Roger’s Rants: My Canon/Nikon Mirrorless Camera Unfanboy Opinion

I’ve gotten about 632 texts and emails over the last couple of weeks. About half of them are “I’m a Nikon / Canon shooter. Should I preorder a Z / RF camera?” My answer is generally no. The reason is simple; unless you just have to play with the new technology and have money to burn, wait until the 2nd generation when the prices are way lower, some of the bugs are worked out, and more native lenses are available. (Yes, I’m aware you can shoot lenses on an adapter. You can also shoot lenses on the SLR you already have.)

The other half of the questions are “Who has the best mirrorless camera.” My answer is generally I refuse to play fanboy wars, which are typically nasty discussions between people who are already committed to a brand and people who have never tried the brand. But, of course, at this moment in time, Sony has the best mirrorless full-frame cameras. They should, they’ve released about a dozen of them, while Canon and Nikon have not quite released 1 and 2 respectively.

A few people, though, knowing I rarely recommend Generation 1 technology and don’t wallow in the fanboy cesspool, have asked “So where do you think this is going.” That’s a good question. And following my usual ‘often wrong but never in doubt‘ philosophy of life, I’m willing to speculate.

Cameras and Electronics

Sony has a much more mature technology which gives them a lead, of course. They also are a huge electronics company and sensor manufacturer, so I think it’s reasonable to believe they’ll keep that lead for the next couple of years. Then again, being Sony, they have a menu that is best described as ‘you get used to it.’ I think Canon and Nikon did really well as far as first generation ergonomics, especially if the goal was to keep their own customers from migrating over to Sony.

Both Canon and Nikon did one thing I hadn’t really expected, and the more I’ve researched it, the more important I think it will be. They went with very wide mounts (54mm for the Canon R, 55mm for the Nikon Z6/7). This is an especially big jump for Nikon (the F-mount was only 44mm) and wider than Sony’s 46.1mm E mount. Why does this matter? Because of optics.

Optical Differences

A wider mount allows lens designers more freedom. Wider aperture lenses are possible. Mount diameter is one reason Canon had f/1.2 lenses, and Nikon hasn’t, for example. Wider lens mounts also allow lens designers more freedom to correct aberrations and do all kinds of cute optical things. I’m sure the designers at both firms are salivating at the fun they’re going to have.

One thing to always remember, though, is lens design is still a compromise. The graphic from Canon’s white paper on the new mount shows it perfectly.

Thorpe, L: A New Lens-Camera System. Canon White Paper. https://downloads.canon.com/nw/camera/misc-pages/eos-r/pdf/canon_eos_r_white_paper.pdf

Designers can reduce the size and weight of a given lens, improve the optical performance, or increase aperture on the new mounts. To a lesser degree, they can do two out of three, and perhaps to a small degree all three in a given lens.

What Will This Mean?

At this moment in time, Sony has a much larger native-mount lens selection. They have also demonstrated the ability to release lenses at a very rapid pace and will have more native-mount lenses for years to come. Canon certainly has the resources to catch Sony if they want but given Canon’s conservative nature and dominant SLR position; I don’t expect that (but remember, I’m often wrong). I don’t think Nikon has the resources to do so for two reasons. First, Nikon has, in recent history, released new lenses at a slower pace. Second, Nikon has downsized significantly, and this has included lens designers. I meet a lot of designers and engineers from a lot of companies and ‘formerly at Nikon’ seems to be part of the introduction most of the time.

But both Nikon and Canon designers will have a significant advantage to work with going forward. Sony, Canon, and Nikon all make some excellent lenses. Going forward I think Canon and Nikon will have the opportunity to perhaps make ‘more exceptionaller’ lenses.

However, when we discuss optics and imaging we have to address the pink elephant in the room; image manipulation. Obviously, in-camera jpgs are strongly influenced by in-camera processing, but more and more we see evidence that RAW files are also manipulated in-camera. Electronic correction of optical aberrations may make optical differences in lenses less apparent, although it will never eliminate it.

What About All That Other Stuff?

Most of that I find rather inconsequential, although it’s obviously life-and-death to many Fanboys. One has in-body stabilization; the other doesn’t. One has probably better focusing than the other. I can’t imagine anyone is going to change from Nikon to Canon based on the mirrorless system.

A lot of people will buy their brand’s 1st Generation mirrorless cameras and use an adapter. Personally, I think Canon’s Control Ring is the most interesting thing I see for right now, and making it available on an adapter was brilliant. Lots of people won’t use it. I will, though, it seems very usable and intuitive to me.

