System thinking

_4D08985 copy
Choices, choices, choices. From the ultimate image quality shootout.

We have a rather strange hardware problem: on casual observation, simultaneously too much choice, but at the same time, when all things are taken into account, a lack of it. It isn’t the problem of the perfect camera not existing, but rather that we have to jump through a lot of hoops for a complete solution. There are digital systems with sensor sizes ranging from 2/3” (Pentax Q) to 645 (Phase One, Hasselblad) – and to make things more confusing, surprising amounts of interchangeability*. So what is a serious photographer to do?

*Practically, this is nothing more than an illusion and a bunch of empty promises: even if you can do it, it doesn’t necessarily mean it’s a good idea.

Before making a choice – either an additional investment or a switch – I think it’s most important for you to know both your own immediate objectives, and have some honest idea of your level of seriousness. I measure this by intellectual and physical commitment, not finances: you may be extremely wealthy, but I wouldn’t recommend a large format scanning back or a Hasselbald Multishot for your family vacation snaps. On the other hand, for somebody who is extremely dedicated in photographing monochrome still lifes and only outputs as large optical prints, perhaps a studio monorail and 8×10” sheet film is still the way to go. If you don’t have the patience to run it properly, you’re simply not going to make any photographs. The objective of a system should be to enhance the photographic experience; compliment your current skill level whilst allowing you space to grow, and certainly not limiting you.

For all consumers, and the vast majority of amateurs, it isn’t the ultimate image quality that matters: it’s the ease of getting there and the size of the shooting envelope. The easier it is to achieve decent to good image quality, the more suitable a system is; it demands little but delivers beyond expectations. Since few bother to print or even share images in any way other than digitally in low resolution via social media, you probably won’t be surprised to hear me recommend no camera at all for consumers: just use your phone. It requires no commitment and almost no effort, yet with a little bit of care yields results that are far beyond what was possible not so many years ago. If there are more specific needs, then something with a 1” or M4/3 sensor and a decent fast lens with a reasonably wide range is probably the order of the day: something like a Sony RX10. Very likely that will have a more accessible shooting envelope than a low end DSLR with a kit lens; the biggest difference will lie in the optics, not the sensor.
Moving up a little, to somebody who’s willing to spend a bit more time on photography as a hobby, may well make it the main purpose of their vacation, and be willing to carry a bit more weight, interchange lenses and invest the time in learning some technical and artistic skills – I believe the optimum sweet spot lies in the region of M4/3 or APSC, but mirrorless. The resolution and shooting envelope is more than sufficient for even reasonably large prints; the range of lenses is diverse, relatively modern (i.e. designed for digital and capable of matching the resolution of the sensor) and for the most part, relatively affordable because the physical size of the glass is much smaller. And the sensors are large enough to satisfy any amateur obsessions with shallow depth of field.

You may be surprised I’m excluding DSLRs from the running, but it’s because technicalities start to come into play: specifically, focusing and exposure. With an EVF or live view, focusing is always accurate because you’re focusing at the sensor plane; exposure can be previewed and taken in conjunction, both of these things lower the miss rate considerably and allow the photographer to focus on composition and timing.

Mainly on image quality and increased control over depth of field grounds, the serious amateur or working studio pro has to really look at full frame – and that means a DSLR of some description, because frankly, the compression and shutter vibration issues of the A7R make it pretty much unusable – that is, if the extremely limited lens selection doesn’t already rule it out. That said, if your job is documentary and about fast turnaround, moving quickly and overall responsiveness, I’d still be looking at Micro Four Thirds: that hits the sweet spot on all fronts for that kind of work, in my mind. And you’ve got a large choice of excellent lenses that mean it’s a viable option to go either the zoom or fast prime route, or a mix of both.

Which system, exactly, you land up selecting depends both on your format choice and your specific needs; personally, I find the lens lineups available for M4/3 much more complete and well thought out in terms of focal length spacing than APSC for the major brands; the APSC cameras are stuck using the full frame primes, which gives them a dearth of fast wide options. Consumer zooms are far too many (pick your flavor of 18-whatever, from 55 to 300mm) the only thing really missing is tilt shifts, but those are rather specialized and not for the majority of users anyway – and if you need those, you’re almost certainly going to be looking at a full frame system or a technical camera solution.

Theoretically, for the missing lenses on all mirrorless systems, you could adapt anything with a longer flange distance; it should work. Ergonomic considerations notwithstanding, It won’t always work well – in fact, it will rarely work well – because of three considerations. Firstly, the optical design of the larger format lenses usually has less resolving power than the smaller mirrorless formats require (they may also not be telecentric); secondly, the adaptors tend not to be precise enough and introduce a whole range of additional planarity issues into the mix – both tilt and decentering, which further ruins resolving power, and finally, the filter packs tend to be of very different thicknesses, sometimes including offset microlenses, and sometimes not: the net upshot is that a lot of lenses simply do not work very well. The Zeiss 1.4/55 Otus, for instance, is a reference lens on any DSLR, but it performs poorly on M4/3 because of the thickness of the filter pack and interference with the optical formula. Never mind trying to hold that combination, either. Remember: just because you can, doesn’t mean you should.

On the full frame front, you have four choices. (I’m going into a bit more detail here because I think this is probably where the majority of audience interest lies.) Sony again is a non-starter because of lens selection, and I’ve already said my piece above about adapted lenses. Leica is the next candidate I’m dismissing: partially because at that price, I can get far better image quality potential in the body, and because unless you’re going to turn it into a frankencamera with EVF, you’re limited to 28-75mm if you care about accurate focusing. It is a single scalpel rather than a full surgical suite – there are some things it may work well for, but for most of us – it isn’t flexible enough. And let’s not even talk about tilt shifts or telephotos.

We’re left with Nikon and Canon. Nikon’s weakness is twofold: firstly, very few of the native AF lenses work well with the D800E/D810 unless stopped down; even then, you still don’t get the same degree of bite and clarity (microcontrast, aberration correction and resolution of fine detail structures) as the Zeiss options. This again means limited focal lengths and worse still, manual focus only – the viewfinder is not at all up to the task, and frankly, almost never even arrives properly calibrated. Good thing we have live view, then. They’re also lacking an ultrawide tilt shift. Fortunately, the flash system is its redeeming factor: it’s excellent; very flexible and very consistent in metering. Canon appears a bit more balanced, because as they still lack a high resolution body, this tends to hide the flaws of the lenses somewhat more; you simply can’t fault what you can’t see. There are some excellent lenses in the lineup, but there are also a lot of surprisingly poor performers, like the 24-70/2.8.

You may be wondering why I haven’t mentioned medium format options. I honeslty think the reasons for using medium format have gotten fewer and fewer, the lower end (30-40MP) especially. Unless you need the leaf shutter and very fast flash sync, or extremely high resolution for large prints, I can’t really think of any reason to bother all. Part of this is pricing: the D800E started the death of that market; part of this is system width. Even for a very established system like Hasselblad or Phase One, there are still far fewer lens options than for Nikon or Canon; there’s no equivalent to the Canon 17 TSE, or Nikon 800/5.6 for instance, or either of the Zeiss Otuses. Medium format lenses may be uniformly reasonably good, but with the exception of the Leica S lenses, they don’t perform at anywhere near the same level as the Otuses unless stopped down considerably: and then your depth of field control and flash power advantages disappear completely. Pentax 645 might seem like a very economical way to get in, with a cheap legacy lens selection, but beware: almost none of them can really match the resolution of the sensor. Only the newer SDM lenses do it justice, and those are not cheap at all.

There are also of course the very specialized matched lens-sensor solutions, like the Sigma Merrills/ Quattros and one-offs like the Ricoh GR; their output results seem to suggest that this is really the way to go in future. However, being matched of course means non-interchangeability; a system this is not. The closest we get is probably a full set of Sigmas, though that’s just 28/45/75mm. I think it’s pretty clear that once you pass what I like to think of as the ‘consumer threshold’, there’s really no such thing as one system: even if you have moderately diverse interests, you’re probably going to have to consider one and a bit. Or worse, if you’re chasing ultimate image quality, probably have one lens and one camera for each focal length – turning the logistics into a bit of a hodgepodge and a ‘did-I-remember-to-bring-the-right-chargers’ nightmare.

I personally have been travelling with the following of late: for street work and fast response, a Ricoh GR; for wide angle work, including some architecture and landscape, a Nikon D810 with 24/3.5 PCE; that same body also mounts the Zeiss 1.4/85 Otus for short tele and cinematic work; the midrange is covered by the Pentax 645Z and 55/2.8 SDM. I leapfrog again with an older manual focus 150/3.5 Pentax lens to provide a ~120mm option. The Zeiss 2/135 APO is definitely a superior lens, but it’s also got superior weight: at this point, with this level of weight, and the relative infrequency which I need something that long, it’s difficult to justify – and even harder to get past airline carry on. As to why I use the 55/2.8 on the 645 instead of the 55 Otus on the D810, it’s mainly because of the rendering style: the former is a wide-normal, the latter is a long-normal/short tele, and I prefer the former. Let’s just say that whilst it produces the results I want, and makes for spectacular prints, it’s not ideal: I’d love to have lenses of Otus quality for the 645, or the resolution and high ISO performance of the 645 on the D810, but neither one exists. Instead, I’m left with about 6kg as my lightest option – and that doesn’t count the tripod I’m also carrying most of the time.

In fact, the advice I’m going to close with is one of considering compromise: be honest with yourself, look back through the back catalog of work, and decide what should be your focus; what types of images do you want to knock out of the park? The rest…well, if you’re not masochistically pushing the envelope, you may want to reconsider. MT

__________________

Limited edition Ultraprints of these images and others are available from mingthein.gallery

__________________

Masterclass Venice (November 2014) now open for booking – click here to book or for more info

____________

Visit the Teaching Store to up your photographic game – including workshop and Photoshop Workflow videos and the customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!

Don’t forget to like us on Facebook and join the reader Flickr group!

appstorebadge

Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved

Comments

  1. I’m returning to this subject after a few weeks. I was pondering the Sony A-mount as a possible FF solution. I handled the A99 at a local store. I don’t know how I missed it before, but the A99 does have an EVF, which is a very deciding factor for me. However, the EVF is nowhere near as good as the E-M1, so it will not be good enough for precise MF.

    That said, the AF seemed dead accurate and fast. It’s basically a hybrid system where a fixed translucent mirror handles the phase detection and the sensor handles contrast detection. This should at least in theory eliminate any AF calibration problems. Image stabilization in body is another big plus.

    I would like to have some more resolution though, that’s partly the reason I would go FF anyway. If Sony releases a Mark II of the A99 with 36+ MP and a state-of-the-art EVF, it just might be the FF solution for me. The Zeiss 85mm f/1.4 would probably suit most of my needs, so I could feasibly get by with just that one lens for quite some time.

