Mirrorless… the one design to rule them all. The master of the full frame is undoubtedly Sony, however has it inadvertently introduced a short-lived shelf life with some in-built obsolescence?
Sony's Slow Burn
Sony, that behemoth of consumer manufacturers who brought us iconic designs such as the Walkman, have a strange camera heritage which includes the 1981 Mavica, the first digital stills video camera. For a company so wedded to consumers, it had long championed video cameras (anyone remember the Betamovie BMC-100P?) but was strangely silent (barring the Mavica range) when it came to stills. It wasn't until it acquired Minolta in 2006 that it's stills camera division was truly born. Minolta manufactured both cameras and lenses and was arguably at the technical leading edge from the 1970s onwards. In this sense they were a good match for Sony, although were late to the digital party, releasing their first DSLR (Maxxum 5D) in 2005 which used the 1980s A-mount for all its lenses. This is Sony's "standard" DSLR mount and can still be found on a few of its SLT cameras.
Sony followed the pivot to mirrorless that was led by Olympus and Panasonic in 2008 with their release of the Micro Four Thirds format. It's worth bearing in mind that this was an exciting time for mirrorless development and the requirement for new lens mounts. Sony released their first mirrorless camera in 2010 (NEX-3; APS-C), along with Nikon in 2011 (Nikon 1 V1; CX), Canon in 2012 (EOS M; APS-C), Fuji in 2012 (X-Pro1; APS-C), and Leica in 2014 (Leica T Typ701; FF). It's also worth remember that the Alpha 7 arrived in November 2013.
Has there ever been such a gold rush of lens mount announcements before or since? It makes me wonder about the nature of shifts in the camera industry. Consumer electronics has increasingly seen short iteration cycles of rapid development and product replacement. This is perhaps counter to traditional camera manufacturing, particularly for lenses which are intended for longevity. As a result, the market now operates on the basis of a rapid turnover of camera bodies and this may well be driving revenue. So why was there a sudden move to new mounts? Was the commercial success of Panasonic and Olympus a key reason or did the other manufacturers already have plans in the pipeline? Was it a case of herd mentality, with each manufacturer offering their own vision of what a mirrorless mount should offer? Remember that this was a period of rapid growth and large sales volumes, which almost certainly allowed for generous R&D budgets.
Sony, Nikon, and Canon were all relatively quick to market with new mounts and lens lineups, yet these were clearly consumer products. In particular, Nikon and Canon were clear that they didn't want to cannibalize their professional DSLR sales and that drove their mentality. It wasn't until Fuji's X-series and Sony's Alpha 7 that higher end mirrorless cameras came to market.
There are two primary design parameters for any new mount: mount-to-sensor distance (flange distance) and mount diameter. The flange distance is a result of the thickness of the camera which has to house all mechanical components. Principle among these is the viewfinder: in analogue cameras, manufacturers have used twin lens reflex and rangefinder designs, but settled upon the single-lens reflex. To accommodate the mirrorbox, the camera needed to be relatively thick, making the flange distance large: the Nikon F-mount (below) is 46.5 mm. By going mirrorless with an EVF, Nikon have reduced this to 16 mm with the Z-mount, whilst still incorporating an IBIS sensor.
The mount diameter needs to be at least as large as the sensor so that light can enter parallel to the optical axis. Leica's legendary M-mount is 40 mm which gives 2 mm of leeway on either side (it's also worth noting that the flange distance was a svelte 27.8 mm due to the rangefinder design). Making the mount diameter larger gives two benefits. Firstly, more light can enter the lens meaning, within design constraints, faster lenses can be manufactured. This is measured as the incidence angle, the angular deviation from the optical axis.The larger the value the better, although sensor microlenses become less efficient at extreme angles which can lead to vignetting.
Secondly, large mounts (by definition) result in larger orifices allowing significantly more space for optics, meaning manufacturers have greater latitude in their lens designs. The trade-off of bigger mounts is bigger (and heavier) lenses.
The combination of a larger mount diameter and shorter flange distance produces the largest incidence angles, but does any of this really matter? Well yes, because in the search for ever more exotic glass, manufacturers ultimately run up against optical physics and the constraints they impose. The flip side is that you can use the in-built advantage of a larger mount with more light, to produce simpler and cheaper lenses.
