The Micro Four Thirds Announcement: More Detail and Analysis
Today's announcement of the "Micro Four Thirds" lens mount standard will doubtless send shockwaves through the camera industry and SLR marketplace. It's arguably the most significant announcement in the camera market this year, as it introduces a whole new body/lens standard, and (at long last) opens the possibility of compact digital cameras with interchangeable lenses.
It's significant enough that I wanted to weigh in with my own thoughts about the system and what it means. Herewith, then, a closer look at what Micro Four Thirds means for camera buyers, other manufacturers, and current Four Thirds system owners.
What's the major change?
As noted in our first news item on the Micro Four Thirds announcement, the biggest change here is that the mirror box has been eliminated (meaning that all Micro Four Thirds cameras will be LiveView-only, there won't be any direct optical path from the viewfinder through the camera lens. They'll thus technically be EVF cameras, similar to most non-SLR digicams.
By eliminating the mirror box, the "back focus distance" (or back focal length) of the lenses can now be reduced, and will be cut roughly in half from the equivalent dimension in the current Four Thirds system. Technically, back focus distance is defined as the distance from the vertex of the lens's rear element to the sensor, but for practical purposes, it's easiest to speak in terms of the distance from the lens mounting flange to the sensor surface. In current Four Thirds systems, this dimension is 40mm, in the Micro Four Thirds system, it will be 20mm.
The illustration at right (credit for this and others:Four-Thirds.org) shows schematically the difference between the current Four Thirds mount and the new Micro Four Thirds mount. On a camera.
Why are they doing this?
One of the first reasons consumers give for not making the move to an SLR is that SLRs are "too big." One of the original marketing points for the Four Thirds concept was that, by designing the system from the ground up to use a smaller sensor, lenses and bodies could be made much smaller than those for conventional SLRs, which are all based on dimensions taken from the 35mm film world. (Even reduced-frame cameras deviate only slightly from the size constraints imposed by the 35mm roots of other current SLR systems.)
Part of the reason that the original Four Thirds system hasn't resulted in much smaller cameras is that the 40mm back focus distance mandates rather large optics. It turns out that the physical size of camera lenses depends somewhat on the size image circle they must produce, but depends even more strongly on the back focus distance they're designed for. Reducing the back focus distance can dramatically decrease the size of the lenses needed to cover the Four Thirds frame. It also (obviously) reduces the thickness of the camera body itself.
The decrease in lens size isn't only in length: Lenses for the new system can be smaller in diameter as well, so the new mount is also 6mm smaller (44mm vs 50mm) than the original Four Thirds system called for. The illustration above compares the size of the new and old lens flanges.
Why didn't they do this in the first place?
When the Four Thirds system was developed, Olympus (one of the founding companies of the Four Thirds coalition) felt that telecentric lens designs were critical to producing the best image quality on electronic sensors. Telecentric lenses are unusual, in that light rays striking the imager surface all arrive perpendicular to the imager's surface. This is important for CCD sensors, because the multiple metallization and insulating layers used for that technology produce a chip surface with a very significant 3-D structure. This structure can actually cast shadows on the light-gathering areas of the chip if light rays arrive at a glancing angle. This leads to shading, and (when the microlenses are considered) possibly chromatic aberration effects and crosstalk between the camera's RGB color channels. Geometric distortion is also generally less in telecentric lens designs, and
Telecentric lenses aren't easy to make though, and having a greater back focal distance really helps in their design. - So the original Four Thirds lens specification dictated a rather large back focus distance.
Since that time though, some of the downsides of non-telecentric lenses have been mitigated by advances in image processing: Current camera CPUs are fast enough to do very sophisticated image processing to greatly reduce some of the image artifacts resulting from the use of non-telecentric lenses. At the same time, the NMOS-based image sensors developed by Panasonic (and others?) have a much flatter surface profile, and so are much less subject to problems with light arriving at higher angles of incidence.
The combination of these two factors means that many of the benefits of telecentric lens design are less compelling than they were only a few years ago.
What does this mean for us as users?
This, of course, is the $64,000 question: What sort of characteristics can we expect from the new cameras, and how will they work (if at all) with existing Four Thirds lenses?
This is the obvious one: The new system will permit substantially smaller system components, most notably the lenses themselves. Expect Micro Four Thirds lenses to be almost half the size of current ones sporting the same focal length/aperture specs. This is a big reduction that will only become truly apparent when we see the first Micro Four Thirds camera and lens prototypes. (No product announcement is being made at this time, but we expect to see product announcements either before or at the Photokina show in late September.) Stay tuned, we think you'll find the reduction in size quite impressive.
