For beginners, it's easy. You focus on your principal point of interest (usually in the middle of the picture) and press the button. You may not even have to set the focus for yourself: autofocus cameras are probably more common than manual-focus nowadays, and of course there are still plenty of 'focus free' (=fixed-focus) cameras in the world, especially disposables.

As you learn more about photography, you start to realize that focus, like just about everything else, can be a creative tool. By controlling what is in focus, and what isn't, you can make quite a difference to the impact of your pictures.





Roger (who took this picture on 8x10 inch Ilford Ortho Plus) has mixed feelings about it: it's a bit too sterile, and the background (dirty white background paper) isn't really ideal. On the other hand, whatever appeal it does have is due in large part to its sharpness and the extreme rendition of texture. If it were not sharply focused, it would be a great deal weaker. To hold sharpness from front to back with an 8x10 inch camera, it is rarely sufficient just to stop down; you normally have to apply camera movements (tilting the front or back or both) as well, but this is a fit subject for another module. The contact print is on Centennial printing-out paper, gold-toned.

Frosty grass

Just as the picture above owes a good deal to its front-to-back sharpness, this one owes a good deal to its very shallow depth of field. If the background were sharper, the simplicity of the line of the frosty grass would be lost. Arguably, though, the depth of field here is too shallow. Roger used his 135/2.8 Elmarit-M at full aperture on the M8, where it is the equivalent of 190mm on full-frame 35mm. At the closest focusing distance of 1.6 metres (5 feet) as used here, depth of field is wafer-thin, just a few millimetres or a small fraction of an inch. Stopping down to f/8 would greatly have enhanced the detail on the grass and probably would not have made all that much difference to the background.

With an SLR you can get a good idea of depth of field via the preview lever, but with a rangefinder camera like the Leica it is more a matter of experience. You can forget about the depth of field scale at this sort of aperture and focused distance.



Acceptable focus

One of the first lessons about focus is that there can be a significant gap between what is acceptable and what is desirable. A picture can quite often 'carry' a greater or lesser departure from optimum focus, especially if the subject matter is particularly interesting, and if the picture is not over-enlarged.


frosty grass


Acceptable focus was how box cameras worked. With the exception of those taken by keen photographers, the great majority of family snapshots from before about 1960 (and plenty after that date) were shot with fixed-focus box cameras. The (very basic) lens on these was normally at its sharpest at about 3 metres/10 feet, though some were optimised for 5 or 6 metres (15-20 feet) with a sliding close-focus lens marked 'Portrait' or something similar for 2 metres or so (6 or 7 feet). Everything else was taken care of via depth of field, which we have always found easiest to understand via a simple diagram (below).


bar chart



Depth of field at different apertures

The imaging rays arrive parallel at the lens (yellow), or nearly so, and are focused on a point, the aptly named point of focus. If the film (or focusing screen, or digital sensor) is not actually at the point of focus, as represented by the two lines either side of the point of sharp focus, then a point is rendered as a blob. At a wide aperture (red lines) the blob is inevitably bigger than at a smaller aperture (blue lines).

The diagram makes it clear (a) why small lens apertures are sharper than big ones and (b) how a bigger enlargement creates less depth of field because the out-of-focus blob will look bigger when you enlarge it more. This drawing is a gross over-simplification, and illustrates depth of focus (permissible inaccuracy of the location of the focal plane) better than depth of field (range of distances at which subjects are acceptably sharp) but it gets the idea across nice and easily.

