What is the Relationship between Film Speed and Camera Exposure?

[Stephen Benskin]

I’m sure almost everyone has experienced this at one time or another. You think you understand a certain concept until you have to explain it to someone, and then you realize there are gaps or questionable assumptions in your thought process. I believe this is a frequent occurrence with teachers early in their careers. From my experience, after the flop sweat has dried, it can prove to be as much of a learning experience for the teacher as for the student. The following concept might just fall into this category.

The classic saying for exposure is the exposure meter wants to make whatever it is pointing at middle gray. In a recent thread there was a discussion about the scene luminance range and the range of Reflectances from a 12% and 18% Zone V model. So, I would like to think we’ve determined what the scene is and what the meter “keys” on.

In order for the exposure meter to work properly with films of differing sensitivities, the speed of the film is entered into the meter’s exposure calculator. There’s been copious discussions on film speed on this forum, so the basics of film speed should be relatively well understood.

In a paper by D. Connelly, he writes, “It is evident that the relationship between the sensitometric measurement of exposure (film speed) and the photographic exposure (camera exposure) requirement must be determine.” There needs to be a way of connecting the various photographic elements together. You determine the film speed and that value somehow has to relate to the exposure meter and camera exposure in order to achieve good repeatable negatives.

Here are the questions to consider, and please keep in mind it's the sensitometric measurement of exposure.
What is the relationship that Connelly writes about?
How can it be determined?
How does the film speed work with the exposure meter to determine the exposure placement of whatever the meter is pointing at?
Also keeping in mind that exposure meters are designed to produce one data value for the scene, what should that exposure be for a given film speed?

I know it seems pretty straight forward, but it involves some fundamental underlying principles. I've uploaded Connelly's paper for a reference. The above quote comes from the section on Film Speed on page 187.

 

[keithwms]

I suppose the ISO must have determined the value. That is or was an organization that did such things, right?

P.S. Ah yes, here you go, it must be in this article:

http://en.wikipedia.org/wiki/Film_speed

 

[spacer]

If I'm understanding the question right, once the ISO value is determined, the camera computes exposure with no other consideration for the film.

 

[ic-racer]

I'd say that paper presents alot of interesting and practical information for those attempting to design a system for dimwit camera operators. However, it seems to me to be irrelevant if one determines one's own EI, especially if one is shooting only negative film.

Having said that I am a fan of the contemporary Nikon matrix system which addresses many of the same issues in a suitable, practical manner: http://www.mir.com.my/rb/photography/hardwares/classics/nikonfa/fatech/matrixflowchart.htm

I'm no Nikon fan, but their matrix metering system seems to be designed by some knowledgable folks to address many of the significant exposure issues illustrated by Steve's posts on exposure and film sensitivity.

 

[Stephen Benskin]

Maybe the questions are too broadly phrased. Let's break it down. And I'm not asking for reference links.

Step 1:
What are the equations used to determine B&W and Color Reversal sensitometric speed?

 

[markbarendt]

Here are the questions to consider, and please keep in mind it's the sensitometric measurement of exposure.
What is the relationship that Connelly writes about?
How can it be determined?

In my head there are only two variables, the volume of light and the physical properties (speed) of the film in a given process.

The only thing that seems to be missing from the equation is a constant.

That constant can only be determined one way, by experimenting to see what effect a/any given volume of light has on that film.

The actual reference point though to key a meter to seems to be pretty arbitrary. Any given density on the negative could be used as long as it is repeatable.

How does the film speed work with the exposure meter to determine the exposure placement of whatever the meter is pointing at?
Also keeping in mind that exposure meters are designed to produce one data value for the scene, what should that exposure be for a given film speed?

Light meters measure the intensity of light, we already know (well we could look up) the speed of light, add in the time a shutter is open and the area through which the light can pass and we have what we need to define the volume of light; at this point we can effectively count how many photons we will get.

The only thing missing is the constant we "found" experientially for a given film's speed.

