Follow the nose of the subject?

[Steve Smith]

I have re-read the section on duplex metering but I'm still not sure if they consider an invercone to be equivalent to a flat receptor or a dome.

And now I have read the section specific to invercones and understand it a bit better. The book says that the invercones on the older meters such as the II and III are to be considered the same as a flat receptor and would benefit from the duplex method of averaging between the pointing at the camera and pointing at the light source readings whereas the newer IV and V meters had invercones which compensated for this and only needed the single reading taken whilst pointing at the camera.


Steve.

 

[Ian Leake]

It seems to me that there are two ways to skin this particular cat. You can either point the dome at the camera to get a 'proper' reading, then compensate a bit to allow for the bright highlights and let the shadows fall where they may. Or alternatively you can point the dome at the light source (so the meter does the compensation for you), and letting the shadows fall where they may.

 

[2F/2F]

Pointing the dome at the camera is merely one way of using the tool. You do that if you want to average all of the light sources falling on the subject, on a plane that is parallel with the film plane. My argument is that there are many situations in which that is not the way to get ideal exposures and control; the way I see it, these are most situations (pretty much any situation with shadows). And getting into the habit of following one method for every situation takes away a lot of the control and benefit of using an incident meter in the first place. They are extremely versatile tools. Don't handicap that versatility by blindly following the universally-accepted directions without thinking about why you are doing so, and certainly not without learning exactly how they influence your results. As far as the technical argument for the "meter the light method," I have already explained that ad nauseam.

 

[Ian Leake]

Don't handicap that versatility by blindly following directions without thinking about why you are doing so.

Wise words

 

[Lee L]

The reason I point out standard usage and the design parameters of an incident light meter is so that people will have a better understanding of the tool and how it is designed to work. Once that is understood, one understands better what's happening when departing from the method for which the tool is designed (i.e. pointed with incident dome toward camera in the same light as the subject). If a non-standard method of use is portrayed as standard usage, it can prevent that understanding of the tool's design and prove confusing, especially to beginners. Of course everyone is free to use their tools as they see fit. Understanding the design will help with making intelligent departures. See Phil Davis and his incident metering method in Beyond the Zone System for one example, where he shows how and why one can determine subject brightness range with an incident meter in many/most circumstances.

In practice, under many circumstances there may be less difference between standard readings pointing an incident dome toward the camera and away from it than many people assume, and film latitude (especially negative film) often easily covers many of those departures. The domed incident meter does a great job of integrating a reading in varying lighting circumstances as Dunn and Wakefield (and many others) have mentioned. It's more comprehensive a tool in standard usage than many people give it credit for. The best thing is for people to point their domed incident meters on and off the lens axis under varying lighting conditions and pay attention what happens to the reading.

The photo linked to by the OP was made with an FA, likely with early matrix metering. If the appearance on screen is indicative of what's on the negative, then I'd say it was underexposed and overdeveloped. The skin goes from blocked up highlights to blocked up shadow in the space of a few inches. To me it looks very much like film that was 'pushed' at least a couple of stops.

Lee

 

[markbarendt]

point the dome at the camera to get a 'proper' reading, then compensate a bit to allow for the bright highlights and let the shadows fall where they may.

The more I think about this the more I fall bak to this idea.

The advantage I see is that a classic reading can tell me essentially where to set the camera to get a normally exposed face on the side facing the camera.

I can then decide how to adjust that reading.

The idea of following the nose was to "tie" the dome angle to the face angle so that the meter reads the average of the light the subject is getting from the direction they are looking.

The face is given precedence and everything else is ignored.

That thought still deserves some experimentation.

 

[benjiboy]

Yes. I re-read mine recently and realised it was for a flat receptor and that the dome style diffusers automatically carry out the duplex calculation.

If I get a bit of spare time tomorrow I might do some experiments with my Weston meter with its invercone and a Zeiss Ikophot with it's flat diffuser to see what the variance is as I thought the invercone behaved as a dome rather than a flat receptor as you suggest. I suppose I should go and read that chapter again!.


Steve.

Western never made a flat incidental light receptor Steve, if you read this link it explains invercones better than I can http://www.johndesq.com/pinhole/invercones.htm, as you will see the invercones for the V and Euro master were bigger than the meters body and were the best incidental light metering receptor ever devised, they had a back leak facility that took into account some of the backlighting and have never been bettered.

 

[Steve Smith]

Western never made a flat incidental light receptor Steve.

I know. It's just in the book they suggest that the earlier invercones are treated in the same way as a flat receptor and the duplex method is used but say that the later invercones already compensate and that one reading pointing at the camera is sufficient.


