Full frame DSLRs or not?

There's lots of discussion these days on whether digital will move to "full frame", meaning sensors which are 24x36mm, or whether a smaller frame size will "be good enough". The smaller frame format is usually called APS-C and is the typical "1.6x multiplier" format used by many manufacturers (Canon, Pentax, Minolta and Nikon - though Nikon have a 1.5x "multiplier").

The use of the term "APS-C" is perhaps unfortunate for a format some are hoping will stay around. It's derived from the similar frame size of the late, but not lamented, APS film format (APS "Classic" format was 23.4 x 16.7 mm) . We all know what a success that was. Smaller, lighter cameras were promised - but not delivered, and image quality, though fine for small prints, couldn't match 35mm. Image quality may have been "good enough" for the vast majority of users (who rarely make a print larger than 5x7), but that didn't really help!

So what about APS-C (ca. 22mm x 15mm) as a digital format. Will it survive?

My guess is that it will be with us for a while, but in the end it will fade out. Why? Because nobody ever got rich by telling the American public that what they had was good enough and that they really didn't need anything better! If 24x36mm sensors had been available for the same price as 22x15mm sensors do you think anyone would have made a camera with the smaller sensor? Would arguments about it being "good enough" have convinced the consumer? Of course not (look at APS again).

The sole reason we have APS-C sized sensors is cost and availability. To make a camera that retails at under $1500 today, you have to use a small sensor. That sensor yields excellent quality, quality good enough to lure vast numbers of photographers away from using film. Did the camera manufacturers pick that size because they thought it was better than full frame 35mm? No. Do most photographers actually want a small sensor? Again the answer is no. They'd rather have a full frame sensor, but they just can't afford it.

You could argue that 35mm was good enough and people didn't move to medium or large format, but that's a cost and convenience issue. Both medium and large format cameras are much larger (and more difficult to use for the amateur) than 35mm cameras and both medium and large format processing is difficult to find and expensive. On the other hand APS-C and full frame DSLRs are about the same size and the cost and ease of processing the images is essentially the same.

So where does the future lie?

Yields on larger sensors are improving and the result is better availability and lower prices. It may take another 5 years before they get to the price point that smaller sensors are today, but eventually they will get there. The major camera makers will then offer them as an "upgrade" path for users of smaller sensor cameras. As I said before, nobody ever got rich by telling people that what they had was good enough and that they didn't need anything better. If you don't believe that, just look at sales ever larger SUVs, or ever faster sports cars, or ever larger and more expensive LCD TVs. Do we need them, no. Do we want them? Of course we do!

There are now a number of manufacturers making lenses that only cover a 22x15mm frame, so there will be a lot of lenses out there that can only be fully used on APS-C sensor bodies, and that will probably keep the production of smaller sensor cameras going for quite a while, but my guess is that they will end up being aimed at the lower end of the market and once full frame sensor cameras drop below $2000, that's where the manufacturers attention is going to be. Marketing and advertising will switch from telling photographers about the benefits of the small sensor cameras, and start telling photographers about the benefits of full frame sensor cameras. There may well be a consumer market for DSLR with 8-10MP APS-C sensors selling for under $750 or so for a long time, but the more serious photographers will certainly move to full frame bodies (with maybe a 12MP sensor) if and when the price falls below $2000.

Another thing that will drive a whole industry move to larger sensors is, of course, when one of the major manufacturers (and my money is on it being Canon) brings out a 24mm x 36mm sensor camera at an affordable price. The others will have to follow (if they can) or lose sales.

If you want another driving force for larger sensors, it's megapixels. There's a limit to the number of pixels you can cram onto an APS-C sized sensor before you start to run into real problems with noise due to small pixel size. You can't make a noiseless sensor and the smaller the pixels, the higher the intrinsic noise. There's a lower limit to noise which is determined by fundamental physics and includes things like photon shot noise which is determined by the statistics of photons hitting a pixel, plus various forms of thermally induced noise. At some point you'll hit a limit to the number of usable pixels on an APS-C sized chip. It's probably in the 10-12MP range. People like more pixels and will usually pick a 12MP camera over a 10MP camera over an 8MP camera. At some point the only way you can up the pixel count without lowering image quality (by increasing image noise) is to go to a bigger sensor. That's why 8MP DSLRs outperform 8MP consumer digicams with 5mm x 7mm sensors. Size matters.

Won't the number of APS-C coverage lenses out there keep the APS-C format in production?

