My little theory on exposure

[siverta]

This is probably easy and, perhaps, a stupid question.

If I measured the brightest highlight in a composition, and then measured the darkest shadow coud I get a "perfect" exposure if I choose an exposure that is directly in the middle of the two?

EXAMPLE:
Lets say if I find that the brightest highlight is 1/125@F/16, and the darkest shadow is 1/4@F/16, could I set the camera at 1/30@F/16 and expect a "perfect" exposure?

I have searched through books and the Internet for answer to this question, but could not find any :bang:


Sivert

 

[ChrisN]

The manual for my Weston V light meter describes exactly that procedure. I guess it just depends on what part of your composition you want to be perfectly exposed.

 

[wdenies]

The same does the Minolta autometer. The averaging of two readings is even done automatically.

Wim

 

[Ralf]

It sound a whole lot better that a letting the camera make a decision for you. Dont forget though that some films will not record the full brightness range of some scenes, therefore you would need to make a decision about whether to expose for the highlights (perhaps some intersting sky in a landscape) or for the shadows (some intersting rocks in the forground perhaps). I'm thinking here of slide film in particular.

B&W and print film will record the deatil accurately with the method you've decribed.

 

[mikule]

I can find out for your exactly, but it depends on the films 'error latency' to over or underexposre when shooting high contrast scenes. Not sure if error latency is the correct term, but I'll try and explain.

Aparently b&w film is good for 5 stops of exposure, meaning there is a range, say from f2.8 to f16 where your highlights and shadows will render reasonably well. The only thing is, shadows have an error margin of about 2 stops, and highlights have an error margin for about 5 stops.

For example, if your brightest highlight was at f16, and your darkest shadow was at f2.8, same shutter speed, then you wouldn't go in the middle (between f5.6 and f8) for a 'perfect' exposure. You would be better over exposing the shadows by on or two stops, to say f4, and underexposing the highlights by 4 stops. That way you'll have a better tonal range over your print.

I hope this makes some kind of sense, I have it written in a book on exposure at home (currenly at work) so unless someone else can better explain this for me, I'll try to repost again once I get my facts and figures straighter 🙂

Hope this helps in the mean time.

 

[Roger Hicks]

There are several different things at work here, which is why I wrote a book called 'Perfect Exposure' (David & Charles/Sterling)

With negative films, all that matters is adequate shadow detail, i.e. the exposure is 'keyed' to the shadows, so you need only take a spot shadow reading. You cannot easily overexpose anything resembling a normal subject: some B+W film can record 10 stops or more, which can then be fitted onto the paper by (1) reduced development, in which case a second, highlight reading will be useful to know how much to reduce it, or (2) a softer paper grade or (3) dodging and burning. With colour neg (3) is the most useful.

With slides, you have to avoid 'blowing' the highlights i.e. the exposure is 'keyed' to these. You therefore need either a spot reading of the brightest highlight or (which is much easier) an incident light reading: the old name for incident light readings was the 'artificial highlight' method.

The first commercially successful spot meter, the SEI Photometer, had only two indices, shadow (for negative) and highlight (for reversal, especially movie). The so-called mid-tone index on more modern spot meters is of very limited use; IRE 1 (shadow) and IRE 10 are more widely used by those who know what they are doing.

The maximum brightness range of a print is about 200:1, so a subject brightness range of 200:1 or lless can be reproduced fairly accurately. More than 200:1 must be compressed (or highlights or shadows discarded) and less than 200:1 can be either expanded or one extreme or the other ignored.

This is all basic sensitometry as devised originally by Hurter and Driffield in 1890 and expanded in the Kodak Research Labs in the late 30s and early 40s with the First Excellent Print experiments, but there's more. Take a look at the Photo School at www.rogerandfrances.com for more -- there's plenty of free information as well as paid-for.

Cheers,

Roger

 

[aizan]

I have been doing what Roger recommends with my Hexar RF. I set the exposure compensatioin dial to -2, meter a shadow area, half depress and recompose. I don't use slides, but that would be similar. I think the meter is a partial meter, not center-weighted averaging. Anyhow...

 

[rover]

I have been successful metering an average lighted area similar to my subject, setting my exposure and recomposing. I tend to favor the shadows (I am shooting B&W) by not allowing any highlight areas into my metering area. Otherwise, I use an incident meter.

 

[Igor.Burshteyn]

This world is soooo small - I just prepared to answer with information that I picked up from "Perfect exposure" of Roger Hicks and then I saw original Hicks' response. Internet and photography are connecting people!