So how do the different mounts stack up? It's worth reviewing incidence angles for traditional DSLR mounts: the Nikon F is 12.14° (44/46.5 mm; mount diameter/flange distance), whilst the Canon EF is 16.8° (50.6/44 mm). This is one of the reasons why Canon shooters have often had a good lens selection. It also goes to show the constraints that the SLR specification places upon lens design.
So what about the new mounts? Here Nikon makes great play of a 41.2° (52/16 mm) incidence angle, in comparison to Canon's RF at 33.62° (50.6/20 mm). As a comparator, the MFT format runs at 32.5° (38/19.25 mm) which is perhaps what originally made other manufacturers sit up and listen: a significantly smaller body, significantly smaller lenses, yet relatively fast glass (not withstanding the effect of the crop factor).
And what about the other mounts? The Leica L is 33.13° (48.8/19 mm) and Fuji X is 35.3° (40.7/17.7 mm). Which brings us back to Sony, the first to the punch with a full frame mirrorless mount which has an incidence angle of 28.6° (43.6/18mm). Except it wasn't designed for full frame mirrorless but its NEX series of APS-C format cameras and camcorders. This raises the question as to the design motivations of Sony when it first developed the E-mount. Was it to terminate the Minolta product line and split their offerings in to an APS-C mirrorless range that could support both stills cameras and camcorders, along with a "professional" A-mount SLT line? If you calculate the incidence angle of the E-mount for an APS-C sensor then it is a much more forward looking 37.9°.
So did Sony intend to produce a full frame mirrorless camera or was it a result of the success of its NEX range and some Sony labs product experimentation? Was there always a plan to disrupt the DSLR market with mirrorless or did they fortuitously expand at a time of DSLR decline? Certainly, pursuing the mirrorless full frame format when camera sales were so high seems a risky strategy in comparison to Nikon and Canon, yet the inroads they have made in to the market have subsequently been significant.
Of course the performance of the E mount is still significantly better than comparable DSLRs and back in 2013 gave them the flexibility to produce competitive quality lenses. However that advantage has now flipped back to Canon and Nikon (and to a certain extent the L-Mount Alliance) with the release of the RF and Z mounts which are designed from the ground up for full frame mirrorless, with Nikon arguably the most aggressive in its pursuit of competitive advantage. Canon is also no stranger to mount development: over the period of Nikon's F-mount, they have had the R, FL, FD, and EF mounts which didn't seem to hinder the success of the brand.
Has Nikon built a lens mount that provides users with the option of exotic glass (such as the 58 mm f/0.95 Noct), whilst giving the flexibility for a wide range of affordable lens designs from third party suppliers? Or has it encumbered itself with another great design that will simply be superseded by another manufacturer in a few years time? Has Sony shot itself in the foot by sticking with the E mount and relinquishing the mirrorless advantage it once had? The willingness of photographers to change brands suggest the "stickiness" of remaining with one manufacturer may be short lived. Or will we see Sony, and then Canon, iterate over mount designs to constantly improve their optical offerings?
Do you think Sony should change mounts?
Lead image courtesy Rainer Knäpper (Free Art License) via Wikimedia. Body image courtesy of emdx and Nebrot via Wikimedia. All used under Creative Commons.
If mount diameter and flange distance are so important, then someone needs to explain why the e-mount has so many class leading lenses.
This is like saying "if [insert any obstacle] matters so much then why is [insert whoever or whatever] so successful?"
It is 100% factual if you understand optics that a shorter flange and larger throat are better (in this case, it's the larger throat that matters, the flanges are very similar). Sony has worked around it to make some stellar lenses, but that doesn't change the science of optics.
Both things can be true: the E mount provides obstacles and Sony overcomes those to produce some great lenses.
... and the larger throat has tradeoffs as mentioned in the article, so each parameter can be optimized for a variety of goals, including size and weight of the final system in-hand.
Are you suggesting the opening should be the same size as the sensor, plus the anticipated dimension of whatever holds the lens elements in place?
"... The lens mount opening should be bigger than the sensor ..."
Is that so hard to say?
Thanks for exploring this.
