April 29, 2010
The essential principle of the single-lens-reflex camera is actually quite old. Most of the SLR’s earliest incarnations are long forgotten today; but it’s interesting that the name “Graflex” derives from that brand’s early, jumbo sheet-film reflexes—dating all the way back to the 1890s.
In the middle of the 20th century, 35mm cameras rapidly gained respect and market share. But if you could time-travel back to 1958 and ask photographers how they felt about 35mm SLRs, you might be surprised at how divided and contentious the answers were.
Most would probably admit to the SLR’s advantages—in principle—over the other viewfinder styles common at the time.
There is no offset between the taking lens and some separate viewing lens. The framing of a shot is previewed exactly—which is particularly useful in macro and telephoto work. And the lens’s depth of field can be seen directly on the groundglass (albeit dimly, when apertures are small).
All these advantages might even cause our 1958 friends to proclaim SLRs as the wave of the future.
But keep in mind: At that time, the leading 35mm SLR brand was the Exakta, from Dresden, East Germany.
This was a nicely-finished camera, capable of taking fine photos. But its operation was rather clunky. For those who wanted to shoot quickly and spontaneously (which after all, was the forte of 35mm compared to larger formats), SLRs could not compete with rangefinder cameras. Decades of improvements had brought rangefinders to quite a high level of refinement, and they fully dominated the era’s 35mm marketplace.
Before the shutter opens, an SLR’s mirror must flip up out of the way. But in these early models, it did not drop down again afterward—not until the film was advanced. The resulting viewfinder blackout was disorienting, and made it quite hard to follow action.
Also, for a clear, bright groundglass image, an SLR’s lens ought to be at its widest aperture. But ordinarily, it must then be stopped down to a smaller working aperture before making the exposure. Doing that manually for every single shot becomes kind of a pain. Early SLRs struggled with this problem, and manufacturers created numerous rather half-baked solutions to it.
In retrospect the answer was obvious: just design an instant-return mirror and an instant-reopen lens diaphragm. Yet throughout the 1950s, a puzzling thing happened: All the elements of the solution existed somewhere; yet no camera maker ever put all the pieces together.
Many Exakta lenses used an external, spring-loaded plunger aligned in front of the shutter release; the photographer’s finger pressure on this closed the diaphragm. The Praktina of 1954 (another East German brand) pioneered the first instant-stopdown linkage built inside a lensmount—though needing a separate, manual lever to reset it.
But when it came to instant-return mirrors, German camera-makers had a strange resistance to them. Exakta would not redesign their SLRs to include one until 1966.
An instant-return mirror appeared in the 1954 Asahiflex, predecessor to the Pentax. Several other Japanese brands quickly adopted the innovation. But they still wrestled with the diaphragm problem. Many brands used a mechanism that stayed closed down after the shot, dimming the viewfinder until it was reset.
So, it’s rather understandible that many 1950s photographers found SLRs exasperating—and assumed they would always stay that way.
Finally in the spring of 1959, three new Japanese SLRs were introduced: the 120-film Zenza Bronica, the Canonflex, and the Nikon F. All made the breakthrough of combining the two essential features—the instant-return mirror and the instant-reopen diaphragm.
Although Canon’s first SLR proved an evolutionary dead end, the Nikon F was an instant classic (and its lens mount lives on, in Nikon’s DSLRs today). It cemented Nikon’s reputation as a top-tier camera maker; and it announced the arrival of the Japanese as the world’s new camera-design leaders. And within a few years, the two SLR innovations were practically universal among Japanese brands.
So—how does all this ancient history relate to the current wave of EVIL cameras? (“Electronic Viewing, Interchangeable Lens.”) Well, I believe EVIL stands at a similar crossroads today.*
In principle, we know electronic viewfinders offer certain advantages:
Potentially, they can give a much larger and brighter image than the cut-down reflex viewfinder in an APS-C camera. Histograms or any other information can be overlaid on the live image (and be easily reconfigured, via menu or firmware updates). Depth-of-field preview can be brightened electronically, for easier viewing. And instantly magnifying a portion of the frame eases focusing manually when desired.
Electronic viewing opens up possibilities for unconventional new body designs—ones that might be innovative, less obtrusive, and more easily pocketable.
And in principle, (relatively) larger sensors ought to offer no-compromise shooting at higher sensitivities—say, ISO 800, at least.
A midsized image format, with no reflex mirror getting in the way, should stimulate nifty lens innovations, too—imagine shrunken-down rangefinder-type designs. An f/1.4 normal lens could be half the size of its 35mm equivalent! (If the purpose of larger sensors is enhanced low-light shooting, why not fully capitalize on this?)
But the EVIL cameras of today are, in their own way, 1958 Exaktas.
We are beginning to glimpse the great potential they offer. But all the current implementations are crippled by maddening omissions and flaws.
The Olympus VF-2 is the nicest electronic viewfinder currently available. But EVFs must continue to improve in speed, clarity, and low-light usability before they can replace optical viewfinders entirely. And ready access to magnified focusing is essential—something the Samsung NX10 seems to have bungled badly.
“Faux-SLR” body shapes are boringly unimaginative, and needlessly large.
To date, Micro Four Thirds has only delivered a single camera with adequate high-ISO performance (the Panasonic GH1)—despite trumpeting this as the key performance advantage of larger sensors. (The GH1 also uses the only µ4/3 sensor to allow 3:2 framing without penalty.)
Likwise, the Samsung NX10 gives sub-par high-ISO performance compared to other APS-C cameras, such as those using 12 Mp Sony sensors. Pixel counts higher than this simply become counterproductive.
Only Panasonic has delivered any native EVIL lens brighter than f/2.0—which is inexcusable, considering the wide apertures of cine and TV lenses covering similar image circles. Adapting legacy lenses to EVIL bodies remains problematic, due to µ4/3’s crop factor and Samsung’s NX design choices.
Shooting quickly and spontaneously requires an eye-level viewfinder—the history of cameras has repeatedly shown it. A touch-screen interface may look whizzy, but it splits attention between the camera and subject. Controls need to be graspable and usable by feel, while looking through the camera. We’ll have to see if Sony’s upcoming EVIL system (rumored to be called “NEX”) makes any concessions on this point.
In short, it may be quite reasonable to claim “EVIL is the future.”
But I say, “the future isn’t here yet.”
I am still waiting for 1959.
*(Some prefer “MILC”—Mirrorless Interchangeable Lens Cameras—or even “SLEV”—Single Lens Electronic Viewfinder. But why name a camera after what it lacks?)
April 1, 2010
Their tests show it having slightly worse high-ISO performance than its competitors in the “compact EVIL” segment.
The entire selling point of Micro Four Thirds is that the larger sensor offers improved picture quality, relative to typical compact cameras. DxO Mark’s “Low-Light ISO” score is expressed in ISO sensitivity numbers; and here they report that noise becomes objectionable at around ISO 500.
But the overall “DxO Mark Sensor” score falls much closer to that of a modern compact camera than to that of a good recent DSLR.
This is a disappointment, given the extra time Olympus had for developing the E-PL1; and also compared to the dramatically-better performance of its Micro Four Thirds cousin, the Panasonic GH1.
As always, note that DxO Labs tests are entirely “numbers oriented” and only analyze the raw sensor data. Handling, price, the quality of in-camera JPEG processing, etc. are not considered.