 

[keithwms]

Step 1:
What are the equations used to determine B&W and Color Reversal sensitometric speed?

Well, I don't think there can be one formula for all films. Even for one film there are different developers and times and temps etc., and different development will affect the density/exposure curve.

I am guessing that the formula would relate some standard optical density (minus the film base density, usually very small) to the exposure.

So, suppose that you have a known light intensity on a standard grey card. We would stipulate that a standard exposure will yield x% of the maximum optical density on the film under standard development. That will determine the nominal box speed. (astute analogue photographers will of course note that this point actually moves around a bit depending on how you develop)

In other words, you'd simply look at the log density / log exposure chart, pick the special point that you want to correlate to standard grey under standard (say, sunny 16) light and exposure, and off you go. There will be a log relationship.

Of course, every standard has its limits, and in this case the standard grey point would say nothing about what happens to the higher and lower tones. But still I suppose it'd be reasonable to subtract the high density from the low density, say grey corresponds to x% of that range, etc.

 

[Bill Burk]

For any one piece or shot of film exposed... There was only one combination of f/stop and shutter speed used.

Set it on the dial of the meter (working backwards).

Develop the film as planned and check backwards to either the meter reading notes or the original scene (if possible).

The relationship between the meter needle and the numbers on the scale -- would be your constant if the exposure had proved to be right.

 

[Stephen Benskin]

If I'm understanding the question right, once the ISO value is determined, the camera computes exposure with no other consideration for the film.

This is a good generalization of it. I have some reservations about the "no other consideration part", but will discuss that later. What I'm asking about is the specifics of the relationship. How does the film speed value interact with the exposure meter and why? So far everyone is being very general or just offering an overview. How about nailing it down?

No one has given the film speed equations for either B&W negative film or color reversal film. I believe this is a good place to start, because we need to know where on the characteristic curve the film speed point is and where the metered exposure point is in order to determine what their relationship is. Most of this can be found on pages 187 and 188 of the Connelly paper. One small note: There has been a change in the equation for color reversal film speed since this paper was published which I believe offers further insight into way in which film speed and exposure placement work.

 

[Kirk Keyes]

Maybe the questions are too broadly phrased. Let's break it down. And I'm not asking for reference links.

Step 1:
What are the equations used to determine B&W and Color Reversal sensitometric speed?

For B&W negative films, arithmetic ISO speed is determined by:
ISO Speed = 0.8 / (Exposure st speed point)
The speed index result is rounded off to the nearest 1/3rd f-stop.

For color reversal films, the the exposure needed to get to the shoulder point (S), which is about 90% maximum density is determined. Then the exposure needed to get to the toe (T) which is 0.2 density above the minumum density is measured. These two exposures are averaged, that is: ((S+T)/2), and this value is the Speed Point. The reciprocal of the exposure at the Speed Point is then multiplied by 10 and the result is the arithmetic Speed Number.

The preceeding was taken from "Sensitometry for Photographers" by Jack Eggleston. Stephen recommended this book to me many years ago, and it is certainly worth getting if you are at all interested in the subjects at hand.

 

[keithwms]

Here's the ISO explanation of how to determine b&w film speed.... if you want to pay for it..

http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=3580

and here's something for image sensors from Kodak...

http://www.kodak.com/ek/uploadedFiles/ISOMeasurements(1).pdf

Note that many things have to be standardized and stipulated before you can measure the speed. You'd need to know the colour temp of the lamp, because different films have different wavelength sensitivities. And bear in mind that if there is no grain standard, you could arguably call any b&w film any speed you want just by dev'ing it to the standard density you want under a given exposure... knee and toe and gamma be damned! And what about gamma.... who decides the standard gamma? Many things to consider....

So... everybody has to come to their own film speed- the box speeds are just (very good) suggestions.