No Optical Viewfinder
This will depend somewhat on the camera form factor manufacturers are aiming for. While you could add a conventional, non-TTL (through the lens) viewfinder to such a camera, the fact that they'll have interchangeable lenses makes that less likely. (Leica's long-standing system of interchangeable-lens 35mm rangefinder cameras not withstanding.) Micro Four Thirds cameras are thus likely to all be EVF (electronic viewfinder) designs. Personally, I view this as a distinctly mixed blessing. EVFs to date have been rather lackluster in their performance in several areas. They generally have poor tonal response in the highlight area, making it difficult to see subtle subject detail in areas of strong highlight. They also often have difficulty in very dim light: To keep the viewfinder refresh rate high enough to be useful, they end up having to clock the sensor chip for relatively short exposure times for each frame. Think about it: You're in a situation where the correct exposure requires a one-second shutter time, yet you expect the camera to be refreshing the viewfinder display 15 times/second. Something has to give, and it's image quality in the viewfinder: EVF cameras tend to show great amounts of noise in their viewfinder displays at low light levels.
There's also the matter of responsiveness of the viewfinder when dealing with fast-moving subjects (sports, kids, some wildlife). Most EVF cameras to date have trouble with smearing of the viewfinder display when dealing with rapid movement, even under conditions of good lighting.
Contrast-detect autofocus only
To my mind, this is potentially one of the most critical areas: One of the biggest reasons people make the move from digicams to an SLR is to get past the shutter lag found in most digicams. The reason SLRs focus so much more quickly than digicams is that they use an entirely different mechanism for determining focus. Rather than clocking data off the image sensor and adjusting the lens until the image is sharpest (so-called contrast-detect AF), they use a separate AF sensor that indicates not only whether the subject is in focus or not, but how much out of focus it is, and in which direction. This is called phase-detect autofocus, and it makes for much faster AF operation. Rather than having to creep up on the final focus setting, most SLRs can take one "look" at the subject, calculate the focus correction needed, and simply set the lens to the correct focal distance.
The catch is that that AF sensor has to go someplace, and in most SLRs, it's in the bottom of the mirror box, with light reflected down to it by a small secondary mirror. No mirror, no mirror box = no phase-detect AF system. (That's perhaps not an absolute, you might be able to do some sort of phase-detect AF still, by inserting a mirror into the light path ahead of the image sensor - But that would interfere with the viewfinder display, somewhat defeating the purpose.)
So Micro Four Thirds cameras will depend on contrast detection for their AF systems. This isn't necessarily a kiss of death on the system: We've seen impressive improvements in digicam shutter response figures over the last couple of years, with the best of them approaching the lag times of entry-level SLRs. The lenses for the Micro Four Thirds system will be much larger than typical digicam lenses (due to the larger sensor), so the focus motors will have more mass to move back and forth. Current Live View SLRs that support contrast-detect focusing range from decidedly slow to only a little slower than phase-detect AF systems. Overall, we've some reason to be hopeful that Micro Four Thirds cameras will compete reasonably well with SLRs, particularly under good lighting.
Lens compatibility: Yes and no
This leads us to the next question, that of compatibility between current Four Thirds lenses and the new bodies. As the illustration above shows, there's backward compatibility between the two systems: Current Four Thirds lenses can be mounted onto Micro Four Thirds bodies via an adapter, but Micro Four Thirds lenses can't be attached to a current Four Thirds body. (The mirror in current Four Thirds cameras won't permit the Micro Four Thirds lenses to be mounted close enough to the body.)
This is good news for current Four Thirds users: Your current lenses will at least be able to be mounted on the new Micro Four Thirds bodies. Note though, that I said "mounted", not necessarily "operate with full functionality." The distinction comes in how fast a given lens can change its focus setting. Normal phase-detect AF systems are fairly forgiving of lens performance, because the correct focal distance setting is determined somewhat independently of the lens elements' motion: The camera looks at the subject, calculates the correct focal distance, commands the lens to move, and then snaps the shot once the lens reports back that it's moved to the correct setting.
With contrast-detect autofocus though, the lens has to move multiple times, and must come to rest before the camera can take each "look" at the subject, to determine whether the focus is better or worse than it was at the previous focal setting. In order for the overall focus cycle to be performed quickly, the lens needs to be able to shift focal settings very quickly, multiple times per second. This is a demanding requirement, and not all lenses will be up to the challenge. Olympus offers contrast-detect AF on some of their recent SLRs, but the feature only works with certain lenses that have focus motors both strong and fast enough to slew the lens elements to a new position very quickly. Other manufacturers offer contrast-detect AF that works with any of their lenses, but at the cost of very slow AF times. (It's for this reason that Nikon refers to the contrast detect AF option in their SLRs as "tripod mode.")
