The Death of the SLR

A completely new camera format is quietly taking over the world.  You heard it here first.

Film camera designers have always been faced by a big problem.  When you take a picture you need light to fall onto a light-sensitive film so that an image can be captured.  But if the light is falling on a film, it can’t simultaneously be falling on the eye of the photographer.  So how does the photographer see what he is photographing?

One solution is to provide an offset viewfinder, positioned as close to the lens as possible.  This was the route chosen most famously by Leica, whose legendary “M” range of cameras are often called rangefinders.  Another design uses two lenses positioned alongside one another, or (more commonly) one above the other, so that the photographer can look through one and take pictures through the second.  The best known cameras of this kind are the magnificent twin lens reflex (TLR) cameras built by Rollei.  But both these solutions can offer only an approximation of the image actually captured on the film.  In particular, both introduce parallax error, meaning that the photographer sees his subject from a slightly different angle than his lens does.  

A single lens reflex (SLR) camera, by contrast, uses a mirror and prism system that allows photographers to see exactly the image that will be captured by their camera.  The light enters the camera through the (single) lens, is reflected upwards through 90° by a mirror, is then refracted back through 90° by a prism, and is finally sent out through the viewfinder to the photographer’s eye.  When you press the shutter button to take a picture the mirror flips up out of the way just before the shutter opens so that the light can fall on the film or sensor.  

The SLR has been the dominant 35mm film camera format for professionals and enthusiasts for about half a century.  Just ask any camera magazine or website:

“The 35mm single lens reflex (SLR) is most folks' idea of a serious camera  . . . Digital single-lens reflex (SLR) cameras are the standard tool for serious photographers” (www.photo.net)

“for serious photography the SLR wins hands down” (www.dpreview.com)

“for serious and professional photographers, digital single lens reflex (DSLR) cameras are the way to go” (www.macworld.com)

So there you go: SLRs win because they are seriously serious.  But seriously, what features make SLRs so seriously suitable for serious photographers?

There are indeed some advantages associated with the SLR’s “through-the-lens” (TTL) viewing.  The ones I can think of are these:

1. Telephoto lenses.  Using a longer lens allows you to take pictures of things a long way away, but it is impossible for a photographer to focus, frame and compose through a long lens unless he can see the faraway object as well as the lens can.    The image in an offset viewfinder is always the same, since it doesn’t change when the lens changes.  So offset viewfinder cameras can’t easily be used for faraway subjects.  TLR cameras tend not to have interchangeable lenses at all, since each new “lens” would in fact have to consist of two lenses, making them very expensive.
2. Macro photography.  SLRs allow extreme close-up or “macro” photography.  Due to parallax error, if you put the lens of an offset viewfinder or twin lens camera just a few millimetres from your subject you are never going to be able to compose and focus the shot accurately.  To do so you have to be able to see what the lens sees.
3. Filters.  If you put a filter on your SLR’s lens, you will see its effect when you look through the viewfinder.  In cameras where you don’t look through the picture-taking lens this is not the case.
4. Depth of field preview.  By changing aperture (i.e., the size of the hole in the diaphragm through which light enters the camera) photographers can vary the “depth of field”, i.e., the range of distances away from the camera within which subjects are acceptably sharp.  A smaller aperture allows a greater depth of field.   Normally when you look through the viewfinder of an SLR the aperture is set to its maximum to let in the maximum amount of light and thus give the photographer the brightest view of the scene in front of him, and so allow him to compose, frame and focus as accurately as possible.  However, many SLRs allow the viewer to preview depth of field by adjusting the diaphragm to the smaller size at which the picture will actually be taken before pressing the shutter button.  Cameras with viewfinders lack this feature.

That’s it.  Plenty of people say there are other advantages of SLR cameras: some of these are wrong, some are arguable, and some are right in a sense but in fact have nothing to do with the fact that the cameras use the SLR design.  I discuss some of them below.  First, however, we should think about the disadvantages of the SLR system.

