Understanding Incident Metering

[Pere Casals]

We could talk about specific circumstances where that's true

R-27 Kodak grey card includes precise instructions about how to use the grey card, it has to be angled depending light direction...

I find that the instructions in the R-27 are very well explained, this was really helpful to me.

but many people find a 1 stop difference between spot meter and incident meter.

For caucasian skin there is just one stop difference, but this is very easy to explain.

1) I take incident metering of light illuminating a wall

2) If I take an spot metering... if the wall is white I'll real +2, it is black I'll read -2.

As we normally place white skin in Z-VI (or VII) we have to overexpose (+1 or +2) what we read with spot meter in the cheek. If it is black skin we will place the face in Z IV (or V).


The spot metering depends on subject's reflectance, so we'll need to correct exposure depending on the zone we what that spot... what is true is that a well done spot metering on the gray card should match the incident meter

I just check it both ways and make an exposure determination based on which type of meter I think was more likely to give me the correct exposure in that situation.

I agree...

IMHO it's just about understanding what we are doing. If we use the incident metering (with uniform light) we know that subjects reflecting 18% like a grey card will be in Z-V, and that white things will be in Z VII or VIII, from that we can take decisions to modify exposure.

Instead, if we explore the scene with the spot meter we will know in what zone we will have each spot, for each exposure we consider, so we can change exposure and/or plan a custom development...

[Bill Burk]

For caucasian skin there is just one stop difference, but this is very easy to explain.

1) I take incident metering of light illuminating a wall

2) If I take an spot metering... if the wall is white I'll real +2, it is black I'll read -2.

As we normally place white skin in Z-VI (or VII) we have to overexpose (+1 or +2) what we read with spot meter in the cheek. If it is black skin we will place the face in Z IV (or V).

The spot metering depends on subject's reflectance, so we'll need to correct exposure depending on the zone we what that spot...

I agree... and it is easy to agree when we talk about reading a subject and placing it on an appropriate Zone. So long as we aren't talking about that dratted gray card.

What is true is that a well done spot metering on the gray card should match the incident meter

This is the point on which I'd like to bring you around to my way of thinking.

I teach that it doesn't necessarily match the incident meter.

We are talking about a minor difference, around 5%.

So it's not something to get up in arms over.

I mean, 5% is a small difference. In terms of Zone System, the gray card remains in Zone V.

But you might want me to answer how can it be possible that a gray card is not Zone V when you meter and place it at Zone V for camera tests?

Isn't a gray card Zone V by definition?

Yeah I guess so. I'm going to have some trouble explaining that.

I think the 2/3 stop difference that people often find in film speed tests has something to do with using the gray card for Zone V.

Try this experiment, for a 400 speed film, such as Tri-X set the incident meter at 400 and the spot meter at 250 and wander around taking meter readings...

Let me know if you don't find readings of the gray card and the incident readings are in agreement more often.

[Alan Klein]

I found that gray card reflective vs. incident measurements are up to a stop difference. I read somewhere that the meter manufacturers design has 18% vs. 12% for gray and that Ansel Adams may have had something to do with that. I don't have time right now but maybe someone could do the research and report the results here.

[Bill Burk]

Isn't a gray card Zone V by definition?

I think the 2/3 stop difference that people often find in film speed tests has something to do with using the gray card for Zone V.

Try this experiment, for a 400 speed film, such as Tri-X set the incident meter at 400 and the spot meter at 250 and wander around taking meter readings...

Let me know if you don't find readings of the gray card and the incident readings are in agreement more often.

Scratch that. Though the experiment would work, it's not the same 2/3 stop.

All three kinds of light meters are calibrated to give the same resulting reading, but they all "intend" to place an amount of light that is 10 times the speed point on film.

Zone System "intends" to place 4 stops more than the speed point (Zone I) on film.

That's where 2/3 stop comes from directly.

The difference of an 18% gray card and the standard scene that all the meters are calibrated to...

That's what we are searching for. It's not the 2/3 stop I mention above.