Steve.

 

[benjiboy]

I know. It's just in the book they suggest that the earlier invercones are treated in the same way as a flat receptor and the duplex method is used but say that the later invercones already compensate and that one reading pointing at the camera is sufficient.


Steve.

As you write Steve with the last versions of the invercone the The Duplex Method isn't necessary, but receptor domes on all other light meters that the majority of people use with the conventional dome including even the current top of the line Sekonic, Kenko and Gossen digital meters don't have this back leaking ability in which case Duplexing is appropriate I.M.O. in backlighting.

 

[markbarendt]

I'm still trying to wrap my head around what benefit duplexing or for that matter ever not pointing the dome at the camera might give me.

Seems to me that both are meant to make a compromise between subject placement and highlight placement and that it normally does that at the "expense" of shadow detail and specific mid-tone placement.

Also seems to me that that compromise is subjective, surely one based on experience, but still subjective.

My thought here is that before we choose duplexing, following the nose, or any "split the difference" method, don't we still have to answer the question "where do I want the subject, highlights, shadows to land on the film curve?"

It seems to me that if I answer that question and take a direct reading, then based on similar experience it would be just as successful and just as subjective to dial in a my own offset, just like I place a zone with a spot meter.

What am I missing here?

 

[Steve Smith]

It seems to me that if I answer that question and take a direct reading, then based on similar experience it would be just as successful and just as subjective to dial in a my own offset, just like I place a zone with a spot meter.

I don't think you're missing anything here. If you spot meter an area and you know where you want to place the exposure of that area then you have arrived at the exposure settings you want.

Incident light reading, if done properly, eliminates any errors or adjustments you might normally need to make from 'difficult' subjects (white dresses, black suits, etc.) if using a reflected light reading.


Steve.

 

[Lee L]

I suspect that the 'back leaking' receptors were not adopted by other manufacturers for several reasons. One is probably licensing. Perhaps the most important reason would be the necessity to provide more complex instructions for meter use, because the user would have to learn to stand out of the way of both the incoming backlight and the forward receptor dome/invercone. I can see that being a potential headache given the attention paid to written instructions by many users.

One thing that is interesting to me in reading Dunn and Wakefield is that they indicate that any reflected light meter can be made into a flat receptor incident meter with a diffuser panel, and that one can construct their own backlight compensating incident dome with half a white table tennis ball if the receptor is small enough. The amount of backlight leakage can be adjusted by increasing/decreasing the amount by which the hemispherical receptor overlaps the meter body. Of course any modifications of a reflectance meter would require some recalibration (an EV or ISO adjustment), but one could calculate that with a few readings compared to a conventional reflected or incident meter. And with a PIC or Arduino, a light sensor, half a table tennis ball, and a little code, one could relatively easily make a back light compensating meter.

It would be nice if the linked BJP article included the illustrations that are referred to, and the character encoding capable of showing a degree symbol, but it appears he's more worried about theft of his article than disseminating information. (See his page source code.)

Lee

 

[Diapositivo]

Mark, I see it this way.

When you use a spot meter, you let's say measure the highlights, read the value, place this 3EV above middle grey, and you know it's going to land, in the film curve, 3EV above middle grey. No matter how contrasted is the scene, you know where that spot lands on the film curve. You also use the spot meter to "scan" the scene and see where other parts of the scene "land" on the film curve. You'll have an idea of where shadows will begin to block. But you will know that your highlights will not burn (supposing you know that in your film 3EV above middle grey give the detail you want).

Now let's imagine you use an incident meter. An incident meter, whatever you put in front of it (disk, sphere, or any other contraption) gives you an exposure that is:
- certainly right for a middle grey object;
- certainly right for any other darker of lighter object, provided that they fall within the capabilities of your film.

The normal way of thinking is that, when you use incident metering, everything will fall right on film: grey will be grey, dark grey will be dark grey, bright white will be bright white. And that's true, until you reach the shoulder and the toe of your film. In shoulder and in toe you will have less tones, less "description" of the subject. After that, you will have blocked shadows and burned highlights.

When you take a picture of, let's say, some flour and you want to maintain the flour texture, if you just use your incident meter you place the flour on the upper part of the curve. You might have a very flat rendering of the flour. If you want the "texture" of the flour to be more visible, you close the exposure a bit so that the flour is described by a "more linear" portion of the film curve. The same for the black cat on a coal carpet (you open this time).

Dome, disc.