 

[jlw]

OK, here is the scoop in probably more detail than you really want!

The procedure of averaging highlight and shadow readings to obtain an exposure WILL work ("work" meaning that it will retain detail in both highlights and shadows) as long as the scene's brightness range falls within the film's recording range. It works really well in studio shooting, where you can adjust the lighting to keep the brightness range within your preferred limits (which vary from film to film, slide films generally having less range and negative films having more.) In fact, many exposure meters made for studio use have a memory feature that will perform the calculation for you. Very handy.

There are two big problems with this method, though, and in many situations they force you to do some decision-making rather than just averaging two readings.

The first problem is that many scenes (including most naturally lit ones) exceed the brightness range that the fim can record. As an extreme example, imagine an outdoor scene with the sun in the frame. You want an exposure that will record the spots on the surface of the sun, and also will record detail in the black fur of a skunk down in a hole in the foreground (not too close to you, let's hope!) Obviously, there's no way you can retain detail in BOTH those extremes! The detail at both ends of the brightness scale will be "clipped" (eliminated) and you won't get the effect you want.

So, what you have to do when the scene brightness range exceeds the film's recording range is decide which end of the scale is more important to you, adjust the exposure to capture detail in it, and accept "clipping" at the other end of the scale. Again, some meters have a "highlight and shadow biasing" feature that will make a rough calculation of this automatically -- but you still have to apply your intelligence to using it, because (again) different films have different recording ranges.

The second kicker is that even if you can pick an exposure value that will put the highlights and shadows within the film's recording range, the midtones (which is where the interesting details are likely to be) are likely to look too light or too dark. The reason for this is that the exposure values you're averaging are in a linear mathematical relationship -- but the film's actual response to light is a NON-linear relationship called its "characteristic curve." The shape of this curve varies from film to film, and also according to how the film is developed.

When you pick an exposure value, you're actually fitting a specific point on this curve against the straight-line progression of exposure values. At places where the curve crosses the straight line, you'll reproduce the brightness values you expect. Where the curve is above the line, the values will look too bright, and where it's below the line, they'll look too dark.

What that means in practice is that if you use highlight and shadow readings to "pin" the upper and lower points of the exposure, you'll need to adjust the slope of the tone curve to get the middle tones to fall where you want them. In b&w you can do this to some extent by altering developing time (this is the heart of the notorious Zone System) and/or by selecting paper contrast; when scanning, you do it by adjusting the gamma (midpoint slope) of the scan. With color your options are somewhat more limited, and that's one reason color exposure is more critical (especially with slide films, which have a smaller recording range and consequently less room for post-exposure adjustment.)

A good compromise procedure when you need to be really careful about exposure is to meter a midtone value normally to find a "base" exposure. Then measure the lightest important highlight and the darkest important shadow to determine the brightness range. Compare these to the base exposure, and if the range exceeds what you're likely to be able to capture on film, bias the midtone exposure toward either the highlight reading or the shadow reading, depending on which areas are more important to your conception of the final picture.

See, I told you this would be more than you wanted to know! Sometimes when I think about all this stuff, I'm amazed that photography works at all!

 

[GeneW]

One of the handiest metering techniques I ever learned (for B&W shooting on the run) is to meter the palm of your hand in the same light as your principle subject, then open up one stop. This is a quick-and-dirty way of setting an average. Looking at the scene you can decide whether to go with the reading as is, or modify it based on experience.

Gene

 

[wdenies]

Measuring for the highlights (slides) needs a compensation of +2 stops,measuring the shadows (negs) needs a compensation of -2 stops (rough values).
Why?
Because light meters are expensive, but dummy devices.
They are calibrated on the so called middle gray value. Everything they "see" will be translated to this gray value, they are not aware of the subject.
So:
metering the shadows: result overexposure
metering the highlights: result underexposure
You can test this by shooting a white, black and middle gray sheet without any exposure compensation : 3 times the same result
Now shoot
black with -2 stops
white with +2 stops
middle gray with 0 stops
Looks better?

The ultimate theory about metering/exposure is the zone system (mainly B&W) invented by Ansel Adams (hard stuff).
For more information search the web for "zone system".
If enough interest I am willing to make a summary of this theory.

Digicam owners have an easier job to find the right exposure : the on-line histogram (never buy a camera without this feature!).
How to use it?
Look at the article "expose to the right" on the luminous landscape site.

Wim

 

[billbarber]

I try to find "middle-gray" area to take my reading, then compensate from there. I have a light meter on the way to help. Jim Zuckerman's book Perfect Exposure was a great help to me. you can get it at Amazon. It sure simplified the way I metered my compositions.

