ISO film speed point relative to in-camera exposure

[Stephen Benskin]

Thanks Bill.

There is a pdf version of The Theory of the Photographic Process on line. https://ia600700.us.archive.org/11/items/TheTheoryOfThePhotographicProcess/aa012.pdf
Chapter XX contains a large chunk of the First Excellent Print Test.

 

[Anon Ymous]

That's quite a lot of reading material, thank you all!

 

[Stephen Benskin]

Hello all...

According to the ISO standard of film speed, the film must meet certain criteria. The characteristic curve must pass from two specific points. The first is a point where the density is 0,1 above film base plus fog. The second point is one that has received 4,33 stops more exposure (ΔE = 1,3) and must have a density 0,8 above the previous point. Based on the exposure that the first point has received one can calculate the ISO film speed with the appropriate equation. That's all fine, but requires some apparatus that most of us don't have. So, the question is, where are these (exposure) points relative to the exposure that one would get from a camera using auto exposure and shooting a gray card?

The difference between the b&W speed ppint and the metered exposure point is 1.0 log-H. Point m os an additional stop. To determine the exposure necessary for a given film speed, just use the equation 0.8 / ISO speed. The exposure at the metered exposure point is then 8.0 / ISO speed.

And remember The Zone System and real sensitometry are not interchangeable. They use different methodologies. The Zone System's Zone I is 4 stop below Zone V, which is considered the metered exposure point. While real sensitometry the speed point is 3 1/3 stops below the metered exposure point. For the statistically average scene that makes the metered subject 12% gray.

 

[Stephen Benskin]

Here is the math behind the ISO film speed contrast parameters. By adhering to the contrast parameters, you are effectively using the Delta-X equation so that the answer will always be 0.296. The Delta-X criterion is an easier and more consistent way to find the fractional gradient point. If the film contrast is outside the ISO contrast parameters, it is necessary to use the Delta-X equation to obtain accurate film speeds. Simply using a fixed density point of 0.10 without the equation will result in progressively incorrect film speeds as the film contrast is increased or decreased.

 

[Stephen Benskin]

But Anon Ymous,

You should not design a test that verifies the 0.72 density at metered point.

All the tests traditionally use a very low density change (for example 0.1 above base+fog) as the speed point because it changes very little with experimental differences. If you tried to key exposure index from 0.72 density you will have some problems that won’t be easy to figure out.

Don't forget film density at the metered exposure point has a lot to do with curve shape. It's going to fall at a lower density with a long toed curve than with a short toed curve.

Also, let's not forget that the 0.72 tends to be connected with the Zone System where they have a four stop difference between the fixed density point of 0.10 over Fb+f and Zone V which isn't the metered exposure point. That's why Zone EI is always 1/2 to 1 stop slower than the ISO speed. Basically, 18% Reflectance is 1/2 stop greater than 12% Reflectance. That is the average difference between Zone EI and ISO speed.

 

[Photo Engineer]

The real speed is based on the "first acceptable print" when measuring the speed of negative film. This is shown in both Haist and Mees.

PE

 

[Craig75]

The real speed is based on the "first acceptable print" when measuring the speed of negative film. This is shown in both Haist and Mees.

PE

Yes i was just reading this exact same thing in Todd and Zakia yesterday.

I'd have thought a bracketed exposure of a scene and choosing the first one you like is the best measure of speed.

 

[ic-racer]

I found W-speed much easier to have a spreadsheet calculate than Delta-X. I wonder if anyone has a spreadsheet that calculates Delta-X. This thread got me thinking about a way to do it. I can do a IF/THEN loop in Excel looking for the two data points that straddle the point on the curve that would be 1.3 units from the 0.1 point. From there connect the two points as a straight line and triangulate to find an approximate value for the Y-value; the density or Delta-D. After finding the Delta-D that way, the spread sheet would easily use the published Look-up Table to find the Delta-X.

That method would avoid having to fit the film curve to a quadratic equation and then try to solve it for X=1.3+n to find the Y value, Delta-D point (where n = the 0.1 point).

 

[ic-racer]

I had another idea, someone with a CNC router, could easily make these templates. A little old fashion but they would be so easy to use. Use your computer to print the H&D curve at a 'standard' size* and just overlay some templates. It shouldn't cost much more than a gray card. Look how many people have gray cards, of somewhat limited value. This template would be incredibly useful.

Just to emphasize the point for those new to the concept of simple methods for approximating the fractional gradient speeds of photographic materials, with this meter a speed determination can be made on film processed to ANY gamma. You don't need a "Standard Development" to satisfy the ASA triangle. It works; its tested in the literature!

*Your graph would have to be scaled so the point from "A" on the template to the right vertical = 1.3 log D units.

 

[Stephen Benskin]

Yes i was just reading this exact same thing in Todd and Zakia yesterday.