But these first-generation cameras feature-for-feature probably aren’t as good as Sony’s multiple current offerings. Both will be more competitive in a year or two, although I suspect Sony’s cameras will have some advantages still. They’ll certainly be good enough for job 1, which is to slow the migration away from their own brand over to Sony.

My own opinion is eventually (3-5 years) mirrorless will be a significant portion of both Nikon and Canon’s business and the lenses, more than the cameras, will be the driving force. The early lens releases probably give us a hint of how each manufacturer plans to go forward.

Nikon Mirrorless Rentals

Nikon started with three very practical native-mount lenses; a 24-70mm f/4 and 50mm f/1.8 and 35mm f/1.8 primes. Given the optical triad of performance, size, and specifications, Nikon seems to be leaning towards the compromise of somewhat smaller and somewhat better optically while fleshing out a practical, useful lineup. Their roadmap looks to emphasize useful and practical, but there is the spectacularly dramatic 58mm f0.95 on deck.

Canon Mirrorless Rentals

Canon started with three show-off lenses (and I don’t mean that in a bad way; I love optical show-offing). The 50mm f/1.2 and 28-70mm f/2.0 are all about amazing optics and big apertures, but they are huge beasts. The 35mm f/1.8 Macro is a bit smaller, a bit wider aperture, good optics, and a Macro feature (not that I’ve ever found 35mm full-frame macro lenses particularly useful), while the 24-105 IS is practical. Canon hasn’t released a roadmap but has said they are working on a series of fast lenses of f/2.8 or more, so I suspect some workhorse zooms are upcoming.

Logic suggests that from a pure optics standpoint upcoming Canon RF and Nikon Z lenses may be better than Sony’s, although it will be years before they have a similarly broad lineup. Logic also suggests Canon RF and Nikon Z lenses will be better than their SLR lenses; perhaps more so for Nikon who has a much larger mount now.

For both companies, the adapted lens route certainly makes it practical to dip a toe into their mirrorless offerings, although for me it will be another generation before I do. It won’t be a long wait for those with patience. Look at how much the Sony A7 series has improved over basically three years. Then again, Sony won’t be standing still over the next couple of years, either.

Interesting times.


Roger Cicala


September, 2018


You can preorder the Canon R and Nikon Z6/Z7 now.


Author: Roger Cicala

I’m Roger and I am the founder of Lensrentals.com. Hailed as one of the optic nerds here, I enjoy shooting collimated light through 30X microscope objectives in my spare time. When I do take real pictures I like using something different: a Medium format, or Pentax K1, or a Sony RX1R.

Posted in Equipment
  • JarnoP

    A very good article. I am very happy to see both Nikon and Canon entering the FF MILC market and saw it more a “system introduction” rather than “camera launch”. I do not know is Nikon’s stopped down focusing a limitation of the current cameras, lenses or the system. Canon did great in many aspects although it let the door open bit too much for Sony fanboys to cry. As if they would buy Canon in any case.

  • bdbender4

    Son of a gun, a discussion between Thom and Roger. You are the two that I look to for insight into the camera market amidst the hurricane of on-line drivel.

    Personally I cleaned house of my accumulated stuff earlier this year (old Nikon lenses, Canon EF setup, Fuji X setup) and kept only a Canon M5 and a few lenses. I was hoping for the rumored M5 Mark II, which still seems to be out there, but now have no confidence in the future of EF-M. Thom’s article on the current plethora of mounts raises the EF-M question, too.

    So, being in a place to start fresh, I have been unable to resist GAS as common sense would dictate, and pre-ordered a Nikon Z6 with the 24-70. Among other things it is a bit less bulky than the equivalent Canon R with 24-105 – and $800 less expensive (!?). We shall see…

  • ching

    I really like your teardown posts. Would you be able to do a teardown or any objective testing of the Profoto B10?

  • ching

    zeiss loxia lenses are substantially smaller than dslr equivalent (milvus).
    12-24 is smaller
    16-35 is lighter
    15mm voigt is smaller
    12mm voigt is smaller
    10mm voigt is smaller

    My belief is that lenses are big because they can be.

  • ching

    To answer his question, the microlenses on the outer perimeter of the sensor are angled inwards, on a Leica sensor, to mimic the behaviour of film. That is the primary reason they’re superior.

    “A wider mount allows lens designers more freedom.”

    I always thought the Sony mount is already bigger than required for full-frame lenses. I’ve read this comment a few times. Do you have any data on this blog to back it up? I am not trying to pick a fight with you, but I’ve read this claim over and over and I’ve yet to see any evidence.

