exposure for slides and digital

The great advantage of conventional neg/pos printing is that you can compensate for shortcomings in the negative at the printing stage. With a slide, you can't, which is why exposure for slides is so critical. Digital sensors are if anything even worse.

Pont des Arts, Paris

If you are shooting on negative, a picture like this is comparatively easy. As long as you give reasonably generous exposure, you can do pretty much anything you want at the printing stage. With slide -- this was shot on the fast, old, grainy Ferrania 640T -- or with digital, there is almost no latitude. Any less exposure and you lose the differentiation of the upper part of the bridge against the sky (which is pretty marginal anyway). Any more exposure and you lose the colour under the bridge. That is quite apart from the difficulty of determining the exposure, which in this case was a straight broad-area reading with a Weston Master 4 reflected light meter, no Invercone, favouring the sky and the Seine to get as much colour there as possible. Roger used an M-series Leica and 35/1.4 Summilux.

With slides, if you over-expose by more than about 1/2 or (at most) 2/3 stop, the colours soon become weak and desaturated. Over-expose still more and the highlights 'blow' to a featureless white. In the other direction, if you under-expose by 2/3 stop or less you can 'pop' or increase the saturation of the colours but further under-exposure makes them increasingly murky and muddy and the shadows rapidly 'block' to a featureless black.

Much the same happens with digital sensors, except that the latitude for over-exposure is even less -- though the latitude for under-exposure (especially with RAW files) may be somewhat greater. Although this module is written primarily with film in mind, almost all of it applies equally to digital metering and where there are significant differences, we say so.

Tarxien temple, Malta

This is a comparatively easy subject to expose, using an incident-light meter. The important brightness range is small -- the darkest shadows in the background can safely be ignored -- and the only real problem might be the light stone fooling a reflected light meter. Roger used a Leica M2 and a 35/1.4 Summilux, shooting on Kodak EBX on a dull, overcast day. The bowl in the middle was carved from a single piece of stone, probably two to four thousand years ago.

Before we look at the mechanics of metering and interpreting meter readings, it is worth looking at some more fundamental considerations about the nature of the 'right' exposure.

post-production manipulation

In the days when slides were shot mainly for projection and the slide show was commonplace, the only real possibility for post-production manipulation was duplicating. Although this could be used within very narrow limits to correct exposure and with rather more effect to change colour balance, it almost invariably meant an increase in contrast unless you used special low-contrast duplicating films or contrast control techniques which mostly came down to controlled flare.

Today, with the option of scanning transparencies, quite wide variations in exposure can be brought back to a more or less common standard. It is also possible to adjust colour saturation and contrast within quite wide limits. With most modern film scanners, even modest under-exposure is less desirable than a 'correctly' exposed or even slightly over-exposed slide. This makes slides quite attractive as originals for some kinds of printing: ink-jet, dye-sublimation or photomechanical.

A slide has the inherent advantage of providing a standard for comparison in a way that a negative cannot. Also, slide films are often sharper than negatives. For photomechanical reproduction, very few printers are willing to work from negatives, even if you supply a reference print: transparencies are vastly more welcome, and indeed, many publishers still prefer them to electronic image files. The only major exception is newspaper photography. Much newspaper photography is of course digital nowadays, but where film is still used, negative is normally preferred because of its far greater latitude: quality and colour accuracy come a poor second to reliability and flexibility.

But as this shows, negatives still offer a lot more latitude and the options of dodging and burning to a much greater extent than slides. With digital images, of course, it is pretty much the norm with all exposures to make the kind of adjustments that you can make to slides after scanning.

Ghar Hasan, Malta

This is just a happy-snap showing Frances at work but you can see that the picture on the left works better.

< /TD>

The image was scanned from a slide on Kodak EBX (a high colour saturation film with average-to-high contrast). Then, in Adobe Photoshop, Roger lightened Frances and some of the darker rock detail using the 'dodge' control and darkened the sea and some of the lighter rock detail using the 'burn' control. We would probably have got better results on colour negative but quite honestly for a happy snap it was easier to use the film and camera on hand; the camera was a Voigtlander Bessa-R2 with a 35/1.7 Ultron.

pushing and pulling

The vast majority of modern slides are processed by the standardized E6 process, or by compatible processes: E6 is of course the proprietary Kodak Ektachrome non-substantive process. See the free Glossary for definitions of 'substantive' and 'non-substantive'.

