Understanding Exposure

A Complex Subject Made Simple for Beginners

History & Moonshots

When I was in my teens (in the mid-1960s) and just starting to makemy way in the world of photography in a serious manner, Pentax broughtout the Spotmatic. This was an SLR, very much in the modern idiom, butit contained the world‘s first behind the lens metering system. One had to manuallyclose down the aperture to take a reading, and notwithstanding the name it wasn‘ta spotmeter, but everyone who bought one thought that they‘d died and goneto heaven. Imagine — a camera that could read exposure through the shootinglens, with filters, bellows extension and all. Damn, technology was advancingso much we might even land a man on the moon one day.

Tables, Charts and Selenium

Prior to this taking accurate exposure readings was problematic. There were,of course, clip on meters, hand-held meters, extinction meters (you‘re definitelyover 50 if you remember these) and various exposure charts and tables. Therewas even that extremely esoteric Zone System from Ansel Adams thateveryone in the know was talking about — and not understanding.

Hand-held meters primarily had selenium cells — large honeycombmatrixes. Downsides were that they could be dazzled by very bright light andthey didn‘t work worth a damn at light levels much lower than a cloudy day. CDScells (cadmium sulphide) were just coming out and their small sizeallowed them to be placed in shutter-coupled clip-on meters and then eventuallyto become in-built, as with the Pentax Spotmatic.

All of this with the purpose of helping photographers expose their filmcorrectly.

Anyone getting into photography today takes having sophisticated matrixmetering systems in their camera for granted. But even these are fallible, andunless one understands what the automation is doing getting accurately exposedphotographs under difficult lighting conditions can be problematic andfrustrating. Let‘s start with some basics.

Sunny 16 & Beyond

Tumbleweed Dunes, 2000

The extremes of brightness that one encounters in the natural world are notthat varied. For this reason there is the so-called Sunny 16 rule. Thissays that on the brightest day normally encountered the proper exposure isroughly the reciprocal of the film speed at f/16. Thus, if you are shooting ISO200 film then the exposure will be 1/250 second @ f/16. This is the same whetheryou‘re in Auckland or Amsterdam, mid-summer or mid-winter. 

From the extremes of a sunny day outdoors down to typical indoor roomlighting covers a range of about 10 stops. With the exception of seldomencountered situations like fireworks, cityscapes and moonlight scenes these 10stops encompass every lighting situation you are ever likely to encounter. Onlyon the ski slopes or at the beach will you need to stop down one more stopbeyond Sunny 16 because of reflections off the snow and sand.

So, that being the case, why is exposure so difficult? Most people shouldhave no problem in recognizing 10 different light levels, shouldn‘t they?

The Eye‘s Autoexposure

Unfortunately (or fortunately, depending on your point of view), thehuman eye and brain have a superb autoexposure mechanism built in. This meansthat once your eyes have adjusted to the current lighting situation, and withoutclues as to what is causing the light level encountered, it is almost impossibleto tell how bright things are on a relative basis. As long as the light levellies somewhere within that 10 stop range for most people it all appears thesame.

This is why light meters, whether built-in or handheld, are such vital tools.But before exploring light meters and how best to use them it‘s worthwhile tohave in ones mind a firm idea of what "proper" exposure settings arefor the ten light levels normally encountered. This way you‘re not a blind slaveto the meter.

Let‘s assume an F stop of f/8 and a ISO (film speed) of 400.Here‘s what these 10 light levels are and the shutter speed that would beneeded.

A Sunny day outdoors — 1/2000 sec A hazy bright day — 1/1000 sec A bright cloudy day without shadows — 1/500 sec An overcast day, or open shade on a sunny day — 1/250 sec A heavily overcast day — 1/125 sec Deep shade. The woods on an bright overcast day — 1/60 sec Just before a thunderstorm or late on a heavily overcast day — 1/30 sec A brightly lit store interior — 1/15th sec A well lit stage or sports arena — 1/8th sec A well lit home interior — 1/4 sec

Of course you would vary the F stop and shutter speed combinations towhatever would be most appropriate. In the case of a home interior, for example,instead of 1/4 second at f/8 you might choose 1/30 sec at f/2.8. The point isthough that these 10 brightness levels represent 95% of the conditions underwhich we all do our shooting. 

What a Meter Does

A light meter does one thing. It tells you what the correct exposureis for13% Gray. This is approximately the tonality of green grass orconcrete. The meters that are built into almost all cameras today arereflectivemeters. This means that they are measuring the light being reflectedoff thesubject. This is a convenient way to determine proper exposure butthere arepotential problems because grass and concrete usually aren‘t our mainsubjects.

Fortunately, most of time the mixture of objects in a scene; grass, sky,people, trees, rocks and so forth, when averaged together usually are pretty close to an13% gray. But, because much of the time the things that we photograph are not soconveniently neutral in tonality manufacturers of TTL metering systems oncameras have had to go to great lengths to design multi-zone patterns andsophisticated software algorithms to enable their metering systems to providepleasing and accurate exposures.

Of Black Cats and Snow Storms

The classic examples of how reflective meters can be lead astray are theexamples of a black cat in a coal mine or a white cat in a snow storm. In bothcases if you were to trustingly use even the most sophisticated multi-zonematrix metering system you would end up with very badly underexposed andoverexposed negatives or transparencies. The reason, of course, is that themeter sees the black cat and coal, or the white cat and snow as being 13% gray.Even the smartest computer algorithms can‘t (yet) understand what the subjectis, and since there is nothing else in the scene they will provide"correct" but inappropriate exposures.

The smart photographer, knowing this, will dial in someexposure compensation. In fact, because most amateurs use colour negative filmwith its wide exposure latitude even this is seldom necessary and no one is thewiser.