  2. What about the new Sony A7s? I hear wonderful reviews on it. What is your take on it, Ming? I am thinking of selling my Sony NEX 6, Nikon D90 and Leica D-Lux 4 to buy the Sony A7s. I have tried so many cameras and they are all annoying in some regard. The Sony NEX 6 is the best for its size, tilt screen, viewfinder, and the files come out pretty good even though I have to convert them. The kit lens is not so good. I thought about a Fuji X-T1, but quite frankly was underwhelmed by my Fuji X-E1. Like the lenses, but am thinking more on the lines of a full frame. Now the nikon D750 is out. I haven’t even thought of that as a viable option, but I am a Nikon shooter from back in the day. Being older, I like smaller cameras due to neck strain and such. I also use a Sigma DP2M which is a pain, but the files are so lovely. I was going to buy the DP3M, but they are hard to find here in the USA and since Sigma dumped the line and went Quattro, it kind of pissed me off. (Not enough to sell my DP2M though.) Anyway, for an artist/hobbyist like me, who has used so many film and digital cameras, I am trying to find that sweet spot system without going totally bonker broke. Any suggestions?

    • 12MP is insufficient for my print requirements, and all of the A7s suffer from lossy raw compression that make for very ‘fragile’ raw files. The lens holes are very slowly being filled in, but not fast enough. If I shot a lot of low light video or wanted 4k, and was looking for something purely for that purpose with little or no stills, I’d probably look there first. I think for the first time since the D700, we actually do have something that hits the sweet spot – it’s the D750.

    • May I chime in about the A7S which I bought a couple of months ago. I wished to use it as a low light stills camera handheld. What I can say is the low ISO performance is insane clean. Black is really black black black. I am wondering if Sony has baked in some extra contrast in the RAW files? That leaves a pretty crispy impression almost like the Merrills but of course with much lower resolution. For color it performs well up to 6400. For B&W up to 12800. Above is pure trash.except one loves digital noise.
      The extremely clean and contrasty files combined with the low resolution leaves an impression of a bit cartoon like images IMO.
      But don’t touch the files too much in PP… the RAW files are ‘thin*.
      The camera is unpleasant to hold and the dials and buttons are not where you expect them to be. Perhaps a thing one could adopt to along the way?
      I have used it little and since the D750 appeared here on the site I would neither need it any longer. The D70 high ISO performance is impressive.

  3. Oh, a lovely essay on cameras, and their uses, this was!

    Personally, I have come to a similar selection, but with fewer components, and somewhat less weight:

    For street photography I use a Nikon 1 camera, with my very own 8 (actually 7.9, made of a Nikon 1 10 + a Panasonic wide adapter) — in 35mm terms a 21mm lens, with very good flare characteristics) to get a wide view, complemented with another Nikon 1, with a longer lens, like the AF-S micro-Nikkor 40, or the AF-S 85/1.8G (and the FT1 adapter, of course).

    For normal photography I use a Nikon D600 with either a 20 (Voigtländer), 35 (Sigma ‘Art’), or the 85, already mentioned.

    Then for birds, I use either the D600 and the AF-S 80-400 VR II N, or my V2 and the wife’s 70-300CX. Naturally, I can use the 80-400 on the V2, but it isn’t an ideal combination, and the image quality isn’t better (especially in the long end). If the wife is using her 70-300 I often stick to the 30-110, a very capable lens, and light!

    I have APS-C cameras, both Pentax, and Sony (NEX), but they get very little use, unless I want to use a manual lens, for some reason.

    I gladly exchange the APS-C cameras, and their lenses, for a Ricoh GR, which still sounds very sensible 🙂 !

    The wife has a somewhat different set-up, mainly using a RX100 for street photography, and OM-D E-M5 for normal photography, and two excellent choices for birding: a Pentax K-30, with the excellent HD DA55-300 (together a light, easy to handle package, which is excellent for birds in flight), or her V1 and the already mentioned 70-300CX (in 35mm terms, about 200-800mm), a brilliant lens, outstanding in its long end, but maybe not the best for birds in flight. In addition to that, she has also a NEX, which gets very little use!

  4. That’s some big heavy gear you’ve got there Ming. As soon as I feel I’m “lugging” gear and the gear is impeding me, all the fun and creativity just disappears for me. These days I’m happy with the A7r and FE 55/1.8 with some B+W and Singh-Ray in the bag, a blower, some lens wipes and a few spare bats.

  5. Great article Ming. I absolutely agree with the output medium being a large factor in system choices.

    I’ve been thinking through my system choices for a while now and the two areas I have always enjoyed most are macro work and landscape. I’d like to focus on those two areas exclusively. I’m probably going to sell almost all of my leica equipment (m and r systems) as 35mm film isn’t my preferred landscape medium and, frankly, I haven’t used any of that gear in over a year.

    I love my om-d for macro work. The image stabilization and depth of field afforded by the smaller sensor size are excellent. The 60mm macro is superb. I’m reluctant to go to asp-c or full frame for macro work because of depth of field difficulties when working with moving subjects.

    That said, for landscapes I find myself craving higher resolution and greater dynamic range than my micro four thirds system offers. I could probably fund most of a 645z with my leica sale and use my hasselblad lenses on the camera, purchasing the 55mm for inclement weather situations (which are common here in the northeastern USA). Or, I could go full frame Nikon with a d750 or d810 and have a whole complement of lenses for the cost of the 645z body. I haven’t figured out which decision makes more sense to me. The 645z seems to offer substantially better image quality, but is much more expensive and I cannot afford the new D lenses as well as the body. That said, I do already have the 50 FLE, 80 CFE, and 150 CF lenses for the hasselblad that can be readily adapted to the 645z. I don’t often order prints larger than 16×20 so either system would be sufficient. Forgive me for thinking out loud in this post. Definitely need to upgrade to a more solid tripod to complement my gitzo traveler.

    At any rate, thank you for the thought provoking article.

    Cheers,

    David

    • Thanks David – gotta be honest, Leica M isn’t really well suited to macro or landscape (R is better). I’d hold on to the R lenses til you determine which you want to use since a lot can be adapted to Nikon F with reputedly good results.

      Hassy V + 645Z: the lenses are a mixed bag, and you’ll be lacking a wide. The 50 FLE is okay; the 80 is not bad, and the 150 is excellent (I have these three also, plus the 120 Macro). Perhaps add a Pentax 35/3.5 to cover the wide end and you’re set. Or a D810 and Leica R lenses, plus the difference in price to a 645Z would get you one Otus…it’s a good problem to have (and please use one of my referral links if you get a chance! 🙂

      • It may take me a few months before I manage to make the purchase, but I will absolutely use your referral links! 🙂

        Thanks for the reply. I agree re Leica M. That is precisely why I haven’t used that gear in so long.

        My Leica R lens collection is rather limited, but that is a suggestion well taken. I only have a late 50 summicron (r-only) and 90 summicron (3-cam). They both are great on film so I’m happy to hear that there is a chance that they hold up well with the d810 sensor.

        With my budget, if I go the d810 and otus route, I would probably get the 85mm. I like a short tele for landscape work. I could use the 50 summicron as a normal lens and, if necessary, pick up a wide later on although I’m not much of a wide angle person. The d810 and otus along with a new tripod would nearly exhaust my budget of $9500 (total cost ~$8690) and leave me with a rather limited range of lenses (50mm, 85mm, 90mm). Of course, your review of the otus was stellar as, clearly, is the lens itself. It is a very tempting option. I really, really wish that the otus lenses were weather sealed. We get a lot of snow and ice where I am living at the moment and that really makes me worry a fair bit. The 645z and 55mm lens would be a nice weather sealed option and a definite plus compared to the d810 and otus combination. I could sacrifice resolution and save money, purchasing a d750 instead, but that seems to be a halfway measure that somewhat defeats the purpose of the purchase in the first place. The less demanding d750 sensor may, however, lead to better results from older leica r lenses. Of course, I doubt with my poor vision that I can manually focus them anyway. For landscape work, I’ll be on a tripod with live view anyway so that’s a nonissue for the main purpose of the camera.

        With the 645z I’d like to have the camera, tripod, and 55/2.8 lens. This is the more expensive option at ~$10500 and is a bit outside of my budget at present. At almost $2000 more it is substantially more expensive. Whether it is a superior option compared to a d810 and an otus is not clear to me at the moment. According to your review, the 645z with 90mm lens doesn’t quite compare to the otus on a d800e/810. However, the 645z has greater dynamic range and yields an overall larger shooting envelope. The photos from it have a bit of a different look as well – not quite medium format film but maybe halfway between full frame digital and medium format film. The lenses are cheaper – either used hasselblad or pentax 645 / 670 lenses, but I feel like I would be playing the lottery to a certain extent in purchasing and using older lenses and figuring out which hold up to the modern sensor.

        With the savings of the d810 option, I’m inclined at present to go that route and simply be one third of the way toward a second otus purchase. I’d probably wait and see if they announce a wide angle otus rather than save for the 55 though, assuming the summicron is a good performer.

        I truly wish that manual focus were easier on these cameras. Optical viewfinders these days are a shadow of the good ole days. Give me a leicaflex finder in a d810 or in the pentax and I’d be sold. I would love to be able to reliably manually focus again. Or, just give me a top of the line EVF, which I prefer anyway.

        • Thank you. Given you have 50/90 already – I’d go with selling one or both if you’re going with Otuses to avoid duplication and to release some funds. I am pretty sure the Otuses will beat all comers! Weather sealing: not a big deal in moderately bad conditions, but not splash proof. I’ve shot with my 85 in a blizzard and the 55 in rain, both held up fine.

          • Sounds like solid advice to me – I agree. Good to know that there isn’t too much to worry about in moderately poor conditions. Quite a relief, frankly.

            Thanks again.

  6. Marcelo Vaz says:

    Ming, you use a custom focusing screen on their nikon bodies? If yes, is easier to focus on manual lenses?

  7. BOEY KEAN PHENG says:

    Nice article, Thank you very much.
    The lesson is simple: Choose the system/tools that you need to get the results you want. No single tool will fulfil all requirements, choose the system/tool based on shooting envelop.

  8. Could you say exactly how a Leica S with your choice of lenses does not meet your needs?

    • Remember, I don’t have a bottomless pit of money: I’m running a business. I can get better results from a D810 and Zeiss lenses at under a quarter of the total system cost with two bodies. The S does not have more resolution, better color, or better high ISO, or dynamic range. I can’t charge clients more because my camera has a red dot on it. The additional ROI is zero.

