What is the Relationship between Film Speed and Camera Exposure?

[Stephen Benskin]

This actually brings up something that I have wondered about for a while.

Why is it that we don't set spot meters to read the camera setting directly from the speed point?

Off the top of my head, I don't think anybody wants to have to buy separate meters for b&w and color reversal. Exposure is more critical with color reversal film, so it won.
Flare makes determining where the exposure would actually fall a nightmare. There's very little flare at the metered exposure point.
Human vision compresses tones in the lower range. We are a lot more sensitive to tonal differences in the middle range.
Many scenes don't have shadows deep enough to even reach the speed point exposure. You'd be metering at a point where there's no subject luminance.
And flare again, because it is just that important.

 

[markbarendt]

Off the top of my head, I don't think anybody wants to have to buy separate meters for b&w and color reversal. Exposure is more critical with color reversal film, so it won.

Who won makes sense.

My point was not in having different meters, it was more just to be able to read the camera setting directly when we use our normal pegging point, whatever that might be.

Flare makes determining where the exposure would actually fall a nightmare. There's very little flare at the metered exposure point.

This is truly interesting.

Essentially you are saying that meters can get/find better readings in the midtones and highlights, and that they struggle with the lower light values, correct?

Human vision compresses tones in the lower range. We are a lot more sensitive to tonal differences in the middle range.
Many scenes don't have shadows deep enough to even reach the speed point exposure. You'd be metering at a point where there's no subject luminance.
And flare again, because it is just that important.

And our own perception struggles at lower light levels.

These two factors seem to provide a really strong argument for pegging something in the middle when we are deciding on settings rather than the shadows.

Really, if we are striving for accuracy and repeatability in our techniques so that we can get what want rather having to take what we get, why would we meter a place where we and our tools have the hardest time providing reliable results?

Seems like a recipe for building in errors.

 

[Stephen Benskin]

So, how do we know the speed point – metered exposure ratio is 1.20 log-H or 16x (15.8) for Zone System testing? It’s right there in the Zone System testing instructions. Meter a card and stop down four stops. Zone I is the speed point and it is at 0.10 over Fb+f. Four stops of exposure is 1.20 log-H.

It should be obvious at this point that for a 125 speed film where the speed point has an exposure value of 0.0064 for Hm has to have an EI lower than 125 for the 1.20 log-H range to fit between the metered exposure and the speed point. What would that EI need to be?

Exposure constant / (Hm * ratio) = EI
or
8 / (0.0064 * 15.8) = 79.2
or
EI 80

A film with an ISO of 125 and a speed point – metered exposure ratio of 16 (1.20 log-H) will have an EI of 80. Of course, Zone System testing gets to this same conclusion differently.

Traditional Zone System testing is done in camera which means people are far more limited in the exposure information available to them than if they used a sensitometric approach. A card is metered and the camera is stopped down four stops. (While the Zone System test uses only a single tone, the graphs illustrate how the the full exposure range would fall.) Below is an example of how the exposure will fall with a 125 speed film metered using an EI of 125.



Zone I exposure will fall 0.20 log-H units below the 0.10 speed point. Exposure needs to be increased by 0.20 log-H in order to bring the Zone I exposure up from 0.0041 lxs to 0.0064 lxs. While the film speed remains at 125, the meter’s EI is now at 80. This is what creates the change from the speed point - metered exposure ratio of 10x (1.0 log-H) to 16X (1.20 log-H).



Doesn’t the first Zone System testing example look a lot like the sensitometric exposure example? The only difference is the choice of scene luminance ranges. The Zone System uses 2.10 and tone reproduction uses 2.20. So the only difference between the two is the sensitometric example has the shadow fall 1/3 stop further down on the curve.



