Incident reading distance between the meter and the camera

[caliber_60]

My question is:

 

when you take the incident reading, how do you account for the

distance between the subject been read and the camera? For example,

do I use the same incident reading if the camera is placed one meter

away or ten meters away? When the distance varies, doesn't the

amount of light received by the camera also varies?

 

Thank you for the input,

 

James L.

[richard_cochran]

You don't change a thing as distance changes.

<p>

Think about looking at or photographing a sunlit field

on a very clear day (so that atmospheric haze isn't a factor).

The grass a foot away from you looks no brighter than the

grass a mile away, or even the mountain several miles away.

The "sunny-16" rule applies equally well for close-ups as

for long distance landscapes.

<p>

If this doesn't seem to mesh well with the idea that light

from a point source falls off with the inverse square of distance,

realize this: The total amount of light reaching the camera from

a particular blade of grass DOES fall off with the inverse square

of distance, but the amount of film area covered by the image of

that blade of grass ALSO falls off with the inverse square

of distance. So the amount of light per square millimeter

of film stays precisely constant as the distance varies.

[david_henderson]

All that is true though of course as distance between subject and meter increases the probability that the light "over there" is different from the light "here" increases and you have to judge whether and to what degree this is occurring. That's one of the reasons why I don't believe that incident readings are the best way to measure exposure for landscapes.

[andreas_carl]

With incident metering, you are not measuring the brightness of the object, you are measuring how much light "falls onto" it. That dependes on the distance between light source and object. Camera position is irrelevant.

 

A simple test can show this: Go outside and take an incident reading from your front door of your house. Based on that reading, take a number of photos (slides, so that you can judge exposure) of that door, from different distances and different angles, WITHOUT changing exposure setting. You will find, that no meter where your camera is position, no matter how far away or from what angle, the door will lokk the same in all of your slides. By the way, this also show, that it is NOT appropriate to meter towards the camera position, you should meter towards the light source. (This of course is controversial, since it goes against long-held myths, but this simple test easily should convince you).

[ted_roberts]

Incidient readings when used properly give the most consistent and accurate exposure for the majority of situations and subjects. As long as you are in the same light that is falling on the subject, distance is not a factor. Just be sure that the white hemishperic diffuser "dome" of the meter is facing the camera lens. Buy "The Hand Exposure Meter Book" by Zuckerman, Silverman and Shell.... Ted

[caliber_60]

Thank you all for the replies.

 

When I take the incident reading, when is the difference if I stand behind, parallel and in front of the subject being taking the reading?

 

How do I avoid to include myself becoming part of the reading?

 

Thanks,

 

James L.

[andy_ongun]

It doesn't matter where you stand as long as you aren't blocking the light falling on the dome of the meter.

[andreas_carl]

Contrary to what is stated above and what I consider "myth", my recommendation is NOT to point the dome towwards the camera lens, but point it at the light source instead!

[david_henderson]

"Incidient readings when used properly give the most consistent and accurate exposure for the majority of situations and subjects"

 

I just don't agree with this. This forum is littered with debates about incident metering technique and there seems to be little agreement amongst practicioners about how to do it accurately in anything like difficult circumstances. Incident metering does not allow you to measure the variations in brightness around the setting you choose in the way that a spotmeter does. Decisions like "If I use f8 at 1/30 will I keep the detail in the shadow/highlight areas?" and "do I need to use a ND grad?" are reduced to estimation whereas with different techniques a conclusion can be reached based on measurement. If todays spotmeters had been available 50 years ago, my guess is that no-one would be using incident metering for ambient light photography today except for a few specialist applications.

 

Even within the narrow confines of this thread there is disagreement as to how best to execute incident metering. Aiming at the lens or aiming at the light -source will give you different exposures, possibly very different, in side-light. It all just seems more trouble than it's worth.