  9. harold1968 says:

    This whole findings thing is very bizarre
    Apart from the Panny GMx I think 4/3s is squeezed between 1″ and APS-c.
    The combination of lenses and IQ and speed for the Fuji X-T1 and Sony A6000 is unbeatable under FF
    In terms of dismissing the A7 series that’s a serious mistake. 90% of what I do is between 35 and 50, and the outstanding Zeiss primes on the A7 cater for this. There is now the excellent wide angle sony FE zoom and the 70-200mm and more coming
    I have found the A7S with its silent shutter and large pixels one of the best cameras I have ever used

    • Well, it’s good it works for you, but we all have different output objectives and requirements.

    • For me, the 1″ and the full format squeezed out both my APS-C, and my longing for a m4/3 set-up (My wife has a very full OM-D kit, with a few Panasonic lenses as well, so I know they’re good, but not for me).

  10. Michiel953 says:

    One DSLR, four primes (24, 35, 58, 85) is what suits my needs (ambient light portraits, street, situations, emptiness) and what I can work up enthusiasm for. 800E now. AF still a bit iffy, weight and size… 750 might be better, if it weren’t for the fact that 36MP and the associated DR is really addictive. Stick with what I have for a few years, see about the 900 then.

  11. Lots of head nodding while reading, but nearly enough thoughts to write an article of my own. Camera enthusiasts are often much like motorcycle and sports car enthusiasts, the choices need to be validated for many, and there is often a desire to have high performance, despite not getting near the technical limits in usage. The other thing is the number of capable cameras hobbled by cheap lenses (i.e. D800 with low cost zoom, and often the lens hood on backwards).

    Choosing based upon output is really the best idea in all of this. The problem is that many people are not really sure about the output they want. In the commercial photography realm, our clients determine out output needs. While I still have a full bagged and ready-to-go system to produce images for large trade show prints, the reality is that I don’t get many requests for that. Invariably single page or two page spread is vastly more common; many systems can achieve this. I’ve tended towards wide to normal, and less often short telephoto, which again opens up a wide range of choices.

    This year, I picked up a Nikon Coolpix A to use as a small back-up camera. It’s a bit tougher to get use to lightweight cameras, but after a while it got me thinking more about other fixed lens lightweight cameras. One advantage of APS-C and 4/3rds is that I can shoot more wide open and get a reasonably good DOF. Another thing I enjoy is not needing to do sensor cleaning. I don’t think I will ever move away from 35mm size digital and medium format, but I’m finding that I want a small camera in the bag with one of those larger systems. I still have the 4×5, but due to a shoulder injury it’s become tougher to get set-up the way I want to use it.

    • Choices need to be validated for those without a spine or ability to decide for themselves whether something works for their vision because they have no vision. They need to feel good about their purchases because for whatever reason, they use these to form their own identity – probably because their own work has no identity in itself.

      Clients may determine the output, but frequently I find that their requirements are far less stringent than my own – and I’d rather that be the case, because it’s much better to deliver more than less…

      • Absolutely. The popular saying in the US is “under-promise and over-deliver”. Most clients are not that technical, and do not really understand the process and details, nor do they really need to understand it. Some do want to see a big (or expensive) camera, but beyond that it’s just results. I found that out once using a Contax 645 with digital back. Since I didn’t want to fully trust in that digital back, I shot quite a bit of that project on film. It turned out that most of the images selected later were film scans.

        Equipment validation is something that confuses me a bit. While many of us like nice things, especially nice cameras, there’s a big difference between appreciation and that odd need for validation. We could probably do quite a bit with just an iPhone, yet technically it’s not that great a camera. I think your exercise of guess which camera for which image was a great example of that.

        Why you use the equipment you do, is something you document quite well (my opinion). If I run into a client who asks me what camera I would use on a project, I usually ask them what they would like me to use; though they don’t really understand it. Clients who ask what camera are usually a red flag for problems, at least in my experience. It’s unfortunate I run into that, but perceptions are tough to change. In the last few days, I’ve read about projects shot on an iPhone, and projects done with studio lighting that cost as much as a new car, but the point was that it was the idea of the photographer that made it all come together to create compelling images. Whether an enthusiast, or a professional, making compelling images should be a goal before buying another camera.

        • Opposite case in Asia; over promising and missing the target by a huge margin means most clients are weary. Can’t blame them, it’s really the fault of people who don’t know their abilities and consequently make the rest of us look bad…

    • Gordon, I’ve committed the “lens hood on backwards” crime every now and then (and felt guilty). There is a reason though. I’ve got four 1.4G primes. Lens hood on they’re big, they do protect, but are not always needed. But they (or rather “it”; I rarely carry more than one or two lenses) are still there, and you need somewhere to put it if you take it off, making the whole combo just that little bit smaller and more friendly. In your trouser pocket? Your bag, yes, if you have a bag with you. So sticking it backwards on the lens is a practical way of storing it, even if ithat makes you look like a dork. Which it does. 😉

      • I’ve actually given up on hoods other than for situations that require impact protection. Either the lenses don’t need them for optical performance, or they don’t help…

      • I have a Velcro tie strap on each camera carry strap, just for such an moment, like when using a polarizer filter. Many of the lens hoods are long enough that they interfere with holding the lens, or using the focus ring on some lenses. Obviously there are exceptions to every situation, but you would be amazed by how many people never knew they could take the lens hood off and mount it properly. 😉

    • Age, and sore shoulders, and a bad leg, made me sell the up-to-date APS-C gear, and opt for a Nikon 1 system, complemented, for use in low light, with a full format system (with a few, light, prime lenses, and a single telephoto zoom).

      I therefore bought the lightest FF camera there was on the market, at that time I could afford, the D600. Since then things have moved on, so I might have opted for another camera today, but many of the F Mount lenses work well on the Nikon 1 cameras, too!

  12. Ian Christie says:

    A nice article, thanks. I’ve been won over by your point in favour of the high-end fixed prime lens cameras with matched sensors. The Sigma DP2 Merrill poses some workflow problems, for sure, and it’s not great in low light/high ISO; but the UI is good and the output from Raw can be spectacular. I’m sticking with the Merrill ‘system’ and would only supplement it with another compact fixed prime camera, one that can do well in low light. The Ricoh GR is probably the one. At the current low prices the Merrills are a supreme bargain.

    • It’s strange that there are not really any fixed lens choices above 35mm other than the Sigmas though – perhaps that’s an untapped market for some of the other manufacturers…

      • Any manufacturer would do. I am still waiting for this to happen. The Merrills are so … challenging 😉

      • Ian Christie says:

        Agreed. When I began taking pictures back in the early 70s, there was a host of advanced fixed lens compacts, with rangefinders and manual exposure control, and focal lengths in the 35-45 range. I’m grateful for the high-end digital digital compacts we have now – the choice is pretty good, even if I could wish for more compacts with built-in OVFs – but there’s a lot more that could be done. Maybe Canon should drop the Eos-M concept and bring out a FF digital version of its old QL17/19 compacts: fixed 40 or 45mm lens, f1.9 or 1.7, OVF with rangefinder. I reckon there’d be takers for it. As for Sigma, if they could just fix the software and put the Merrill sensor in a body resembling the Epson RD1, I’d be set for life…Dream on, I think.

      • I agree 150% with you here. It started with the original X100 prototype under glass at Paris photo fair just after the Kina : love at first sight, except I want a 50mm. And/or longer. Seriously, they all have their shortcomings but the first one to come out (A, GR, RX, X-Fuji, X-Leica) I’ll buy with no hesitation. Not the Sigma though, because as good as it seems to be in some situations, it is everything but an all-rounder.

        I’m still pondering to grab a DP3M, but that would be for a very specific usage. Sort of an affordable/reasonable/portable alternative to the 645Z with its 90mm. Just as you wrote in your conclusion: “decide what should be your focus”.

        • The longer compacts really need viewfinders (or those LCD hoods, but they wouldn’t then be compact) for stability, or IS systems – which means it won’t be Ricoh or Leica, Fuji maybe, Nikon maybe, Sony probably (but not that effective).

          • I would be surprised should Sony develop a 3 focal fixed lens *system* with OVC and other delicacies. They did the 7 series and they are indeed compact. But hey then again .. why not a new A ILCE series with non-compressed RAW, say around not too crazy many MP, a bit more logic layout and comfort to hold, get 3 Zeiss Loxia lenses and keep an A body sticked onto each lens. That would make a nice compact always ready to shoot set. Not a too bad idea actually 🙂

  13. O_o equipment dreams

  14. Seems like what’s happening is a change in the notion of what one’s camera system is, from one camera with multiple lens combinations to multiple cameras. With gradual miniaturization over the last 10 years, a line was crossed when a small camera like the GR or even the Sigma Merrill’s that combine a great single fixed lens with a great sensor have become as light or even lighter than an average lens on a DSLR cameras (whose sizes have mushroomed compared to the film DSLR’s). That means in the same (small) shoulder bag that you would cary a camera with zoom lens and say two extra fixed lenses, you can now cary the same camera with one lens attached and two complete small cameras with fixed lenses, like the GR. Or just put the GR in your pocket and throw in a different small camera with a different lens, getting you now up to four cameras. No need to changes lenses at all, just grab a different camera. Your discussion and use of the GR on this site convinced me to buy one (I’m biased mainly towards the best sensors, especially if the lens is also great). I’ve always had trouble with going below a 35mm lens, having spent so many years with a 50mm “standard” lens. Well the GR changed everything for me. I am now enjoying shooting with the 28mm lens, finally getting the idea of getting closer, or framing wider and cropping if necessary. And what a great camera to hold and use, immediately makes you wonder what’s wrong with all the other camera companies and why Ricoh doesn’t go full throttle with this and get allow an attachable viewfinder (put it in the other pocket!). So, I skipped buying the new 28mm equivalent lens for my Olympus M1 after trying it for a while. A great lens and “system,” but why not take the money for that lens and just buy a GR? Use the Olympus for longer lenses, grab the GR out of your pocket on a moments notice and switch to 28 without sitting down to switch lenses. What else? No more “I’m a Nikon guy,” or a “Leica guy.” I’m a X-Y-Z guy, and like all three as a “system.” This, of course, leave out the issue of similarity across images based on a common sensor, type of lens, etc. Not sure if that’s what your timely essay on camera systems is getting at or not. But you keep bringing up the GR every time you write about something.

    • The engineering requirements have gotten a lot more stringent now that resolutions have effectively gone up a size – 35mm in MF territory; MF in LF territory etc. And I don’t think all of the components of the system have caught up, for any system. Usually it’s the legacy lenses that are the issue, not so much the newly designed ones. (Or in the case of Leica S, the lens are great but the body doesn’t make full use of them).