1. Blackout at the crucial moment.  As explained above, before you press an SLR’s shutter button you can see the subject you are going to photograph exactly as your camera sees it, and even preview the depth of field you will use.  But you can’t see the subject at the moment you actually photograph it.  This is because when that mirror flips up out of the way in order to let the light pass through the shutter to the film, the viewfinder (obviously) goes blank.  So in a way the SLR doesn’t even solve the original problem of letting the photographer see what he is photographing.  Viewfinder and twin-lens cameras don’t have this problem.  Of course, the mirror flips up just a fraction of a second before the shutter opens.  But a fraction of a second can be crucial to a photograph.
2. Shutter lag.  As explained above, before you press the shutter button the mirror is in front of the shutter and the aperture is at its maximum setting.  When you pull the trigger the mirror has to get out of the way and the diaphragm has to be shrunk to the photographer’s preferred size before the shutter can be opened.  Again, all this happens in a fraction of a second.  And again, a fraction of a second can be everything.
3. Vibration.  When that mirror folds up into the camera it has to be set in motion and then has to be stopped before the shutter opens.  The resulting “mirror slap” causes the camera to shake, and can blur the image, especially when the camera is being used at longer exposures or with longer lenses.
4. Lens design.  For optimal performance some lenses, and particularly wide angle lenses, need to be positioned very close to the surface of the film.  This is impossible where a mirror has to be interposed between lens and shutter.  
5. Mechanical complexity. All cameras suffer from this to some extent, but an SLR with its mirror, diaphragm and prism is a particularly complex instrument with lots of moving parts that can go wrong (or just out of alignment).
6. Size and weight.  An SLR has to have enough room for the mirror and prism arrangement and is therefore always a sizeable and weighty box.  Not only does this make it cumbersome to use, it puts photographers off carrying it, which means they miss pictures.  This is bad.

In other words, the SLR has numerous major drawbacks.  What is more, all the putative advantages of the SLR design can be matched using other formats.

1. Macro / parallax. If it is absolutely essential to have the photographed image exactly the same as that seen through the viewing lens, gadgets exist that can move a twin lens camera so that the taking lens goes to the exact position that the viewing lens occupied when the picture was composed and framed.
2. Telephoto. Contax G cameras have an offset viewfinder that shows a different image when the lens changes, meaning the image is magnified to give a view equivalent to that through the attached lens.  And as early as 1935 – before the first SLR cameras were commercially available – Leica offered an SLR housing that fits onto their M series rangefinders to allow TTL viewing and therefore the use of telephoto lenses.  Telephoto lenses have also been offered for TLRs, though rarely with satisfactory results.
3. Filters. Two-part filters are available, allowing both the photographing lens and either the offset viewfinder or second lens to be covered.  This allows the photographer to see the impact of the filter before shooting.  
4. Depth of field preview. TLR cameras exist with this feature.

So how did the SLR take over the market?  Simple: the real reason for the success and near-ubiquity of SLRs is neither the overall technical superiority of the format nor any particular feature unique to of the SLR design.  The reason is the application to the SLR design of the effort, ingenuity, vision and above all marketing acumen of Japanese cameramakers, and particularly Nikon.  

SLR wasn’t new when the Japanese took to it.  The first patent for an SLR camera design was filed in England in 1861.  The first 35mm SLR prototype came from the USSR in 1934 and was called the “Спорт” (“Sport”).  Germany’s Exakta was first to produce a 35mm SLR commercially (from 1936), and their compatriots at Zeiss introduced the Contax S in 1949 using a design that was essentially the foundation for everything that followed.   The first Japanese 35mm SLR came from Pentax (going into production in 1952), and this opened the door to other products from firms including Miranda, Periflex, Minolta and Canon, which generally met with modest success.  If you wanted a serious camera you bought a Leica rangefinder; if you absolutely had to have an SLR you bought a Contax; and if you wanted an inexpensive alternative you bought a Japanese SLR.  That was it.

What changed everything was the launch of the now-legendary Nikon F in 1959.  It was affordable, very tough & practical, made pretty good photos and had a very bright, clear viewfinder.  However, what set it apart from the Pentaxes, Yashicas, Canons, and every other comparable camera in the world was that it was a system camera, meaning that from the day it was released Nikon offered a huge range of accessories including lenses of all lengths, flash units and motor drives, making it appealing to professionals (who, for example, used it widely in Vietnam) and amateur enthusiasts (who typically love taking pictures much less than they love buying new camera equipment).  So successful was this camera that Nikon based all their subsequent 35mm SLRs on the same design, right up to the last of their top-end 35mm SLRs, 2004’s F6.  This meant that many accessories could be used on almost any Nikon SLR, and in particular that any Nikon F camera could take any Nikon lens from the late 1950s onward.

This feature – “SLRs are system cameras” – is often touted as a reason why serious photographers should prefer SLRs.  And it is certainly true that there is an astonishing array of bits and pieces available for the SLR owner, far more than for any other format, so that if you really care about all that flexibility and equipment you more or less have to have one.  The best SLR cameras have phenomenal autofocus, metering and flash technology.  Their motordrives mean you can take an amazing number of photos per second (which can be important for nature or sports photography, for example).  But none of these “advantages of SLRs” have anything to do with the SLR design per se.  They are the results of years of focused product development by highly skilled engineers at Japanese camera companies who happen to have decided that the SLR system was the one they were going to devote their time and energy to.  