The final step is where a lot of people get tripped up. There’s a difference between the scene’s luminance range and the camera’s exposure range. While the original scene might have a luminance range of 2.20, by the time it gets to the film plane, the exposure range averages at least a stop smaller: 1.90. Sensitometric testing is done in non-flare conditions, so the interpretation of the sensitometric results has to incorporate flare. While the difference between the metered exposure point of the subject and the shadow has a difference of 4 1/3 stops, on the film plane it is only 3 1/3 stops. It's not a coincidence that the ISO standard has a speed point - metered exposure ratio of 1.0 (3 1/3 stops) and not 1.30 (4 1/3 stops).

While the Zone System testing utilizes a camera, the flare is practically non-existent by the way the test is conducted, so it too can be considered non-flare testing conditions. So, just like what was done with the earlier sensitometric example, the exposure range needs to be adjusted to account for flare. To simplify thing, I’ve reduced the exposure range to 1.90 log-H.



It’s interesting how similar this example is to the fractional gradient example, but it shouldn’t surprising. Both the Zone System and 0.3G were developed in the early 1940s. At that time they had a good correlation between the two results. Take a look at their speed point – metered exposure ratios. While at first the fractional gradient’s 1.50 might not seem to match the Zone System’s 1.20, if you adjust for the 0.3G having a speed point approximately one stop (0.30 log-H) units to the left of the Zone System’s speed point, they are the same.



It wasn’t until the 1960 standard reduced the safety factor and changed the speed point – metered exposure ratio which then caused the resulting speeds from the Zone System and the ANSI/ISO method to diverge.

 

[Bill Burk]

These are great diagrams, your final diagram fits exactly Zone System benchmarks - Zone 0 at 0.3G, Zone I at 0.1 speed point, Zone V at the metered point.

You say the safety factor was removed in 1960? Does that change the final diagram, moving everything one stop to the left?

 

[Stephen Benskin]

This is truly interesting.

Essentially you are saying that meters can get/find better readings in the midtones and highlights, and that they struggle with the lower light values, correct?

Not better readings. Better confidence in exposure placement.

 

[Bill Burk]

I'd like to see three graphs marked like this that show

What the Scene Reflects

What the Meter Indicates

What arrives at the Film Plane

This model is wrong, because it shows what arrives at the film plane, including the one stop of flare in the shadow.

So for example a Scene Zone 0 came up to Film Plane Zone I.

But did a Scene Zone 0 come up to a Meter Zone I? The meter readings are not without flare. Whenever I aim a spotmeter at the darkness under a car for example, a lot of flare influences the meter reading.

 

[Bill Burk]

This actually brings up something that I have wondered about for a while.

Why is it that we don't set spot meters to read the camera setting directly from the speed point?

The math isn't the issue since the formula can be adapted.

The SEI Photometer does what you are thinking. There is a black spot and a white spot. If you put the black spot on ASA 4, the white spot is on ASA 400 - exactly 6 2/3 stops apart.

You are expected to meter a shadow, and set the ASA to the black spot. Or for transparencies meter a highlight and set ASA to the white spot.

But the value of using a Zone System sticker on a meter is that you can pick any recognizable value and spot it and place it on that Zone.

 

[Stephen Benskin]

Below is basically the same as the current 10X ratio example except that I've added on the camera image curve. I don't know if this helps clear things up.

 

[markbarendt]

Not better readings. Better confidence in exposure placement.

Your words are better than mine.

The question/problem remains though, even when reframed with fresh words.

Why would we (or a meter or film manufacturer) want to find our exposure settings based on points that should be trusted less?

It seems to me that the old adage "shoot to the shadows and develop for the highlights" like most simplifications, can lead us a bit astray here. In this case by, on its surface at least, encouraging us to measure a reference point that should rightfully inspire less confidence of exposure success.

It seems to me also that, the decision to use a mid-tone as the reference point/standard in concert with a tested constant (ISO or EI) to know where the speed point for a film falls when finding exposure makes real sense.

This thought also seems to me to answer why the data sheets, from say Ilford, suggest differing EI's dependent on the development planned even though the true speed point changes much less.

The manufacturers seem to be adjusting the constant to place the mid-tone point on the alternative curve at roughly the same density as on the normal curve, instead of pegging the speed point. So for example faces should print at nearly the same average brightness in each prescribed process. Is that a fair assessment?