By varying the time in the first developer, it is possible to 'push' the film (increase its effective speed) or 'pull' it (decrease the effective speed). This is normally done purely as a means of varying speed but it also affects contrast and saturation to some extent. Pushing usually increases contrast and may increase saturation, while pulling often decreases both contrast and saturation.

With most modern films, these effects are far less obvious than they used to be, and their importance is steadily declining as manufacturers find better and better ways to overcome them: they are regarded as undesirable side-effects of pushing and pulling, rather than tools in their own right.

It is also disputable whether pushing and pulling can be called post-production techniques, because you have to plan in advance to use them, varying the exposure accordingly. But they are worth knowing about and they do not really fit in anywhere else.

Mdina, Malta, spring

Now that ultra-fast emulsions like the ISO 1000 Ferrania film used for this shot are no longer available, push processing is often the only answer. Roger used a 21/4.5 Zeiss Biogon on a Leica M2 for this shot.

All good digital cameras of course give the option of various ISO sensor speeds, but more speed normally means more noise (in many ways the digital equivalent of grain) and a general deterioration of image quality. On the other hand the loss of quality is often less than the more precious pundits would have you believe, and you do have the advantage of shorter shutter speeds (less subject movement or camera shake) or smaller apertures (greater depth of field) or both.

the 'correct' exposure

Although we have talked about 'over-' and 'under-' exposure we have not yet discussed what we mean by 'correct' exposure, and this is because the only 'correct' exposure is the one that produces the effect you want. An exposure that one photographer regards as 'correct' might be too dark or too light for another photographer -- or indeed, too dark or too light for another subject shot by the same photographer. Some subjects, after all, look best with richer, more saturated colours, and others with lighter, airier colours.

This is why we called one of our books 'Perfect Exposure'. In it we argue that there is no such thing as a correct exposure -- there is too much scope for personal taste -- there is such a thing as a perfect exposure, which is the exposure that conveys exactly the effect you wanted.

The only way to assure yourself of perfect exposures is to shoot a lot of a particular kind of film (or of course familiarize yourself with a particular digital camera), so that you can visualize the probable effect of a particular exposure level. The way we do it is to think of a hypothetical 'correct' exposure -- the one that would be recommended by a perfect exposure meter -- and then to think about whether this particular shot would look better with a little more or a little less exposure. When we say 'a little more' or 'a little less' we generally mean up to a stop in either direction, though 2/3 stop or even 1/2 stop is likelier and a difference of as little as 1/3 stop can materially affect the mood of a picture. You may find the free module on bracketing useful in this regard.

Stations of the Cross, Ta'Pinu, Gozo

Even with a high-saturation film (Kodak Elite Chrome EBX ISO 100) Roger chose to under-expose by 2/3 stop as compared with an incident light reading. This was to get more texture and tone in the sculptures; deepening the blue of the Mediterranean sky and making the statues stand out all the more was a side-benefit. Voigtländer Bessa-R, 90/3.5 Voigtländer Apo-Lanthar.


Closely related to the question of the 'correct' or 'perfect' exposure is the question of latitude; that is, how much the image can be over- or under-exposed as compared with the 'perfect' exposure and still remain acceptable.

The weasel word here is of course 'acceptable'. To whom? And for what purpose? In what picture? To the uncritical beginner, especially one that is trying to economise on film, very large deviations from the optimum may be fully acceptable: up to a stop under, and as much as a stop and a half over, though this is at least partly because the novice does not know what a really good slide looks like, or attributes good slides to luck rather than judgement. At the other extreme, when Roger was working in advertising photography in the early to mid 1970s, deviations of more than +/- 1/3 stop could be completely unacceptable in a sequence of matched slides. We have even met photographers who aim for variations of 1/4 stop or less -- one aimed for 1/6 stop -- but we believe that this is looking for more accuracy and consistency than is realistically attainable.