[Pere Casals]

I teach that it doesn't necessarily match the incident meter.

We are talking about a minor difference, around 5%.

So it's not something to get up in arms over.

I mean, 5% is a small difference. In terms of Zone System, the gray card remains in Zone V.

But you might want me to answer how can it be possible that a gray card is not Zone V when you meter and place it at Zone V for camera tests?

Isn't a gray card Zone V by definition?

Yeah I guess so. I'm going to have some trouble explaining that.

I think the 2/3 stop difference that people often find in film speed tests has something to do with using the gray card for Zone V.

Try this experiment, for a 400 speed film, such as Tri-X set the incident meter at 400 and the spot meter at 250 and wander around taking meter readings...

Let me know if you don't find readings of the gray card and the incident readings are in agreement more often.

Bill,

First, the calibration of photographic light meters is covered by ISO 2720:1974. Calibration of in camera internal meters is specified by ISO 2721:1982.

Recommended ISO calibration factors spread 1/6 EV, with recommended K in the range 10.6 to 13.4 (cd/m²). A manufacturer is free to use what K he wants, but there is that recommended range.


IMHO with on axis illumination the incident metering and the spot metering on a card should match... The ISO 2720:1974 recommendations are aiming that.

...but we can have practical situations that would make the readings differ because illumination is not like the one in the calibration conditions:

> Light is more or less directional/diffuse, and the imperfect lambertian card reacts different than the imperfect half dome in the incident meter.

> Light is not on axis, for example sunlight illuminating a face in a portrait or a building can come from sun elevated 30º, or 45º or even 90º. Because that Kodak recommends increasing the exposure by 1/2 stop when sun is high in the sky. Again a half dome and the card react different to out of axis illumination.

the meter manufacturers design has 18% vs. 12% for gray

What I understood from the reading of the ISO norms is that incident light meters are calibrated to a 12% reflectance in the formulas while reflective meters are calibrated to around 18% (16-18%), in that way both kinds of meter deliver the same reading with on axis illumination on a 18% lambertian reflective subject.

So IMHO that 18% vs 12% difference in the calibrations is introduced just to make the incident reading match the spot reading, but this is with on-axis illumination, when illumination is well not on-axis we'll have a difference coming from different reaction to the off-axis light depending on if we read throught a dome or from a naked photocell.

please correct me if I'm wrong...

____________________________


Then we have spectral sensitivity of the sensors, some are spectrally flatter than other... and depending on the light, subject color and filtering...

[Alan Klein]

Bill,

First, the calibration of photographic light meters is covered by ISO 2720:1974. Calibration of in camera internal meters is specified by ISO 2721:1982.

Recommended ISO calibration factors spread 1/6 EV, with recommended K in the range 10.6 to 13.4 (cd/m²). A manufacturer is free to use what K he wants, but there is that recommended range.


IMHO with on axis illumination the incident metering and the spot metering on a card should match... The ISO 2720:1974 recommendations are aiming that.

...but we can have practical situations that would make the readings differ because illumination is not like the one in the calibration conditions:

> Light is more or less directional/diffuse, and the imperfect lambertian card reacts different than the imperfect half dome in the incident meter.

> Light is not on axis, for example sunlight illuminating a face in a portrait or a building can come from sun elevated 30º, or 45º or even 90º. Because that Kodak recommends increasing the exposure by 1/2 stop when sun is high in the sky. Again a half dome and the card react different to out of axis illumination.




What I understood from the reading of the ISO norms is that incident light meters are calibrated to a 12% reflectance in the formulas while reflective meters are calibrated to around 18% (16-18%), in that way both kinds of meter deliver the same reading with on axis illumination on a 18% lambertian reflective subject.

So IMHO that 18% vs 12% difference in the calibrations is introduced just to make the incident reading match the spot reading, but this is with on-axis illumination, when illumination is well not on-axis we'll have a difference coming from different reaction to the off-axis light depending on if we read throught a dome or from a naked photocell.

please correct me if I'm wrong...