So, while some current Four Thirds lenses will be able to work fine on the new Micro Four Thirds system, supporting full AF operation, other lenses will likely be reduced to "guided manual" focus, in which the user will need to adjust the focus setting, looking to either the viewfinder (less accurate) or the camera (more accurate) for focus confirmation.
This isn't the first time I've kicked myself for not sharing some of my musings with the outside world: A few months ago, I found myself thinking "Hmm... Live View SLRs, can movie recording be far behind?" I'm perpetually short on time though, so I passed on the opportunity, only to find that I'd missed yet another chance at prophecy: Not included in the press release, but prominently mentioned on the Four-Thirds.org website, is the possibility of movie recording from Micro Four Thirds cameras. The fact that the press release made no mention of movie recording, and that the Four-Thirds website uses the phrase "soon" in describing its availability (rather than just saying "Users will be able to..." suggests that a) not all Micro Four Thirds cameras will be movie-capable, and b) it may be later models that support this feature. Actually, it's not entirely clear why current Live View SLRs don't offer movie recording, but I suspect it's due to limitations with current contrast-detect autofocus systems: Continuous AF is a requirement for movie recording, and would in turn demand fast-operating AF systems and lenses. I won't be surprised to see this capability arrive in some non-Micro Olympus models, given that they've already got contrast-detect AF working reasonably quickly. But I also suspect that the Micro Four Thirds lenses and cameras will focus much faster, and therefore be more-capable movie platforms.
Whither the current Four Thirds system?
What this will mean to the current Four Thirds system is an open question. In the joint announcement today, Olympus was careful to state that they'll continue to "develop Four Thirds System interchangeable lens type digital camera system products," while they're working on new Micro Four Thirds products, and Panasonic said essentially the same. Resources are scarce and corporate budgets tight though, so it's an open question where each manufacturer's primary focus will lie: I don't personally expect either to maintain a strong position in both product lines simultaneously. I also see the two product lines aimed at somewhat different groups of photographers. While it will doubtless be popular with enthusiasts interested in its small size and resulting portability (not to mention its likely stealth: No whap-slap mirror noise to give away the candid shooter), the Micro Four Thirds line seems clearly aimed at current digicam users who've yet to be enticed into the world of SLRs. With their optical viewfinders and (probably) faster phase-detect AF, cameras based on the current Four Thirds standard are likely to be more appealing to advanced users and/or those most concerned with photographing rapidly-moving subjects. They'll also compete more directly with current SLR models having similar optical underpinnings.
The competition is pretty severe in the mainstream SLR market, and will only become more so as time goes on. A lot will obviously depend on how consumers react to the first Micro Four Thirds products. If the new models do poorly, the issue of Micro vs mainstream Four Thirds will be moot. If they're successful (and I think they have an excellent chance of being so), the market equation will likely be such that both Olympus and Panasonic will be motivated to pursue Micro Four Thirds more strongly. Among other things the pool of potential new adopters of the Micro Four Thirds format is much larger than the group of users currently engaged in the SLR market.
Of the two players, Olympus has more to potentially lose, and perhaps less to gain, while Panasonic appears to have less to lose and potentially much to gain. Olympus currently has a much more developed SLR lineup, with three current models, a very extensive array of lenses supporting them (21 lens models, plus two teleconverters and an extension tube), and a large existing user base. By contrast, Panasonic has only one current model (the DMC-L1 appears to be largely out of the market now, at least here in the US), and far fewer existing users. Given this balance, it's reasonable to expect to see more activity from Panasonic under the new standard. (I think it's also significant that, while it's a joint press release, and Olympus' name is mentioned first in it, we received the press release here in the US from the Panasonic organization.) If they're successful though, we think Olympus will follow with models of their own. Olympus has to be very careful how they position any involvement with Micro Four Thirds, to avoid their current, loyal user base from feeling left behind. Ultimately though, the market will decide whether one or both standards persist into the future. (I don't view the fourth option, of Four Thirds going away entirely, as being very likely, at least not over the next several years.)
All in all, a very interesting announcement for the digital camera market, clearly one of the more significant of 2008. Stay tuned for our video coverage of the Photokina trade show, we'll surely be taking a look at any Micro Four Thirds products announced there, and interviewing key executives and camera designers!