APS-C Emerges as “the new full frame quality” + the Sagging Momentum of Micro Four Thirds (m4/3)
Well, 2013 might be the year of the compact camera with APS-C sensor (generally with fixed lens, but I’d like to see a 28/50 “bi elmarit” myself). That trend might veer and split off into a full-frame variant too as per the existence proof, the Sony RX1.
Witness the Nikon Coolpix A (in the wild already), Ricoh GR V (about to arrive) on top of the Sigma DP Merrill and Sony NEX and Fuji X cameras and probably others that I’m forgetting: all compacts with very high image quality, most of them with a fixed lens but NEX and Fuji X offering interchangeable lenses.
This game is moving fast.
Micro Four Thirds
Micro Four Thirds cannot touch these APS-C cameras in terms of image quality, certainly not the sensational noise and color rendition of the Nikon Coolpix A and the oddball but bitingly sharp sensor of the Sigma DP Merrill series. And you can bet that a next-gen Sony sensor will knock your socks off.
Which makes me wonder about the viability of the Micro Four Thirds format: what is the point of insisting on such a small sensor when APS-C is now proven to afford very compact form factor possibilities? And with the industry’s heavy hitters behind APS-C and full frame (Nikon, Canon, Sony).
The M4/3 sensor is only 60% of the size of APS-C, yet cameras like the Olympus OM-D E-M5 suffer from noise at embarrassingly low ISO values and the Olympus E-M5 is actually a lot larger than entrants like the Coolpix A. There are smaller M4/3 camera bodies, but they are not materially smaller in a practical way, nor all that appealing.
The Sony RX100 is as small as I’d want to go, and there’s only a little wiggle room between the Coolpix A form factor and the Sony RX100 form factor. Where does that leave M4/3?
As for image quality, the Sony RX100 does amazingly well and arguably as good as the E-M5 sensor (maybe even better in some ways)—even though it’s smaller in area: as sensors improve APS-C is going to really rock, and the RX100 sensor size will get even better. So the M4/3 gang had better burn the midnight oil and this business about “new camera next fall” is a dangerous game plan for Olympus. But of course Olympus now has Sony as an ally, so a state of the art Sony sensor in M4/3 is a pleasing thing to contemplate.
But assuming that M4/3 achieves the same sensor quality gains as elsewhere, where is the broad-based industry support? Nikon and Canon don’t support it, and Zeiss has not committed to building M4/3 lenses, but has committed to Sony NEX and Fuji X. Ponder that—think 'traction'.
No fast lenses exist for M4/3. Lens speed is a relative measure: relative to format.
An ƒ/1.0 lens on M4/3 is about equivalent in DoF terms to ƒ/2.0 on full frame (depending on how one accounts for the aspect ratio difference). Hence most lenses for M4/3 are actually SLOW compared to APS-C or full frame. The flip side is that deep depth of field is more easily obtained.
ƒ/1.0 ~= ƒ/2.0
ƒ/1.4 ~= ƒ/2.8
25mm ƒ/1.0 ~= 50mm ƒ/2.0
12mm ƒ/2.0 ~= 24mm ƒ/4
24mm ƒ/1.4 ~= 50mm ƒ/2.8
50mm ƒ/2.0 ~= 100mm ƒ/4
75mm ƒ/1.8 ~= 150mm ƒ/3.5 “~=” means “approximately equals vs full frame”
Background blur possibilities with M4/3 are thus non-existent when compared to full-frame and there are very few options even with respect to APS-C. And the “look” of shorter focal lengths is less pleasing for some purposes, just a medium format has long been favored for full frame, or large format over smaller formats.
Take for example the full frame Sony RX1 with its 35mm f/2 lens. M4/3 could offer an equivalent which be smaller by perhaps 1/3, but it would need to offer a 17mm ƒ/1.0 lens. And if you’ve seen that Zeiss 35/2 lens on the RX1, I state bluntly that equivalent performance in a 17mm ƒ/1.0 on M4/3 is extremely unlikely.
Zeiss is offering Zeiss lenses for Sony NEX and Fuji X and presumably any other APS-C format that gains popularity. But as yet, not M4/3: to really exploit the M4/3 sensor size (keeping quality up and size/weight down), separate designs are required for the format. And yet the savings in lens size and weight will not scale down as much as the sensor size difference.