  • Eamon Hickey

    Yes, maybe an A6500 successor will tell us something. I guess I’m not sure what Sony would gain by using a version of the X-T3’s sensor with stacked architecture? The X-T3 already does 20fps with AF and no blackout (30fps with a slight crop), plus 4k 60P video. (I have an X-T3 for testing right now, and all of this works quite well.)

    So Sony could presumably beat those numbers with a stacked sensor, but do they really think that would move very many buyers? And of course, there would be a big cost penalty. A stacked sensor is essentially two sensors sandwiched together. I’m sure that does not double the cost of the device, but it certainly raises it significantly. So I’m not sure I’m seeing a value proposition in this scenario.

  • Derek W

    It’s not just Leica, Voigtlander also just released a 50mm 1.2 for the Leica M-mount that is small and weighs less than 350g. A couple things Roger didn’t mention that are important:

    1. The lenses are manual focus only, so no space needed for electronics or motors.
    2. The minimum focus on most of these rangefinder lenses is 70cm or 2.3′ which is a smaller operating range than most lenses which probably makes lens design easier.

  • Thom Hogan

    It’s a relative thing. In CES the difference between a 35% GPM and a 40% GPM is meaningful.

    But to answer your question, the Nikon DL decision is pretty illustrative of how tight things are. Nikon cancelled the DLs because they fell under their margin/demand curves. That’s the thing about the RX100, for instance. The models that are selling best are the older ones at discount, not the new one at above US$1000.

    This is the tricky play that Nikon eventually decided against continuing: as overall demand keeps going down, high prices also push that demand further down (or you can take less margin). You can leave older generation gear in the market and try to milk out the tail, but I’m pretty sure that’s turned unprofitable for a couple of the Sony products now. It did for a couple of Nikon products and Olympus and others.

    Sony Semiconductor has “some sort of agreement” with Fujifilm, Nikon, Olympus, Panasonic, Pentax, and Sony imaging groups. It may even have one with Canon for the 1″ and smaller sensors. It seems very clear that every one of those companies is driving different technologies at their core. I’ve reported, for instance, that Sony Semiconductor warned Nikon about potential yield issues with the 45mp sensor and offered the 42mp sensor only seen in the A7Rm3 as an alternative. Likewise, the 36mp sensor used originally only by Nikon eventually found its way into Pentax and Sony products. There’s a ton of pollination going on with the camera companies buzzing about the flower at the center, Sony Semiconductor.

    And no, I don’t think Sony Imaging would incur a loss if they “had to pay for everything themselves.” Because of the way those companies are established, they actually DO have to pay for everything. Failure to do so and not reporting it in their public financials would be a huge fraud liability.

  • Adam Fo

    Roger, when you get the chance measure the diameter of the R 50mm f1.2 front element. It looks smaller than the EF equivalent.

  • Adam Fo

    Many fast lenses exhibit focus shift when stopped down. The increased depth of field doesn’t cover that shift in many instances. An SLR is focusing with the lens open.
    The Canon R is focusing stopped down up to f5.6 making focusing more accurate on fast telephotos.

  • Adam Fo

    No, it’s almost twice as fast. The 1/3 f-stop chart is:


  • Adam Fo

    Nikon still make the 50mm f1.2 Ais

  • Athanasius Kirchner

    We’ll see. I’m operating on the basis of the rumored A6500 successor, which would equip the same 26MP chip, except stacked. If it pans out, it’d be a strong confirmation of SI’s exclusivity regarding that technology.

  • Athanasius Kirchner

    Why do you assume that the margins on the 1” cameras are ‘tight’? Do you have a source for that? The model proliferation alone should be a good indicator of the profits made on those cameras, and I personally doubt that the margins are slim (especially since they don’t move much volume).
    Sony Semi doesn’t play favorites (I mean, the X-T3 just got their newest APS-C sensor earlier), but it’s evident that they do have some sort of agreement with Imaging. Developing sensors like the ones on the A7RM2 and A7S isn’t cheap, and those two cameras represent a tiny amount of overall sales; at their current prices, Sony Imaging would probably incur a loss if they had to pay for everything themselves.

  • Benz Oberst

    While hitting the education field sounds like a fine idea, I wonder if you’ve ever seen what a camera that came out of a photography class looks like. You need products with Craftsman ruggedness and Craftsman-type service to survive there.

  • Benz Oberst

    That would have a side benefit of being able to make lighter body and still maintain weight balance.

  • Benz Oberst

    Umm, who do you think buys 99.95% of these cameras?