      I think the conclusion is you have to be an image guy: know what you want to produce, and then figure out what you ned to get there…

      The GR is something that just goes with me, is very low-overhead in terms of size and processing, and just works. Not many cameras like that, so we appreciate them where we can – it’s a natural consequence that so many of my images are produced with it.

  15. Ming,

    To follow up my last message, I have done a “rough” comparison of 4 cameras (OMD EM5, Fuji XT1(borrowed), Sony A7, and Leica M240). All shooting at roughly the same focal length, 27mm to 30mm. All wide open. OMD EM5, using Panasonic Leica 15mm, 1.7, XT1 Fuji 18mm F2, A7 and M240 using Voigtlander Ultron 28mm F2. All handheld shot.All were converted from JPEG from RAW with the latest lightroom. ISO 200 to ISO 3200 were shot with one focusing on the back model and one focusing on the front model. All uploaded with full resolution. You will see the advantages of IBIS and the disappointment of FUJI. FUJI smudges all the details…. And with the adjustable M mount adapter, the sony can focuses much closer, where the Leica can’t. This is just a rough comparison…..

    https://plus.google.com/photos/101191223694267273293/albums/6057306767850916449?authkey=CNTE_8vY_JiWKg

    One of the Lecia file is corrupted, but you should get the point.

  16. Ming,

    Thanks for the great article. I am a gear freak. And would like to share my experience.

    I used to own a D600, and still own a A7, a m240, OMD EM5, and Pana GM1. I always come back to all my m43 due to its shooting envelop. I am getting rid of my A7 and keep my m240 (because I enjoy shooting a range finder). I realize that I can carry 2 bodies (OMD and GM1) with ease and the flexibility of the OMD IBIS give me. With 2 bodies, I can shoot at 30mm F1.7 while also be shooting portrait at 85mm F1.8. The IBIS give me 2 to 3 stops advantages which negates all the benefits of a FF. The only thing I lose is shallow DOF, which I don’t miss too much. I found myself needing to stop down more often then opening up. The key part is the weight. It allows me to go anywhere and shoot at any angle at ease. The selection of Lens, as you said it, allows me to choose what I need for different occasions. A slow pancake zoom for my GM1 maybe all I need for a daytime walk around.

    I still find a lot of people demanding FF, and I get the looked down for using my m43 gear. But this is how the world, or at least in Hong Kong, right now. People judge you by the gear you use. And btw, not all Hong Kong people agree with the current demonstrations (40+% vs. 30+%). We are getting a step change (universial suffrage but nomination is selected by a selected group) but some demand an open nomination as well. I am one of those that is against the demonstration. Don’t believe all you read from the newspaper. Sorry for digressing.

    • Thanks for sharing, Derrick. I think we should put the gear aside and just look at the results first…

      As for the situation in HK…well, I don’t claim to understand what’s going on, but it certainly seems like change is afoot – the question is whether the majority of those demonstrating understand why they’re doing it.

  17. mosswings says:

    I walked in to my local pro shop yesterday and told my long-time salesperson “My D7100 was great, but I never want to schlep it around the world again. If you were downsizing, what would you do?” He reached for an XT-1. Quite small for an APS-C body, and a viewfinder that is truly eyeglass-friendly. Then I compared it with an EM-1. Almost there on many fronts, but again…that viewfinder. Then I played with the controls, and realized that I really didn’t like retro cameras all that much. What was I most taken by? The GM-1…which suggests that the GM-5 might be a great travel camera and more than adequate.

    The only problem? The XT-1, EM-1, etc. are all rather stupidly expensive. These days, you can get all you want in IQ in a very convenient package. You have to give up some combo of viewfinder and AF to get it, but that for most is not a problem. Would I like something like an RX-10? Sure, but at that sensor size I begin thinking that the RX100 or Canon G7X or Panny LX100 make more sense as a body form factor.

    Time to take up painting.

    • And then there’s still the whole workflow issue which most people don’t seem to understand or care about…

      • What do you mean ‘the workflow issue’ ? I use Lightroom and it seems fine. When I print I go into PS. where is the problem?

        • Firstly, dodging and burning control is terrible. Secondly, operations are not sequential, they’re cumulative, and that produces visibly different results. Finally, if I have to take it into PS anyway to prepare for print or retouching, then that’s an extra step in the file transfer. Remember, I’m dealing with at least hundreds and usually thousands of files a week. Add ten seconds extra to every file – that’s optimistic – and I’ve lost nearly three hours already even if I just have a thousand files. I work pretty much non stop every day to keep this site and the rest of my business running. Every hour saved is precious.

          • Ok , I see that in your highly pressured circumstances time is everything. I was not thinking of such a dense work schedule.

            For my purposes dodging and burning seem to work well with the brush or radial tools, though, of course, if an image needs surgery, it has to go into PS anyway, so I can do that there. But how often do images need such fundamental work?

            I’m not sure I know what you mean by the difference between sequential and cumulative changes. How does this pan out in practice?

            • Quite frequently, in the commercial world – dust removal on watches, for example.

              It’s more difficult to achieve a nonlinear, natural-looking ‘shoulder’ without sequential changes.

  18. I’m a happy M43 user although I have experienced the dynamic range difference betweren M43 and the D800. For what I currently do I don’t need FF but if I was indeed selling large prints then I would consider the D810 purely for that reason.

  19. Errata: the Pentax Q has a type 1/2.3- inch sensor not a type 2/3-inch sensor.

    It doesn’t affect the argument just emphasizes how large a range of sensor sizes we have. In reality though, in the future, type 1-inch is going to be the bottom end for CSC as no one else thinks like Pentax.

    • Curiously though, the Q sells well in Japan. Can’t imagine why, as image quality is terrible and not even close to say an LX7…let alone the 1″ cameras. And it’s really not much smaller than M4/3 and one of those pancake zooms, like the GM1.

    • The Q7 and Q-S1 have a 1/1.7-inch sensor.. As to why they sell, I suspect it is because they are well designed and fun to use and cute.

  20. Martin Fritter says:

    I got a Sigma Dp3 recently, and it is truly extraordinary. It is rather absurd as well. And SIGMA Photo Pro is a PITA, but livable. Still, totally worth the effort. However none of the Merrills represent a complete solution. Even if you had all three. You just can’t shoot fast, for one thing. My Leica M6 handles faster. So does a Rolleiflex! But, does anybody have any experience with the SD1? I realize this may be a bit off topic, but in a way you’re talking about achieving state-of-the-art for your various envelopes (none of which includes sports, action, wildlife, war correspondent…)

    • Martin.. Perhaps one should just image these little beasts are huge old fashion film cameras. I think the pocket sizes fools one to think they are as responsive as a pocket camera.
      Seems to be an attitude question when picking up those cameras and go out and shoot.

  21. Erling Maartmann-Moe says:

    Interesting article, Ming. What strikes me is that there is no “perfect” system. What it comes down to, in my opinion, is what type of equipment that gives you joy of shooting – most kinds of equipment is able to deliver decent picture.

    I think you dismiss the Leica M system too easily – it has flaws both in lenses and in mechanical adjustments, but for those who have converted to the rangefinder way of shooting, it gives immense joy (and some frustrations) – the manual everything – focus, framing, metering – makes you consider your photos in a completely different way than firing off a 12 fps DSLR. It has a certain resistance that for me is pure joy (and I have been through the DSLR phase, in all variants).

    EVF just does not work for me, a good finder is key.

    And then: The Leica S system. You say the lenses are extremely good, isn’t that the ultimate thing? I have been so happy with the Leica S2 (bought used at a decent price, with a couple of lenses), beautiful colors, almost no postprocessing needed, perfect ergonomics. 37 or 50 mpix? Not sure I care, it is such a joy to work with. OK, it is a daylight camera. But I have the Monochrom for high ISO.

    Too expensive? Yes, but instead of buying a new system every year, mixing components that really are not made for each other, and trying to get to the moon by climbing one and one tree, go slow, buy used, build your system, get to know the quirks and strengths and live with it, take it as a challenge – there is a lot of creative potential in some resistance.

    • **there is a lot of creative potential in some resistance**

      That’s true.

    • Nope, I used to own Leica M for several years – I was more frustrated than overjoyed, especially with the unreliability. A 645Z or Hassy is definitely not 12fps! 🙂

      The lenses of the S system are good, but the sensor isn’t. Nor is reliability (again) – the one I used ate cards. And yes, low light is poor, which isn’t the case with the D810 or 645Z, both of which offer better overall sensor image quality, and there are lenses to match now.

      Then there’s the price – for the cost of an S body alone, I can get both 645 and D810 systems and lenses…at the end of the day, I still have economic considerations to factor in because I am running a business, after all.

      • I respect your needs and requirements as a pro. But I assume most of your readers are amateurs, who seek not just good pictures, but also joy and challenges from the photographic process itself. By dismissing the M system, you also dismiss the whole rangefinder experience, which, although a minority in the photographic world, has a dedicated set of followers, and I dare say also some top-notch photographers. And some find the rangefinder experience gratifying, after having been around the block once or twice, with SLRs, DSLR, compacts and mirrorless.

        As for the quality issues, I remember in my Canon years all the talk about “bad copy” on lenses, people buying and returning to B&H, driving themselves crazy trying to be at ease with what they got, so there’s always something.

        As for anecdotical evidence, I have had none of the card-eating problems you refer to, neither from the M nor the S system – being an early adopter sometimes has its price, and I think they have evolved. Pricey, yes, but it is not cheap to upgrade to new models all the time, and there is a used marked which can be rather attractive. I buy and sell M lenses basically for the same price, bodies are a different game. As with cars, one cannot find an objective value, otherwise they would be no BMWs or Mercedeses on the roads. How good we have choices.

        • Erling, I’ve owned three Ms, all of which ate cards, and seven lenses, five of which had to be replaced – some multiple times.

          There are plenty of people writing pro-Leica and pro-whatever brand. I’m pro-image, I don’t care how I get there.

          • Erling Maartmann-Moe says:

            Oh, I thought this whole discussion was about how to get there. I tried to argue against totally dismissing the rangefinder approach, but you have some negative experiences with it. OK, I still think you dismiss a very interesting approach that works for a significant niche of photographers, but …whatever.

            • It is, but I still have reasons against it. Not because of the concept or technology, but because of the execution. There’s no point in using a tool for a result if the tool doesn’t work half the time, or there is a significantly better one. I’m not the only one who’s had bad experiences with Leica digital; you’re lucky and honestly my thoughts might have been different if that was the case. I just cannot recommend buying something that expensive and unreliable when there’s something else that does the job as well or better and cheaper.

              Here’s my biggest stumbling block though: what set of capabilities it you get from the RF that you can’t get from another system? I honestly cannot for the life of me define anything.