But the bottom line is the M4/3 has much more limited vendor support and hence its long term viability is in question: Nikon, Canon, Sony, Leica do not support M4/3. This is a very serious risk to the format.
Peter C writes:
I have always said - Olympus painted itself into a corner with 4/3. By designing their entire lens line for this size, it forever limited itself to this format. It cannot go bigger. Very foolish in my opinion. As a former OM system owner with quite a big investment, I was angered by their rejection of digital and abandonment of users like me.
DIGLLOYD: My statements above are not so much about megapixels as about total image quality and system value and options and the entire ecosystem. I do expect M4/3 to get to 32 megapixels or so, and it will probably cross a quality threshold in so doing as sensor technology evolves. But I don’t expect it to offer a superior form factor than APS-C except some (not that much really) advantage on lens size. And vendor support is also lagging.
Roy P writes:
Agree 100%. In fact, that was true two years ago – that’s why I went with a Sony NEX-5 instead of a M4/3 as an alternative back for my Leica M lenses. Since then, the progression to the 5N, 6 and 7 (or 7 and 6, to be more precise) have all been very impressive.
The rumored second version of the NEX-7 will likely surpass all but the highest end of the DSLR line ups!
DIGLLOYD: There are many things I really like about the E-M5 (super image stabilization, EVF + rear LCD, just fun to shoot in general), and several I really dislike (menus and buttons and high noise even at ISO 200 under some types of lighting).
Rumored cameras can have attributed to them magical properties. Will it have camera-like ergonomics or still feel toy-like as with current NEX?
But Sony is relentless and so NEX holds a lot of potential, as do spin-offs like the Sony RX1 and Sony RX100. And maybe Olympus will fill in the mid-range (imagine a Sony RX100 style camera with a M4/3 sensor now that Sony has invested in Olympus).
Steward Logie writes:
Micro four thirds is about lens size, not body size. Your examples quote single lens cameras and body only.
For low light I choose my full frame canon kit, but the Olympus OMD and its suite of nice fixed focal length lenses comes with me on travels.
DIGLLOYD: that’s why the APS-C Coolpix A body lens is no larger and probably smaller than even the Panasonic M4/3 pancake lenses. Ditto for Sigma DP Merrill, whose lenses are little different in size from comparable M4/3 lenses.
There really is not much difference in the size requirements for a high performing M4/3 lens versus a high performing APS-C lens. And M4/3 has not proven its potential at all: compared to full frame, lenses need to be 2 stops faster for equivalent DoF blur. Where are the ƒ/0.7 and ƒ/1.0 lenses in a variety of focal lengths?
APS-C ALLVIEW cameras will become relevant to me (and I suspect others) when the manufacturers start shipping weather-sealed bodies and lenses with in-body image stabilization. My EM-5 + 12–50 kit lens shrugged off sleet and volcanic ash driven by 50+ knot gusts during our last trip to Iceland. Try that with a Sigma DP Merrill!
The samples you provide online leave no doubt that the EM-5 trails the pack in terms of ultimate image quality. But I know that it's the camera that will bring home the most compelling images for my purposes. And I'd argue that the EM-5's sensor is good enough where previous m43 sensors were comically bad with obvious noise present even at the base ISO.
I'm looking forward to your continued coverage of this product space!
DIGLLOYD: weather sealing has never mattered to me in the outdoors, but I do avoid volcanic ash. As for “lenses with in-body” IS I presume that Andres means simply “in-body image stabilization” which is awesomely good with the E-M5 and a huge plus beccause any lens benefits.
I agree the E-M5 sensor is the first M4/3 sensor I would deem acceptable. But it shows its weakness nonetheless. Under some lighting conditions, even ISO 200 disappoints.
Markus H writes:
How can a RX100-sized Sony sensor be viable but a larger m43 Sony sensor not be viable?
Well, the answer is that there is no fixed-lens m43 camera (yet). Such a camera would slot in between the RX100 and the Ricoh GR and Coolpix A, both in IQ as in regard to size (of course a prime will normally be better and smaller than a zoom).