[conrad_hoffman]

For subjects directly illuminated by a light source, say a front illuminated ten step gray scale, an incident meter will be completely reliable. The problem occurs when some important part of the scene is in extreme shadow, or is lit in some different fashion. Then reflected readings (plus brain) will give better results. As an example, I recently had to shoot some people in front of a building with simple direct lighting. I carefully metered the white building, the dark areas that I wanted detail in, and various other things. I arrived at an exposure that I thought would place everything where it belonged on the film curve. I never take incident readings, but just out of curiosity, took one. The suggested exposure was *exactly* the same as all my fiddling around arrived at. I might add that this exposure was over a stop different than the center weighted camera meter suggested. I have to agree that for many subjects, an incident reading is the fastest way to get a reliable answer, but (like everything) you have to know when the subject doesn't match the assumptions on which the meter reading is based.

[Rob F.]

Conrad has a good point when he introduces the concept "brain" to the use of a meter. Only "brain" enhances not only the reflected reading, but also the incident one. Hollywood cinematographers shooting $100 million movies use incident light meters (Spectra). But they move them around to check highlight and shadow, trying how the light falls on different areas of the set.

 

In the final analysis, use of the brain will maximize accuracy with either technique. My tentative belief after only 45 years of experience is that if the final exposure choice varies from the incident reading by more than 1/2 stop, something's wrong in 90% of the cases.

[andreas_carl]

Yes, of course, you have to use an incident meter in a "smart" way. And you may have to compensate from the reading depending on what effect you want. But same is true for spot metering. If you use Zone system and can control the contrast range of your film, then of course the spot meter is far superior. If you don't have control over the contrast range of your materials, then it makes sense to determine what part of the picture is going to be the most important and take an incident reading of that.

 

Take David's example for example, people in front of a building. If you base your exposure on the contrast range of the building, you may up with great shadow detail, but the people's faces may be on a "wrong" zone. Assuming, that the people are the important elements, an incident reading off the people's faces would give proper exposure.

[andreas_carl]

Here is my preferred approach to incident metering:

 

I believe that the advice usually given (to point the meter towards the camera) is wrong. Take another example: You stand on an open meadow with afternoon sun. You take 4 pictures, pointing the camera north, east, south and west. If you incident meter by pointing the meter towards the camera, you will get 4 different exposures and the meadow will look different in each. If you take an incident reading of the main light source (by pointing the meter towards the sun) you will get consistent exposure in this case.

 

Of course, you must calibrate your system, depending on the technique you use.

 

So why does the "usual advice" (to point the meter towards the camera) work at all? Since the meter has a dome, it averages light from many directions, so directionality of the main light sources is reduced, but it is not removed completely, as the examples above show. (If you don't believe it, take your meter outside and try it). My preferred approach is to take incident readings with a flat diffuser disk instead of the dome. If lighting is straightforward, a single reading (pointing the meter towards the dominant light source) gives proper exposure. In more complex lighting situations, I take readings of light areas as well as shadow areas and determine my exposure depending on my vision for the subject.

[peter walker]

This seems to be a good case for the old "horses for courses"

adage.

 

Use an incident meter for situations where an incident meter will

perform well (e.g. portrait work, studio shots, product shots). In

other words, incident meters work well when the subject to

camera distance is not great, when you can get close to the

subject and when you want to measure the light falling on

different parts of the image.

 

Use a spot meter for situations where it peforms well (e.g.

landscapes, public performances, scenes with a lot of distance

variability). In other words a reflected or spot meter works well

when it is not easy to approach the subject due to distance or

barriers and when you want to measue the amount of light

reaching the camera - taking into account haze, etc.).

 

Today's meters usually come equipped for both roles so this

should not be a problem.

 

Regards

Peter

[cpj]