____________________________


Then we have spectral sensitivity of the sensors, some are spectrally flatter than other... and depending on the light, subject color and filtering...

I'm not familiar with the technical. So I assume you're correct. What I am more interested in is practical application. So when I'm trying to determine exposure, what would be your recommendations when using my meter that has a dome for incident and a 10% angle for reflectance? How to aim and how to interpret? Any suggestions would be appreciated.

[Bill Burk]

Recommended ISO calibration factors spread 1/6 EV, with recommended K in the range 10.6 to 13.4 (cd/m²). A manufacturer is free to use what K he wants, but there is that recommended range.


IMHO with on axis illumination the incident metering and the spot metering on a card should match...

So IMHO that 18% vs 12% difference ...please correct me if I'm wrong...

____________________________


Then we have spectral sensitivity of the sensors, some are spectrally flatter than other... and depending on the light, subject color and filtering...

I think you will find manufacturers and standards never talk of 18%. That is just a card that we have for our own use.

As I understand the manufacturers use the spread for K and C to try to make the exposure recommendations identical. They use the spread to account for sensor differences, etc.


10 times the speed point is the target amount of light that a meter will try to put on film. And they all want to do that. Unless you do Zone System tests. Then the aim is four stops above the speed point instead of ten times.

[Pere Casals]

I think you will find manufacturers and standards never talk of 18%. That is just a card that we have for our own use.
As I understand the manufacturers use the spread for K and C to try to make the exposure recommendations identical. They use the spread to account for sensor differences, etc.

My guess is that manufacturers follow the ISO recommendation that specifies a range, so the we have that 1/6 EV differences between meters, so they are in fact aiming a 16% to 18% reflectance.

10 times the speed point is the target amount of light that a meter will try to put on film. And they all want to do that. Unless you do Zone System tests. Then the aim is four stops above the speed point instead of ten times.

You are right, this comes from the ASA PH2.5-1960 speed change.

In 1960 Box Speed was doubled without any film manufacturing change, so in theory the Zone System experimented a shift.

Divine ZS inspiration came to AA while shooting the Half Dome with an Adon, this was 1927, 33 years later ASA changed the 20x to 10x.


Anyway IMHO it can be debated if we should or not rate film at its half speed to use the ZS, now we have accurate meters and we don't need that additional safety factor of one stop that was removed in 1960.

AA says Z-II is "Textured black; the darkest part of the image in which slight detail is recorded". Z-II is -3 stops in the meter, so outside the toe if using the post 1960 box speed.


My view is that ZS is not a recipe but a methodology & visualization and we need to ajust it to our process, depending on how we meter and develop we have an impact, so IMHO at the end we just need to consider a personal ISO for a film that makes our Z-II "the darkest part of the image in which slight detail is recorded".

This is what I personally concluded...

[Pere Casals]

meter that has a dome for incident and a 10% angle for reflectance? How to aim and how to interpret?

It should deliver a good exposure for normal conditions. The Problem with incident metering is when not all the scene is bathed with the same incident light, then may take more readings to decide what you over/under expose.

Also while with spot metering you know what over/under exposure you have in each spot of the scene, with incident metering (when illumination is challenging) you have to guess how the deep shadows and highlights will be, and particular film toe/shoulder has an impact.

One thing it can be done is shooting a 35mm test roll of the same film you use for LF. You just select some meaningful scenes to make bracketings, and taking notes of the incident metering, the SLR metering in matrix, ponderated and spot modes, with the spot readings of highlights and shadows. This would give you a solid criterion about when you should modify the indicent metering to suit your visualization, from the results you bracketed.

[IanBarber]

It should deliver a good exposure for normal conditions. The Problem with incident metering is when not all the scene is bathed with the same incident light, then may take more readings to decide what you over/under expose.

Very interesting thread, this to me sounds like even with roll film when "not all the scene is bathed with the same incident light" it may be just as easy to spot meter.