Still and yet it seems that the shadow point can be found more reliably by measuring a mid-tone, it simply requires a bit of familiarity with the tools at hand.

 

[Bill Burk]

Why would we (or a meter or film manufacturer) want to find our exposure settings based on points that should be trusted less?

It seems to me that the old adage "shoot to the shadows and develop for the highlights" like most simplifications, can lead us a bit astray here.

I would say underexposing shadows can quickly ruin a negative - so it is often the most important exposure to determine. This applies mostly to scenic/landscape photography.

Other kinds of photography where a tripod is not used, will be more quickly ruined by slow shutter speeds - so for majority of purposes it is more important to emphasize speed.

 

[Stephen Benskin]

This model is wrong, because it shows what arrives at the film plane, including the one stop of flare in the shadow.

So for example a Scene Zone 0 came up to Film Plane Zone I.

But did a Scene Zone 0 come up to a Meter Zone I? The meter readings are not without flare. Whenever I aim a spotmeter at the darkness under a car for example, a lot of flare influences the meter reading.

Bill, I'm really curious as to what you see being wrong. We have to be getting our lines crossed somewhere. I'm interested to learn if there is any miscommunication with the way I'm explaining this. I've added Zone markings to the standard exposure model with one stop of flare so you can compare it with your version.



Flare does have an effect with the optics of spot meters, but just like with the camera, it's larger in the shadow areas than around the mid-tones. The exposure constant of 8 has already factored flare in. You're answered Mark's question about metering the shadows. Flare is too variable, hard to measure, and impossible to predict for the shadows to be a viable option. The next set of examples are three different flare models: no flare, one stop of flare, and two stops of flare. The exposure around the metered exposure point changes little while the shadow placement varies considerably.

 

[Bill Burk]

Bill, I'm really curious as to what you see being wrong.

I only mean the diagram I drew was wrong, I didn't want it to be taken as a solution.

My diagram is too ideal. Zone 0 at the 0.3G, Zone I at speed point and Zone V at the meter point is too perfect to be right - so it's wrong.

That's all. The rest of this is good.

---
Since spotmeters are subject to flare, could it be (since the sensor of a spotmeter got similar flare to film), you really do get a reasonable prediction of shadow exposure?

 

[markbarendt]

I would say underexposing shadows can quickly ruin a negative - so it is often the most important exposure to determine. This applies mostly to scenic/landscape photography.

Other kinds of photography where a tripod is not used, will be more quickly ruined by slow shutter speeds - so for majority of purposes it is more important to emphasize speed.

I'm not suggesting that anyone changes the exposure point they want to peg. What I'm suggesting is that it could be more reliable finding it indirectly.

As Stephen suggests the offset to the speed point is a known and any personal preference we may have for exposure offset can be applied to that offset.

So, how do we know the speed point – metered exposure ratio is 1.20 log-H or 16x (15.8) for Zone System testing? It’s right there in the Zone System testing instructions. Meter a card and stop down four stops. Zone I is the speed point and it is at 0.10 over Fb+f. Four stops of exposure is 1.20 log-H.

This is a relationship, the offset or constant, between meter reading and film speed point is fixed.

A given reading to a known mid-tone point gives us our shadow.

I am also not suggesting that measuring the shadow point is without value just that it may be better used as a check and in the SBR determination rather than as the primary reference.

Because

Flare makes determining where the exposure would actually fall a nightmare. There's very little flare at the metered exposure point.
Human vision compresses tones in the lower range. We are a lot more sensitive to tonal differences in the middle range.
Many scenes don't have shadows deep enough to even reach the speed point exposure. You'd be metering at a point where there's no subject luminance.
And flare again, because it is just that important.

Our meters and our tone perception work better in the mid-tone range, we can differentiate better and flare doesn't affect mid-tone readings like it does for zone 1 readings.

 

[Bill Burk]

I'm not suggesting that anyone changes the exposure point they want to peg. What I'm suggesting is that it could be more reliable finding it indirectly.