Of course, a lot of the foregoing is unfair.  Just as other manufacturers have engineered around the disadvantages of their chosen system, SLR manufacturers have found inventive ways to address many of the shortcomings of their own.  The best SLRs have minimal shutter lag (though still not as good as the best examples of other designs: for example, Nikon’s top-of-the-range F6 35mm film camera has shutter lag of 0.037s, while the figure for Leica’s M7 rangefinder is 0.025s).  It is also true that there are some phenomenal lenses, including wide angle lenses, for SLRs – though due to the way these have to be designed and built they tend to be extremely large and heavy.  The damping of mirror slap is excellent in modern cameras, and in case this is not enough (for very long exposures, say) since the Nikon F there has always been an option to lock the mirror up out of the way before the shutter is opened (though this of course gives rise to an extreme version of the “blackout at the crucial moment” problem).  And though they remain mechanically complex, Japanese SLRs are built to very high standards and are invariably extremely reliable. But two key points remain: firstly, that in principle the SLR system has no real technical advantage over any other film camera format; and secondly, that SLR users have to put up with a lot of compromises to get around the essential flaws of the basic design, most importantly the immense size and bulk of SLR cameras and their accessories. 

When digital came along not much changed.  The big Japanese manufacturers continued making their 35mm SLRs but also started making digital SLRs (DSLRs).  These were extremely expensive and offered very ordinary image quality, but were immediately adopted by professional news photographers who cared less about the fine detail of the pictures than about the convenience and speed digital offered, and the fact that they were able to continue using all their old lenses, flashes and other paraphernalia. As the technology improved so did quality, even while prices fell to make DSLRs affordable to amateurs.  So enthusiasts everywhere bought a DSLR and carried on shooting with their old accessories.

This made absolutely no sense.  All digital cameras offer the photographer an image exactly like that reaching the sensor but displayed on an LCD screen, making a viewfinder otiose.  And indeed, the majority of pocket cameras (disparagingly referred to by the “serious” crowd as “point and shoot” or PAS models) have no viewfinder at all.  To begin with, these LCD screens were pretty dull and slow, which may explain why the photographing public didn’t realize their full meaning. But now LCD technology has now developed to a point where the picture can be bright and accurate, and updates extremely rapidly.  And so there is no excuse for not realizing that all alternative methods of viewing the photographic subject– offset viewfinder, TLR, SLR, whatever – are completely obsolete.  

Here’s why.  A digital camera’s LCD display shows you exactly what the camera sees right up to and including the moment when you push the shutter button.  (Remember that no film system can do this because it is simply not possible to have the light fall on a piece of film and the photographer’s eye at exactly the same moment.)  There is therefore no parallax error, but also no restriction on the use of telephoto or macro lenses, or filters, and depth of field preview is easy.  Since there need be no silly mirror flapping around inside the box, and no prism the size of a satsuma sitting above the lens, there is no blackout at the crucial moment, there is no vibration from “mirror slap”, lenses can be located as close to the sensor as the designer likes, there is no risk of mechanical failure or misalignment and – best of all – the box (even with a high quality lens attached) can be small enough to fit in your trouser pocket.  What is more, while many early digital pocket cameras had appalling shutter lag (this above all is what put me off buying one for a long time), the technology has improved and now mirrorless digital cameras have better shutter lag times than any film camera. Sony’s digital compact NEX-5N has a shorter shutter lag – 0.22s – than even the legendarily quick Leica M7 35mm rangefinder, and Sony claim that the newly announced NEX-7 will be faster still, at an amazing 0.20s.

But still camera-makers remained amazingly ignorant of the consequences of their invention.  As the technology improved, a couple of manufacturers started making pocket cameras with good lenses, high-performance sensors and offering the photographer lots of control over the camera’s operation.  They suggested that professional photographers might pick up one of these for personal photography, or perhaps drop one into their camera bag for taking preliminary pictures before getting down to the serious stuff with their high-end DSLR.  They appear also to have thought that this format would serve as a “bridge camera”, offering the aspiring photographer a “step up” from a compact on his inevitable journey to a “serious” SLR.  But, to the sound of palms slapping foreheads and jaws hitting laptop keyboards from one end of Japan to the other, the pros started using these little cameras all the time and leaving their huge bags of DSLR equipment at home.  The pros loved the little digital cameras, because they were freed from the tyranny of their immense heavy kitbags to go and do what they really want to: take pictures.