Typically, the lens on a 6x9cm (8-on-120, 2¼ x 3¼ inch) box camera operated at f/11 to f/14 and gave (barely) acceptable sharpness in a contact print or maybe a 2x enlargement. Go up more than about 2x and you got the familiar 'glassy' look of an over-enlarged, low-resolution picture with limited depth of field.


big parents





Frances's parents

Frances took this picture in 1956 with a fixed-focus 127 box camera. The focus is perfectly adequate -- if you already know who they are. You can just about see that Marion is holding a cat in the sectional blow-up (which Roger cleaned up in Adobe Photoshop), but this is mostly deduction based on the fact that few people cuddle a frozen rabbit this way. The best one can say is that a picture as unsharp as this is still a lot better than having no pictures of one's parents, who were in their 40s when this was taken, or indeed of one's childhood home. For those who do not recall 127, it was like a miniature version of 120, 46mm wide and giving 8-on at '645' (roughly 58 x 42mm), 12-on at '4x4' (40-43mm square) or more rarely 16-on at '3x4' (28-32mm x 40-43mm). The 'en-print' (enlarged print) from which this was taken is 77mm (3 inches) square, a 1.8x enlargement. This was the maximum it could stand...

Curiously, a poorly focused image from a better camera is often more acceptable. This is perhaps our first inkling of bokeh, the quality of the out-of-focus image, to which we shall return later. Equally, though, it is more than arguable that a picture taken with a better camera is likely to be taken by a more experienced photographer, and therefore to be better exposed. To quote Mike Gristwood, 'below a certain level of exposure, quality falls off a cliff'.




Cutter, all in

This image is just about acceptably sharp, but it is poorly focused; Roger was still getting used to the 135/2.8 Elmarit-M on the Leica M8. There's also a tiny bit of camera shake. It is, however, quite surprisingly three-dimensional, partly as a result of the lighting and partly because the grassy bank behind the cutter is well out of focus.

cutter crop 1

cutter crop 2

On the left, you can count the individual spicules of frost on the sharper blades of grass, but on the right you can see that the guides are not as sharply focused as they might be, and the frost is downright fuzzy.

Point of focus

The point of focus is literally the point (or perhaps more accurately, the plane) on which the camera is focused -- which is not necessarily the same as the focal point.

What you choose as your point of focus will depend on the subject. Normally, it is the principal subject, i.e., the thing in the picture to which you wish to draw attention. Within the principal subject, if some parts are in focus and others are not, the same rule applies: you will normally want the eyes to be the point of focus (as well as the focal point) in a portrait.


There is a convention in large format portraits that depth of field is very shallow and centred on the eyes.

We strongly suspect that this convention arose as a matter of necessity, in Hollywood in the great days of 8x10 inch portraiture. If you stop and think about it, a 'large head' portrait on 8x10 inch is very nearly a macro photograph (1:1) and the effective aperture is going to be pretty slow.

With the 21 inch (533mm) f/7.7 Ross that Roger used here on our De Vere 8x10 inch monorail the effective aperture at full bore was probably around f/16 and even with ISO 400 film (Ilford HP5 Plus) this translates to a long-ish shutter speed unless you have very powerful lights. All right, in Hollywood in the 1930s they did have very powerful lights, but the fastest films available in the mid-to-late 30s were about ISO 200 and many were slower, especially in the early 30s.

This is why so many Hollywood-style portraits show the stars in very relaxed poses. It is not just so the subject wouldn't move between the time the photographer focused on the ground-glass and the time he took the picture. It was also to prevent subject movement during the time the shutter was open, typically 1/10 to 1/2 second.





vince portrait



Even in his teens -- he must have been about 16 or 17 when this was taken -- Vincent Donlin was a brilliant artist, so Roger chose to focus (literally) on the brush rather than on his subject's face.

Technically and indeed compositionally the picture leaves a certain amount to be desired -- Roger's excuse is that he took it in about 1974 -- but it's a good illustration of how the eyes are not necessarily the optimum point of focus in a portrait.