We saw that when using a spot reflected meter, if you know the film curve, you know you will have no surprises on where the highlight fall.
When using an incident meter on a very contrasted subject (as in the case of extreme lateral lighting of the OP), and pointing the dome toward the camera you are averaging the luminous side and the dark side. Whatever this average is, it is not guaranteed not to burn your highlights or to place them where you will have enough detail.
If your goal is, as in this case, not to burn the highlights, you don't care how dark shadows will be and you don't want the shadows to skew your calculation. An "average" makes sense only until you know that all your subject is contained in the film curve. If the film curve is not broad enough to contain all your subject, "averaging" might just lead to have both blocked shadows and burned highlights. You place the center of your blanket just over your navel, and you have both your head, and your feet, in the cold.

In this case, you point the dome toward the light source, so that you know that your exposure will be right in the sense that your highlights will not be burned.

The rule of thumb is:
If the subject range is within film dynamic range, point the dome toward the camera. It will nicely average the light values, and you will place your exposure where both highlights and shadows will be correct.

If the subject range is larger than dynamic range, forget about averaging. Choose what you really want to salvage. If it's highlights you want to preserve, point the dome toward the light source, so that you'll be certain that you will not block the highlights.

IMO

Fabrizio

 

[markbarendt]

What I'm trying to get at is that turning the incident meters head toward the light source seems to be just as much of a guess as reading the meter normally and closing down a stop to protect the highlights.

With regard to placement with the incident meter I know that a face will be placed normally if I use a classic reading. If I want what the camera sees to be darker or more dramatic like the example in the OP then I can just close down a stop.

 

[2F/2F]

I would set up a hair stylist's mannequin, or even a basketball or something similar, with two hot lamps on either side. Take pictures at various lighting ratios from 1:1 to as contrasty as you can. For each ratio, make an exposure 1) for the brighter lamp, 2) toward the camera, 3) for the darker lamp (or no lamp) 4) for an averaged reading of the bright and dark sides, and 5) for a duplex reading – and averaged reading of the bright side and the side toward the camera. Let the results inform your decision making in the future.

You should see that for the exposures numbers 1 - 5 above, respectively speaking: 1) The bright side of the face is well exposed in ratio lighting. In 1:1 lighting, everything is well exposed. 2) The bright side of the face is overexposed in ratio lighting, but if using negative film, there is more on the neg to dodge out of the shadows if you print the bright side down to the "normal" print density. In 1:1 lighting, everything is well exposed. 3) The shadow side is well exposed and the bright side overexposed in ratio lighting. In 1:1 lighting, everything is well exposed. 4) The bright side of the face is overexposed by half the difference between the two readings in ratio lighting. In 1:1 lighting, everything is well exposed. 5) The bright side of the face is overexposed by one-quarter the difference between the two readings. In 1:1 lighting, everything is well exposed.

You will see that 1) in 1:1 lighting, you can meter with any one of these methods and get good exposures. 2) In ratio lighting, the only way to place the skin in a place on the negative such that it prints to the "right" tone when making a "normal" print (i.e. the only way to get the "correct" exposure for the skin) is to point the dome in the direction of the light. This does not mean, however, that you will have enough shadow detail to suit your taste. In cases in which the lighting ratio is wide, one of the averaging methods (down the middle, average dark/light, and duplex) will compromise the proper exposure for the bright side and the one for the dark side.

 

[Diapositivo]

What I'm trying to get at is that turning the incident meters head toward the light source seems to be just as much of a guess as reading the meter normally and closing down a stop to protect the highlights.

If you turn the incident lightmeter toward the light source, in a situation like that of the OP, you will have a reading that's guaranteed to preserve highlights. No guessing game there*.

If you instead point the light meter toward the camera, in that exact situation, and then close down 1 stop, you are "guessing" that 1 stop will be enough to avoid highlights deterioration. 1 stop might be too much, too little of a correction.

Fabrizio

* You can take the first exposure with this value, knowing that your highlights will be preserved, which is in our assumption our goal. If you don't want to bracket and want to be sure of an acceptable result, this is the one "correct" exposure to make.

From this firm ground, you can take further exposures, opening more and more, and see which is the exposure where you have more open shadows while still having acceptable highlights. That's a bit of "bracketing" (on one side only) to try to have the best possible results.

 

[markbarendt]

Well we need to square up on the assumption.

My goal in the OP is to simply to place the face, in this case I don't really care about where the rest of the exposure falls.

So I did a down and dirty test, in camera JPEGs uploaded to Flickr

http://www.flickr.com/photos/30056819@N00/5723538245/in/set-72157626728718220/

Like the example in the op the subject is in open shade camera looking out of the shade.