In one sense, latitude has increased significantly over the years: modern films are likely to prove more acceptable to the beginner over a wider range of exposures. On the other hand, to a photographer who knows exactly what he or she wants, latitude has never really changed much: the exposure that 'sings' is likely to be within 1/3 or at most 1/2 stop of optimum.

Festa San Giuzepp, Rabat, Malta

Which of these is more 'correct'? Latitude allows both to look reasonably natural, but with different moods. Arguably the lighter shot is 'better exposed', but we believe that the darker shot better reflects the way we saw the place. The base exposure (for the image on the left) was an incident light reading with a sensitive Gossen meter at maybe 4 metres from the nearest street-light. The other shot received a full stop more exposure. Roger used a Leica M4-P and 35/1.4 Summilux, shooting on Kodak Elite Chrome 100.

Latitude for digital cameras is another matter entirely, and is almost all in the direction of under-exposure: in a subject with a long brightness range, even automation with the cleverest multi-sector meters can lead to 'blown' highlights and 1/2 stop over is likely to be excessive if you are metering incident light. You should however be able to go 1 stop (or more) under, especially if you are shooting RAW files.

film contrast and saturation

Older photographers will remember when contrast and saturation were far more intimately entwined than they are today: more saturation meant more contrast and less latitude. The worst example was probably the old ORWO colour film, but early Fuji Velvia was somewhat the same way. When pushed, these films were even worse. But films like today's Kodak E100VS or EBX deliver remarkable saturation without a significant penalty in contrast.

On the other hand, high-saturation films should be treated with caution when you are photographing people. Even slight under exposure can result in excessively florid complexions and anything more than slight under-exposure can result in faces the colour of liver on a butcher's slab. As a general rule it is best to avoid such films for photographing people, at least if their faces take up much of the image area; if you are wedded to them (as we are) then err on the side of over-exposure rather than under-exposure.


Two rusty cannon, Grand Harbour, Malta

Modern high-saturation films can affect exposure technique. In the days of Kodachrome the only way to get this sort of saturation was to under-expose as far as you dared, risking 'losing' the edges of the cannon. Today you can be quite a lot more generous with the exposure, though in this case you would not want to go too far for fear of 'blowing' Fort St. Elmo in the background or the salt rim on the cannon. Roger shot this (as far as he recalls) with an M-series Leica and a 35/1.4 Summilux, cutting exposure 2/3 stop as compared with an incident light reading.

subject brightness range

There is of course a free module on this subject, but it deals mainly with black and white. As a rough guide, colours can be represented reasonably well across a range of about 5 stops, a log brightness range (free module) of 1.5.

By definition, the exposure must be keyed to any highlights that occupy more than a tiny area of the picture. If exposure is too generous, these will 'blow' to a featureless white, which normally looks awful. It therefore follows that the colours in anything that is more than about 5 stops down from the highlight will be increasingly murky, though there may still be a surprising amount of detail and texture in those dark, desaturated colours.


The highlights on the pavement are very slightly 'blown' but as they are a small part of the image it does not matter all that much. If exposure were cut 1/2 stop to recover them, the carousel and the rest of the scene would be too dark. As it is, the most important highlight is the white horse in the centre of the picture. Roger used a Leica M4-P and 35/1.4 Summilux, shooting on Kodak EBX ISO 100

From the importance of the highlights it further follows that subjects with a short brightness range (below 5 stops, and indeed preferably less) are easier to meter, and allow a greater latitude in exposure, than ones with a long brightness range. This is especially true of digital images or of slides that are scanned: slides for projection must still be within 1/3 stop or so of the optimum exposure in most cases if they are to 'sing'.

Paradoxically, the range of acceptable exposures of night scenes may often be surprisingly great, despite a very long brightness range. The reason for this is that often, the brightness range of the important part of the subject is quite small, as little as 3 stops, so the main effect of increasing or decreasing exposure is to bring in more or less shadow detail. Also, we are used to night scenes being contrasty and even to perceiving some of the highlights as 'blown': when the iris of the eye is wide open, as it may well be at night, the highlights may indeed look featureless or almost so, as do the shadows.

controlled lighting and fill flash

In the studio you can control the brightness range via the positioning and intensity of lights, and when you are working with available light you may be able to use on-camera 'fill flash' to brighten the darker areas and reduce the overall brightness range. Some cameras even have special fill-flash features.