It was probably shot on grievously outdated (1963 expiration date) ex-government Ilford FP3 using a Pentax SV and 55/1.8 Super Takumar at full bore. In fact, nothing in the picture is really sharp, though the tip of the brush (below) comes close.

vince brush

Depth of field

Depth of field is a measure of the depth either side of the focused point that is acceptably sharp. As we all know (and as the diagram towards the beginning of the module illustrates) it depends very much on aperture, but this is far from the sole variable. In fact, a staggering amount of nonsense has been written about depth of field. The biggest single piece of misinformation is that it is in some way constant, for a given aperture. Of course, it isn't. It depends on how big the enlargement is, and how far away you are when you look at that enlargement. It also depends on what is 'acceptable' in a given image, for a particular kind of subject matter, to a particular photographer.

The degree of enlargement also depends on the size of the image on the film (or sensor): obviously, a small image has to be enlarged more than a bigger one to get a given size of print. Perhaps counter-intuitively, though, you will see more depth of field in a small image that is enlarged more, and less in a big image that is enlarged less. To minimize depth of field, you want the largest possible format and the widest possible aperture. To maximize depth of field, you want the smallest possible format and the smallest possible aperture. The sharpness of the subject is also important.


cutter 1



cutter 2


In the left-hand picture, the point of focus is the cutter itself; the lamp, post and wire behind are left to go soft. In the right-hand picture, the point of focus is the lamp post. At this size, you have to look fairly hard to see the difference: the relative sharpness of the grassy bank in front of the cutter is probably the clearest indicator. But this demonstrates that apparent sharpness and depth of field are heavily dependent on image size; that more unsharpness is tolerable in a subject that is already hazy or fuzzy; and that normally, it looks more natural or convincing to have the foreground in focus, and let the background go soft. Technical information as for the earlier shot of the cutter.


windmill, leaves


Backgrounds and foregrounds

Although (as noted above) it is normal to have the foreground sharp and to let the background go out of focus, there are plenty of exceptions. Almost all pictures that are 'framed' (by a door, a window, foliage or some other foreground feature) will look better if the foreground is sacrificed out of necessity, rather than the background. Arguably a 'deep field' composition works better in many cases, but there are also times when an excess of sharpness in the foreground might distract attention from the principal subject -- which is, almost always, what you want to be sharpest.


Windmill, Mazeuil

Frances used our then-new Summicron 75/2 for this picture, though we have forgotten which camera body it was on. Obviously it would have been impossible to stop down far enough to keep both the foreground and the background in focus, but in any case she wanted to create the impression of moving through woods, and seeing the windmill in the clearing. You would naturally concentrate on the mill and only half-notice the foliage. Film was Kodak Tri-X and the print is on Ilford Multigrade Warmtone.

Image size and depth of field

As soon as you think about it, you can see very easily why the smaller formats have more depth of field. A frame filling image on a bigger format is necessarily bigger on the film (or sensor) than a frame filling image on a smaller format.

An inevitable result of this is that selective focus can be difficult or impossible if you use a very small format (especially the smaller digital formats) or slow lenses (especially slow zooms on digital cameras). At the other extreme, with a long enough lens on a big enough format, especially close up, depth of field can be tiny even at f/8 -- or f/7.7, as in the portrait of Holly, above.

Focal length and depth of field

'Everyone knows' that wide-angle lenses have more depth of field -- and everyone is pretty much right. Strictly and pedantically, depth of field depends on the degree of magnification on the film or sensor. An image that is recorded at (say) one-tenth life size, at a given aperture of (say) f5.6, will have the same depth of field regardless of the focal length of the lens.

In the real world, however, wide-angles are normally used to 'get more in', rather than to allow the same picture to be taken from closer to the subject. As a result, they do indeed have more depth of field. Likewise, longer lenses are used to 'pull things in', rather than to allow the same picture to be taken from further away.