Used EI 400 and f/5.6 for all shots.

Classic metering got 1/250

Follow the nose 1/320

Meter the Main got 1/400

None of these methods came close. None had blown highlights on the subject.

From there I shot 1/500, 1/1000

At 1/1000, two stops reduced exposure from classic, the subject is starting to look a bit like the op example.

Could probably go 3 stops.

 

[2F/2F]

That is very different lighting than in the OP, which is far more contrasty. It is very flat. None of those three exposure readings should be expected to be much different, or to blow out the subject.

The underexposed ones ('500 and '1000) look underexposed to me.

The lighting is so flat that all of them could be called "acceptable" exposures except the last one ('1000 shutter speed), which is too dark to use without some manipulation.

The '500 shot looks much different than the '400 shot, which is odd, since the '250, '320, and '400 shots don't look all that different from each other.

My favorite is the '400, followed by the '320. I believe the '400 shot best represents a good density and renders the lighting ratio most accurately. '250 and '320 are too bright, and '500 and '1000 are too dark. But, as I mentioned, the very flat lighting does a lot to cover slop, so all but the '1000 shot are probably good enough to use.

I really suggest a more controlled test, done with artificial lighting and a tripod in various lighting ratios.

 

[markbarendt]

I agree that the test was far from perfect. Camera was in manual.

I do think that the lighting is close enough to further the discussion.

One variable we don't know about is how much manipulation was done to the example used in the op, I didn't take that shot. Hard paper, burn & dodge, PS, who knows?

My fav as just a straight portrait is actually the classic metering shot. This is just a mood/personal pref thing though.

I do think the one done @ -2 from classic is underexposed too.

 

[Rudeofus]

Incident meters, pointed from subject towards the camera, give perfect meter readings for reflective surfaces which observe the Lambert law. As it just so happens human faces are sometimes more specular than the perfect diffuse reflector, and that's where incident meters may end up reporting incorrect values, i.e. highlight sections burn out. The method of pointing the incident meter at the light source is a crutch: of course it will show a higher EV that way and avoid burned out highlights, but simply applying an educated exposure compensation does exactly the same, and probably better.

 

[markbarendt]

The method of pointing the incident meter at the light source is a crutch: of course it will show a higher EV that way and avoid burned out highlights, but simply applying an educated exposure compensation does exactly the same, and probably better.

I agree.

I'm not above using shortcuts when they work and save me time and metering the main sure could/can where it fits and is practiced in someone's system.

Intellectually the biggest problem that I have with the meter the main method is that it seems to be a hold-over from slide shooting and it is intentionally mis-placing the subject to protect the highlights.

First, with negatives that's not normally a problem on the film, and second, the need for protecting the highlights is a subjective/artistic choice anyway.

Third, and this for me is the deal killer, before I make the decision to compensate in any manner I need to decide, if it is even needed? If it is then I need to decide, how much? From experience before I turn the meter's head to the main I need to know how that will change my reading and if I know that, I don't need to even turn the head.

 

[Diapositivo]

Mark I think we all agree. In post #12 I stated pointing the dome toward the light source makes sense when using slides, or digital. That's because in that case one undergoes a serious risk, in such a situation, to burn the highlights and make a mess of the scene.

When using negatives one has a lot of spare room toward the highlights. Actually in a situation like that of the original post, if using a negative that I know has a lot of dynamic range I might even point the light meter at the camera and then open one stop, or more, to try to recover more shadows, or more probably just place the meter in an important shade zone, pointing it toward the camera, and use that value as is. With many negatives burning highlights is not easy even if done intentionally.

Also in post #12 I said that in that situation I think a negative could have been printed with less contrast and probably more details both in shadows and highlights. But I am not a printer so this is just an opinion based on what I see around. The printer presumably wanted that high-contrast rendering of the scene.

 

[markbarendt]

Mark I think we all agree.

We do.

When using negatives one has a lot of spare room toward the highlights. Actually in a situation like that of the original post, if using a negative that I know has a lot of dynamic range I might even point the light meter at the camera and then open one stop, or more, to try to recover more shadows, or more probably just place the meter in an important shade zone, pointing it toward the camera, and use that value as is. With many negatives burning highlights is not easy even if done intentionally.

Your thought is an artistic one and is exactly what I'm getting at.

Also in post #12 I said that in that situation I think a negative could have been printed with less contrast and probably more details both in shadows and highlights. But I am not a printer so this is just an opinion based on what I see around. The printer presumably wanted that high-contrast rendering of the scene.

You are surely right there too. The person who did that example made artistic choices of his own.