Freelance Dreams, 1936

Roger was prompted to shoot this by acquiring a 1936 copy of the Writers' and Artists' Yearbook. Most of the rest of the stuff would have been available in 1936 though the Graflex is actually (just) post-war and we are not sure about the vintage of the Kodak film can. The thing is that with controlled studio lighting there is no real excuse for 'blown' highlights or 'blocked' shadows unless they are a deliberate part of the composition. This was shot on several formats; this shot is from a 6x7cm Fuji transparency shot using a roll-film back on a Linhof Technikardan 4x5 inch/9x12cm and a 210/5.6 lens, probably a Schneider Symmar.

There are however two significant technical risks with fill flash. One is that it is quite easy to overdo, so that (for example) a shadowed figure in front of a bright background is over-brightened, so that it becomes a bright figure in front of a strangely subdued background, and the other is that fill flash is often the wrong way around anyway: you need to lighten the background, not the foreground, but the foreground is by definition nearer the camera so the end result is an overlit figure against an adequately lit background.

There is also the aesthetic risk that by introducing extra light you will lose precisely the lighting that makes the scene so attractive in the first place, but this is a general problem with all on-camera flash, not just fill flash.


Boy in tree

Arguably a perfect case for fill-flash, but Roger preferred to use a compromise exposure, 'blowing' the background slightly, to create a more natural effect. Of course you need a lens/camera combination with very low flare and high contrast or the whole shot will be flat and muddy. Roger used a Leica M4-P and 35/1.4 Summilux, shooting on Kodak EBX ISO 100; he would have had slightly more contrast if he had removed the protective UV filter before shooting.

ultra contrast filters

This remarkable series of filters from Tiffen is somewhat mis-named, as they actually reduce contrast (and therefore effective brightness range) rather than increasing it, which is what 'ultra contrast' sounds as though it should do. In effect they introduce flare which reduces the image brightness range: the remarkable thing is that they have little or no effect on resolution. They are expensive, but they are so good that when they were introduced a few years ago they actually won an Academy Award in Hollywood.

exposure determination

At last, after all the other considerations, we come to the actual machinery of exposure determination -- and it is surprisingly easy. We look at in-camera meters and other broad-area reflected-light meters first, as the most convenient approach; at incident-light meters second, as the easiest and most reliable approach (though admittedly less convenient); third, at spot meters used for highlight readings, which are accurate if you know what you are doing, but inconvenient and tricky to master; and fourth at grey cards (free module), which we believe are completely superseded by other methods. Finally there is a separate section about establishing personal exposure indices.

in-camera and other broad-area reflected-light meters

Pretty much by definition, almost any meter will suffice for an 'average' subject. An 'average' subject has a modest brightness range: ideally between 64:1, 6 stops, log range 1.8 and 128:1, 7 stops, log range 2.1. The tones within that brightness range are more or less evenly distributed, and as a result, it reflects about 12 to 14 per cent of the light falling on it (not 18 per cent as many believe). It consists of green grass, blue sky, and perhaps a few buildings of yellow stone or red brick. Point the meter slightly downwards (so it is not unduly influenced by sky-light) and you should get a very good exposure.

Ferry Port, Rhodes

This is the kind of scene that just about any meter should be able to handle, reflected, incident or spot. You could even use the eccentric and little-known approach of spot-metering the sky and using the spot meter's mid-tone index. A deep blue sky should be an excellent 'mid-tone' in a slide. Roger used a Leica M4-P with 35/1.4 Summilux for this shot on Kodak EBX ISO 100.

The fun starts when the subject deviates significantly from 'average'. The classic extremes are a snow scene in bright sun and a black cat in a coal cellar.

snow scenes

A snow scene may reflect as much as 90 per cent of the light falling on it, and more than 60 per cent is entirely likely. If you rely on a simple reflected-light reading, it will therefore reproduce as a leaden grey: it is likely to be under-exposed by at least two stops. Of course you want to under-expose slightly, or the snow may blow to a featureless white, so a two stop compensation (open up two stops) is a good rule of thumb.


