January 19, 2003
Weekly Column By
Pentax Under Glass
I'm hearing dribs and drabs from various sources that makes it seem 90% certain that Pentax will show something concrete at PMA in the way of a digital SLR. Although the camera will likely not ship until later in the year, there will probably be at least "under glass" mockups and, at most, multiple working prototypes and detailed (if perhaps tentative) specs.
This is exciting to a number of people for several reasons. Firstly, because we love to speculate about the future and its possibilities, and anything that doesn't exist yet is a handy repository for all our fond hopes. Secondly (and less tongue-in-cheek), because Pentax has a history of making relatively basic, no-nonsense cameras in reasonably small packages, usually for down-to-earth prices. And thirdly, because Pentax's DSLR promises to accept the famous "K" lensmount, with reasonable hopes for backwards compatibility. Nikon PR notwithstanding, Pentax actually has better MF-to-AF and historical-to-current lens and lensmount compatibility than any other 35mm camera brand.
More Megapixels = Better?
One thing that concerns me, however, is the persistent prevalence of the idea that "more megapixels = better." People are saying that a 6-mp DSLR will be outdated before it's launched, and expressing hopes for a 10-mp Foveon sensor.
Why? In my opinion, more megapixels are not necessarily a good thing. The idea is a "horsepower" argument. It's oversimplified.
This is, of course, a time-honored failing of American males — we love power, and we love to simplify things down to brute force measurements (Tim Allen made a whole career out of lampooning this tendency, and not entirely ironically, either). Cubic inches, caliber of bore, clock speed, whatever. The hell with truth! Americans want the bigger number, and promptly clap their hands over their ears when the argument gets any more sophisticated than that. Never mind that today's off-the-shelf multi-valve café-racer four-bangers are faster than a 1960s muscle-car, with half the cubic inches. Real men know which is more real-manly. Bigger is better.
Traditional photographers reflexively conceive of image quality in analog film terms. The smaller and more closely-packed the grain clumps, the better the image quality. A contact print is best, a small enlargement is better than a bigger enlargement. The bigger the piece of film, the better. Fair enough.
The trouble is that digital doesn't work that way. Your inkjet printer can lay a certain number of dpi down on a piece of paper, and that's the end of it. All other things (sensor size, pixel density, post-processing algorithms) being equal, number of pixels is the measurement of an optimum print size. Not potential image quality in small prints. (Some people have difficulty with this idea, although it's like confusing the term "light year" with a measurement of time.) Make the print smaller, and you simply have to toss out the excess information. So if what you want is to make 4x6 prints, for instance, you don't gain any quality, necessarily, by having 11 megapixels instead of 3 megapixels. Three is enough. Starting with the 11 megapixel file just means you throw away a heck of a lot more information when you size the image for a 4x6 print and lay the ink dots down on the paper.
Where's the harm in this? What's wrong with having the ability to make a 14x17-inch print even if what you normally make are 5x7s? Well, simply because it's not efficient. You'll either waste a whole lot of money paying for an 11-mp camera when all you're going to do with it is record on a lower resolution, or else you're going to be recording files that are a whole lot bigger than you need most of the time. The trouble with this is that you then have to pay for a) more storage media capacity, and b) more hard drive space and RAM with which to handle those monster files.
The Ravings of a Troglodyte
The fact is, well over 99% of the photographic prints made in the world are 9x12 inches in size or smaller. If you have the larger of the desktop printers, say an Epson 2200 or a Canon 9000, making 9x12s floated in the large sheet-size is an elegant and presentable maximum print size for all but the most diehard of amateurs. So do you really need 11 mp, or 14 mp, or a 10-mp Foveon-type sensor to make prints this size?
It depends on a lot of factors, of course. Execution matters. But personally, I never hit the limit of print size with the 5-mp Sony I enjoyed owning for six months or so. Since all of my own prints are 8x10 or smaller, I'm pretty sure a Canon D-30 would serve my needs adequately if I were to switch over to digital-only. You might dismiss this out of hand as the ravings of a troglodyte, but the fact remains — most of us don't need an infinite number of pixels, and the sordid truth is that however much we like to crow and preen about big numbers, most of us actually don't want to lay down the cold hard cash for the camera that can record them or for the other capabilities we'd need in order to support them.
The Pentax Way
I'm looking forward to seeing Pentax's particular slant on the DSLR. If a corporation can be said to have a character, Pentax tends towards "enough" quality without a lot of sizzle and pizazz. The Pentax DSLR is still vaporware as far as any of us know, and as soon as it's revealed, the inevitable bitching about its designers' choices will begin. But one thing I will personally not be complaining about will be if the sensor size is no more than 6 mp. I know, if this happens we'll all read anguished outcries from the Chicken Little spec-sheet-readers of the world about how 6-mp is yesterday's news, and not competitive with Canon, and outdated before it hits the shelves, and blah blah blah. It's inevitable. But I say "bull." Six well-cooked megapixels are plenty for the market any true Pentax will take aim at.
Me? I'm sure I'll find something else to complain about!
— Mike Johnston
Mike Johnston writes and publishes an independent quarterly ink-on-paper magazine called The 37th Frame for people who are really "into" photography. His book, The Empirical Photographer, is scheduled to be published in 2003.