Well, bucket, frost

Yet another shot taken (by Roger) with the 135/2.8 Elmarit-M on the Leica M8: we had just got the lens when we wrote this module and had only had the camera for a few weeks. The perspective on the well-head is slightly compressed, but perhaps more importantly, it stands out all the more clearly against the out-of-focus background, even at f/5.6 or so. The apparent 'sharpening' which is most apparent on the winding handle is actually nothing but a frosty edge. Contrast was somewhat increased in Adobe Photoshop but that is the only real change to the DNG (RAW) file.


well close-up


Perspective, viewpoint and depth of field

Another truism peddled by the pedants is that perspective depends solely on viewpoint. Well, yes, sort of -- provided the pictures are enlarged to the appropriate degree and viewed from the appropriate distance (which may not always be possible). There is a whole paid module on perspective and the illusion of depth, and once again, in the real world, both perspective and depth of field are heavily influenced by choice of focal length.

Using depth of field scales

Depth of field scales on lenses used to be taken for granted, but they should always be treated with some reserve until you are familiar with a particular lens. There are, after all, no absolute standards for depth of field: the manufacturer must make certain assumptions about enlargement size and what is acceptable.

Reputedly -- we have never verified this scientifically, but it certainly looks as if it was the case -- the d-o-f scales on 1930s Contax rangefinder cameras were based on considerably less stringent criteria than those on Leicas of the same period. The Contax, it is said, assumed postcard-sized enlargements (3½ x 5 inches, 89 x 127mm), while Leicas assumed whole-plate enlargements (6½ x 8½ inches or 165 x 216mm). Contax lenses therefore had more generous d-o-f scales, though of course, the d-o-f was identical in both cases.


roger, film box



As a rule of thumb, if we want a 'deep field' composition with a rangefinder camera, we normally use the next stop down. In other words, if we are shooting at f/8, we'll use the d-o-f scale for f/5.6, or if we're shooting at f/11, we'll use the d-o-f scales for f/8.


Depth of field previews

Another way to handle depth of field is to stop the lens down to the working aperture. This is a lot more reliable than using scales, but even then, it is easy to be misled or to fool yourself: it's a guide, not a true preview. If you have time and facilities, a very useful trick is to use bar-codes as a means of checking focus.


Roger and film box

Bar-codes provide a wonderful high-contrast focusing target, whether you're going for minimum depth of field (as Frances was here with our 8x10 in De Vere and 21 inch f/7.7 Ross on Ilford HP5 Plus) or 'deep field' in a still life. Just remember to remove the bar-codes before you shoot the final picture...

Depth of field and digital sensors

The 'normal' rules of depth of field sometimes seem not to apply to digitally captured photographs. The only convincing argument we have heard for this was advanced by Geoffrey Crawley. He reckons that a lot depends on the edge sharpening algorithms used in processing the captured image. In some circumstances, he suggests, an edge that would appear soft in an unprocessed picture is rendered sufficiently sharp that the depth of field appears greater -- and given that most digital sensors are smaller than 35mm film anyway, this means still more depth of field. With 'raw' files and cameras that do not have aggressive sharpening algorithms, this effect is less obvious and may not exist at all.


atelier doll


Doll, Atelier du Buissonier

The half-walnut-shell gives a pretty good idea of the scale of this picture; the doll's head is a poppy seed pod. The lens was a 65/3.5 Elmar on a Visoflex, used on a Leica M8 where its equivalent focal length is near enough 85mm. The lens is stopped well down, probably to f/16, but the depth of field still seems slightly greater than one would expect.

Pinhole focus

An intriguing thing about pinhole shots is that depth of field is infinite. Everything is equally sharp -- and, of course, equally unsharp. As far as we are concerned, with pinhole photography, the bigger the original format, the better. We have seen (though never made) seriously large pinhole pictures that are quite remarkably sharp, while at the other extreme, enlarged 35mm pinhole shots normally seem to us to be far too soft. Generally, we prefer 4x5 inch or 5x7 inch for pinhole pictures: big enough to be reasonably sharp, but small enough that the unsharpness is clearly deliberate.


sepia lamp shop


Lamp shop, Margate

Polaroid Sepia lent itself well to pinhole photography; Roger shot this with a Rigby pinhole camera. It well illustrates the difference between soft focus and out-of-focus; there is no plane of focus here, and as noted above, everything is equally sharp (or unsharp).