 

[Stephen Benskin]

Note that many things have to be standardized and stipulated before you can measure the speed. You'd need to know the colour temp of the lamp, because different films have different wavelength sensitivities. And bear in mind that if there is no grain standard, you could arguably call any b&w film any speed you want just by dev'ing it to the standard density you want under a given exposure... knee and toe and gamma be damned! And what about gamma.... who decides the standard gamma? Many things to consider....

So... everybody has to come to their own film speed- the box speeds are just (very good) suggestions.

Don't forget hold time. The testing parameters are all part of the important information associated with film speeds. And this is a topic which I love to discuss in great detail. But determining the film speed, isn't what this thread is about. It's not the final goal, but a place to start in how to link the sensitivity of the film to the exposure meter. I'm currently putting together the next part to be discussed. In the mean time consider this, B&W film speed and color reversal film speed are determined at different points, B&W film speed in the toe and color reversal in the middle of the curve, yet they can have the same film speed and work with a the same meter. You don't need a meter for color reversal and a meter for B&W.

The reason why it's important to understand the film speed equations is because this is how the sensitivity of the film is translated into a number that then links to the exposure meter that then connects back onto the film with exposure placement.

Side note: the standard gamma of the B&W film speed standard is to create a good correlation between the speed point of the fixed density method of the standard and the fractional gradient speed point. In actuality, the fixed density method of the standard is just a proxy for the fractional gradient method. To nit-pick, it can't be called Gamma or even contrast index because the standard's parameters don't adhere to the testing methodology associated with either method.

 

[Bill Burk]

Your PDF implied that the needs of black and white negatives took a back seat to color (hey it has lots of latitude and people already know how to expose it well).

It's interesting that Super-8 was deliberately overexposed (this was done in camera, meters and ISO weren't changed) so that amateur projections would be bright enough to see.

And wasn't color slide film standard exposure chosen to be suitable for projection? I think this inspired photographers to give less exposure when intended for print production (Galen Rowell underexposed Velvia).

 

[Stephen Benskin]

Your PDF implied that the needs of black and white negatives took a back seat to color (hey it has lots of latitude and people already know how to expose it well).

It's interesting that Super-8 was deliberately overexposed (this was done in camera, meters and ISO weren't changed) so that amateur projections would be bright enough to see.

And wasn't color slide film standard exposure chosen to be suitable for projection? I think this inspired photographers to give less exposure when intended for print production (Galen Rowell underexposed Velvia).

I love looking behind the curtain. With slide film, the preferred density range changes depending on the way it is viewed. Projected requires a greater range than viewing on a light table.

 

[keithwms]

Determining the film speed, isn't what this thread is about. It's not the final goal, but a place to start in how to link the sensitivity of the film to the exposure meter.

This may be useful, then....

Dead Link Removed

 

[paul ron]

I really do enjoy all this technical information explaining what I've taken for granted over the years. Perhpas I once knew what all this was but some things just become a blur with experiance. I've found I can explain it very simply boiled down to this....

If you were to fill a standarized sized glass with water, how long will it take at a certain flow rate? The glass can be the various ISOs, flow is the brightness vs fill time, exposure.

But of course I have leared so much from you guys, so please don't stop, just give me a little gigle.


Happy New year everyone.

.

 

[Stephen Benskin]

This may be useful, then....

Dead Link Removed

I've written rather extensively on this. In fact, there's a current thread on this very issue "Comparison of Reflectance from 18% and 12% Zone Models." If you happen to read the Thom piece, you will find very little there. Where's his proof? I'm not saying he's wrong about the average reflectance, he just doesn't explain how or why. This is what I'm wanting to do with this thread. We might think we are on sound ground with certain concepts, but are we. We've all been there. "I read it in an article by Thom. He says Exposure meters don't see 18%, they see 12%." Great, what's the next sentence? How well can this concept be explained to someone else based on Thom's article?

I used to have a lot of fun with the articles from the early years of View Camera Magazine. The authors would present a new technique or some evaluation of a new film or something like that then they would come to that point where they didn't have enough information or theory to continue and the articles would quickly devolve into a meandering series of music analogies.

