True ISO rating of Tri-x 400 ?

[Sirius Glass]

Is this the book you are referring to, or a different edition?

Fifth Edition.

 

[Sirius Glass]

Isn't that like shooting it at ASA 100?

Yes, but other than overcast conditions or snow, I use the box speed, 400, with the built in meters.

Steve

 

[gregmacc]

Thanks all for the replies so far .... A lot of interesting information and opinions .... I think I'll leave the camera speed set to 400 ISO and give the TriX 20% longer development and see what happens.

 

[RalphLambrecht]

...The fractional gradient method of speed determination has always been regarded as the most accurate method...

The fractional gradient method is around since at least the 1940s. It overcame a few issues inherent to earlier (much older) systems that were based on threshold, fixed density and inertia. In the 1960s the fractional gradient system was dropped in favor of the 0.1 density>b&f system, because the new system delivered virtually the same results and is much easier to use than the fractional gradient system. Who knows, if they had PCs in the 1960s, the decision might have been different.

The current ISO standard is still based on the simple 0.1 density>b&f system, but it was improved by the implementation of a standard developing scheme to eliminate the influence of negative contrast and increase accuracy even further.

As far as I know, nobody is seriously considering to go back to the fractional gradient method.

 

[Colin Corneau]

Thanks all for the replies so far .... A lot of interesting information and opinions .... I think I'll leave the camera speed set to 400 ISO and give the TriX 20% longer development and see what happens.

This has always worked very very well for me, in terms of increasing contrast.

 

[Stephen Benskin]

The fractional gradient method is around since at least the 1940s. It overcame a few issues inherent to earlier (much older) systems that were based on threshold, fixed density and inertia. In the 1960s the fractional gradient system was dropped in favor of the 0.1 density>b&f system, because the new system delivered virtually the same results and is much easier to use than the fractional gradient system. Who knows, if they had PCs in the 1960s, the decision might have been different.

The current ISO standard is still based on the simple 0.1 density>b&f system, but it was improved by the implementation of a standard developing scheme to eliminate the influence of negative contrast and increase accuracy even further.

As far as I know, nobody is seriously considering to go back to the fractional gradient method.

Only the method in which to determine the fractional gradient method was dropped. The fractional gradient concept, technically, was never dropped. That is an urban myth. They found that under certain contrast parameters, you could achieve the equivalent of the fractional gradient method using a fixed density method. For anything outside of those parameters, you need to incorporate the Delta-X criterion formula. The problem with the fractional gradient method is not its accuracy. Jones' testing and his exposure equation still holds up. The problem was a problem in implementing the fractional gradient method. It was found that finding the point where the gradient was 0.3X the average gradient was prone to error. The fixed density method fixes it, but it doesn't change the fundamental principles of the fraction gradient concept. Just the method. So, the fractional gradient method is alive and well hidden within a fixed density method.

The information is out there. You just have to know where to look. There are two papers that came out around the time of the change. Generally, the guys at Kodak would publish a paper to detail the theory and reasoning why changes were made to a standard. For the big change in film speed determination method, they are:

Nelson, C.N. and Simonds, J.L., Simple Methods for Approximating the Fractional Gradient Speeds of Photographic Materials, Journal of the Optical Society of America, V. 46, N. 5, May 1956.

Nelson, C.N., Safety Factors in Camera Exposures, Photographic Science and Engineering, V. 4, N. 1, Jan-Feb 1960.

According to Nelson in Safety Factors, "If a specification is adopted requiring development to a delta D or 0.80 or an average gradient or 0.62, for example, the log H difference (Delta X) between the two types of speed becomes 0.29; and the exposure, Hm, at a density of 0.10 above fog becomes 1.9 times greater than the exposure Hs at the fractional gradient speed point...If this interval were constant, the two types of speed would correlate perfectly."

Every once in a while the topic of why the ISO standard has for its target a gradient of 0.61 to 0.62 which is about 1/2 stop above normal processing. The above explains it. And when you look at how speed works with the fraction gradient method or the Delta-X Criterion, you will find that it doesn't make a difference in film speeds.