Snow on the road to the Drama School, Dharamsala

This is what is likely to happen if you rely on a through-lens meter in a snow scene on a misty day. The whole scene is grievously under-exposed. Although in this case it was deliberate -- Roger was trying to capture the gloom of falling night on a February evening -- it is still not outstandingly successful. The camera was probably a Leica M2 with a 35/1.4 Summilux; the actual exposure determination would have been with a Gossen Lunasix in incident light mode, minus two stops. Because the light is very flat, this is an easier shot to meter (i.e. with more latitude) than the one below.


The shot on the left is a straight through-lens meter reading and is about a stop and a half too dark. The centre shot is 2 stops lighter and is fractionally too light: there is very little texture in the snow. And the shot on the right splits the difference, and is fractionally too dark. The optimum would be between the centre shot and the shot on the right, so in this sort of contrasty, high-key light, exposure is very critical indeed. Of course with a scanned image or a digital shot it would be possible to extract all the information you needed from the picture that is just slightly too dark. This also shows the colour shifts that are normal in snow: the shadows reflect a great deal of blue sky light and are therefore much more blue than we remember them. Voigtlander Bessa-R2, 35/1.7 Ultron, Kodak EBX (the first time Roger had tried it in snow).

dark subjects and dark backgrounds

Sorry, we can't actually muster a picture of a black cat in a coal cellar but it's fair to say that the overall scene might reflect as little as 2 per cent of the light falling on it, and depending on the depth of the coal-dust, the rest of the cellar may not be much better.

If you rely on a simple reflected-light reading the scene will be over-exposed by about two stops and everything will once again come out as a grey; not as light as the snow, but still a lot lighter than it looks to the eye. Again you want to compensate a little in the opposite direction, around a stop to a stop and a half less than the meter indicates: minus two stops to compensate for the meter error, but plus half a stop to a stop to get texture in the cat's fur

A rather more common problem, at least for us, is brightly lit subjects against dark backgrounds. If you rely on the meter, the likelihood is that the background will still be somewhat under-exposed, but the brightly lit subject will be hopelessly over-exposed.

Coffee shop, Benares

A through-lens reading (or indeed any broad-area reflected-light reading) would result in hopeless over-exposure of the two figures as the meter strained to get detail in the background. Actually, Roger's camera (Leica M2 with 35/1.4 Summilux) didn't have a through-lens meter so the problem didn't arise: this is the exposure recommended by an incident-light meter (Weston Master IV with Invercone) without any further modification, shooting on Kodachrome 64.


This is all a simple illustration of the truth that all meters are designed to read 'average' subjects and will therefore tend to try to reproduce them as the same sort of 'average' print. Where they get clever is in trying to compensate for non-'average' subjects.

multi-sector meters

The cleverest multi-sector meters take numerous readings at different points on the scene and then, in accordance with empirically derived algorithms, computer-process the results in order to get the best possible exposure. These algorithms are all proprietary, and the various manufacturers will not let on what they are. But it is not hard to imagine an algorithm which, faced with six low readings and one high reading, concludes that the high reading is probably the one that matters -- a spotlit figure against a theatre curtain, perhaps -- and sets the exposure accordingly.

The problem is that all meters, no matter how clever, will sooner or later be fooled, and the cleverer they are, the harder it is to guess what degree of compensation will be needed to get the optimum exposure. Thus, although such a meter will give a higher number of optimum exposures to the novice, it may actually be less reliable and predictable for the experienced photographer. Of course, one option is to rely on the in-camera meter for easy, average subjects and then switch to another metering method for difficult ones. But, of course, that means you need the experience to know what is easy and what isn't.

incident light readings

The old name for incident light metering gives the game away: it used to be known as artificial highlight metering. The dome, Invercone or even flat receptor is translucent white plastic (opal or frosted glass on some very old incident light meters) and the effect is somewhat akin to sticking the meter cell inside a white plastic cup and basing your exposure reading on that.