They are also the most expensive for the most part. Not saying that’s the only reason, but I know it’s a factor somewhere.
Not when compared to Canon or Nikon mirrorless lenses. And really, they are only $100-$200 more for the comparative lenses, but offer a bunch of newer and better features.
Idk man. If you compare the G-Masters to L glass or Nikons top lenses, Sony is consistently more expensive. I’m a Canon to Sony convert, so I’m not one of these guys out here dogging on Sony cuz it’s not my brand. It’s more like $100-$300 difference, and as a photographer or filmmaker, you don’t just have one lens. So even if you have a basic 3 lens Sony kit, it’s gonna cost you $300-$900 more than the alternative.
Also, if you are going to make a point about Sony lenses being $100-$300 more than EF mount, but not bring up the fact that RF is $100-$300 more than Sony, that seems like you want your cake and to eat it too.
Well they all have IS... I’m referring to their trinity.
Also, no need to get defensive. Just trying to have a conversation and give some of my thoughts.
Also, not sure what you’re referring to as far as “a bunch of newer and better features.” They’re optically awesome (and often superior) just because of the build quality, but I don’t know if any features they have that other lenses don’t. The Canon RF 24-70 is $100 more than the Sony G, but it also has incredible IS where the Sony doesn’t.
Why would Sony put IS in the lens when they already have it in the body, though?
Lol. This seems like a silly statement considering how many camera manufacturers make both IBIS camera and IS lenses.
1. Sony Alpha
2. Nikon Z
6. Canon’s rumored R cameras
Really the only company that can stand alone with IBIS without making IS lenses is Olympus which has arguably the best stabilization ever fitted to an interchangeable lens system. And even then, you can use Panasonic lenses which do sometimes have IS.
Actually, on some of their newer lenses, Olympus also incorporates IS which can be used in combination with their IBIS for truly amazing results.
Also, I believe it’s pretty widely accepted that Sony’s full frame IBIS is pretty mediocre. I tend to agree. When I use my Sony 24-105 with IS, it’s noticeably more stable at similar focal lengths with other non-IS lenses.
A few reasons:
1. Not all E mount bodies have IBIS
2. ILIS can theoretically provide better stabilization for longer lenses
3. Cooperative stabilization between IBIS and ILIS works very well for all kinds of movement
Off the top of my head... both the 90mm and 135mm have two focusing groups for faster auto focus. The 24mm G Master has a specially designed element to reduce the amount of CA when wide open which is a huge bonus for astro photographers. In general the focus motors used by Sony are designed to be as close to silent as possible (not a big factor for photo, but critical in video), Sony lenses have "stepless" iris capability (again not critical in photo, but good for video).
I guess I’m just comparing to the trinity of lenses from the DSLR ranges.
Make that "... where the Sony does not NEED in-lens anti shake ..."
No, you have it backwards. You're still stuck in pre Canon/Nikon mirrorless days. Fast forward to today, Sony in general is about $100 less.
Ah. Good point.
To be fair, Sony's lenses have been around for some time while Nikon / Canon are still riding on the early adopter curve (within those who are loyal t their Brands). I do expect the prices there to drop slowly by 15-30%).
How slowly? By the time a newer version comes out? If we took the $2699 lens as an example at 15%/30% off, that's $2294 (-$405) and $1889 (-$810) respectively. I find it not likely it'll retail that low unless it's being replaced.
Low 2000 is possible. I would hope. But only Gawd knows
Class leading? LOL.
... not sure about the Emount becoming obsolete... but high paid 'pro" photographers are...
I disagree - you obviously have no idea what you are talking about. As someone who used to hire pros for major project work, I can say for sure that market isn't going anywhere. Indeed it is quite the opposite. More images are needed and there's general lack of quality photographers that can deliver consistently.
Don't confuse say, wedding, with actual professional work. What is dropping away are the "man with a camera" type of work. Indeed if you consider those pro then you are right. High end Fashion, Product, Architecture, Advertising photography isn't going anywhere and requires a lot of experience no matter how good the gear becomes, because it goes far beyond just "technical skill" - although all of these genres require far more technical skills than a, say, wedding shoot.
The amount photographers are paid today what is driving so many away. Major journals are dying by the wayside and offer pennies on the dollar for quality work. The few papers that survive have drastically cut their photography budgets.
