Rumors abound – these you can be certain of.

Teaser ad on the Canon site

With all that speculation about the iminent replacement for the Canon 5D, here’s my list of things I can pretty much guarantee will not be in the 5D Mark II:

• A 21mp sensor. No way. That would immediately cannibalize sales of the ultra-high margin 21mp 1Ds Mark III which goes for $8,000 a pop. And with the 11 mp in the 5D being as good as it is, reckon on no more than 16mp. Let’s hope they don’t muck up resolution in the process.
• Ultra-high framing rates like in the 40D and 50D. Same reason as above.
• Full weather sealing. No way no how. This camera is aimed at the advanced amateur snapper, not the pro in rain forests, even if a few rubber gaskets cost $1.50 to add.
• Eye controlled focus. That’s the fabulous technology available in some late Canon film SLRs. The camera focuses where you look. (This still seems like magic to me). For some reason Canon have never added it to any of their DSLRs. Just imagine using something like Helicon Focus with eye controlled focus. Sight down your subject – click. Look a little further – click. Oh! wow. But not to be in the Mark II.
• A smaller body. That would cost too much to re-engineer. Canon will add already mature and developed technologies like sensor dust removal and live view (ugh!) but a comprehensive re-engineering of the body would cost too much.
• A change for that dumb Print button to make it useful – such as a mirror lock-up control. Someone at Canon has a real axe to grind for their printers (surprise!), so expect more of this silliness.
• Lens aberration correction inside the camera’s software. See the first bullet point above.
• A permanently attached vertical hand grip. That would look too ‘professional’ and adds needless bulk to an amateur’s camera. Indeed, one of the appealing aspects of the 5D is that it does not look professional – especially if you add some electrician’s tape to all those gauche logos.

Expect the announcement in late September at Photokina in Cologne.

The law of diminishing returns kicks in.

A couple of years ago I wrote, with something approaching amazement, about the longevity of the Epson 1270 ink jet printer, dubbing it a Ten Year Digital Device. Indeed, that printer’s current owner will testify to the Epson’s longevity having just picked up a prize for one of his pictures printed on it. Sure, the nozzles clog if you don’t use it frequently and the inks fade in bright sun, but the quality of the prints cannot be disputed.

Canon 5D and friends. A ten year kit?

All of which prompts the question whether the Canon 5D has a similar life expectancy. Sure, it remains a current model and certainly it is not as fast or as slick as newer offerings from DSLR makers. It coasts along at a modest three frames/second, has no dust removal and lacks silly features like live previews. Now given that 3 fps is meaningless to me as I take one picture at a time and avoid sports photography, I can only question who really needs the insane framing rates available today, sports and fashion snappers apart? Live previews are a solution looking for a problem with DSLRs but, yes, dust removal from the sensor would be nice to have. But I can live without it, just as I learned to live with the 1270’s clogging nozzles.

Wear is not an issue for me. After 30 months with the 5D it reports that I am on frame 6,873. That figures to some 25,000 frames over ten years, well below the 100,000 life expectancy of the 5D’s shutter.

Definition is not an issue. The law of diminishing returns suggests that all those latest pixel-heavy sensors are running into noise issues, and that the modest 12.8 megapixels of the 5D make for a perfect compromise between definition and noise.

Sensor size is an issue. I like what I have. As I want my 20mm lens to be 20mm, not the 32mm that I would get with a cropped sensor, and I like the depth of field a standard lens offers on the big sensor, my alternatives are limited to full frame cameras of which there are but two from each of Nikon and Canon. It’s clear we will have more large sensor DSLRs (Sony is rumored to be releasing one soon) and choice is always a good thing but the bottom line is that the images from the 5D’s sensor are so crisp, noise free and well defined that trading for more pixels or a medium format sensor make no sense.

Build quality is fine, too. Doubtless the big Canon and Nikon offerings are tougher but I’m an amateur snapper, for heaven’s sake, and not a photojournalist in a war zone.

Lens choice is fine and will only get better. A really good 20mm would be nice, Canon’s wide primes being less than thrilling unless you get the ridiculously bulky and expensive ‘L’ variants. Unless Canon does something truly dumb – like changing the lens mount – I am set.

Dynamic range, the biggest bugaboo of digital cameras (as in they have too little), is something I have worked around. Under-expose 1/2-1 stop and bring things back as needed in Lightroom, and all is well. Further, there will have to be some serious breakthroughs in sensor technology before DSLRs start exhibiting enhanced dynamic range. So for now I watch the highlights and let the shadows look after themselves at the exposure stage. Much as in the Kodachrome days….

Given that digital was a joke ten years ago and has now plateaued at a level significantly higher than film, it’s foolish to try to predict what will be on offer ten years hence. That plateau was reached a few years back by the Canon 1Ds Mark I and the 5D. So until some shattering new technology comes along that offers the image quality of the 5D in a package half the size, weight and noise – and I’m not holding my breath – I’m going to stick my neck out and suggest that maybe the Canon 5D really is a ten year digital device. That’s assuming I am not completely gaga 7 years hence and can still lift a camera to eye level without wetting myself. No calling that one.

Old but still topical.

I came across these two on my hard drive – both white papers were released by Canon about the time of the 5D’s debut and still make fascinating reading.

Download either PDF by clicking the pictures below.

Both are fairly lengthy, with the first having an especially interesting discussion of the development of CCD and CMOS sensors, clearly explaining the differences.

Why publish this now when the 5D is about to be updated and obsoleted? Because if you can live with the modest maximum framing rate and the absence of dust removal, I believe the 5D lightly used market will be flooded with cameras from upgraders and will, as a result, offer an enticing opportunity to enter the world of full frame imaging at a very attractive price. If your print size is limited to 24″ x 30″, you cannot go wrong with this body and sensor – just don’t use Canon’s truly execrable cheap zooms on it. A good sensor deserves the best glass – primes if you can swing it as no zoom compares to a like prime for definition and overall performance. Plus most of Canon’s primes are far cheaper than their L zooms and are generally superb – I use the 20mm (not so superb), the 15mm fish eye, the 50mm f/1.4, the 85mm f/1.8 and the 100mm macro. The last four are as good as anything I have used at any price and run $300-400 new, with lots of used bargains out there.