In fact it is slightly more complicated than this, but only because most meters are designed for both incident and reflected use. The meter is calibrated for reflected-light use on the assumption that the subject reflects 12 to 14 per cent of the light falling on it. When used for incident light readings, the incident-light attachment must cut down the light to a controlled extent or the reading would be too high (leading to under-exposure). This is easily overcome at the design stage.

Red Saree, banks of the Ganges

With an incident light meter on an open river it is easy to take an equivalent light reading even if you are several score metres away. Of course you cannot read the shadows directly but if the river bank is in full sun you do not want to. Roger used a Nikon F and 200/3 Vivitar Series 1 to shoot this on Kodachrome some time in the 1980s; the meter would have been a selenium-cell Weston Master IV with Invercone incident-light attachment.

The shape of the receptor is important. This is why the Invercone of the old Weston Master meters is so highly regarded. Not only is it more compact than a dome: it also makes very good compensation for side-lighting and even for lighting that comes partially from the rear. Domes are just as useful for most applications but flat-plate receptors are quite a bit trickier.

With an Invercone or dome, the normal procedure is to take the reading on the subject-camera axis, with the dome pointing straight at the camera. With a flat plate, there are two possibilities. One is to angle the receptor so it splits the difference between the subject-camera axis and the subject-light axis (assuming a strong key light such as the sun), and the other is to take two readings, one on the subject-camera axis and the other on the subject-key light axis, and average the two. This is sometimes known as dual metering and can be used with domes and Invercones if the lighting is particularly tricky. A further interesting possibility is to take one incident-light reading and one reflected-light reading and split the difference between those two.

The great advantage of an incident light reading is that it directly reads the highlights, to which (as we have seen) the exposure is keyed. It is therefore perfectly matched to slide films and digital media. It is much less suitable for negative film readings, because if there is important shadow detail that is significantly more than 5 stops darker than the highlight, shadow detail will be inadequate.

The big disadvantage of incident light readings is that they must be taken from the subject position or under equivalent lighting. If you are standing in the middle of a landscape under a clear blue sky this is no problem, but if there are fast-moving clouds and you are shaded while your subject is not, or vice versa, you have a problem. Likewise if you are in the shade on one side of a river and your subject is in the sun on the other -- or again, vice versa.



Girls on a hot day

The glaring pale stone would fool many reflected-light meters, leading to under-exposure. An incident-light reading allows a much better assessment of the true light levels and makes it easy to cut exposure slightly for extra saturation or (as here) to over-expose very slightly to capture the heat. This is very much a grab shot -- Roger used a Leica M4-P and 90/2 Summicron -- but we decided to include it despite its defects as it captures the moment.

spot highlight readings

The first point to make here is that we are talking about true spot meters, not in-camera 'spot modes'. A true spot meter normally reads 1 degree, which makes it vastly easier to meter specific areas. The angle of coverage with in-camera 'spot modes' varies with the lens in use, and you normally have to use a pretty long lens before they cover the same angle as a true spot meter.

True spot meters such as the Pentax on the right are invaluable for high-precision metering for negatives because you can directly read the darkest area in which you want shadow and detail and base your exposure on that: exposure for negatives is keyed to the shadows.

A similar approach can be used for high-precision metering for slides and digital, by taking a spot reading of the lightest area in which you want shadow and detail and basing your exposure on that. This is arguably the best approach when you cannot approach your subject closely or you cannot take a reading in equivalent lighting, as noted above. On the other hand, it takes a little practice before you learn to recognize the brightest highlight in which you want texture and detail: it sounds easier than it is.

Of course you have to use the appropriate highlight index on the meter. The very first commercially successful spot meter, and one that is still highly regarded today, was the S.E.I. half-degree comparison photometer. It had only two indices, shadow and highlight: there was no 'mid-tone' index, because no exposure criterion is based on a mid-tone and no-one who really understands what they are doing will have many occasions to read a mid-tone index. Almost all subsequent meters have a mid-tone index, however, and many photographers understand so little about spot metering that they use it for the majority of their readings. If you used the mid-tone index to read highlights then your pictures would be grossly over-exposed (typically 2 1/3 stops), and if you used the shadow index they would be even more grossly exposed (typically 5 stops).