A little later, Nelson concludes, "The fractional gradient speed criterion (and it's approximate equivalent, the simpler Delta-X speed criterion) will continue to be useful as a supplement to the fixed density speed criterion when an evaluation is desired of the effective speeds of films that have been developed to average gradients higher or lower than the proposed standard average gradient. The fixed density criterion tends to underrate films that are developed to a lower average gradient and to overrate films that are developed to a higher average gradient. A new constant in the formula for fractional gradient speed is desirable for this non-standard application in order to provide a safety factor of about 1.2 and thus make the speed comparable with the proposed fixed density speeds."

 

[RalphLambrecht]

...Every once in a while the topic of why the ISO standard has for its target a gradient of 0.61 to 0.62 which is about 1/2 stop above normal processing...

Stephen

I agree with what you're saying. I have the papers you mentioned and read them. I also discussed this subject many times with Phil Davis, when I tried to develop a Mac version of his software. Now and then, I just don't think that the current ISO standard would greatly benefit from a change. By fixing the negative gradient, the method delivers reliable results.

It is, however, unfortunate that the standard picked a gradient, which is a bit harder than what most people prefer. This does indeed lead to about 1/2 stop above normal processing, as you point out. I have fixed that for me by modifying the standard to a gradient of 0.57 (1.2 density range/2.1 exposure range) and standardizing at a foot-speed density of 0.17 (instead of 0.1) at Zone I.5 (instead of Zone I).

Maybe we should compare both systems for modern films and typical contrast ranges to see if the fractional gradient method is worth the effort. With modern computers, the effort is not much of an issue anymore anyway. It would be simple enough to take a few characteristic curves and run speed calculations with both methods.

 

[df cardwell]

With all respect to Phil, bless him, and your effort,
it is a wonderful thing that we can get to the truth by different routes.

The 'correct' method is simply nonfunctional for about half photo population,
(and we really are no more stupid than those for whom BTZS is The Way.)

Have we totally hijacked this from the O.P. ?

 

[Stephen Benskin]

Stephen

I agree with what you're saying. I have the papers you mentioned and read them. I also discussed this subject many times with Phil Davis, when I tried to develop a Mac version of his software. Now and then, I just don't think that the current ISO standard would greatly benefit from a change. By fixing the negative gradient, the method delivers reliable results.

It is, however, unfortunate that the standard picked a gradient, which is a bit harder than what most people prefer. This does indeed lead to about 1/2 stop above normal processing, as you point out. I have fixed that for me by modifying the standard to a gradient of 0.57 (1.2 density range/2.1 exposure range) and standardizing at a foot-speed density of 0.17 (instead of 0.1) at Zone I.5 (instead of Zone I).

Maybe we should compare both systems for modern films and typical contrast ranges to see if the fractional gradient method is worth the effort. With modern computers, the effort is not much of an issue anymore anyway. It would be simple enough to take a few characteristic curves and run speed calculations with both methods.

I never liked the Davis programs, so I wrote one myself. My biggest problem is that he tended to combine the camera image/flare curve with the film curve and that his flare factor was a guesstimate calculation. His program also only dealt with relative film speeds and I use a sensitometer. I also wanted some specific tools to help evaluate the curves and tone reproduction.

Concerning the current ISO standard, I believe you are misunderstanding me. I'm not proposing making changes to it. I'm saying that the Delta-X criterion is an integral part of the fixed density method from the very beginning. Using the fixed density method in conjunction with the Delta X equation is the same as using the fractional gradient method. The only reason why you don't need to use the formula with the ISO standard is because the standard has it built into the parameters for normal. One of the reasons why the 1993 version eliminated the range of Delta X from +- 0.05 is to increase the accuracy of Delta X which is the difference between the 0.10 fixed density speed point and the fractional gradient speed point (Δ 02.28). I've done a lot of research on this and have written a paper which PHOTO Techniques think is too esoteric. While doing the research, I played a lot with the math and discovered some interesting results.

ΔX Equation

ΔX = 0.83 - (0.86 * ΔD) + (0.24 * ΔD^2)


ΔD CI Fixed Density ΔX
0.70 0.54 108 121
0.80 0.62 125 125
0.90 0.69 144 129

Rounded you get 100, 125, and 160 for the fixed density speeds and 125 for all ΔX speeds. As you can see, based on the ΔX criterion, the film gradient doesn't have that much of an affect on film speed, so having the standard's gradient higher than pictorial normal isn't really a factor. This is why the standards using the fractional gradient method didn't have specific contrast parameters except for a minimum requirement of over a gamma of around 0.50. I think Nelson, who worked with Jones, understood this and didn't make a misjudgment that requires adjustment. I can't be certain about this, but the person who wrote the scientific paper that accompanied the release of the standard usually chaired the subcommittee.