I'd have thought a bracketed exposure of a scene and choosing the first one you like is the best measure of speed.

In a way you are correct. Jones First Excellent Print test asked observers to select the best prints made from negatives of differing exposure. What made this test special is that it used psychophysics. Jones related the perception of quality to the sensitometric response of the photographic process and the natural world.

But don't confuse desired exposure with film speed. The key is in the title of the test, "First Excellent." Jones found that the perceived quality of a black and white print quickly falls off when the shadow gradient is below a certain value when compared to the overall film gradient. This limiting criteria defines the point of minimal exposure. Jones also found that prints were considered excellent when shadow exposure fell above the point of minimum exposure over a range of stops depending on film size and degree of enlargement. The limiting criteria defines the point of minimal exposure for a film. Any shadow exposure above this point, within a reasonable degree, will produce a quality image. So the idea is to use the limiting criteria to define film speed and then to use a working EI in practice. And by working EI, I'm not necessarily talking about personal EI. Both the pre 1960 and post ASA speeds uses this concept to varying degrees.

 

[Anon Ymous]

... Finally, in practice, don't be so concerned about determining your own film speed.

Basically, I'm not that much concerned. It is more of a curiosity thing than real need. In this particular case, I wanted to see how close I could get to the ISO standard, while using a developer that is supposedly very close to the real thing. That and curve shape as well. In the end, the negatives speak for themselves. This TMY2, shot at box speed and developed in my homebrew Xtol turned out fine, remarkably fine grained and sharp.

... Personal testing without a calibrated sensitometer has too many uncontrollable variables to have any real and repeatable precision. Film speed with general purpose developers is pretty much at the ISO speed. The metering tastes and methods of the photographer have a greater influence on exposure placement than small differences in film speed. Focus more on contrast.

Ah, yes, I was afraid you'd say that such work without a sensitometer is plagued by too many uncontrollable variables. Still it is quite fun to do and IMHO has some value, albeit not too much. I also agree that metering tastes and methods have far greater influence.

 

[Nodda Duma]

Sensitometers can be found on ebay.

Transmission Densitometers are cheap on ebay...you can use those for sensitometry, too. Preference is to get one with a calibration card (basically two pieces of film with known densities). From there, get a calibrated step wedge and uniform light source (i.e your enlarger), and then you can look at whatever part of the curve you want for whatever speed film you have.

 

[Anon Ymous]

Sensitometers can be found on ebay...

But sadly, from the US. Whatever cheapish device will become much more expensive if you take into account any shipping and import duties/taxes.

 

[Stephen Benskin]

But sadly, from the US. Whatever cheapish device will become much more expensive if you take into account any shipping and import duties/taxes.

All you need is a consistent light source if you want something to monitor your developer testing. Even with a sensitometer, you will need to get the light source tested if you want actual exposure. Find one online, play with it to find an exposure that works, and then use it to compare developer batches. You might also want to use a gradient that is more inline with usage. The ISO standard is for film speed testing and does not represent an ideal gradient in many cases.

Long toed films need to be developed to a higher average gradient to fit the ISO contrast parameters than short toed curves. The idea that the film will have a gradient of 0.61 because of the 0.80 / 1.30 aims only applies to short toed curves. Long toed films generally are developed to around 0.65 - 0.67.

 

[Bill Burk]

I had another idea, someone with a CNC router, could easily make these templates. A little old fashion but they would be so easy to use. Use your computer to print the H&D curve at a 'standard' size* and just overlay some templates. It shouldn't cost much more than a gray card. Look how many people have gray cards, of somewhat limited value. This template would be incredibly useful.

Just to emphasize the point for those new to the concept of simple methods for approximating the fractional gradient speeds of photographic materials, with this meter a speed determination can be made on film processed to ANY gamma. You don't need a "Standard Development" to satisfy the ASA triangle. It works; its tested in the literature!

*Your graph would have to be scaled so the point from "A" on the template to the right vertical = 1.3 log D units.
View attachment 214430

I have a matching set of graphs and this template in pdf if anyone is interested. I just print them on laser printer same size

Easy enough to print the template on overhead material

 

[Bill Burk]

http://beefalobill.com/images/DeltaX_Meter_Shifted.pdf

http://beefalobill.com/images/DeltaX_Meter.pdf

http://beefalobill.com/imgs/sensitometry.pdf

Here’s a set. I played with the Delta-X template to make it “direct reading”.

 

[ic-racer]

Nice, thanks for posting.

 

[Stephen Benskin]

Nice job on the template!

 

[Kodiak]

TO OP:
the first point relates to the Film speed. Measure Light from the uniform surface ( I take A4 size white paper and stick it on the window glass) and see the aperture. Now close aperture for 3 stops (e.g. F5.6 to F16) and shoot the surface. When develop the Film look at that frame. It should be with density c.0.1 above not exposed part of the Film (or it is just a little less transparent than not exposed part of the Film).
The second point should evaluated in connection with the paper you use and is out on my time.