But, the pro and the serious fine-art photographer typically shoots colourtransparency material which has a much more limited exposure latitude. Withthese, if yourexposure is more than a half-stop off, particularly in terms of overexposure,you‘ve forfeited the shot.

Incident to The Rescue

Usingthe L508 as an incident meter

An meter capable of taking incident light readings, like the Sekonic L508 reviewed on these pages, features what looks like a half ofa golf-ball-sized hemisphere, usually on a swiveling support. To take an exposurereading instead of pointing the meter at the subject, as one does within-camera and reflected meters, you instead place the meter in the samelight as the subject.

The hemisphere, or lumisphere as some call it, is designed as a 13%gray object and thus provides a reading equivalent to that which you would getif you took a reflected reading off a theoretically perfectly integrated scene,or a Kodak 18% gray card. (It really should be 13%, but don‘t ask!?)

The beauty of the incident metering approach is that you needn‘t carry alarge gray card around with you on location, and you don‘t have to worry thatyour subject matter— whether because of its colour or reflectance characteristics, will givean erroneous reading.

On The Spot

A third approach after straight reflective and incident metering is spotmetering. Strictly speaking spot metering is no different than other formsof reflective metering. You point the camera or handheld spotmeter at thesubject to be photographed and take a reading. But interpreting the reading andknowing what to read are the trick. From a practical point of view Iregard it as the most useful form of light metering practice.

A spot meter reads a very small area of the scene being viewed. In the caseof a meter like the Sekonic L508 it can read a variable area as small as 1degree. While using a standard reflective meter, a camera‘s built in meter or anincident meter is straightforward and doesn‘t requite much thought on the partof the photographer, using a spot meter implies having an understanding of film latitude,contrast ranges and other aspects of exposure control. 

Most frequently a spot meter is used in conjunction with the Zone System.Rather that delve too deeply into this sometimes arcane topic, here‘s an explanationof how to use a spot meter to both understand and control exposure using thesystem‘s basic concepts. If this sparks your interest you may wish to readthe article on this site on a simplified zone systemapproach.

In The Zone

The above example photograph consists of three images. On the left is astraightforward photograph taken on a hazy bright morning. At ISO 100 my CanonD30‘s built in meter read exactly 1/125 second at f/8 taking into account theentire scene.

As you can see from the histogram at the bottom right of the frame theexposure is technically perfect. The range from shadows to highlights is smoothand the shadows are just clipped, while there are clearly some extremehighlights at the far right.

Now look at the image of the Sekonic meter‘s LCD screen. I used it at thesame time to take 1 degree spot readings of various parts of the scene. It showsa digital readout of 1/125 second between f/8 and f/11 when reading the cedarshingles on the building‘s roof; about a half stop off from the Canon‘sintegrated reading of the whole scene.

If you look at the meter‘s reading again you‘ll see a small square above thepoint between f/2.8 and f/4 and another one above the point between f/22 andf/32. These resulted from taking readings at the points shown — in thesky and in the shadow area on the side of the building.

If you count the number of F stops between these two extreme readings you‘llsee that the range is 6 stops. The square at the center just above thepoint between f/8 and f/11 is the result of pressing the Average buttonon the meter, integrating the two extreme readings. It matches exactly thereading taken from the cedar roof, which is close to medium gray in tone.

What does this all tell us? First, you have to understand that most colourtransparency films have a contrast range of about 5 stops. The best currentdigital SLRs about 6 stops and colour negative film about 7 stops. Carefullyprocessed B&W film can have 8 stops.

The shot that you see above, because it was taken with a moderately high-enddigital SLR can just encompass the range needed for this exposure. Thereis just a hint of detail in the hazy sky and some modest detail in the deepshadows. The frame you see above has been imported in 8 bit mode and the onlyadjustment made was a bit of USM. No Levels or Curves adjustmentswere made. If I had imported it as a 16 bit file and spent some time with it in PhotoshopI could have extracted worthwhile detail in both the highlight and shadowareas.

Working with What Nature Provides

Clingmans Dome Sun. Great SmokyMountains NP. September, 2000
Photographed with a Rollei 6008 and 300mm Schneider lens onProvia 100F.

In the studio one can adjust lighting ratios and be in control of the entirelighting strategy. But, for us landscape and nature photographers we have totake what‘s served up and sometimes it‘s a dog‘s breakfast.

For this reason most serious photographers working outdoors carry a handheldmeter in addition to whatever metering capability their camera may offer. Ameter like the Sekonic 508 or new 608 is ideal because it providestwo meters in one, an incident and a spot meter. (In fact they are also flashmeters, but that‘s outside the scope of the article).

I personally prefer to use my Sekonic in incident mode for most types ofphotography. When this isn‘t possible I use the meter in spot mode and do what Idid in the above example — take highlight, shadow and mid-rangereadings and average them. Of course this presupposes that I have the luxury oftaking the time to do so. 

An Experiment

Try an experiment. Turn off the autoexposure mode of your camera and turn itto manual. If you don‘t have a handheld meter use your camera‘s meter to takeclose-up readings of various parts of a scene — the highlight areasthat you want to hold detail and the shadow areas where you also want to havedetail — and then see how many stops range are encompassed. Can yourfilm handle it? What‘s the average of the two readings? How does it compare tothe camera‘s integrated reading of the whole scene?

Now, take several exposures, at 1 stop intervals, making notes of what you‘redoing. Use transparency film because you‘ll then be able to judge directly whatthe effects of varying exposure are. 

Why do this? Because it will make you think about light, contrast and whatfilm is capable of. Just remember, don‘t let the highlights become overexposed.Clear film contains no information and there‘s little to nothing that can bedone to salvage an overexposed frame. A bit of underexposure is always to bepreferred.