I'm not going to discourage you from exploring your idea.

But I have trouble gauging gray. When I meter off what I think should be the gray of my scene, say the middle of an old paved street, I am surprised to find it is my highlight!

BTZS uses the incident meter to indirectly determine shadow exposure. You are right that there are benefits to using indirect metering.

But I still think the spotmeter, from camera position, will give a good indication of what will reach the film plane thanks to flare. I just did a check and saw the needle move at least a stop or two between spotmeter from camera position and walking right up to what I was metering. When I was close, the reading was less influenced by flare - so it may have been a more accurate reading of the subject luminance. But at camera position, the spotmeter could have been giving me a good shadow reading.

 

[Chuck_P]

But at camera position, the spotmeter could have been giving me a good shadow reading.

I agree with you 100%.

I do my ZS testing with a 4x5 step tablet contacted to the film targeting a white matboard with lens at infinity in uniform shade (I use a Pentax V spot meter), I have yet to be disappointed with my resulting EIs and subsequent shadow placements. I barely give flare a moments worth of thought, except just prior to releasing the shutter, when I try to shield the lens the best that I can. But I still find this thread interesting enough.

 

[Stephen Benskin]

As Stephen suggests the offset to the speed point is a known and any personal preference we may have for exposure offset can be applied to that offset.

This is a relationship, the offset or constant, between meter reading and film speed point is fixed.

A given reading to a known mid-tone point gives us our shadow.

Mark,

That's exactly right. I was worried that nobody was going to appreciate the significance of the speed point-metered exposure ratio.

I also hope the idea that the speed point isn't an arbitrary target density didn't go unnoticed. That the process of determining the speed point also defines important aspects of the film curve. With b&w film it's the relationship with the minimum useful gradient point or 0.3G. Keeping that in mind, the 0.3G point tends to change in relation to the 0.10 speed point with changes in development. It's only has a exposure difference of around one stop when the ISO contrast parameters are followed. With increased or decreased development, the one stop relationship on longer exists which means the important connection between the 0.10 speed point and the lower useful end of the film curve is lost as well as the direct correlation connecting the exposure to the quality of the practical picture results of the first excellent print tests.

There is a fix which still enables the continued use of the 0.10 speed point. It just requires using an equation that is in the ISO standard but is built into the ISO contrast parameters. Because the 0.3G point has an inverse relation with the film gradient, it tends to shift to the right with extended development and to the left with decreased development.



This tends to mute the degree of speed change in regard to development comparied to just using the 0.10 speed point without the “fix.” This fix is known as the Delta-X Criterion.

Another concept that I hope came across was that if the photographer is doing sensitometric film testing, it’s better to test using well determined speed methods such as the ISO standard, fractional gradient, or Delta-X and to either adjust the meter’s EI or the speed constant, to change the speed point-metered exposure ratio, if more or less exposure is desired.

 

[markbarendt]

But I have trouble gauging gray. When I meter off what I think should be the gray of my scene, say the middle of an old paved street, I am surprised to find it is my highlight!

Bill you make "my" point perfectly.

I use a Beseler PM2L to set enlarger exposure and when measuring negatives I constantly get surprised on where I find the darkest and lightest points.

But alas it's not really "my" point. It's a point that I finally "got" when reading Dunn and Wakefield's, Exposure Manual, and that point has been reinforced by Stephen and others and in practice.

Finding or putting a "known" of some type into the scene is key to finding camera settings in a quick, reliable, and repeatable manner.

And it's not an incident vs spot meter thing. Dunn and Wakefield teach that they are fully interchangeable.

An incident meter brings along it's own "known" target but we still need to know what we are reading. The first question with an incident meter should be, is the reading being taken in the same light as the scene/subject? If not, an adjustment needs to be applied, the same exact adjustment as a spot meter would need for a known target in that same light.

Knowing "What point/zone are we reading" is always key.

Cute story. Time has passed so the details may be remembered differently by others but I attended a workshop with Jim Galli and Per Volquartz and a bunch of great guys a few years back and one of the things we did was a portrait of Per in the Goldfield NV court room.