              • this would be:
                -seeing around the frame can help to choose the important part of the scene
                -small lens size
                -acurate focusing for wides and normal lengths (If you dont life view I would expect that it is harder to focus a otus 55mm on a Nikon than a 50 APO on a rangefinder)
                -you see the subject while you take the image

                of course there are also many advantages of DSLRs like fast AF for example

                • 1. Readily achievable with any Nikon with a crop mode (all FX bodies, D2X(s), D7100). D810 in 1.2x crop mode is ideal for this and still exceeds any RF for IQ/resolution.
                  2. Readily achievable with other mirrorless solutions, generally for the cost of one of those Leica lenses, and with comparable IQ.
                  3. Myth, while better than an SLR OVF in this case, RF’s are notably inferior to LV/EVF in accuracy (the only optical method which can match LV is a view camera with a magnifier, which always was the most accurate method of focusing before modern LV). Given the D810 already requires LV for accurate focus, it will beat any RF here. And given the limitations of the Type 240’s LV implementation, the RF’s can’t even match performance in LV.
                  4. Nice, but not significant given modern blackout times.

  22. Great Post, followed by a great discussion… thank you Ming!

  23. Typo: Hasselbald. Funny. Is that one of the Sony clones?

  24. So I’m trying to rationalize this. Based on your essay here, I’m a prime candidate to move from Canon over to M43. I currently shoot older Canon APS-C cameras (40D and 50D), largely kept them this long because I also use them underwater and the replacement cost for the UW gear is high. I’d love to have something smaller and with better IQ. Smaller is very attractive as smaller translates from the camera and lenses to the housing, etc. So, I lean toward the Olympus M43 offerings like the E-M1 and the E-M10. Seem like good cameras. Reasonable lens offerings for above and underwater. Housing support for underwater use. I just keep reading things like “fatally flawed” and “not quite there” in other posts on your site as well as other places. Shutter resonance, slow buffers, etc. I don’t print huge and don’t have a ton of time to spend on post processing, but enjoy equipment that works well. What is it that keeps me from entering the world of M43?

    • Probably the cost and availability of the UW gear, I’d guess – but that’s a very specific use case.

      My output requirements are very different to most people’s, so you might well find that M4/3 is just fine for you – and will be better than your current starting point, probably. But you need to remember it’s all relative: I’m coming the other way, from medium format…

  25. A strange thing happened to me yesterday. I erroneously set up my laptop screen saver to run all of my photographs. As the computer ran image after image, all from a wide assortment of cameras and lenses, from different manufacturers, and various resolution and size sensors, I became mesmerized by the flow of images. As it was a screen saver, I could not quickly zoom in for pixel level analysis, but rather, had to just view the images for what they are. It amazed me that there was no camera/lens relationship between between the great, good, and mediocre images. While the images went by on that 15 inch display, it became apparent that the important factors were, in this order, composition, color/tonality, and exposure; resolution did not seem to matter. Great photos were rendered by any number of camera/lens combinations. The lowest resolution camera in the bunch would have been the D700, the one camera, if any, that seemed to render the greatest number great images.

  26. Hi Ming

    Great post and comment thread. I see the Nikon 24/3.5 PC-E has made your regular equipment list. Have you considered doing a post on your experiences with this lens (or tilt shift in general)? I have this lens and struggle with getting good results. Things focused carefully using LiveView on my D800 in the field are disappointing when viewed back at my computer. I also have the 85 PC-E and it is terrific. I know other bloggers like Tim Ashley have had issues with the 24 PC-E and would appreciate your comments.

    Steve

    • I reviewed it some time ago here. Firstly, there’s a lot of sample variation; secondly, it has field curvature AND focus shift, which makes things even trickier. I find that in general if you focus on the most distant objection your scene wide open and stop down to f8-11, then the plane comes forward and ends up about where you’d expect it. The corners are poor though when shifted or tilted, and I wouldn’t use anything beyond about 7mm on a D800/D810. And that’s before we even begin to talk about how disastrous the ergonomics are (tiny ineffective lock knobs, too-coarse gearing on the tilt axis). The 28/3.5 PC is actually not a bad alternative.

  27. Fun read. Sometimes I really don’t understand the general negative mood among photo forum frequenters about the state of camera systems. It’s as if up until the zeiss otus lenses were created, all images were inferior and high resolution prints were not possible. Some people seem unhappy until they can create gigapan (http://gigapan.com/) images from a single exposure. I agree that there are certainly advances to be made for professionals and enthusiasts in the camera world, but, by and large, several professional systems offer an abundance of lenses to choose from (for whatever you want, 8mm to 800mm) and any professional would be fine using a 5 year old DSLR with lenses from the last decade. I’m not sure why there is this general consensus that there is a dearth of complete photography systems.

    • Well, you can’t stitch moving objects, for starters 🙂 Also, higher pixel level quality will usually appear perceptually better than more but weaker pixels.

  28. Ming, if you have the time, I would like to hear your thoughts on the Sony A-mount system. I have no personal experience with it, but on paper it looks rather good. The lens selection is also quite extensive. Maybe in the future Sony will make a mirrorless A-mount body with a good EVF, I believe that would make one great camera.

    • Lots of legacy lenses – but no idea how many of those perform adequately for high resolution digital. I suspect it’s much like Nikon’s legacy lens selection…a mostly mixed bag and don’t expect great sharpness wide open, but there are probably some pretty good exceptions. I haven’t had enough experience with this system to quote specific lenses.

      • Thanks for your comment. Sony does make great sensors and from what I hear, the A99 and A77 are pretty good cameras. At least until today, Sigma Art series lenses are available for the system, so at least primes are covered when Sigma completes the series with a 24mm and a 85mm. Past these, all I require is a good macro and a good zoom up to 400mm.

        If there are any A-mount users here that would like to chime in, I’d like to hear what you have to say.

        • I would disagree with Ming here, imho he is dismissing Sony way too quickly for lacking lenses, yet in m43 land, they have “a large choice”?! In any case, the way I see it, a large choice is not very relevant unless you have no idea what you are going to shoot in the future. If you know your preferences, then it surely is straightforward to choose the system that fits your needs and for some this would be Sony. There are excellent lenses in a-mount (Zeiss and G) and you have access to some very good Minolta ones. Here you can see all the options:
          http://www.mhohner.de/sony-minolta/lenses.php

          • They are lacking in NATIVE lenses that perform adequately for the 36MP sensor. Nikon is in this boat too. And they still haven’t resolved the shutter vibration issue.

            • Vlad, thanks for the comments and the link, I will look into it. At least Sigma and Zeiss are there to cover all my prime needs, then it’s pretty much down to finding a good zoom. I’ll have to look at some second hand options, I don’t want to spend a fortune on another system. I’m currently using the E-M1 and I have all the lenses I need, but I would like to have a FF camera for some specific needs, and the possibility to expand in the future if needed.

            • You are either moving the goal posts for Sony and putting random conditions such as “NATIVE lenses that perform adequately for the 36MP sensor”, or you are dismissing the whole a-mount for some unknown reason.

              • No, I think we just have very different requirements for image quality. I don’t see why I need to buy and use something just because everybody else seems to like it.

                • Ming, I think you are talking about E-mount and Vlad is talking about A-mount. If you read back, Vlad originally responded to me about A-mount, when I was asking about it. I think lack of lenses and shutter shock are E-mount issues, not A-mount issues. A-mount must be evaluated as a completely separate system. Again, I have no experience with A-mount, but on paper it has some promise. The lens selection has modern Sigma and Zeiss lenses, so at least the primes should be very capable. Then it all comes down how good the camera bodies are. I wonder if the SLT mirror design eliminates or at least alleviates the focus issues one may encounter with traditional SLR mirrors?

                • iovokitangra says:

                  Huh? I am questioning your logic, this has nothing to see with your needs. You are writing an article that is supposed to be read as an advice and while quite good overall, I do not understand the reasons for excluding a-mount. Here is what you are saying: “On the full frame front, you have four choices. (…) Sony again is a non-starter because of lens selection.” There is a problem here.
                  1. If you are speaking about e-mount, sure, they do lack lenses, but then where is a-mount? Did you forget about it?
                  2. If you are speaking about a-mount, then you are dismissing Sony for reasons that do not apply to M43, because you state that M43 has a “large choice of excellent lenses”, yet it is not any larger than Sony’s. So, you are moving the goal posts. Or, maybe you have some other reasons for excluding a-mount, but you didn’t write them down…
                  3. Then, in your comment above, you are telling me that you are excluding them because “they are lacking in NATIVE lenses that perform adequately for the 36MP sensor. Nikon is in this boat too.” If Nikon is in this boat too, why isn’t it excluded too?
                  For me, it is unclear what you are trying to do and say. You are presenting the article as an advice and then defend what you wrote in it with your personal needs. And, on top of that, you don’t follow your own logic.

                  • No, I don’t think you understand the article or the comments. There are a lot more native mount M4/3 lenses with full functionality than sony. Go count them. Nikon lacks Nikon branded lenses that perform adequately, but Zeiss fills most of the holes with the right mount and no adapters. Sony does not.

                    Besides, if you need validation for your Sony purchase, there are plenty of sites for that. I care about the means only as far as it enables the end result, and I can say categorically that it does not enable the end result I want. I don’t see why I need to waste time or money buying equipment to prove or refute it to you. If it works for you, great. That’s why there are choices, both in equipment and photography sites. We are free to decide.

                    • Ming, Vlad is talking A-mount, you are talking E-mount, therein lies the rub. A-mount does not require adapters either, it has native Zeiss and Sigma glass. So, you guys are arguing for no reason, you are talking about two different systems!

                    • I’m not sure that’s the case. He’s arguing for 36MP options, of which none exist except in E mount.

                    • iovokitangra says:

                      I don’t understand the article, for sure. I merely disagreed with you on something and look where we are. I explained my arguments in a rather civil manner, yet you are being patronizing and entirely ignoring almost everything I said. You do not need to waste money or prove anything to me and if this article is simply a description of what fits best your needs, then great. But you present it as an advice for others. Which is it? Certainly, it can’t be both.
                      M43 has 43 native lenses, not counting third parties, out of which eight are basically the same kit zoom (14-42/45), another four are basically the same long zoom (14-140/50), another three are the same long zoom (40-150) and yet another three are the same even longer zoom (75/100-300). And add to that even more doubled lenses between Panasonic and Olympus. Is that the “large choice of excellent lenses”?

                    • Well, it IS both. If you want to produce the results I do, and judging from traffic and emails and everything else, there are some who do, who happen to read the site, then it is also advice.