The most common highlight index today is the I.R.E. (Institute of Radio Engineers) scale, which spans 5 stops. I.R.E. 1 is the shadow index and I.R.E. 10 is the highlight index. If you are metering shadow detail for negatives you use I.R.E. 1, thereby guaranteeing adequate shadow detail, and if you are metering highlight detail for transparencies you use I.R.E. 10, thereby guaranteeing that the highlights won't 'blow'. You also know, from the I.R.E. scale, that any shadow detail that reads much below I.R.E. 1 will block up to a featureless black.

grey cards

The easiest thing here is to refer you to the free module on grey cards. At best, a grey card is a substitute for an incident light reading: the only real difference is that you are reading a mid-grey instead of an artificial highlight. Because a grey card is a flat receptor, and because it reflects 18 per cent of the light falling on it instead of 12 to 14 per cent (1/3 to 1/2 stop different) it should normally be positioned at an angle that splits the difference between the subject-camera axis and the subject-key light axis. Then you need to know whether your meter is calibrated to 12-14 per cent (non-spot meters) or 18 per cent (some spot meters, apparently). But this leads us to our final consideration.

personal exposure indices

If you compare two meters of the same make and model, and take readings of the same subject, you will often find variations of 1/3 stop and sometimes more; if they agree to within 1/6 stop, it is unusual. Compare two meters of different make or model, and variations are even more likely. Many meters, of course, cannot be read to greater accuracy than 1/3 stop anyway.

Now give the very same meter to two different photographers and ask them to meter the very same subject, without comparing notes. Once again, you are quite likely to find variations of 1/3 stop or more -- possibly quite a lot more.

And yet, most meters allow most photographers to get the exposures they want, most of the time. More remarkable still, they do so even when they are set to the nominal ISO speed. The more you think about this, the more puzzling it becomes: we are indebted to Garry Coward-Williams, editor of "Amateur Photographer" magazine, for this intriguing observation.

Sunset, Malta

Remember that deliberate exposure variation for a particular effect -- here, slight under exposure to capture the mood of the setting sun -- is not the same as re-rating a film; you do not change the meter setting or the development time. Slightly more exposure might have been more 'accurate' but it wouldn't have had the same emotional effect. Roger used am M-series Leica with 35/1.4 Summilux, shooting on Kodachrome 64.

If you are not getting the exposures you want, there are two possibilities.

One is that your exposures are consistently darker or lighter than you would like. If they are consistently darker, it means you are consistently under-exposing and you need to re-set the ISO speed on your meter a little lower: this will lead to your camera recommending more exposure. If they are consistently lighter, it means that you are over-exposing and you need to re-set the ISO speed on your meter a little higher: this will lead to your camera recommending less exposure. If you cannot easily re-set the ISO speed, use the exposure compensation dial or switch instead.

The other is that your exposures are not consistent. In this case, it should be no more than a matter of refining your metering technique.

the bottom line


As in so many modules, there is not much to understand -- though there may be a certain amount of deep-rooted misapprehension to be overcome, such as the belief that meters are calibrated to an 18 per cent grey (we used to believe it ourselves) and the belief that grey cards have some particular merit. The main thing to do is to think clearly and avoid the more overheated proclamations of 'experts'. If you understand what you are doing, you can see what advice is good, and what is bad; if you accept things on faith, you can be royally misled.

And yet again, the royal road to improvement is simply to take more pictures, armed with the knowledge above. We do not believe that you need to take notes of all your exposures, but if you feel more comfortable doing that, them by all means do so. That's all!


Graffiti and poster, Arles

In Arles, home of the Rencontres Photographiques, even the graffiti tend to have style. Exposure is simple: this is effectively flat copying so you don't have to worry about shadows, and a simple incident light reading is all you need. Roger used a Leica MP and 35/1.4 Summilux to capture this on Kodak EBX.

Go to the list of modules

or go to the home page

or support the site with a small donation.

© 2009 Roger W. Hicks