During the initial set up there was a scramble by everyone around to grab their meters and most were spot meters. This was followed by lots of readings and fun discussion of the minutia and Maths.

Yes my incident meter was among the fray.

In the end, even though there was a bit of variation individually the consensus of the spot meter reading group matched my incident meter reading exactly.

Jim Galli was one of the guys stepping up to the plate to take a shot after all the fuss over metering.

He used an old barrel lens with two dark slides as his shutter. (there was a url link here which no longer exists)

Truly a case of measure with a micrometer, mark it with a crayon, and cut it with an axe.

 

[Stephen Benskin]

It's a point that I finally "got" when reading Dunn and Wakefield's, Exposure Manual, and that point has been reinforced by Stephen and others and in practice.

Finding or putting a "known" of some type into the scene is key to finding camera settings in a quick, reliable, and repeatable manner.

And it's not an incident vs spot meter thing. Dunn and Wakefield teach that they are fully interchangeable.

I suggest anyone who owns Exposure Manual to take another look at Appendix B. Much of it is exactly what has been discussed on this thread. Figure 8.1 shows the relationships between the fractional gradient speed point, the 0.10 density speed point, and the metered exposure among others.



Under the heading (i) Scene Luminance, Film Speed, and Camera Exposure Relationships, Dunn writes:

"Thus when an exposure meter which meets the quoted standard specifications is used to measure the reflected light from the simulated "average scene" gray test-surface in daylight, and the computer dial of the meter is used in conjunction with the rated speed of the film to determine the camera settings, the resulting exposure Hg, on the film for that scene element is expected to be 10 times greater than the exposure Hm involved in the sensitometric derivation of the film speed.

The ratio Hg/Hm is important because it indicates the level of exposure which will be obtained by the use of the rated speed with a calibrated exposure meter in normal daylight photography...This is the situation at the time of writing, and it may be regarded as wholely satisfactory as after the removal of the larger safety factor included in earlier days it is now left to the photographer himself to apply any safety factor he may consider desirable for the work he is doing. Using a large format camera he can, if he wishes, play safe and revert to a safety factor of 2 or even 4 without endangering print quality. If on the other hand he is working with a 35mm camera and is seeking negatives capable of big enlargement he can follow his exposure meter intelligently with only such modifications as may be suggested by unusual scene contrast or unbalanced luminance distribution."

 

[markbarendt]

I suggest anyone who owns Exposure Manual to take another look at Appendix B.

I'd suggest that anyone who is really interested in understanding the how's and why's of exposure and metering should get and read as late an edition as they can find, if they don't already have it.

It is not just another "Zone System" knock off.

 

[Stephen Benskin]

I agree. I have the fourth edition from 1981.

 

[Bill Burk]

... I attended a workshop with Jim Galli and Per Volquartz and a bunch of great guys a few years back ...

Great story.

 

[MattKing]

I think this is relevant to the discussion here.

I have one of those fragments of memory that has stuck with me. It was an article about using the Zone System, and was related by someone who had run a workshop with a number of relatively inexperienced LF photographers.

Apparently, at the end of a day of shooting, they all developed their negatives and then got together to have them evaluated. The workshop instructor noted a common and consistent problem - while everybody seemed to have a good technical understanding of the Zone System principles, in a large number of cases the areas of the scene chosen by the students to base their exposure choices on were either too dark, or too light, or both.

In other words, they gauged very deep shadows as being "important shadow areas showing substance and detail" and very bright highlights as being "important highlight areas showing substance and detail", and they were adjusting their development to contract almost everything.

The problem wasn't with the theory, it was with training the perception.

 

[Chuck_P]

The problem wasn't with the theory, it was with training the perception.

I agree, having a good technical understanding of ZS principles doesn't always translate to a visualization that satisfies the intended final aesthetic.....speaking from experience, of course. Often it can turn out that a different visualization is needed.

 

[markbarendt]

I agree that over visualization/idealization can be a problem, but I also think the solution is easy; printing and practice.