Or maybe the real answer is that m43's role as the most compact camera with an acceptable IQ is getting challenged. And while that is an important role for m43, it is not the only one. There are also the wide-focal-length-range, the high-IQ-lens-range, and the fast-ish-prime-lens all in comparatively small sizes roles.
DIGLLOYD: I never used the word “viable”. All of these cameras are way beyond viable.
The topic is where the future leads and whether M4/3 has a significant advantage over APS-C in reality. My hypothesis is that the theoretical advantages of M4/3 over APS-C are small, and even smaller in marketplace reality. And that this might spell the doom of M4/3 which already lags in multiple ways: slow lenses (for the format), no world-class lenses that are not also huge and heavy, no lens support from key vendors like Zeiss (not yet at least), aspect ratio of 4:3 instead of the commonplace 3:2, getting eaten from below by cameras like the Sony RX100, which deliver the pocketability and image quality for that type of use.
There are no fast lenses for M4/3, here M4/3 has failed in promise, see the discussion above.
The latest sensor technology (RX100) paired with a fixed lens that is designed optimally for the sensor can do wonders (though the RX100 lens is pretty weak by the poor choice of being a zoom). But the main thing is that the sensor in the RX100 is really good.
Were the Olympus E-M5 to use the same sensor technologyas the RX100 scaled up to M4/3, it would be ~26 megapixels and make M4/3 much more attractive. But most of the lenses are not up to this demands, which would require new lenses of nearly the size of APS-C lenses.
Richard J writes:
In this argument that arises between sensor sizes, the prevailing comparison always seems to be the image quality of the sensor and that you can get more detail and less noise out of a larger sensor. Although this is not an inaccurate statement I believe people are missing an equal benefit if not, in my mind, a greater benefit to having a larger sensor over a smaller one.
This benefit is not about how well the sensor sees a scene but how much of that scene it can see. A M4/3 size chip is just not able to see as much of a scene as is a 4x5 camera, at least not without taking a wide angle lens and bending the light as much as possible to fit that whole scene on to the smaller chip. Also a 4x5 size is able to show a greater level of 3 dimensionality then a smaller chip, due in part to a narrower depth of field. I have wanted to do a visual demonstration of this to show the same scene on a M4/3 camera thru 35mm, medium format and 4x5 with both the same focal length and then with the equivalent focal length, however I don't have great access to the larger formats.
I think this would be a great comparison for you to do and add it to the DAP (which I am now a very pleased member) as I think it's as valuable a comparison between the formats as image quality is. I would love to see the results myself as I have always put how an image feels and the look of the image before outright image quality and I think doing this test would demonstrate that greatly. And I hope that it might go towards slowing the increases of smaller sensors and promoted a renewed desire for even larger sensors then we have now, even if I doubt I could afford a 4x5 digital back.
DIGLLOYD: As stated these claims appear to be erroneous, unless I misunderstand them.
It is M4/3 that has a much wider range of focal lengths, and 4X5 lenses also have to “bend” the light in order to cover a very large format— this is very hard to do at high quality and with a fast aperture using a very large chunk of glass (area, grinding precision, etc), hence we see ƒ/5.6 or ƒ/8 or ƒ/9 or even ƒ/11 view camera lenses. And of course, 4X5 must be stopped down for more for equivalent depth of field.
Depth of Field (DoF) is certainly something that delivers a very different look across formats. A 4X5 camera lens needs 4-5 stops more stopping down for equivalent depth of field (e.g. M4/3 at ƒ/1.4 ~= ƒ/8 on 4X5), but the subject perspective is wildly different (same idea with full frame versus medium format), and this “look” is the quality attractive to some shooters. M4/3 completely loses this look, APS-C mutes it.
Bill B writes:
This sensor / film size / flavor of the day has been going on in the photographic industry for decades. In the 60's 35mm became the standard camera for general use. Kodak started the 126 cartridge in order to expand the ease of photography to the masses. Loading 35mm was a challenge but 126 was easy. Then the 110 format in order to make the cameras smaller and lighter. Keeping in mind that Kodak was the mover of all these formats in order sell more film. New cameras, back when Kodak actually made the camera, meant more Kodak product in the hands of the consumer. Photofinishing labs had to buy new machines and new paper sizes. Inevitably the paper was designed to help optimize what the film was able, or not able, to resolve. When things really started to stagnate in the '80's a consortium of companies (Kodak working with smaller point and shoot camera companies who had suffered the reduced popularity of 110) developed the Disc camera. Notice the trend to smaller and smaller formats. And lets not forget that some of the leaps in optical technology were developed on those lowly plastic cameras. The aspherical plastic "poured" (composite) lens on the Disc camera was ground breaking.
