlarger as a calibrated light source. There's no camera or lens involved, of course, but, to some extent, the lens performance (e.g., flare) can be accounted for by the data analysis software, if need be.

 

[Bill Burk]

Photographing a backlit step wedge also introduces flare, which can distort the curve you end up plotting.

That's putting it mildly. It does distort the curve.

Taping a step wedge to a window and photographing it creates a test negative that includes camera flare. It's hard to interpret the curves and the interpretation is inaccurate.

Taping a step wedge in contact with the film and exposing it with light formed by the lens through the shutter of the camera... does not include flare.
This sound like the way Doc W does it.
That's the right way to do it.
Also aparat's way is the right way too.

 

[Stephen Benskin]

One shortcoming of this type of EI test (fixed density criterion) is that it overestimates the change in speed with changes in gradient.

Michael has packed a lot into a single sentence. What he is referring to is the methodology of film speed determination. Accurate film speed has to do with the toe gradient in comparison with the average gradient. The ISO's fixed density of 0.01 over Fb+f is only accurate when adhering to the contrast parameters specified. If the film is developed outside those parameters, using only the fixed density point is no longer accurate. The ISO uses the Delta-X Criterion. While it uses a fixed density method to determine film speed, it is only part of a larger equation. When using the ISO's contrast parameters, the Delta-X equation is built into the method and brings into agreement the fixed density method of 0.10 over Fb+f with the fractional gradient speed point. Delta-X can be regarded as a simplified way of determining the fractional gradient method.



Given that only a small portion of people realize this and have made the necessary adjustments, yet still obtain quality results that seeming support their testing method brings into question the necessary accuracy required in any methodology. Adams developed the Zone System speed method when ASA film speeds averaged 2/3 to 1 stop slower than what they would test at today only do to using a different methodology prior to 1960. Adams never m ade changes post 1960 to adjust for the ASA speed method, and no one noticed. In effect, the difference in the relationship between the meter's exposure point and the b&w speed point with the Zone System and the ISO standard for the fixed density point of 0.10 is 2/3 stop. Zone System stops down four stops from the metered point. The relationship between the fixed density point of 0.10 and the metered exposure point is 3 1/3 stops. So even if a given film is tested perfectly using each method, eliminating all variables, the resulting speed values will differ by 2/3 stops.

Many common testing methods start with the ISO to determine the individual's working speed. For most general purpose developers, any discrepancy from the ISO speed is more likely the result of testing errors or the methodology used. Speeds outside the ISO contrast parameters as Michael states overestimate the changes in film speed resulting from changes in processing. According to the Delta-X Criterion, there is an inverse relationship between the average gradient and the Delta-X speed point. This is caused by changes in the toe's gradient in relationship to the average gradient. While the density of the fixed density point will increase or decrease in direct proportion to processing, the change of the toe's gradient to the average gradient with processing tends to dampen changes to the resulting film speed compared to a strict fixed density point.

Choice of methodology determines results. The question is if the all the varying results resulting from the different common methods used apparently aren't enough of a factor to be obvious to the average tester or affects quality enough to be noticeable, is there a need for the average person to test for speed?

 

[Stephen Benskin]

Photographing a backlit step wedge also introduces flare, which can distort the curve you end up plotting.

Agreed. Always exclude flare from the test and then to factor it in when interpreting.

 

[Bill Burk]

Doc W, getting to your original question about varying effective speeds at different contrast gradients...

The fixed 0.1 density "criterion" is just one way of many to determine film speed. It's easy to calculate and Ralph Lambrecht used it in his spreadsheet.

When you take the fixed 0.1 density criterion literally, effective film speed changes exactly as you observed, at different contrast gradients.

What I see in the following graph is that you will tend to get 1/3 to 1 1/3 stop additional exposure above minimum required exposure for an excellent print... when you use 0.1 density criterion.

There are other criteria you could use, and get closer to the correct exposure indication, for example Delta-X which is +/- 1/3 stop of the minimum exposure for an excellent print.

I host Stephen Benskin’s collection of some relevant information here

http://beefalobill.com/benskin/

 

[Stephen Benskin]

The best way to determine the conditions that will produce a quality print is by psychophysical methods. The print judgement speed is based on the prints a group of viewers thought had the highest quality and then working backwards to determine how the negatives were produced. The task is then to find a sensitometric speed method that most closely agrees with the print judgement speed in the majority of situations with the majority of film / developer combinations. The smaller the differences, the better the speed method, as Bill has shown with the spread function graphs. Something else to keep in mind, speed point isn't necessarily where the shadow of the photographic exposure should fall. The speed point may only be a point that is sensitometrically determinable from which the exposure can be derived. One idea is to find the minimum useful point of exposure. Anything below this point returns prints of reduced levels of quality. This isn't necessarily the ideal point to place exposure. The minimum useful point of exposure defines the limits of the film, any exposure above it, within a definable range, will produce quality prints. It's easy enough to take the minimum useful point and add an adjustment of one stop for example. So instead of a speed of 200, the film has a speed of 100 which has the exposure point fall 0.30 log-H to the right of the speed point.