Stopping down too far

A pinhole is of course the limiting case of a small aperture; if you stop any lens down far enough, you get, in effect, a pinhole. Pinholes 'focus' the light by diffraction, and diffraction is what sets a limit to the resolution of any lens at any aperture. A very good rule of thumb is to divide the aperture in use into 1500. This will give you an approximate limit, in line pairs per millimetre, to the resolution attainable. Thus, for example, at f/16 you cannot exceed 1500/16 = 94 lp/mm. This also explains why 35mm lenses rarely stop down as low as f/22 (68 lp/mm) and almost never as far as f/32 (49 lp/mm). It is also worth remembering that if you are shooting close up, a marked f/16 is effectively around f/22 at half life size and f/32 at life size.

With bigger formats, that are enlarged less, smaller apertures are less of a problem. This is why lenses for 6x9cm commonly stop down to f/22, sometimes f/32 and rarely f/45. If the image is not enlarged at all (a contact print), then 20 lp/mm is as sharp as you can normally see and this is attainable at f/75; f/90 still allows 17 lp/mm and even f/128 will still look pretty sharp at around 12 lp/mm.

Compromise focus

Very often, you gave to compromise on focus, especially if you would ideally like a 'deep field' composition with everything in focus from front to back. It may be impossible to stop down far enough to get a sharp picture (see 'Diffraction limits', above); it may be impracticable to stop down as far as you would like, because doing so would entail an insupportably long exposure time; and even if you have a large format camera with movements, the Scheimpflug rule (see Glossary) allows you only to hold a receding plane in focus. If the subject departs too far from a single plane, then even at minimum aperture, you may not be able to hold everything.

Likewise, if you want to use differential focus, you may be limited because your lens isn't fast enough, even at full aperture, to create a very narrow depth of field.

There are two possibilities in such situations. One is to come as close as you can to your intended use of focus, whether by working at the minimum practicable aperture or maximum aperture, or by the use of movements -- and remember that movements can be used to restrict focus as well as to add apparent depth of field. The other is to rethink your composition, asking yourself whether there is not some other way to use the focus (or indeed arrange the composition).

Rodchenko's Toast

Both the angle and the focus were essential for the composition, but the only way to hold focus on both the glass and the vodka bottle was by the use of 'reverse movements', creating a plane of focus that passed through the glass and the lower part of the bottle, but not the upper part. Roger used a 4x5 inch camera (as far as we recall, our Linhof Technikardan) to shoot this on Ilford FP4 film. When scanned as 4-colour, interesting colours came in. Lens was probably the 210/5.6 Schneider Symmar.



r's toast


atelier dec 6


Atelier du Buissonier

If you can't hold everything in focus, simply because you can't stop down far enough (or, if you are using a large format camera, because the movements don't allow you to focus on a receding plane) you have to decide whether or not you can let things go out of focus; what you can let go out of focus; and how far you can let things go out of focus. An out-of-focus teazle is often curiously distressing, simply because it is so sharp in real life; but because the teazle and the hairy string are so sharp, the loss of focus on the table, the tools and even parts of the wire are acceptable. It's also worth remembering that if part of something is in focus (the wire, the hairy string) the eye latches on to this and remembers more sharpness than is really there, e.g. in the other teazle stem. Roger used the Leica M8 and 65/3.5 Elmar (on a Visoflex).


Bokeh or boke, pronounced 'boh kay', appeared quite suddenly in the English language in the 1990s. Apparently the original meaning in Japanese is something like 'soft' or 'fuzzy', but it was eagerly adopted (especially in the United States) to mean 'the quality of the out of focus image'.