 

[keithwms]

I'd have to get further into it to figure out the argument for 1; at some point it's probably an issue of QE, which depends on the spectrum, so I am not sure how the math shakes out. Anyway I just thought it interesting that something this basic is still contested.

 

[Stephen Benskin]

Anyway I just thought it interesting that something this basic is still contested.

Totally agree. Although it's not contested in any scientific capacity, just misunderstood at the average user's level. I believe this is because of a lack of access to solid information. And by that I mean "the maths." Have you had a chance to look through the Connelly paper? Admittedly, the paper is not something that can be understood at a single reading. I always read these types of papers with Mathcad running on the computer.

 

[keithwms]

Nope, haven't read it in detail. I get the gist, but frankly it isn't written in the first-principles style that I'd prefer. Relationships are simply presented and that's that. An annoying style, frankly. I am a photons and electrons kind of guy

 

[Stephen Benskin]

For B&W negative films, arithmetic ISO speed is determined by:
ISO Speed = 0.8 / (Exposure st speed point)
The speed index result is rounded off to the nearest 1/3rd f-stop.

For color reversal films, the the exposure needed to get to the shoulder point (S), which is about 90% maximum density is determined. Then the exposure needed to get to the toe (T) which is 0.2 density above the minumum density is measured. These two exposures are averaged, that is: ((S+T)/2), and this value is the Speed Point. The reciprocal of the exposure at the Speed Point is then multiplied by 10 and the result is the arithmetic Speed Number.

Thanks Kirk,

For B&W it's 0.8 / Hm
For Color Reversal it's 10 / Hm
Connelly uses HR for color reversal films to avoid confusion. I’ll be doing the same.

Hm and HR is the exposure in meter candle seconds at the measured point for the respective film speeds. The reason why it needs to be a sensitometric exposure is to produce known values to work with. Few people have this capability, but it’s possible to determine the necessary values for Hm and HR for any film speed by dividing the aim film speed into the constant.

100 speed - .8/100 = 0.0080 mcs 10/100 = 0.10 mcs
125 speed - .8/125 = 0.0064 mcs 10/125 = 0.80 mcs
200 speed - .8/200 = 0.0040 mcs 10/200 = 0.05 mcs
400 speed - .8/400 = 0.0020 mcs 10/400 = 0.025 mcs

For a black and white negative with a speed of 125, an exposure of 0.0064 mcs at the film plane will produce a negative density of 0.10 over Fb+f when all the other conditions of the standard are met. With a transparency an exposure of 0.80 mcs with a 125 speed film will fall half way between the points HS and HT.

But how can the same film speed produce two different exposure values? As you can see from the examples of the two graphs from the speed standards, the speed points for black and white negative films and color reversal films fall at very different points on the characteristic curve. This is accomplished by having different values for the film speed constants. For black and white negative film it’s 0.80 and for color reversal film, it’s 10.0.

Sometime in the 70s or 80s, the constant for color reversal film changed from 8.0 to 10.0 while the method of determining the speed point didn’t. The film speed determined from the same point on the film curve went from 8/0.064 = 125 to 10/0.064 = 156. Changing the value of the film speed constant changed the resulting film speed value. In the case of color reversal film, the film speed increased by 1/3 stop which decreases the film exposure by 1/3 stop.

Why change the constant and not the speed point? Why not just shift the speed point to the right by 0.10 logs? One of the purposes in film speed testing is to define the limiting points on the film curve that are critical to the perception of quality in the finished image. For color reversal film, it is the two defined end points on the curve (S and T). For black and white negative film, it is the point where the minimum gradient it .3x the average gradient of the film curve. By adhering to the contrast parameters of the black and white film standard, this point will always fall 0.29 log-H units to the left of the speed point (attachment - B W film speed standard and Fractional Gradient point.jpg). Knowing the position of these points is important toward producing a film speed that has any relevance to the film and to quality results. By adjusting the exposure placement through a change in the constant and not a change in the speed point maintains this relevance.

