Shooting Film at Box Speed

[Bill C]

Bill, I guess my question comes down to which of the two is the correct reading?

 

Just so no one has to go searching, here's what Alan is asking about:

I find a 1/2 to 1 stop difference when reading a gray card with a reflectance meter and an incidence reading both from the same meter. Am I nuts?

 

So I'd say there are several ways to look at this. First is the question of what is the "correct" reading, and what is the tolerance. You're giving a tolerance of sorts in your question - you say your difference is either 1/2 or 1 stop. If it IS 1/2 stop, well perhaps one method is 1/4 stop high and the other is 1/4 stop low. If this is the case I'd say that both are in an acceptable tolerance, which essentially would mean that BOTH are "right" (within the metering tolerance).

 

I get that you might not like this answer, that I'm waffling on the answer. Ok, I agree; I think that modern equipment ought to be more accurate than this. Ok, two more angles. First, if you really insist on knowing which method is "right," I'd say you gotta go with the incident reading. Why? Well, the metering standards DON'T SAY ANYTHING ABOUT AN AVERAGE SCENE REFLECTANCE OR AN 18% GRAY CARD. They only establish what is a reference light source that the exposure meter is calibrated against. So strictly speaking the incident reading seems like it ought to be a better metering method. But... we know that the incident reading can be affected by the angle of the light, so you have to use your judgment.

 

The second angle has to do with the "constants" in the metering equations (you need to go to the ANSI or ISO standard for this). There is an allowable range for each of these, and it allows the meter maker to bias for various reasons. You would think that the maker would stay near the center of the range, but who knows? If you wanna check your meter manual and give the "constants" I'll look at the standard and see where they fall. They should be in a "specifications" (or similar) section, and be called "K" for reflectance, "C" for incident.

 

Of course, the easy thing is to just do what the meter says, and if things are working ok, then you're probably fine; if not you can adjust accordingly.

[glen_h]

Just so no one has to go searching, here's what Alan is asking about:

 

 

(snip)

 

Ok, two more angles. First, if you really insist on knowing which method is "right," I'd say you gotta go with the incident reading. Why? Well, the metering standards DON'T SAY ANYTHING ABOUT AN AVERAGE SCENE REFLECTANCE OR AN 18% GRAY CARD. They only establish what is a reference light source that the exposure meter is calibrated against. So strictly speaking the incident reading seems like it ought to be a better metering method. But... we know that the incident reading can be affected by the angle of the light, so you have to use your judgment.

 

(snip)

 

Now that you mention angles...

 

Consider that a common incident light source is the sun, close enough

(for photographic purposes) to a point source. But incident light sources other

than the sun might shine on the subject from many different angles.

 

In the case of a flat subject, a flat meter that would automatically include

the cosine correction for off-axis light sources would be right.

 

But most subjects aren't flat. Many incident meters now have a white dome to allow

light from many directions, but still makes an assumption about the subject shape.

 

A reflected light meter should give the right exposure for sunlight close to directly

onto an 18% gray card, though watch out for the shadow of the meter. That should

also work for a subject of average (18%) reflectance, which is why averaging meters

work as well as they do.

 

Since a typical subject isn't flat, I am not so sure what to say about a typical

incident light meter. Also, much of the time the light source is in some direction

other than directly behind the camera. (But not causing a shadow on the subject.)

 

For all those reasons, I usually figure close enough if I am within one stop.

[rodeo_joe1]

In the case of a flat subject, a flat meter that would automatically include

the cosine correction for off-axis light sources would be right.

Except that the cosine 'law' rarely fully applies to real-world surfaces Glen.

 

The closer a surface approximates to a perfect Lambertian reflector, the less the cosine thing applies.

 

A flat Lambertian reflective surface re-radiates the light falling on it perfectly evenly over a 180 degree solid angle, and likewise accepts light equally evenly over the same angle. (It has to do with the surface micro-structure, which is like a miniature mountain terrain.)

 

Such surfaces are not uncommon in real life. A piece of plain white copier paper approximates a Lambertian reflector very closely. As should any reference grey-card worth having. Matt emulsion paint, cloud surfaces, non-greasy or sweaty human skin, and almost anything else that doesn't show a noticeable surface sheen from a shallow angle.

 

So take those commonly-held school textbook 'rules' and 'laws' with a big pinch of salt!

[glen_h]

Except that the cosine 'law' rarely fully applies to real-world surfaces Glen.

 

The closer a surface approximates to a perfect Lambertian reflector, the less the cosine thing applies.