With 4/3 it was easy to see what was going on with Sony, Olympus, Samsung, Panasonic, Ricoh and others. Develop another format which will allow the sales of cameras and accessories. By making the 4/3 mount "universal" the exposure to having to produce a huge series of lenses is reduced and manufacturers can focus on what they can do best, market and build boxes for sensors. This new sensor approach permits greater margins, which is of course the end goal, as smaller sensors do not cost as much just as smaller pieces of film cost less. The price to the consumer allows for more profit and if the "pro" 35mm format increases in price then just keeping the perception of price value on the smaller format is easy. (Manufacturers can keep the price margin on full frame sensors high and that perceived value on the smaller system seems like a bargain).
Change body styles to keep the marketing alive and push the body (and/or new sensor) upgrade path.
Once again the industry looks for ways to reinvent sales but it all comes back around. Back in "the day", Pentax, not satisfied with the smaller formats, created the Sport 35 full 35 mm point and shoot and Ricoh & Minolta followed immediately while Canon and Nikon 35 p&s took a while. Now you could have the fantastic image quality of 35mm with the vast assortment of films (slide, negative, B&W, pro, consumer) in an autofocus point and shoot. In the digital age we have micro 4/3rds, evolving to APS-C evolving back to 35. Guaranteed someone is going to come out with an interim APS-C / 35 middle size to market more new product. Sony as a sensor company might just be the Kodak of the 21st century driving camera changes to sell sensor formats. Judging by past history that might just happen again.
DIGLLOYD: Agreed, this is a natural evolution of the marketplace, with consumers picking the winners and loser. It is not image quality by itself, but image quality and the entire camera experience and industry support (e.g. lens options and variations on form factor). But as the other factors converge, the pressures mount on M4/3 and the differences shrink. That is the problem for M4/3.
Smaller film was a failure for good reasons: it was awful for anything but a snapsnot. M4/3 is not a failure for that reason; it is not a failure in any technical sense; it is merely a question of medium and long term market viability. M4/3 offers quality well beyond that necessary for most consumer use and is arguably superior for over what most consumers could ever get out of 35mm film.
The basic problem with going smaller with a digital sensor is that the fixed costs remain relatively inflexible: the body, the EVF, the LCD, the card slot, the tripod screw hole, the buttons and dials, the CPU and so on all are relative fixed costs regardless of sensor size.
Which leaves the sensor and optics as cost issues. Sensors are semiconductors and keep getting better and cheaper for any particular size sensor. What happens when a full-frame sensor approaches the cost of APS-C or even M4/3 does not? That cost has shrunk, but the other costs have not, so what is the point of the smaller sensor option?
Lenses also have fixed costs; an M4/3 lens doesn’t really cost much less than an APS-C lens: witness the quite high prices of Olympus M4/3 lenses. These costs become quite high for high quality M4/3 systems.
When all this is added up, the costs are largely fixed with only modest variation: witness the price of the Olympus OM-D E-M5 and assorted lenses; compare that to the APS-C alternatives. The truth is that M4/3 is not winning on price. Not on the camera body and not on the lenses.
With the sensor price on an inexorable dowward trajectory, an APS-C or full-frame sensor used in a mirrorless camera does not incur the expensive optical viewfinder and hence the total build cost asymptotically approaches that of smaller format cameras.
Hendrik M writes:
As a backup system for my M9, I use Olympus EPL5.
For people portrait (1 person), i used M9 for ultimate separation between subject and background.
As for the m4/3 system, i use it for:
1. group shots where i don't have to stop down too much so i can use lower power flash indoor.
2. outdoor environmental/landscape shots
3. travel, because of the low weight of lenses (14-42mm + 25mm summilux)
DIGLLOYD: the right or preferred tool for the right job.