And you are correct: all over books are only logarithms which one out of million photographers can understand and use to make the curve. Things are that all you need as just any Lightmeter but to get to that simplicity you will have to work really hard along time...
Do not buy densitometer...

 

[Bill Burk]

VueScan may be all you need as a densitometer, at least for the purpose of supporting this discussion and to help your understanding of sensitometry.

Logarithms are pretty intuitive when you build on your understanding of f/stop and shutter speed series’.

To aid your memory, keep a few key concepts in mind... 0 is 1, 10 is 2, 100 is 3, each f/stop is 0.3, each film speed change is 0.1

I find it harder to work in arithmetic series’, so I tend to convert to logarithm when I can.

 

[Kodiak]

Bill Burk
Can you please skip logarithms and stay with F-stop on the Lens instead, and how to measure density using just any Lightmeter, even one from his camera.?
All he need is how to find the film speed and Film developing time.

 

[ic-racer]

how to measure density using just any Lightmeter, even one from his camera.?
All he need is how to find the film speed and Film developing time.

Zone 1 negative, when placed over a light meter will lower the reading by 1/3 stop. A zone VIII negative that shows a faint outline of a coin on a print is properly developed.

 

[RalphLambrecht]

Thanks Bill.

There is a pdf version of The Theory of the Photographic Process on line. https://ia600700.us.archive.org/11/items/TheTheoryOfThePhotographicProcess/aa012.pdf
Chapter XX contains a large chunk of the First Excellent Print Test.

great link; the freeware version sells for about$300

 

[Stephen Benskin]

Bill Burk
Can you please skip logarithms and stay with F-stop on the Lens instead, and how to measure density using just any Lightmeter, even one from his camera.?
All he need is how to find the film speed and Film developing time.

Without proper testing the determination of accurate film speed isn't possible and is mostly a waste of time. Non-standardized techniques generally produce a false sense of accuracy. If the OP wants something for batch testing, relative log-H from a consistent light source will identify any variance between batches, but can only offer relative speeds.

Here is a comparison of two film speed determination methods: the fixed density method and the Delta-X criterion.

In terms of contrast, it's rise over run no matter what you use, plus the OP appears to be using logarithms. The hard part is to define what the aim values should be. Here is a table with different development models. I'm partial to The Practical Flare Model that I worked out.

 

[Kodiak]

Bill and Stephen Benskin

The system you are practicing is important in scientifical study of the Film.

Your densitometer cell is different than the cell on your Lightmeter. You pack the densitometer and set the Film-speed on the LightMeter be it handheld or in the Camera. This leads to an error due to reading from different cell. Seconic shows different reading from Gossen, different from Minolta, … Even from Gossen to Gossen is different. There are no two identical cells.

The point is to use the same Photo-cell throughout the process, and the same (well close enough) Light temperature during the test and your Photography, and make correction when you sniff the Light temperature change. Film speed is not the same at evening and noon and at the room-bulb Light. The difference can be two times.

What you do is better than nothing, but it is by far not the best system in practical Photography.

I have Kodak grey card which shows 14 Zones, some use I … IX, some 0…X and so on.

Zone system is invented and is good starting point. It is just so confusable when one says Zone I. What that means to one that made the Kodak grey card and to you is different.

Another think is you work with two decimal places (e.g. 7.25). How you set it on the Lens? How you set on the camera Film speed iso148.96? Here your accuracy is gone. The next error is water for your developer which WILL change chemical composition of the developer. The next is the thermometer, and do you think that all thermometers shows exactly the same temperature? The next think is the broken liquid in the thermometer and you did not notice it so you read 20 deg.C but actually is 21 deg.C, and so on and so on.

All of this just remains me on Jansen’s comment (an art historian): “… they make so large paintings that the size itself only matters, but what happens when that art is shown in illustrations for showing it …?”

You are knowledgeable guy and should concentrate on one new system in Photography, and it is using the same LightMeter for

1. Film exposure during the Film testing
2. reading film densities
3. testing Photo papers
4. during enlargement
5. for exposure reading to set the Camera

And if you want, get a cheap Gossen sixtycolour-colourMeter.

NOW when you assign 0.1 density to 3 F-stops below 18% shot, and when in field and you read from the darkest part and close aperture 3 F-stops you know: it will appear close to 0.1 density on your Film (CLOSE because there are variables you just cannot control 100%, as developer concentration, water, temperature, agitation, Lens aperture accuracy, … The more elements you have in the system the LOWER accuracy is, just opposite of what you expect.

This is where you should concentrate your energy. No oil there where you dig now. Still there are sooooo many things to discover in Photography
Good luck