 

A flat Lambertian reflective surface re-radiates the light falling on it perfectly evenly over a 180 degree solid angle, and likewise accepts light equally evenly over the same angle. (It has to do with the surface micro-structure, which is like a miniature mountain terrain.)

 

Such surfaces are not uncommon in real life. A piece of plain white copier paper approximates a Lambertian reflector very closely. As should any reference grey-card worth having. Matt emulsion paint, cloud surfaces, non-greasy or sweaty human skin, and almost anything else that doesn't show a noticeable surface sheen from a shallow angle.

 

So take those commonly-held school textbook 'rules' and 'laws' with a big pinch of salt!

 

For a flat surface, even with surface texture, there has to be some cosine effect, as a smaller solid angle

of the source covers the whole surface of the subject. Or another way, as the angle changes, there will be shadows

of the microstructure on nearby microstructure. Well, this applies to a source coming mostly from one direction,

that is, an approximation to a point source. (Like the sun.)

 

But flat subjects are somewhat rare, outside art museums. Inside art museums, likely there is

light from many directions. (And likely not a lot of it.). I suppose street art (graffiti) could be close

to flat and sun lit.

 

But anyway, yet it is rare, but useful for understanding less flat subjects and

less flat (such as hemisphere) meters in comparison.

[rodeo_joe1]

For a flat surface, even with surface texture, there has to be some cosine effect, as a smaller solid angle......

Doh!

Just read up on Lambertian reflectance Glen. Or simply play with a sheet of copier paper and see if you can get a significantly higher or lower reflectance from it by altering its angle to a light source.

[Bill C]

Except that the cosine 'law' rarely fully applies to real-world surfaces Glen.

 

The closer a surface approximates to a perfect Lambertian reflector, the less the cosine thing applies.

 

For a flat surface, even with surface texture, there has to be some cosine effect, as a smaller solid angle of the source covers the whole surface of the subject.

 

Doh!

Just read up on Lambertian reflectance Glen. Or simply play with a sheet of copier paper and see if you can get a significantly higher or lower reflectance from it by altering its angle to a light source.

 

Ok, I did the test. Glen is right.

 

The metered light falls off in proportion to how large the tilted card appears to the light source. (How can it be otherwise?)

 

Details: I used a Kodak gray card (on a stand) and measured it with a spot meter. The card was "normal" to the light source, or as most photographers would say, it was "facing directly toward the light." The card measures the same (to the spot meter) whether I read it from "head on" or from just about any direction (within reason). This suggests that it approximates a "Lambertian surface," or in photographer talk, it seems to be a "good diffuse reflector." In more plain language, it is not "shiny."

 

Then I turned the card to a (roughly, by eye) 45 degree angle. What this is gonna do is to present a narrower crossection to the light. In particular, it will be reduced to 0.707 of the original (in a 45 degree right triangle the hypotenuse represents the actual width of the card, and one of the sides represents the apparent width of the rotated card, as seen by the light source). Converted to equivalent f-stops this is a half f-stop, which is about what my meter reading showed.

 

So Glen doesn't have to re-educate himself, and that law of thermodynamics doesn't have to be rewritten.

 

Ps, for the casual reader, this whole thing is mostly limited to flat artwork, and how to meter the exposure.

[glen_h]

Ok, I did the test. Glen is right.

 

The metered light falls off in proportion to how large the tilted card appears to the light source. (How can it be otherwise?)

 

(snip)

 

Ps, for the casual reader, this whole thing is mostly limited to flat artwork, and how to meter the exposure.

 

Even more, diffuse flat artwork with close to a point source. Fortunately not so many outside art museums.

 

Inside, the light source will usually be more diffuse, with reflections off walls and such.

[rodeo_joe1]

Then I turned the card to a (roughly, by eye) 45 degree angle. What this is gonna do is to present a narrower crossection to the light.

But where did you measure it from?

 

OK the same 'amount' of light is spread across the card, at an angle. But at the same time, the apparent area of the card is reduced from the viewpoint of the light-source. With the nett effect that, viewed from the position of the light-source, the cosine-reduced illumination is emitted from a cosine-reduced area, and thus appears the same brightness.

[BeBu Lamar]

The reason to tilt the card is to simulate the spherical dome on the incident meter which is used when photographing 3D object.

[royfisher]

A bit of an aside, but here's a screenshot of a Zone System advocate making an exposure reading using a spot meter.

 

That's a diffuser he's holding in front of the meter.

 

The context is that he is demonstrating composition, not the Zone System, and is exposing Polaroid; any errors would have been easily corrected by the process now known as "chimping."

 

From Ansel Adams: Photographer (1981), a PBS documentary (recorded on VHS back when it first aired).