  • CameraCrazy

    Both new cameras are garbage with 1 card slot. Redundancy is GOD, and no serious pro is buying EITHER except as a weekender / vacation TOY. the Z7 is a D5000-ish mirrorless camera with poor interface and inaccurate AF

  • TurtleCat

    A lot of people aren’t interested in facts. They just want to believe they picked the best horse.

  • Brandon Dube

    I think the slowness of Canon and Nikon to “real” (FF) mirrorless has less to do with milking DSLRs and more to do with aversion to failure. I doubt they were waiting because they were betting against mirrorless;
    they both witnessed firsthand the death of Kodak caused by betting against the new technology. When is the last time you saw Canon or Nikon release a lens that breaks in half in peoples’ bags? Or a camera that overheats when you shoot 6 minutes of video? A serious camera that gets 300 shots of battery life?

    There are hiccups to be sure (D600, 24-70/2.8L II coatings, etc) but so rarely failures. My intuition is that Canon is in a pivotal time where they are bringing fully automatic lens (and camera?) production online in more than “artisanal, small batch prototype” volumes and that their R&D budget, which is centralized among their business units, was heavily devoted to production engineering. They were also expanding into cinema, which would monopolize optical and opto-mechanical designer time, as well as refining DPAF. In addition, their digic CPUs probably are not as good as what Sony can do, so they had to wait some time for more processing power / efficiency, and likely battery technology too.

    So now in late 2018, the time is right where they can economically mass manufacture these optics and cameras, get decent battery life, release lens designs at a steady clip, and have good autofocus instead of releasing a more half-baked camera.

    Canon seems comparatively bad at electronics, so they may have been more hamstrung on that side of things. Nikon seems comparatively bad at robust, high volume production or optical design, whichever causes their lens releases to be much fewer inbetween than Canon’s. Perhaps their talent was only available in time to reach a release now. Perhaps they were playing chicken and waiting to see who would take on Sony.

    It seems to me that the bottom of the camera market is continuously eroding. First P&S cameras went because smartphones were as good. I suspect now that there is reasonable quality fake DoF control in phones, the prospect of selling a $500 APS-C DSLR or MILC kit with an f/5.6 kit lens which only gets you more megapixels is not so good. Or selling a M4/3 camera with a $200 f/2.4 pancake lens. All of the market segments have moved up in price more recently, likely because the margins are better and the sales #s on the lower end stuff are dwindling. Even in the FF ring, the days of the affordable lenses are over. Even the cheaper models, like a 24-105 f/4, are north of $1000 now.

    I wonder how much of this has to do with the end-of-an-era change in optical manufacturing. A spindle spherical polisher from e.g. Strasburgh can be had for, say, $5k. A single optician can man up to 8 or so spindles, each of which could handle up to, say, 36 lens elements. A new pad or pitch pour is only needed intermittently, perhaps once a week. The optician has a cost to the company of maybe $60k/yr after benefits, and may have a cycle time for their step of, say, 4 hours. The machines last a lifetime with minimal maintenance. This is extremely economical and has minimal capital equipment spending. In contrast, the type of CNC optical grinders and polishers that would be deployed in such high volume production are likely in the 250-500k range, work individual pieces at a time, and an operator can only run say, 4 machines at a time. The machines cost more, and the production per machine is much lower than that of a spindle. Granted, you do not need to find master opticians to tend the machines, an extremely rare commodity these days, but the cost of production inevitably rises. So perhaps the low end of DSLR and MILC is not disappearing because the market doesn’t want it, but simply because it is not economical in the new status quo.

  • odddave

    Nikon is coming out with a Z mount 58mm f0.95. How many thousands will that cost, for what? another third of a stop? Here is how you own the night without breaking the bank.


  • Actually, it seems likely that a larger rear element should allow a lot less glass in the front of the lens. That being said, I am not a highly trained lens designer, just read a lot about it and understand some basics.

  • Eamon Hickey

    Actually, I think the X-T3 argues against your main point, assuming its sensor is made by Sony. The X-T3’s sensor is the highest performance APS-C sensor on the market right now—none of Sony Imaging’s own APS-C cameras can match it.

    So if Sony Semiconductor were trying to put Sony Imaging’s competitors at a disadvantage, why would they give their most cutting edge APS-C sensor to Fuji before they gave it to Sony Imaging?

    The X-T3 sensor is definitely not a stacked sensor (I asked Fuji), but I’m guessing that’s exactly what Fuji wanted. They got A9-level performance in a sensor that is undoubtedly significantly cheaper than a stacked sensor would have been. Sony gave Fuji technology that allowed them to match the A9 (albeit in APS-C) and beat all of Sony Imaging’s own APS-C cameras, at conventional sensor prices.