 

[tetrisattack]

JLW needs some recognition for turning a pretty difficult concept and explaining it in simple terms - his is the response I would have written. 🙂

Practically speaking I know that some film/developer combinations yield more highlight contrast, for example, or longer shadows, so when shooting large format I'll often meter both extremes of a given scene, take a nice average, then fudge it a little in one direction or the other. That's encouraged, of course, by the weird shutter speeds on my graflex - 1/10th, 1/25th, etc. 🙂

 

[xcapekey]

hmm...my experience is that the idea of "perfect exposure" or "correct exposure" (in the categorical sense) is a bit elusive....

exposures, like composition, is something that helps interpret a scene...for example shooting someone that is backlit by a window...you can expose for the outside (f16@125) and get a strong graphical silhouette....or expose for the subject (f5.6@125) and get a nice window lit portrait...or expose to blow out the window completely (f2.8@125) and get a highkey portrait...

none of these are inherently perfect or correct, but they do interpret the scene differently for certain effects...

essentially, it depends on what you want something to look like...do u want an average looking scene or something with a bit more dynamic lighting, etc., exposure is then a tool to help acheive that type of look u want...

 

[Roger Hicks]

I must take issue with Wim's advice to search the 'net for information about the Zone System. This is not a personal attack on Wim : it's just a few notes about why you should approach the Zone System in general with some caution, and the Zone System on the 'net with even more caution.

It was devised by Ansel Adams -- who was, remember, a teacher of photography, as well as a brilliant photographer -- as a simplification of basic sensitometry, in the days before cheap, reliable densitometers. As a result, it is more complicated than it needs to be in some ways and insufficiently rigorous in others. It is based on the work of many others (especially Hurter and Driffield in the 19th century and Jones, Condit and others in the 1930s; the only important addition by AA (which indeed was brilliant) was the naming of Zones. The rest of the Zone jargon is best avoided.

Its biggest single drawback is that it is often followed blindly, like a religion, by those who do not really understand it. Again, I'n not attacking Wim: many good photographers use it too. But my experience is that once you understand enough sensitometry to make use of it, you no longer need it.

And of course, the 'net is a prize breeding-ground for cranks and the partially witted. AA's original books are not too bad, though they are outdated in places, but many subsequent works -- the most noble exception being Phil Davis's Beyond the Zone System -- are hagiographies and religious tracts rather than books about photography.

If the ZS suits you, great. If it doesn't, then walk away from it without a backward glance. Clearly it is far from essential -- many of the finest photographers in the world DON'T use it -- and the sky will not fall if you ignore it completely or even point out its shortcomings. For an expansion of this view go to www.rogerandfrances.com; to the photo school section, and then to the free section about why we don't use the Zone System.

Incidentally, on the IRE (Institute of Radio Engineers) scale, IRE 1 (shadow) is 2-1/3 stops below a mid-point and IRE 10 is 2-2/3 stops above.

Cheers,

Roger

 

[Roger Hicks]

I'll also appear argue with Xcapekey, though actually I am agreeing with him.

There is no such thing as a CORRECT exposure, but there is such a thing as a PERFECT exposure, which is why I gave my book the title I did -- I was quite annoyed when Jim Z. used almost the same title.

A perfect exposure is the one that gives the result you want. Learning to get that result is a matter of both theory and practice, and indeed that explains the sub-title of the book: it's full title is 'Perfect Exposure from theory to practice'.

Cheers,

Roger

 

[wdenies]

I agree
The zone system is not a bible, but gives you a good introduction in the problem of light meetring

 

[pshinkaw]

Roger:

Thank you for that explanation. I studied the Zone System very carefully about 25 years ago, and to this day it always seemed to to be unnecessarily complicated. The better I seemed to understand it, the more unnecessarily complex it seemed.

-Paul

 

[jlw]

I must take issue with Wim's advice to search the 'net for information about the Zone System. This is not a personal attack on Wim : it's just a few notes about why you should approach the Zone System in general with some caution, and the Zone System on the 'net with even more caution...

Bravo! Hear, Hear! ...etc.

Adams, to be fair, never claimed to have invented anything -- in one of his introductions, he explicitly stated that the system was simply his personal method for applying the principles of sensitometry.

In one of his later writings, he even admitted that the Zone System doesn't work as well as it used to, because modern films are less responsive to developing-time variations!

But a lot of the Zone cultists on the 'Net never got the message...