This was hardly a new concept: the original Voigtländer Apo-Lanthar, introduced around 1950, was often praised for the quality of its out-of-focus image; we just didn't call it bokeh in those days.


mertola down


Bokeh seems to be something to which some photographers are more sensitive than others. There are those who get very excited about 'good bokeh' and 'bad bokeh', the latter often being described as 'wiry'. Well, yes, we can see bad bokeh -- the legendary Leica Thambar has awful bokeh when used with its centre stop in place -- but we do not recall ever looking at a picture and thinking, 'What excellent bokeh!'

Also, bokeh is very much a matter of composition, subject matter and focus. The nasty bokeh of the Thambar is completely concealed unless there is something immediately behind the main subject that is of the right (or possibly wrong) shape to show up the shortcomings. One example, it must be said, is foliage (especially grass) immediately behind a flower study. If you know that you are using a lens with this sort of bokeh, you need to be a bit more careful with composition -- or of course switch lenses!


Mertola, Portugal

Roger shot this with his 35/1.4 pre-aspheric Summilux, probably on his M4-P, almost certainly on Kodak Elite Chrome 100 Extra Color EBX. The very closest parts of the picture, and the most distant, are not in particularly sharp focus; but lichen never looks very sharp anyway, and the distant part of the picture is hazy anyway; and the main interest is in those areas that are in focus, so how important is bokeh?

Perhaps one clue to bokeh is that it was most enthusiastically adopted by American large-format photographers who shoot a lot of pseudo-wilderness, often with a fairly strong foreground and a lot more landscape (including foliage) in the background. It may well be that under these circumstances, you become especially sensitized to bokeh -- and as it's a style of photography we almost never do, we've never become sensitized to it.

Lens design and depth of field

Some lenses appear to have more depth of field than others, and indeed a few lenses have been marketed on this premise: the 'Deep Field Panchro' was an example. Our own suspicion is that there are two factors at work: field curvature and (still more importantly) low sharpness. For fairly obvious reasons, if the sharpest parts of a picture aren't very sharp, then they contrast less with the parts that even less sharp. In fact, with portraiture, we have sometimes had better results with truly awful lenses than with purpose-designed soft focus lenses.


Nicholas Andrew Gill Hayes

Roger took this when Miklos (who is half-Hungarian) was 15 or 16; he is a year younger than Roger. The lens was a truly awful 90-190mm f/5.8 Yashica zoom from the early-to-mid 1960s, mounted on a Pentax SV. Film would have been outdated (1963, used in about 1967) Ilford FP3. Using the Yashinon wide open at or near the longer limit of its focal length has given a very shallow depth of field.

Roger sold the lens in the late 1970s because it was so awful, but by the late '80s he had grown to regret the decision because it really was a good portrait lens.


hayes 1967


The Bottom Line: Shoot First

...and ask questions afterwards. This is the usual rule in photography unless you have plenty of time -- as, for example, with a still life. But if you've thought about depth of field and differential focus before, and experimented with shots where there is no time pressure, you will be much better placed to get the effects you want, when you want them.


louise, mirror


Louise in the mirror

We don't do weddings -- except for friends and (nowadays, under even greater pressure) for the children of our oldest and closest friends: Louise is Nick Hayes's daughter. As part of our coverage of Louise and Tony's wedding, Frances shot a 'bride dressing' sequence that was indeed a sort of reverse strip-tease.

She used a mixture of Ilford Delta 3200 (for maximum speed) and Ilford XP2 (because it's one of her favourite films) and shot mostly with her Voigtlander Bessa-T and (as here) Bessa-R2; the lens she used most was (again as here) the 50/1.5 Voigtlander Nokton.

Of course, at full aperture at this distance, the frame of the mirror and Louise's hand are both completely out of focus. So? They still provide plenty of context, and without them, there would not be much of a picture.

The question is this: is it a deliberate, planned use of shallow depth of field, or a compromise forced upon Frances by the circumstances? The answer, of course, is "Both." She thought it would work -- and it did. But if it hadn't, she would never have included it among the pictures.

Go to the list of modules

or go to the home page

or support the site with a small donation.

© 2007 Roger W. Hicks