Since then things have got worse still for the mainstream SLR makers.  Other companies not so deeply invested in SLR technology, and with more of a focus on electronics than on mechanical cameras, have rapidly improved their product offering, including (for example) high quality interchangeable lenses, and very accurate and complex metering, focusing and flash systems.  Most importantly of all they have started building extremely high quality and large sensors into small cameras that photographers can take wherever they go (even a full kit with multiple lenses will fit in a small bag).  And for all the reasons laid out above, there is no reason why all the remaining amazing features of existing DSLRs cannot be replicated in the new format.  Sony are clearly determined to prove exactly that with their new NEX-7, which will have a 24 megapixel sensor (exactly the same as Nikon’s top-of-the-range D3s), plus 10 frames-per-second “burst” shooting, customisable manual controls, the shortest shutter lag of any comparable camera and numerous other features clearly aimed at the “serious” market.  “Point and shoot” it ain’t.  Panasonic and Samsung are not far behind.  The big DSLR makers Nikon and Canon have yet to respond. 

So why would anyone stick with the clearly inferior, indeed obviously obsolete SLR technology?  One answer is that photographers have a lot of existing kit to go with their DSLRs and won’t give up their existing platform easily.  But that just means it is a matter of time.  No new photographer with any sense is going to buy a big DSLR and all the associated paraphernalia when he doesn’t need to.  A more powerful consideration is the extreme conservatism of the camera-buying public and the media that serve it.  This is reflected in the ridiculous insistence of press and photographers alike that the new generation of cameras should be called “hybrids”, as if they were a mix of SLR and compact camera technology, or, sillier still, “mirrorless SLRs”, as if these new cameras were just a variant on or refinement of the good old SLR format.  They are not.  They have nothing in common with SLRs except the high image quality, manual controllability and “system camera” status, none of which – as argued at length above – has anything to do with the SLR format per se.  

But who cares?  Call them what you like – hybrid, mirrorless, interchangeable lens compact, high-end PAS, whatever.  These new digital cameras have the potential to solve all the design problems inherited from film cameras, and put small, powerful cameras into the hands of experts and serious amateurs alike.  This will be the biggest change in serious mainstream camera design in a generation.  In other words, these new, high-quality, user-friendly, highly flexible, and above all small digital system cameras will consign the flappy-mirrored, oversized, ugly SLR to photographic history once and for all.  

And I for one will not miss it one little bit.

Cars are Stupid

Come on, face it. Cars are stupid.

Take Italian cars: too stupid to go where you point them, and unable to move without bits dropping off them. At the other end of the spectrum, German cars are over-engineered to extraordinary heights of idiocy: the new Mercedes S-Class, for example, has a control which allows you to decide how long the light over the rear numberplate stays on after you switch off the engine. French cars are perfectly designed for their intended purpose, which is to provide accommodation for chickens. Japanese cars have names like “Cherry”, “Sunny” and “Cedric”, so they are already silly even before the Asian cabbie gets his hands on them and adds the fluffy dice and gilt-edged tissue box cover in the rear window. Stupidest of all, of course, are American cars, the most representative of which are the lumbering monsters known as “sports utility vehicles”. Slumberdown produce sportier models than these: and the utility of these vehicles is restricted to transporting people the size and shape of sofas, in a straight line, at 50mph and 8mpg, forever.

Exclusivity is, of course, no defence against imbecility. Anyone who buys a 200mph Porsche to drive it on the road at a maximum of 70mph is clearly a boob; anyone who takes such a vehicle on a race track is guaranteed to do thousands of pounds worth of damage to his brakes and suspension and is therefore a boob on stilts. He is however still more rational than the owner of the Ferrari 355, which needs its cam belt changing every 10,000 miles, necessitating the complete removal of the engine from the body every 12 months or so. Supercar? Superstupid more like.

As exclusivity, so ubiquity: the unavoidable Ford Mondeo and Vauxhall Vectra, for example, are stupid because there are so stupidly many of them, with so many stupid options; 1.6, 1.8. 2.0, 2.0i, 2.5, 3.0, V6, V8, turbo, diesel, turbo diesel, a thousand infinitesimal steps up the ladder of luxury to match a thousand fiercely defended meaningless niches on your company’s car policy. The manufacturers have extended this idea with the notion of “common platforms” according to which any number of models are built using all the same parts except the wrapper. General Motors builds Saabs, Vauxhalls, Opels, Holdens and so on and so on around the same bendy chassis.