Well, there are millions of Pentax Screw mount lenses out there and nobody makes a camera body that uses them. There are probably millions of Canon FD mount lenses, including very expensive telephotos, out there and Canon no longer make a body that can take them. There are other examples of lens systems that have died out. I see no real reason why the existance of APS-C format lenses should mean the format will control the market or prevent the introduction of full frame sensors in consumer DSLRs. Most owners of APS-C format DSLRs will also have full frame lenses - probably more full frame lens than APS-C coverage lenses - so upgrading to a full frame sensor isn't an insurmountable problem.

An exception may be the Olympus 4/3 system (E1, E300). It has an even smaller sensor than APS-C (18mm x 13.5mm vs 22mm x 15mm), however it's an integrated system with all lenses specifically designed for 18x13.5mm coverage. You can't use a bigger sensor, therefore there's no obvious "upgrade" path. I imagaine it will stay with that format as long as it's in production.

Should we avoid buying APS-C DSLRs and lenses because one day they may be "obsolete"?

Certainly not. I have an APS-C sensor DSLR myself and I have one APS-C lens. I might well buy another one. One day I might not be able to fully use them on my full frame DSLR, but I buy lenses to use today, not in 5 years time. I wouldn't buy a $7000 super telephoto lens with APS-C coverage, but them again I don't expect Canon or Nikon will ever sell one! We buy all sorts of things that become "obsolete". How much did you pay for your first computer, and where is it now? How about that 10MB hard drive you paid $500 for in 1990, or that 64Kbyte (yes, kilobyte) S100 memory card that cost you $200 in 1978? How much did you pay for your car, what's it worth today and what will it be worth in another 5 years? Some people seem to look on lenses as investment vehicles that should hold their value and utility forever. I'm not quite sure why they apply this logic to lenses, but not to cars, TVs, VCRs etc. Cameras and lenses are tools. If you use them, you'll get value from them. If you leave them sitting on a shelf, you won't. If you want to invest, buy gold or real estate! Of course it would be nice if every lens and every body worked with every other lens and every other body from now until the end of time. It would be nice if all Canon lenses worked on all Nikon cameras and vice-versa. However neither is likely to happen.

Just as an aside, it's possible to built an APS-C to full frame 35mm converter which would act as a 1.6x multiplier and which would cover the full 24mm x 36mm frame. So your 10mm APS-C lens (which has the same field of view as a 16mm lens on a full frame camera) would become, with the 1.6x converter, a 16mm lens when mounted on a full frame DSLR! And so you get back to where you started! It defeats the point of APS-C lenses of course, but I'm just pointing out that you could, in fact, use reduced format (APS-C) lenses on 35mm format DSLRs with an appropriate optical interface.

The Bottom Line

I don't think you can make predictions based mainly on photographic "needs". You have to take into account photographic "wants" and the influence that marketing and advertising have to make the consumer want what technology can deliver. Given the existing base of full frame 35mm cameras and lenses with full frame 35mm coverage, my money is on full frame 35mm DSLRs eventually becoming dominant in the market. I'd say that $2000 is the magic number right now, and today nobody is near that, though the Kodak DCS/Pro 14MP full frame DSLR has recently been selling (new) for $3500. Now it has its problems and it's certainly not a Canon EOS 1Ds mark II, but it is a DSLR with a 14MP 24mm x 36mm sensor. I think it's clear that in a few years, a full frame camera under $2000 is certainly possible, even likely. Things move much faster than many people expect. Remember that only 9 years ago you'd have had to pay $20,000 for a 1.3MP DSLR with a 16mm x 20mm sensor (EOS DCS 3). Even just 4 years ago a 3MP DSLR (APS-C) was selling for $3000 (EOS D30). Today you can buy a 6.3MP APS-C camera for $800 (actually $500 with the current Canon Rebates!)

What will the DSLR lineup look like in 5 years time? I'm guessing a pro level DSLR, 36x24mm sensor and 24MP, a mid level DSLR 36x24mm sensor with 16MP, a prosumer level DSLR 36x24mm with a 12MP sensor and a consumer level DSLR, which maybe will still have a 22x15mm (APS-C) sensor but with 8MP. Each level will not only have more pixels, but more features, faster frame rates, buffer sizes etc. Prices around $5000, $3000, $1500 and $750 (in terms of 2004 dollars).

Could I be wrong? Yes, I certainly could, it would be remarkable if I got everything right. In 5 years time I guess we'll find out.