I would add, to be more specific, that a larger mount (please never call it an orifice again) allows for better telecentricity and the potential for smaller lenses as they don't require rear correction elements. I say potential because lens size is always a trade off between IQ, price, and size. We'll see how small the Nikon 28 and 40 pancakes end up being. I think Nikon has leaned toward ultimate image quality (at least based on some of their lenses so far, like the 50 and 85) while keeping the size in a sort of middle ground - though some of their lenses like the 24-70/4 and 14-30/4 are excellent and ridiculously compact. Much more so than similar lenses from Sony.
As for the E mount, I mean, absolutely no they should not change it at this point. Would have been best to do that with the original a7/a7R releases but absolutely not now. They've shown they can make some amazing lenses - just because the mount probably makes that more difficult doesn't mean it's impossible. They're all in now, so it's a moot point for us consumers.
e mount has all the ultra wide aperture lenses you could want 35mm 1.2 from sigma, 3-4 different MF 50mm 1.0 options etc.
Mitakon has a couple of f0.95
"Making the mount diameter larger gives two benefits. Firstly more light can enter the lens meaning, within design constraints, faster lenses can be manufactured"
I guess you missed those lenses with larger front diameters than rear diameters. That's a pretty fundamental error for someone presenting themself as a subject matter expert.
Yes, the way the author said this is not correct/not clear. The larger mount makes it *easier* to manufacturer fast lenses. A bigger exit pupil will do that. But it has nothing to do with light entering the lens or whatever, otherwise lenses like a 135/1.8 or 105/1.4 or any of the f/2.8 or f/4 long tele lenses wouldn't be possible.
Sony are doing fine and this story, while an interesting theoretical story, is just a story. Canikon will love it but for the majority of the people, there isn't probably much relevance.É
The hate for Sony under the umbrella of the news and speculation community never ends does it? I'm sorry, but No, no it is not.
They are putting in an awful lot of R&D and adopting the mount onto their high end video cameras just for them to let it go obsolete....
That should tell you plenty about whether they are finding it hard to design for...as should the fact that Zeiss, Sigma, Tamron, Tokina, Mitakon, Voightlander are all in the E mount lens making game...
In the short term adapters will fill the gap between 'A' body and 'B' glass. The body has become the least element of the equation. All the body has to do hold the lens mount/adapter, IBS, chip and menu system. And for now adapters are KING. The body is all about software and that changes; just look at Nikon improvements in software for the Z mount body. Canon not so much Sony not so much. Who is learning faster?
Sony has and has had the best AF in the camera world since the release of the A9...they improve that on the A9, A7III and A7r3 with a massive software update last year ...and then bested even that with the A9II and A7r4... Nikon has been trying to catch up even before all that.... from the testing I’ve seen they are close now, but still not all the way there...
E isn’t perfect, but it would do just fine for the next decade or so before some new tech would completely overhaul camera design. Lets consider the possibilities with “flat lenses” for example.
Monjurul Meem, Sourangsu Banerji, Apratim Majumder, Fernando Guevara Vasquez, Berardi Sensale-Rodriguez, Rajesh Menon. Broadband lightweight flat lenses for long-wave infrared imaging. Proceedings of the National Academy of Sciences, 2019; 201908447 DOI: 10.1073/pnas.1908447116
The E-Mount is fine.
Isnt technology reaching the limits of pixel density on camera sensors? If so, the only way forward is to use bigger sensors, like Fuji did.
Not yet. Plus, increased pixel density is not the only way to improve IQ... it's not even the only way to improve resolution.
Right... Sony’s bigger sensors...like Fuji did.... so, I’d assume if Sony felt it was reaching the end of viability of FF or the Emount they would do what they felt that had to.... but IQ at FF is very good and they have full visibility into the sensor tech roadmap...and so maybe the don’t need to do anything but wait for the next generation of FF sensor to finish development ....
What about Leica lenses? They are wonderful, especially given the small lens mount.
Camera makers do seem to believe that the market is going upscale and that in particular, serious photographers will buy very large, very expensive lenses. If that's really happening and isn't just online hype, then maybe Nikon Z mount is looking good right now.
This is just plain silly.