Peter Walker has the answer that makes the most sense. Just remember, in 99-percent of the cases you are not photographing "incident" light but "reflected light." An incident light meter in a studio (where the light does fall off with the square of the distance) can be useful as it can also be in cinematography where you need consistent exposure regardless of the brightness of the day because you have to edit the film into a seamless "print." In still photography--both color and black and white--you are photographing REFLECTED light. Your meter is going to produce a Zone V exposure for whatever part of the subject you are reading. Thus, if you know the Zone System, it is wise to read the darkest area where you want to see detail and the brightest area where you want to see detail and then make sure your exposure is such that the dark area is two stops under your selected exposure and the light area is two stops above that same selected exposure. If it's not, oiyu have to make a decision as to which end of the "brightness range" your are going to favor with your exposure. (With b&w and color negative film this standard way of measuring reflected light gives you a useable brightness range of 5 stops which, with b&w can be extended by push/pull development. With color reversal film--slides--you do not have that much latitude. At best you are working with a half-stop under to a full stop over to produce a useable chrome. So you have to meter the REFLECTANCE level of the most important part of the subject, or, if available, put an 18% gray card in the scene and measure it. [And yes, this should be the same reading you'd get with an incident meter held in front of the gray card.) But if you're trying to figure out the best exposure for a slide of the face of an open pit coal mine, you are forced to deal with reflectance and expose accordingly. A reading from the coal and a reading from the gray card should give you a good starting point for interpolating the correct exposure. Dark subjects need less exposure if they are going to look "dark" in the final slide and light subjects need MORE exposure if they are going to look "bright." This is why the "average" amateur photographer who relies on a built-in light meter always gets photographs of "gray" snow and washed out foliage when deep in the woods. Making good color transparencies of a scene with a wide brightness range--say 9 stops--is not easy--but that's why you pay $350 for a good, hand-held spotmeter. You just have to learn how to use it. As my lawyer friend always says, "F-11, Cloudy-Bright." And that's why I only depend on him for legal advice and I take my own spot meter readings. F-11, cloudy-bright, is an incident light reading without regard for the subject matter. And no, I don't want to start a long thread with incident light meter users jumping to the defense of a system to which they have adapted. But unless you are shooting a silhouette against a bright sky, you are photographing REFLECTED light and not INCIDENT light. So measure it and expose accordingly. As I said at the outset, Peter Walker put it best--use and incident light meter for the purposes for which it was designed. Good luck.

[Rob F.]

Today I took some shots in an industrial part of town. One shot included a lot of white painted trailer trucks. I took an incident reading because all that white might have inflated a reflected reading. Then I took a shot of a factory building. I used a reflected reading because the subject contained a mix of tonal values; and because a building behind me would have blocked the light falling on my meter dome, had I taken an incident reading.

 

Horses for Courses it is, Peter!

[garry edwards]

1. I agree with Peter Walker on this. Where the lighting can be controlled incident readings are normally 100% reliable, but that does not mean that reflected light readings are always going to be wrong.

 

2. the meter should be pointed at the camera, not the light source. What you want to measure is the amount of light falling on the subject from the camera position, and not the amount of light reaching the subject from a particular light source. Also, by pointing the sensor at the camera the sensor will go someway towards measuring the cosine effect (the amount of light reflected back towards the camera) and this is highly relevent with some highly-reflective subjects, such as stainless steel. Personally I find that flat receptors are far more accurate that domed receptors, which tend to take in their information from the light source as well as from the camera position.

 

3. Distance from camera to subject certainly is relevant in artificial light, due to the effect of the inverse square law, although of course this only applies in absolute terms when the light source is a point. When it is diffused there is still a fall-off in light reaching the camera, but to a much lesser extent than with a point source of light. You can test this by taking a REFLECTED reading from the camera position and comparing it with a REFLECTED reading from close to the subject. The sun is also a point source of course, but because of the distance involved the fall-off is purely theoretical.

 

4. Finally, it should remembered that no light meter or metering method is a complete answer, especially with 'difficult' subjects like, once agan, stainless steel. It is a guide, and needs to be tempered by experience.

[pete_andrews]

An incident light reading should be taken at the subject position. It reads the amount of light falling on the subject from the direction of the camera. The very reason for using a hemisperical diffuser is to automatically account for light falling obliquely on the subject, as well as light from other than the main source; light being additive.<br>The subject luminance does not 'fall off' as you move away from it, barring atmospheric haze of course.<br>The whole premise of incident light metering is that it will consistently and faithfully reproduce the tones of the subject on a correctly developed film, regardless of the actual reflectance of the subject. Pure white will always be 2.47 stops up from 18% grey, and 18% grey will always reflect 18% of the light that's falling on it, by definition. Shadow detail will then disappear into blackness approximately another 4.5 stops below that.<br>That's the whole basis for incident light measurement, and it works. It works very well, as evidenced by it having been the preferred metering method of demanding cinematographers and lighting cameramen for about 80 years.