The Fractional Gradient Method has a range of densities for the exposure point depending on the film's toe and average gradient. This means that something like a true proof is based on a false assumption. Ralph's 0.17 is about what the shadow exposure point was with the Fractional Gradient Method and in reality, the Zone System. It will produce excellent prints and probably slightly higher quality results than with 0.10 as the exposure point. The main problem with 0.17 is when using it as a speed point. Because it is higher on the curve, changes in processing will affect it to a greater extent than a speed point at a lower density. Resulting film speeds for higher or lower gradients will tend to fall further from the ideal print judgement speed for the same gradient, but as most people only use one or two films, they will not notice the discrepancies. Does that then make the point academic? Maybe, but as the ISO film speed already has an accurately defined speed and exposure point, if you like the idea of a little more exposure, why not use the ISO speed and open up 1/2 to 1 stop.

 

[Chuck_P]

Taping a step wedge in contact with the film and exposing it with light formed by the lens through the shutter of the camera... does not include flare.
This sound like the way Doc W does it.
That's the right way to do it.
Also aparat's way is the right way too.

I am now ready to carry out my film tests for EFS and development times with TMX100 and XTOL.....I have been doing some searching to try and find just this type of confirmation, on contacting the tablet to film via the enlarger or camera system.

Also I'm now using the BTZS WinPlotter program, I have been wondering why Davis recommends using the enlarger to contact the step tablet to film versus using the camera system. I lean toward contacting the film in the film holder like I've done in the past (Schafer) when I did not use the BTZS analysis program.......it was cool learning that using that program, it'll let you input a flare factor of up to a full stop. Also, it appears you're able to select the final curve analysis using, either, the 0.1 over fb+f fixed density criterion, or, CI method. So I'm interested in seeing the changing EI's with changing development times.

Also, I can't see, at least in Adam's later book, "The Negative", where he actually uses different EI's for different development times. However he clearly alludes to changes to the toe region, at the threshold exposure point with with N- and N+ development......suggesting that, necessary exposure adjustments should be guided by experience.....as the way I read it.

 

[Bill Burk]

Some teachers overemphasize the importance of performing sensitometry using the same light source as the photographs you intend to shoot with.

Don’t get me wrong, it’s important. But consistent exposure is more important.

Most sensitometry studies are exposed using tungsten or electronic flash light sources as an approximation of sunlight because the sun is very difficult to control.

An enlarger with tungsten light filtered with a blue filter (I think 80B), would be great. A flash as simple as maybe an old Vivitar 283, might serve well.

It would be good if you can get the exposure to within the range of reciprocity (e.g., 1/100 to 1 second).

That said, I have studied your curves and the ones you make taping the Stouffer scale to the film in the holder are very good. You have the technique down so if you are comfortable doing it in camera, you will get good results.

If you had the space and an enlarger that you could dedicate to the task, that would be great too.

 

[Bill Burk]

You can download and use Ralph Lambrecht’s spreadsheet to input your results. It’s a great tesoirce.

BTZS method of sliding the EI with development time is due to a fallacy.

 

[Chuck_P]

That said, I have studied your curves and the ones you make taping the Stouffer scale to the film in the holder are very good. You have the technique down so if you are comfortable doing it in camera, you will get good results

Thanks Bill, that boosts my confidence level a bit. I initially did it by taping a 1/2 inch x 5 inch step tablet to a sheet, then changed to a 4x5 sheet by sliding both into the holder, it's snug but doable.

 

[Chuck_P]

BTZS method of sliding the EI with development time is due to a fallacy.

Could you please elaborate on this more, thanks.

 

[Bill Burk]

The fallacy with BTZS sliding EI with development comes from it’s assumption that the 0.1 density above base+fog is the point where exposure is to be indexed.

It’s not. It’s just an easy point to find.

The point to index from is the point where the curve reaches a third of its average gradient.

Our friends at Kodak went to a lot of trouble to find out this tidbit of info, but finding this third of average gradient was hard. DIN campaigned with the standards organizations to call the speed point 0.1 above base plus fog. Any darkroom tech could find that.

They compromised in the end and said “fine”. You can have the 0.1 speed point but only when you develop the film to a specific contrast. Any other contrast doesn’t count.

When film is developed to ASA parameters the 0.1 “speed point” is 0.29 away from the one-third average gradient point.

That’s really the point under which you don’t want any shadow of your picture to fall, and the ASA/ISO speeds all make it so (for black and white negative film).

This actual point Kodak says we’re looking for is further down the toe, where the curve is turning so flat you can’t get anything more out of it.

If you develop film less, the whole curve gets flatter. You print on higher contrast paper that can get detail out of the slight detail, you can slide further down from the 0.1 speed point.

If you develop more, then you need to use lower contrast paper (assuming normal contrast subject and the goal to make an excellent print - not talking of pushing when you don’t care about print quality so much). Now your detail falls off closer to the 0.1 speed point.

In other words, effective speed of the film is fairly inflexible, it’s pretty close to the rated film speed under a wide range of development times.

Stephen Benskin wrote about this in threads about Delta-X.

If you accept this idea then you can just call the speed only when it’s met ASA parameters and don’t change the speed for different development times.

 

[Doremus Scudder]

Bill,

A question for you (and maybe Stephen Benskin if he feels like addressing it):

Both you and Stephen advocate a testing method that removes flare from the equation. What, then, do we do with our new-found film speed when we go out to make photographs in real-life situations? Do we just ignore the flare, which is sure to be there to some extent or other? Or do we compensate for it somehow?

My more empirical testing, as you likely know, uses all the flare inherent in my system: taking lens, enlarger lens, etc. I'm doing just fine with exposure and development using my methods, I'm just curious how the sensitometry under controlled circumstances gets translated into practice. Are we really just ending up with a lot of pre-flashed negatives due to camera flare regardless of how we rate our film?

TIA,

Doremus