    Obviously, if the X-T3 sensor is not made by Sony (I have no information either way) it’s a different story.

  • Thom Hogan

    I doubt that Sony Imaging is keeping an “iron grip” on stacked sensors. Instead, they opted for an expensive sensor development earlier than the other camera companies. Sony is a company that tends to promote leading edge technologies; sometimes it works out well, sometimes it doesn’t. But betting on stacked early was a financial risk. No one else really wanted to make US$1200 compact cameras with tight margins.

    Nikon remains promiscuous with sensors. They continue to do their own R&D, as well. Sony Semiconductor still gets more money from Nikon for sensor production than they do from Sony Imaging, though that number has narrowed as Nikon’s product line has narrowed. I think it would be at great financial risk if Sony Semiconductor tried to play favorites now.

  • Thom Hogan

    You’re making assumptions here, not reporting how it actually happened. And you’re making a statement about Nikon mirrorless which is clearly wrong, since they now have new cameras that have sensor PD.

    Also, why is it so important who gets “credit”? We currently have Canon, Fujifilm, Nikon, Olympus, and Sony with on-sensor PD cameras. Likewise, we have Fujifilm, Nikon, Olympus, and Sony with sensor IS (which was invented at HP, by the way). In each case, the companies are doing somewhat different things with underlying tech.

  • Carl Eberhart

    Roger, as usual you make sense, you bring the right perspective, one of balance. For me, I think mirrorless cameras are a bit of a joke.

    From the standpoint of optics, is not a shorter flange distance ALWAYS going to be subject to stronger ghosting via the reflection off the sensor onto the rear lens element?

    I get that casual stills and video photographers like the idea of a light weight camera / lens. What annoys me about mirrorless fanchildren though, is the idea that getting rid of the mirror and prism, is “advancing camera technology”. It’s not doing that. It’s downsizing it, but it’s not simplifying or advancing it. It’s trading one set of problems, for a different set of problems.

    While I agree these larger lens mounts ARE progress…the idea that getting rid of the advantages of a mirror (really two mirrors) has for autofocus, over combining it all into one sensor (both imaging and autofocus)…seems very false to me.

    Has Sony’s A9 managed continuous autofocus as well as the D5 or 1DX2 ? Has it REALLY? (I realize the focus fine tuning issues might seem like a hassle, but honestly…other than those who have gone from an entry level DSLR to a mirrorless camera…I don’t see an inherent advantage…Logically it’s fine to have autofocus on the sensor plane itself, but focus fine tuning found in all the more serious DSLR’s, compensates for the front/back focus anomalies that can and do occur).

    So other than casual video (I say casual vs. the serious video of a cinema camera), why is there an advantage into miniaturizing a camera body, and then mounting it onto a huge lens?

  • Natt Lin

    Thanks for your explanation!

  • Geoff C. Bassett

    Sony’s mount was purely accidental. They had their crop sensor bodies with E-mount, upon realization that they could cram in a full frame sensor to the same mount that is what they went with. They chose that instead of going with another new mount shortly after introducing E-mount, that is the only reason.

  • Athanasius Kirchner

    Sure, and it changes what I pointed out how..?
    Even if Aptina pioneered some of this stuff with Nikon, for some doggone reason they refused to spread the love to larger sensors – that was Sony’s work.
    If the Z’s have OSPDAF (well, the fact that they even exist) it’s thanks to Sony. Nikon would’ve let that die on the vine in their 1 system. Oftentimes it’s not the inventor of a technology, but the entity that makes said tech usable and widespread, that (rightfully) gets the credit.

  • Athanasius Kirchner

    It’s, as you indicate, the stacked sensors that SI have been using exclusively for over four year now, even in a category with lots of competitors (1″ sensors). Apparently, the Fujifilm X-T3 doesn’t use a stacked sensor either, meaning that Sony Imaging are keeping an iron grip on that piece of tech.
    BSI tech for 35mm sensors was also an exclusive for at least two years for Sony; it’s conceivable that Nikon might’ve used it earlier if they had refreshed the D810 sooner, but given the current development cycles it looks far more likely that they started the design of the D850 once the A7RM2 was finalized and operating as a test bed for the current chip. And this, in turn, means that SI were using prototype BSI sensors much earlier than that.
    It’s also worth noting that Nikon have been forced to stick with Sony now, after they gobbled the competition up. Nikon used to be quite “promiscuous” in terms of sensor fabs, and it’s quite likely that their relationship with Sony Semi is a little strained or guarded because of that.

Follow on Feedly