Now that the critical points with each of the film types are known, the next step is to determine where to place the point of exposure averaged from the scene being photographed where all the values from the scene will fall within the tested limits.

The luminance range of the statistically average scene is 2.20 logs (7 1/3 stops). The average luminance of the scene will fall where there are just a little over three stops above the average luminance (0.92) and just a little under 4 1/3 stops below (1.28).

I hope that I was able to begin to convey how the film speed is important to exposure placement. Now, we need to determine the placement of the exposure. This involves understanding something about meter calibration and camera exposure.

Here are the next set of questions:
What is the calibration luminance for an reflection meter?
What is the calibration illuminance for an incident meter?
What is the camera exposure equation and how does the calibration luminance plug into it?
And from my first post:
Also keeping in mind that exposure meters are designed to produce one data value for the scene, what should that exposure be for a given film speed?

 

[Kirk Keyes]

Hm and HR is the exposure in meter candle seconds at the measured point for the respective film speeds. The reason why it needs to be a sensitometric exposure is to produce known values to work with. Few people have this capability, but it’s possible to determine the necessary values for Hm and HR for any film speed by dividing the aim film speed into the constant.

Meter-candle-seconds is also known as lux-seconds.

I know several of the Minolta Flashmeter series can make incident measurements that can be converted into lux. You then multiply the exposure by lenght of the exposure, and then you have an exposure that has a known lux-seconds value.

 

[Kirk Keyes]

Why change the constant and not the speed point? Why not just shift the speed point to the right by 0.10 logs? One of the purposes in film speed testing is to define the limiting points on the film curve that are critical to the perception of quality in the finished image. For color reversal film, it is the two defined end points on the curve (S and T). For black and white negative film, it is the point where the minimum gradient it .3x the average gradient of the film curve.

The speed point is a function of the characteristics of the film - as you say, "the limiting points".

For reversal films, going from the 60s to the 70s, reversal films were able to capture a much large range of exposure. That is, the distance between S and T would have increased over films of a few years earlier.

That means the film speed number could be determined with a differenent equation to better account for this increased range.

For B&W neg films, the the standard printing paper LER did not change, so the speed point calcs would have remained the same.

 

[Kirk Keyes]

View attachment Fujichrome Family.pdf

Just for fun, here's a set of curves I measured for Provia RDP III, Velvia 100, Astia RAP 100F, and RTP II 64T (Tungsten film). I'm afraid I didn't plot speed points (you can see the S and T points of these films on the graphs), but I plotted just a composite curve for all 3 colors layers of each film is plotted instead. Note how there are slight differences in speed, despite 3 out of the 4 being 100 speed films.

Note that Velvia has the highest contrasta of all 4, RTP the lowest. And that Velvia 100 and RDP III have nearly the same highlight response, but then Velvia kicks up the contrast in the shadows. Also note that while Astia and Provia are very similar in the shadows/darker areas, Astia is much lower in contrast in the highlights with a much longer "toe".

 

[Stephen Benskin]

We've discussed how the determination of film speed is not just about the sensitivity of the film, but also defines the critical limits of exposure placement. We've also discussed how the speed constant can be used to adjust the exposure placement so that the average exposure range falls in a desire portion of the curve.

The next step is to determine where the exposure meter wants to place the exposure regardless of the film speed, and then with the film speed calculated in. First, we need to determine the average Luminance for the reflection meter and the average Illuminance for the incident meter. These can be considered the calibration values and as such can be determine using the exposure meter calibration equations.

Even though the refection meter reads Luminances, it possible to determine what is considered the average scene Reflectance in one of two ways. After converting the average Luminance to Footlamberts (* pi), divide the Luminance value by the Illuminance value. The second way is to divide the reflection meter's constant (converted) by the incident meter's constant.



Now, that we have the average Luminance value, all that is necessary is to plug it into the camera exposure equation to determine the exposure value.