                      M4/3 has a lot of consumer zooms, which is true for every system, so let’s disregard these. But it also has: 10/0.95, 12/2, 17/0.95, 25/0.95, 20/1.7. 25/1.4, 25/1.8, 42.5/1.2, 42.5/0.95, 45/2.8 Macro, 45/1.8, 60/2.8 Macro, 75/1.8, 12-40/2.8. 12-35/2.8 IS, 35-100/2.8 IS, 45-150/2.8 lenses. That’s 16 lenses that cover pretty much everything you need, and I’ve probably missed a couple, too. There are different reasons to have the 0.95s or 1.8s or 2.8s (macro, price, video, size). Where are the Sony equivalents? Where is the choice?

                    • You are getting confused. You might try reading the above posts again, they are talking about the Sony A mount, where there are many more native lenses than are available for M43 (if you include all the old Minolta AF lenses). They are not talking about the Sony E Mount, which is the mount on the A7 series of cameras, for which, obviously there is a dearth of lenses. Sony’s mount and camera nomenclature here is also confusing.

                    • Where are the 36MP options the poster mentions for A mount? Only E.

                    • Well, you are addressing consumers, amateurs etc., so I assumed that you are giving advice in general, not only for those wishing your results.
                      “But it also has: 10/0.95, 12/2, 17/0.95, 25/0.95, 20/1.7. 25/1.4, 25/1.8, 42.5/1.2, 42.5/0.95, 45/2.8 Macro, 45/1.8, 60/2.8 Macro, 75/1.8, 12-40/2.8. 12-35/2.8 IS, 35-100/2.8 IS, 45-150/2.8 lenses. That’s 16 lenses that cover pretty much everything you need, and I’ve probably missed a couple, too. There are different reasons to have the 0.95s or 1.8s or 2.8s (macro, price, video, size). Where are the Sony equivalents?”
                      Before I start, M43 is an amazing system. Now, here are some Sony lenses – 16/2.8 fisheye, 20/2.8, 24/2Z, 35/1.4G, 50/1.4, 50/1.4Z, 50/2.8 macro, 85/1.4Z, 85/2.8, 100/2.8 Macro, 135/1.8Z, 135/2.8 STF, 300/2.8G, 500/4G, 16-35/2.8Z, 24-70/2.8Z, 70-200/2.8G, 70-300G, 70-400G. That’s 19 lenses, also covering pretty much everything one needs and stabilized. You also added Voigtlander above, so I will add third party too. Schneider TS 50/2.8, Schneider TS 90/4.5, Sigma 35/1.4 Art, Sigma 50/1.4 Art, Sigma 105/2.8 macro, Sigma 150/2.8 APO macro, Sigma 300/2.8, Sigma 500/4.5. We are at 27 lenses here and I haven’t even finished with all of Sigma’s offering or even mentioned Samyang or Tamron. And there are some excellent Minolta lenses that one can find used and which won’t require an adapter, such as the 85/1.4G, 200 APO G, 300 APO G etc. As you can see, I still disagree, although it is only about excluding Sony from the FF possibilities.

                    • I’m now pretty sure the other posters are right and I’ve misinterpreted you: this isn’t E mount, but where are the 36MP full frame possibilities? Yes, there’s adequacy for the 24MP cameras (I haven’t used the lenses, so I wouldn’t know for sure) – but why downgrade?

  29. As a happy MFT user who fits the description well I couldn’t agree more. However I think Pentax dslr’s would deserve a mention in such article given your comments on lack of apsc primes. I have no idea about quality though.

    • If you’re carrying a K-3 sized body, you’re really not far off a D810. And there’s an enormous difference in image quality there…

      • Ming, could you elaborate a bit on what is the difference in image quality between K-3 and D810? Is this something inherent to the ASP-C sensor, or caused by the available lenses, or both? Perhaps you could also compare image quality of two cameras you have recently used, GR and D810?

  30. @Ming Thein
    What would you think of a sleeker design for the Pentax Q… with say a set of theoretical F/0.5 lenses? It should approximate F/1.8 on FX in light collection abilities (by my very rough calculations) and be much more compact than any comparable M4/3.

    Would it make a good travel companion for pros/semi pros with aperture fever? 😀

    • Lens speed doesn’t overcome sensor limitations such as dynamic range and acuity, and they’d be very difficult/expensive to design with adequate optical quality. Actually, very few pros have aperture fever…

      • Agreed on the dynamic range, but acuity/resolution is actually correlated with aperture size and focal ratio – assuming the lens is properly designed of course…

        Regarding the aperture fever I get what you mean… I’m only freelancing and I have 2 Sigma ART primes. I find excuses not to bring them to shoots. I guess ARThritis gets to me before the fever does.

        I was hoping to get the large aperture and low light performance without a DSLR body actually

        • Acuity only correlates with aperture and focal length as far as it affects the design properties of the lens. You don’t get more acuity from a planar subject with an Otus at f1.4 or f5.6, for example.

      • Agreed on the dynamic range, but acuity/resolution is actually correlated with aperture size and focal ratio – assuming the lens is properly designed of course…

        Regarding the aperture fever I get what you mean… I’m only freelancing and I have 2 Sigma ART primes. I find excuses not to bring them to shoots. I guess ARThritis gets to me before the fever does.

        I was hoping to get the large aperture and low light performance without a DSLR body actually

        • Greytan, you also have to consider the acceptance angle of the sensor too. Even BSI sensors don’t have that big an acceptance angle though they have better acceptance angle than FSI sensors.

          Even if you built a 0.5/f or even 1/f lens it would vignette (and worse, false color) in the corners on current sensors.

          • Hey Kevin, that’s an interesting thought. Do you have a resource to reference that? I’d like to find out more. I was under the impression that if the flange focal distance is suitably short, it should alleviate that problem. Speed Booster made a focal reducer for the BMPCC, I wonder how they got around the problem, since they claim to be able to achieve F0.7 maximum aperture when attached.

  31. Ron Scubadiver says:

    Very thoughtful, but a bit depressing about the DSLR’s.

  32. There’s probably one more category for the ones that might like that kind of ‘fun’ – film photography 😉

  33. “…. because frankly, the compression and shutter vibration issues of the A7R make it pretty much unusable” I have been using this camera for professional work all this year and it has turned out to deliver amazing results. Is this statement put out with after actually using this camera ?

    • I spent a week shooting with it. I wrote a long and detailed review with the conclusion that the idea was great, but for what I do, there’s no resolution gain over a 16-24MP camera because most of the time my shutter speeds fall into the danger zone and produce double images. I also have tonality issues when processing because of the raw compression. These are very visible in print as posterization.

  34. Someone needs to make a tripod that unfolds into a crutch. You can bring this into the plane with no penalty and it frees weight.

    • Actually, you can being tripods on with no penalty now so long as they aren’t too big and clearly not pieces of luggage in their own right….

      • Interesting. I’m curious how you’ll carry the RRS 24L. I was considering the regular 24 because it fits inside hand-carry luggage, but it doesn’t have a good working height for me. So do you strap the tripod to the outside of a bag, for example?

        • I’d just carry it inside a separate tripod bag (theirs are pretty good, actually) or check it in inside a larger suitcase. I agree the regular 24 doesn’t quite have enough working height, especially if you need to have one leg downhill. This was/is one of the things I love about my Gitzo 5562, but that thing is seriously heavy!

          • Thanks. I’m going to be in their area in about a month, and I’ll drop in to take a look at all of their tripods. It could be an expensive visit …

  35. Ming.. thank you, thank you, thank you for shedding some light upon the photographers technical requirements and his/her sane choices sustaining the ambitions and goals with the whole thing.
    Over the last year perhaps I’ve found myself drowning in thoughts about the right system for me. Not to heavy to lug around and sufficient technically to cover my needs. Thinking wow, what if I had a FF DSLR as in the past that could offer the insane high transparency the OTUS lenses and perhaps Sigma ART too, wouldn’t I just droll more over my prints? Wouldn’t I just be able to make more impressive images? Perhaps your article here just cement it finally: No I don’t need 10 kg of new gear and new shooting challenges than I already have today. I honestly do not have the need when I am looking at the typical print size I make at present.

    My EM1 is truly a bastard to get the best out of, but it is possible. Working intensively to match the JPEG based histogram to the RAW output allows for very precise ETTR. Using the highlight warning in live view or through the OVF, I can output a RAW file that is not only showing low noise, but also pretty satisfying raw material to withstand even heavy PP if required for the outcome.
    Thank you Andre mentioning my EM1 images on my Flickr, but there are actually also a few examples of muddy or extremely noisy images. These are mainly the images that are shot in extremely contrasty light where no highlight clipping were allowed and (blocked) shadows falling where they fall and then brought up by dodging in PP They are pretty noisy and and shows posterization and halos. These are examples when you ask for more DR than the camera actually can handle. OK these are made for training spot metering mainly. Normally I would not challenge the cameras DR that much.

    As for meeting my wishes for > A3 prints I did invest in a DP 1, 2 and 3 Merrill system a short time ago. I was reluctant to buy this set due to the extremely narrow shooting envelope and not at least the worse ever RAW developer SW. But now since I got these cameras I am also going to use them from time to time. There is no available 3 or 4 focal mirrorless high quality system with fixed optics weighing at max 1,5 kg or so which is so regrettable.

    At the end of the day I am pretty satisfied with the EM1 and the lenses I normally use. I love the 43 format since I shoot 70% of my work vertically. Should I wish to use super transparent glass I mount my Pan Leica 42,5 f1.2 Noct wich has it in spades and partly so does the Pan Leica 25 f1.4. My 24 f2 is another ever-go lens too but it is no high end lens at all.

    Thank you Ming for helping me clarifying I do not need a 810 and 3 or 4 OTUS lenses 🙂

    • **Thinking wow, what if I had a FF DSLR as in the past WITH GLASS that could offer the insane high transparency the OTUS lenses and perhaps Sigma ART too..**

    • Hi Gerner,

      It is great that the E-M1 is meeting your requirements. I wish I were in the same boat. However, I found that even with ETTL the sensor just couldn’t reveal enough detail in the shadows when shooting tricky landscapes or high contrast scenes (which seems to be quite frequently). I longed for something with much better dynamic range, better noise handling and a little higher resolution.

      I checked landscape images online to see if others were getting better results and it soon became apparent that no landscape photographers were using the E-M1. All were using full-frame systems, or higher.

      I was about to spend money on the Olympus 75mm f1.8 and the up-coming 40-150mm f2.8 PRO zoom when I stopped to consider my investment. I was at a point I had been before I had purchased the E-M1. I was again considering full-frame.