Delta-X is just one of those things which have been lost over the years, but only to the general public. Since the subcommittee didn't change the parameters during the last major revision, I tend to believe it is alive and well in the scientific community.

This shouldn't be surprising. How many times have people written about how film speeds were a stop slower before the 1960 standard because of a higher safety factor (safety factor is a simplification of the reason but) yet never realized that the Zone System testing never changed its testing method? Or that while the difference between the meter calibration point and the ISO speed point is 3 1/3 stops. The Zone system stops down 4. Or don't realize flare plays a part in film speed when the shadow exposure for a statistically average 7 1/3 stop scene falls 4 1/3 stops below the meter calibration point and the ISO point is only 3 1/3 stops. Or that average flare is really 1 1/3 stops which would then have the actual exposure fall 1/3 stop above the ISO speed point. Which also means that considering where exposure falls from flare and considering the concept of ΔX and how the speed doesn't move as much as with fixed density method, the speed point of 0.10 isn't really the target density for exposure. Much like the fractional gradient speed point which falls almost a full stop below the ISO fixed density speed point when processed normally, which one would logically assume would make fractional gradient film speeds one stop higher than the ISO fixed density speeds, yet before 1960, they were one stop slower than what would be obtained by the fixed density method. Heck, how many people think the meter sees or wants to expose for 18% gray?

 

[RalphLambrecht]

Stephen

After all this, I would be interested in seeing this paper you offered to Photo Techniques.

 

[sanking]

I never liked the Davis programs, so I wrote one myself. My biggest problem is that he tended to combine the camera image/flare curve with the film curve and that his flare factor was a guesstimate calculation. His program also only dealt with relative film speeds and I use a sensitometer. I also wanted some specific tools to help evaluate the curves and tone reproduction.

With the Davis program program one chooses whether to include a flare factor in the curve calculator. If you choose to use it you then enter a value.

In the book Beyond the Zone System there is a fairly detailed procedure for determining flare with individual lenses. Having gone through the procedure several times my own conclusion is that for practical photography it is over-kill and one can do just as well by applying a factor according to lens type.

I also use a sensitometer and often wondered why Davis did not. But I guess it is simply due to the fact that he was teaching a practical system to folks not likely to have access to a good sensitometer.

I would be curious to know if your program is available, and if so, what operating system it uses.

Sandy King

 

[Stephen Benskin]

Sandy,

Not chosing Phil Davis' program was mostly personal taste and need on my part. I felt it was important to have a four quadrant reproduction curve so that I have a stand alone curve for the camera image. For my purposes, I wanted to know meter candle seconds and where values fell so I could research exposure theory. I wrote Phil and asked him about how he calculated flare and he basically said that he used an equation that tended to give acceptable flare values. I needed to know how flare was calculated and all the values for the variables for my research so I used the Jones exposure equation with a heavily over researched value for Illuminance.

The program is basically defunct. I was never a good programmer to begin with and the program was constantly in need of tweaking, improving the file handling, and making it robust. I really haven't worked with it for years do to the loss of interest by magazines with traditional photography. I've also answered most of the questions I've had on photographic theory and for my personal photography and don't really need it any more. What few programming skills I had are all but gone and with changing computers and operating systems and forgetting what components are required to make it work, it is sadly nonfunctional. Kirk Keys and I worked on a project in which we used the program. I don't know what he thought about it. The project hit a road block when I had to adjust the program to perform a specific task and I had forgotten too much about programming by that point to make it happen. Haven't touch it since.

You know, I believe Davis' approximate fractional gradient feature in his program uses the Delta X equation. I know he is aware of it because I remember an article he wrote for PHOTO Techniques back in the early 90s which was based on the 1956 Nelson paper. I don't know if he knew all of the ramifications of film speed that Nelson's paper suggests or how Delta X was incorporated into the ASA/ISO standards. He probably did, but because he wrote for a more general readership and had to simplify the concepts about most of the more esoteric and really technical ideas probably never made it to print.

Steve