So, cars are stupid, though in jaw-droppingly many and various ways. But why is this? Whence this rampant nit-wittery?  The answer is simple: people. People are very, very stupid, which is why modern cars are – it must also be admitted – so brilliant. People thus demand features on their cars – like cruise control, automatic gearboxes – which allow even twits to operate them. People being not only stupid but also frit, it goes without saying that total accident protection is a sine qua non; and because, should any problems arise, they would no more dream of opening the bonnet themselves than they would of treating a sore throat by removing their own tonsils, iron-clad reliability should be included in the package. And being above all greedy and mean, they don’t see any reason why car manufacturers should not provide all these things at affordable prices.

The manufacturers, arguably the greatest nincompoops of the lot, are too stupid to think of an argument to the contrary. Which is why most of them are in such dire financial straits.  And I don't just mean the union-ruined basket-case GM.  Even the normally prudent Germans are not immune: BMW’s exit from Rover cost them hundreds of millions, and the takeover of the shambolic Chrysler by the normally clinically efficient Daimler-Mercedes was an unmitigated Katastrophe. And in once-mighty Japan, the once-very-mighty-indeed Nissan made such a mess of its business that it was forced into the ultimate humiliation of being taken over by the French, in the form of Renault.

To summarise: cars are simultaneously stupid and brilliant, largely because people are driven in their actions by a mix of idiocy, ignorance, laziness, fear and greed. In all the world, only a tiny handful of people remain unconvinced of this self-evident truth. They are called Marxists. I once found myself at Speakers’ Corner listening to one. For minutes on end, and without a murmur, I endured his fatuous historical analysis, non-existent sociology, feeble-minded political claptrap, ludicrous economics and pie-in-the-sky hopes of imminent social upheaval. But when he got onto cars I could restrain myself no longer. “How can it make sense,” he asked, “to build more than one kind of car? We don’t need more than one kind!” Able to contain myself no longer, I inquired with feeling – what about big families? Don’t they need big cars? “Well, OK, maybe a couple of . . .” And what about courting couples? Don’t they need something with a capacious and comfortable rear seat? How are we to motivate the travelling salesmen in his Mondeo 1.8L, if not by means of the Mondeo 2.0i V6 Ghia? Unthinkable the estate agent without her dinky little Mini: absurd the African minicab driver absent his Korean charabanc. And comrade, surely you are forgetting your old friends the lumpenproletariat: what of the builders, locksmiths, plumbers, glaziers in their white Transits? And with an eye to your own advancement, tovarish: however would the senior party apparatchiks achieve clarity of historical consciousness without their Zils?

I hardly need to point out the irony in this particular case, which is of course that the most stupid cars of them all were built by Marxists. It is tempting to think of Marx’s notorious “scrap-heap of history” consisting entirely of abandoned Skodas, Yugos and Ladas. The Trabant was slightly less absurd because it was at least put together by Germans. Nonetheless, it made a noise like a model tractor, and was rumoured to have been built out of lacquered cardboard. Stupider still, there was a waiting list longer than one of Fidel Castro’s speeches, because communist working practices and methods of distribution meant that it took ten years to get your hands on one. The fall of the Berlin wall produced no more poignant picture than that of a many-mile long queue of overladen Trabis waiting at the border to just get the hell out of East Germany. Of course, there is a perfectly sensible explanation for all this. Even a car isn’t as stupid as a communist.

Cut it any way you like: people are stupid, and since cars are designed, built, bought and driven by and for people, their nonsensicality should come as no surprise.

But the stupidity of people only explains particular forms of stupidity; my point is quite general. In addition to these myriad forms of nonsensicality, every car is stupid in the same way.

Let me say it slowly so you get it. Depreciation. Servicing. Tax. Petrol (which means tax again). Insurance. Parking. Congestion charging (London only – for now). And interest on the loan you took out to buy it. Even on the most sensible of Saabs, the most mediocre of Mondeos, the combined expense runs into thousands every year. Take a chauffeured limo everywhere, have your shopping delivered by Harrods and you’ll still end up laying out less than just the cost of renting a garage to keep your Fiesta in.

And that’s before you count the cost to your health. I’m not talking about the canard of pollution; modern cars are about fifty times cleaner than they used to be and these days emit toxins less dangerous and in smaller quantities than, say, the average Frenchman. I’m talking about the damage to the nerves caused by the inadequacy of the roads, and by the eternal struggle to find somewhere, anywhere, to park. In London, the second rule of the road is that when you want to move you will be stationary and when you want to park you won’t be able to stop (the first rule of motoring in the capital, as all Londoners know, is “if you put your hazard lights on,  you can park wherever you like”). Add to the brew the inevitable car thieves and what the idiot authorities laughingly refer to as “traffic calming” measures – speed bumps and width restrictions are about as “calming” as a pinch of Colman’s up the anus – and it’s a miracle that there are any sane motorists left in the entire country.