      It was then that I longed for an E-M1 with a full-frame sensor (albeit probably slightly bigger to accommodate the IBIS). This system would do me perfectly. Full-frame OM-class lenses without image stabilisation could be smaller and simpler, and a whole host of old glass could be used and be stabilised. Wonderful. Flexible.

      As Thom Hogan said, a full-frame E-M1 would be a logical progression for Olympus. They have a history of making excellent full-frame glass that would help them in this endeavour.

      Until then it might have to be a Canon DSLR for me (if I can get past the OVF).

      • err, ETTR. Where’s the edit comment button for those with WordPress accounts?

      • I am afraid a FF EM1 spec’ed camera will become a lot bigger. IBIS and FF will result in huge .. very huge lenses.

        As for landscape I thought my Merrill system will do on a pano head 🙂

        • I’d love to take a look at the IBIS technology in the E-M1. Maybe I should have just pulled it apart instead of putting it up for sale…

          • 🙂

            A complex magnetic gyro system that even in MFT format takes up quite some place. Hence a FF IBIS will be enormous and require a big mount.

            BTW I had a 6D with great Zeiss and ART optics and loved it.

            • It’s funny. I am hesitating to jump into a DSLR because of the OVF (and size), whereas most experienced photographers are hesitating to jump into a mirrorless because of the EVF (among other things).

              Did you bail from the 6D because of size primarily? That Sigma Art glass looks awesome and would be a sure thing if I wasn’t chasing small size for street (Voigtlander) at least on the a7/r.

              • Yes I bailed from the 6D due to the wish of easing my total body weight. Excellent FF DSLR lenses easily reach 4-5 kg.

                • You must mean total system weight, Gerner – even if I have a D810 and two Otii, that’s a little under 4kg…

                  • Ming.. I carried the 6D + 5 big glass lenses and some of the time a tripod so I could shoot without my shivering hands spoiled the shot. Handshake belongs to the past with the EM1 … I don’t have any lens longer than the 75mm.
                    Hmm.. 810 + 2 OTIS = 4 kg, doesn’t sound too heavy though. When I lug my whole EM1 system (which I actually never does) it’s about the same. Normally I carry the 12, 25 and 42,5. That’s an easy basket to handle.
                    I would really love that 810 machine and 3 OTUS’s to be mine, but I am unfortunately not getting much younger or stronger along the road. Such a dreamset would cover all my needs for larger prints and I wouldn’t be tortured and robbed off time by using those Merrills. Those cameras have now been staying unboxed on a shelf for quite long. I can’t pull myself together to rig them up 🙂

                    • The Sigmas are worth all their trouble for the picture quality — you’d have to use a Nikon D8xx/Otus to get anything appreciably better. But if you thought the E-M1 was tricky to expose correctly, the Sigmas’ almost random metering combined with their unforgiving exposure latitude will drive you crazy!

                    • **The Sigmas are worth all their trouble for the picture quality — you’d have to use a Nikon D8xx/Otus to get anything appreciably better. But if you thought the E-M1 was tricky to expose correctly, the Sigmas’ almost random metering combined with their unforgiving exposure latitude will drive you crazy!**

                      The FF ART lenses are stunning transparent. They also outresolve the sensor in the 6D by far. My Zeiss 85, Sigma 85 and the Zeiss 21 Distagon (which I still have) are really awesome performers too.
                      If I ever will unbox those Merrills, I guess I will, I’ve been taught I will have a hard time with them. I truly believe it. Some have said better not to take them into the harsh light zone but stay in softer light conditions. The SPP RAW developer and the quirks it has, the following work with the TIFF’s in Lightroom and finally PS will probably also drive me mad. So if the award isn’t stunning IQ afterwards they’ll go to another hopeful soul that is gifted with a far more patient mind than mine 🙂

                    • Manual plus eyeball, or manual plus Lumu

                    • Ah yes, the tripod – forgot that…

                      Surprised your E-M1 kit is that heavy though. I would have thought closer to 2kg.

                      I’d also like the 3rd Otus, but there isn’t one yet 🙂

          • You could pick up a used E-M5 for cheap for that purpose. It’s quite impressive and magnetically suspended…

            • Sounds like you’ve already dissected one, Ming 🙂 I trust you got it back together again, err, mostly 😀

              • Actually, Oly had a demo sensor/stabilizer mechanism out of its cradle at an expo here – no cameras were sacrificed in the acquisition of knowledge…

        • I think the same way – a 35mm System wouldn’t make any sense for Olympus at the moment because they are one of very few who really follow their principals of keeping the system about the half size and weight of a 35mm equivalent and they’ve developed a great lens selection for it.
          Over the years they’ll benefit from the technology that gets better and so do the sensors. They’re already better than many of the larger sensors we’ve had a few years ago. When I try to get some details out of my old Canon files I very quickly run into the typical banding issues they had – an Issue I don’t have with my Olympus. But our goal should always be to nail the exposure as perfect as possible – on any system. I think to some degree the benefits of raw made many of us also a bit lazy – me too. Sometimes It’s good to look back what outstanding images photographers have done with the digital cameras we’ve had a few years ago, or what impressive work they’ve done on slide film which had much less dynamic range than most of today’s cameras. In most cases we are the limiting factor of our own work, not our tools.

          We are living in a wonderful time with many different systems/tools that cover different needs and each is great in it’s own way. Personally I appreciate that a lot and I would hope for even more diversity rather than hoping that a company, that truly makes something different, just follows the others.

          • I agree, Leif. Although, I’m not suggesting that they replace the m43. Instead I think they might be able to introduce full-frame for those that would like to use the Olympus system that they may have gotten to know from using the m43 system. In much the same way that Canon and Nikon have both APS-C and full-frame systems. Heck, they even have 1″ systems (Nikon 1).

            I don’t think it is beyond Olympus to create a full-frame IBIS that could be only marginally bigger than its m43 counterpart. Only time will tell I guess. It is fun to speculate (and dream), too.

      • That would be nice, but it would also mean new mount time…and those lenses aren’t going to be small or cheap.

        • Yeah, I see what you mean. Both Nikon and Canon both use the same mount for their APS-C and FF cameras.

          However, Sony has gone ahead with two mounts. And, of course, Canon has had to design a new mount (EF-M) for the EOS M APS-C camera. Here is the question: Will Canon use the EF-M mount for their eventual FF mirrorless, or can they integrate their existing EF mount into a mirrorless body by keeping the flange focal distance the same so that EF lens owners can continue to use their investment on new mirrorless systems without requiring an adapter?

          If Olympus does create a new FF mount surely they can build a simpler and somewhat smaller lens by not needing to include image stabilization like everyone else?

          • I don’t think the EOS-M spec is provisioned for full frame. As for Olympus, the three things that affect lens size are telecentricity, maximum aperture and image circle – so a fast wide prime that has to clear a mirror is always going to be pretty massive regardless of IS or not.

            • Yet Olympus f2 lenses – 21/2, 24/2, 28/2 and 85/2 (L48mm x D60mm 260g) were all comparatively small, and certainly not massive. All of these lenses are pretty good across the frame on Sony’s 36mp sensor, though their widest stop is usually low in contrast. Technology is now 20 years on, at least, and if new lens design techniques and new glass (aspherics, ED) were used to improve/tweak those nicely diminutive lenses, without going in the Otus direction of super sizing, the technology could well improve on already good designs, keep them small and produce some excellent glass.

              • The Otus is huge because it performs at that level everywhere in the frame at f1.4. If it were f2.8, it’d be significantly smaller – but I was told that the market won’t accept an f2.8 prime lens even if it’s perfect, even if you have to stop down normal f1.4s that far or more to get the same results and the wider stops aren’t that useful. I don’t consider a lens adequate if I can’t use all of the apertures – you might as well not have them.

                Sadly, the lack of smaller, slower, but excellent lenses is something which doesn’t sit well with the current trend for marketing hyperbole. Personally, I would love to see smaller, slower lenses of the same quality since I rarely use f1.4 anyway – that would save my back! I’ve talked to several companies including Zeiss about this, but was given the same above reply…

                • I haven’t mentioned f2.8 lenses, only f2 lenses.

                  Admittedly, it is now possible to have faster lenses at the focal lengths mentioned – Canon, Nikon? And Leica have 24/1.4s, and Leica has 1.4 lenses at 21 and 28mm. Nevertheless the OM f2 lenses are small and good, with modern technology it would not be impossible to keep them to the same size and improve them.

                  An alternative, of course is to develop an M RF that has a beautifully integrated EVF for consistent, fast and accurate focusing of the small, fast, and very good M lenses. The Fuji Pro 2 will probably have something like this integrated into its hybrid viewfinder. Put that technology into a Canon made M FF RF and providing Canon does not go overboard with the sizing and it could be a reliable, ergonomic IQ machine. Of course, it is much more likely that Leica goes in that direction, which would be very nice, but then we will still have Leica electronics, stubborn and idiosyncratic ergonomic decisions and all the unreliabilty and annoyance that comes with that.

                  Either way, though, it is certainly even a practical possibilty to keep the size of a high IQ machine down. Leica just needs to get its act together with a fully functional hybrid viewfinder – that would improve their edge and raise their sales.

      • First, Olympus isn’t going to introduce a full frame camera. Period. There’s too much competition, their imaging division is already hemorrhaging money, and frankly, I don’t think they need to.

        IMO, Olympus has made their bet. These new forthcoming PRO lenses are part of that bet. What’s the bet, you ask? New sensor technologies that are, by all accounts, right around the corner. Whether it’s Fuji and Panasonic’s organic sensor, or something similar Sony is working on … or equivalent technology, the next revolution in sensor development will likely provide all the resolution, DR, and high ISO performance in m4/3 sized sensors that 90% of serious amateurs and professionals will ever require … in a compact form factor. That’s almost certainly the bet they’re making.

        Since you quoted Tom Hogan, I’ll quote him again, since it dovetails with the aforementioned point rather well, I think >>

        “Each bump in sensor size adds parts cost and ultimately adds to the consumer price. We’re already to the point where additional improvements in full frame sensors really don’t add much to the performance as cameras are generally used, especially for JPEG shooters. If the need is 24mp, the differential between APS/DX and full-frame/FX for most users in most situations is marginal. Not that there aren’t any, but of the 16 million folk that bought DSLRs last year, I’ll bet less than a million would be getting anything different in their shots from a 24mp FX camera than with a 24mp DX camera.

        And that’s part of the problem. Much like CPU clock frequencies eventually didn’t make for easily distinguishable differences between most customers, the same thing happens with sensors at some point. Price then becomes more of a driving factor, and smaller sensor size means lower price.” [Not to mention much greater portability.]

        Second, understand that Canon sensor architecture is about 5 years old and hasn’t really been updated. I wouldn’t jump into Canon full frame right now. If you can afford to wait, I’d hold off for another 6 months to a year.

        • I believe Olympus and Panasonic are very aware what their bets are and I also believe we’ll see a new generation of sensors that outperforms the existing. It makes me feel comfort having invested in many nice MFT lenses. My EM1 will neither last forever because of physical wearout so hopefully 2015 we’ll see some IQ breakthrough with MFT sensors.
          As for all formats we do not know where it all ends, but to my humble measures cameras still have to be kept of a certain size to be comfortable to use and EM1 in this respect have all my UI and comfortability wishes checked.

          • Next year’s rumored E-M5 successor will be an iterative improvement, I think. We’ll have to wait one more generation, I suspect, to see a serious leap forward in sensor tech. But I’m convinced it is coming. Probably in 2016 or 2017.

          • Well, that didn’t take long >>

            http://www.photographybay.com/2014/06/23/canon-is-developing-a-new-organic-compound-for-sensor-and-optics-technology/

            Pretty fascinating read. My takeaway: Fuji & Panasonic, Sony, and Canon are all quietly in an ‘arms race’ to develop, perfect, and introduce this new tech. It should also be a way to claw back some sales in a flagging market … and further differentiate themselves from smartphones.

            I tend to agree that increased DR and low light performance will be the new megapixel race, since we’re already reaching the saturation point for resolution in the sweet spot of the purchasing market.

            In the meantime, backside illumination might be something Olympus implements in the E-M5 successor.

            As I said: lots of good reasons why Olympus/Panasonic will stick with m4/3.

    • You will if you start printing 🙂

    • Thanks Gerner. I guess it all comes down what you do show and don’t show. 🙂 I’m really hesitant to raise shadows very much — maybe a bit to reveal a tiny bit of detail. I much prefer to block them up completely. That doesn’t work for everyone’s style, which is why we all have permission to use 5 different camera systems!

    • One more question, Gerner: how do you have your E-M1 setup for ETTR? I’m using Pekka Potka’s method of setting the camera up for the flattest JPEG (muted picture style, -2 contrast and saturation), and setting the highlight warning to the highest tone that I still want detail in (245 for Pekka’s printing method, 240 in practice for the paper I use). UniWB seems like a giant pain to setup and use.

      • I have followed Pekka Potka’s advices for setting the EM-5. I don’t believe there’s any difference between the 1 and 5 in this respect. I’ve set the highlights to 245 and the rest comes from practice. I assume one could achieve as clean files without adjusting the camera just by getting to know your JPEG meter well. I was too nervous and impatient to exercise that and found his settings excellent.
        I haven’t tried to set up is UniWB method for RAW development, perhaps there’s a tad more to gain?
        Anyhow metering to the bleeding edge is a deed. (I found out lately).

        • If you look at RawDigger graphs, then UniWB will get you closer to the edge than Pekka’s method. Whether that’s useful or not depends on so many factors. The DPR discussions I’ve read don’t talk much about the metering mode (I spot meter a neutral tone object in a scene that’s got the right light for my intended composition) or the exposure intent of the frames they’ve been showing, so it’s not clear what they mean by correct exposure other than the artistically vacant goal of getting RawDigger graphs all the way to the right.

          Sure you could get less noise if you ETTR all the way to the margins, but in practice, for my uses, if I ETTR slightly more than the meter says without going over, then I’m happy. This may not work for people like Ming whose work depends on achieving the highest technical levels, but it’s good enough for me. I have not found that kind of noise to be very noticeable in the prints that I make (A4-sized).

          What’s more noticeable for me is the separation of tones at the extreme ranges. If you haven’t done so for your printing yet, you may want to construct a printer threshold test chart as described here: http://www.luminous-landscape.com/tutorials/beyond_calibration_2.shtml (The article as well as his first Beyond Calibration article are great reads, too.) They’ve really helped me start to get a handle on my shadow tones, which tend to be a bit muddy, and how to process them for printing. Basically, I’ve found that I need fairly aggressive steepening of the tone curve in the areas where I need separation, which then requires multiple curves. I’m now discovering that LR alone is unfortunately not going to be good enough for printing, which is too bad since it’s got great facilities for organizing print trials.

          If you like, I think I have the PSD file for my test chart somewhere, and I can put it on Dropbox for you. I didn’t do the full tone series, but just the extremes, and even then it takes the good part of an hour or so to construct. It’s also enlightening as to how different printers really are.

          • Andre… yes this pretty wizardry. I’ve got as much under my skin as I could reading Lloyd’s ETTR articles, so his RAW-digger scribbles.
            I’ve not come across LuLa’s article, so thank you for that.

            Using spot metering as I do for most scenes demands care. I. e. metering at the adjacent area just between highlight and shadow goes often either straight over blown or to heavy low key, if we can’t muster the control holding our lenses still. I hold my breath when metering for such eventual contrasty scenes. The highlight warning is my only religion here. But I am not so bleeding edge oriented that I shoot maybe 10 shots of the same scene just to envelope the most ETTR’ed shot. I find boring to sort the images out coming back for the PP. I perhaps compromise a bit. Most important is to avoid highlight clipping, just as you say.

            I would be pleased to receive your PSD file if you have it around. Don’t tear your disk apart if you can’t find it. Than you Andre.

            • I understand that it is easier to darken shadows and mid-tones than lighten them, but if the light is right and you can expose correctly for the final result why would you ETTR? And, if the highlight is strong, as in direct light strong, then you are forced to deal with the clipping and pull it back down from blowing past the right.

              Many times you cannot expose to the right because the exposure would be too long. If the only option is to raise ISO then I would choose not to.

              So, realistically, when would you get to use it? I guess you can use it if you don’t have a high contrast scene. When the light isn’t making the magic happen. When there is a lot of shadow and not much highlight. When you don’t have much contrast to begin with. So, it is a limited form of HDR to get a muddled image to look better?

              Clearly, I don’t get it. 🙂

            • Hi Gerner, hopefully you’re still following this thread! Here is a link to the threshold chart: https://www.dropbox.com/sh/f0x02u80dib83a1/AAAITXni9QbPJJIlmppiFvjha?dl=0 Anyone else should feel free to download it too if you find it useful. I’ve made it for a 300DPI 8×10 inches printout. Resolution isn’t too important here.

              The numbers are the luminance values in the tone curve tool in PS. 255 is pure white, and 0 is pure black.

      • I do the same, but you can actually go all the way to 255 because there’s some – about half a stop to 2/3rds – headroom in the raw files for recovery. I think you might find that makes quite a bit of difference.

        • I’ll give that a try, thanks. The common justification I’ve read for Olympus’s (and others’ too) overrating of their ISO is to leave some highlight headroom, so it makes sense that there’s still a bit above 255, assuming the JPEG processing hasn’t eaten up all of that margin.

  36. Ahh, what a timely article, Ming.

    I’ve just put my E-M1, with 12-40 f2.8 PRO, up for sale so that I can fund an upgrade to a full-frame camera.

    I am chasing image quality and flexibility for landscape, cityscape, and street shooting (day and night). A fairly general set of requirements one might think except for the fact that landscapes and cityscapes are at the opposite end of requirements to those of street shooting (especially hand-held night shooting).

    My first thought was to get the Sony A7 and match it with Canon L lenses. This thinking was influenced by the number of landscape photographers that are using the Metabones adapter to mount their Canon L lenses quite successfully to the A7r. The problem is that the glass is a bit big on such a small body for street shooting. Even the smaller L primes.

    So, I added the Canon 6D to my short-list, as this DSLR is relatively small and light and has truly excellent high ISO capability . I’m not sure about dynamic range.

    I had thought that the ideal setup would be the A7 with fully manual lenses for street (Canon FDn, Olympus OM, or Voigtlander M-mount), and a Canon zoom or two for landscape (the Canon 70-200 f4 L IS USM comes to mind).

    That was before realising that the A7 doesn’t have very good high ISO performance (compared to the 6D, which is noticeably better than both the D610 and A7, or so it seems).

    So, I then consider the A7R due to its excellent dynamic range and reportedly good noise handling, which seems to outperform both the A7 and 6D, resolution notwithstanding (not sure if the A7R is better than the A7):
    1. I considered using Voigtlander primes adapted with a Metabones Speedbooster on the A7R to shoot in cropped mode to improve performance for street work and keep the file size down for these types of images, and using the Canon lenses in non-cropped mode for landscapes and cityscapes;
    2. Then I considered the setup above without the Speedbooster. But this adds more complications around crop-factor focal lengths and the question about whether or not the ISO and DR benefits of the 36MP sensor are still effective in this mode;
    3. And then there were questions about lens speed requirements due to there being no stabilisation on MF lenses, and, of course, shutter shock.

    So, I go back to the Canon 6D as an overall solution and it seems to be a great compromise, except for the fact that it is a DSLR and, as you pointed out, Ming, it doesn’t have an EVF that can show the new user how their adjustments affect the shot in real time. No blinking highlight and shadow clipping, no live histogram, no live DOF preview (and when you use the DOF preview button it just tends to make the OVF darker), no manual focus assist… there are probably a few other things, too.

    However, I figured that these issues would fade away fairly quickly as I gained experience and moved away from what I knew (and loved) about the E-M1.

    Another consideration I had was that when Canon release a worthy mirrorless system camera as a DSLR alternative my L lenses will still need to be adapted until they are replaced with mirrorless equivalents. The alternative is that I sell them (along with many others who may wish to jump onto the new technology). Although, adaption of Canon onto Canon might not be so bad. It certainly has to be better than through a third party like Metabones.

    So, without buying and trying, it looks impossible to theoretically put together a system to cover my needs (which are also still evolving).

    The end result is that perhaps I should buy, as much as I can afford, both cameras (the A7r and 6D) and a range of Voigtlander and Canon L zooms to match, and just go for it until one reveals the solution. Then sell the rest or use them alternatively as circumstances require.

    Cheers,
    Nick

    • E-M1 with the range of excellent primes available can surely cover that range can it not? Buying 2 different systems to to fight it out in an arena seems like a waste of money, but mainly time which could be better focussed on shooting.
      Sounds complicated.

      • I agree, Ben. It is definitely a waste of money and at this point just a consideration.

        I did consider that perhaps multi-shot HDR might be a viable way to get around the dynamic range and noise issues and allow me to take some good landscape images. However, as Ming says, there is much more to IQ than just dynamic range or sharpness. If sharpness was the primary factor for IQ then all those not-quite-sharp classic lenses from yesteryear wouldn’t resolve images that people adore.

        The thing that kicked off this upgrade was the thought that if I’m going to pay $2K (which, comparatively, isn’t that expensive) for an equivalent 80-300mm f2.8 lens with a 1.4x TC (and another $800 for the 75mm f1.8) then I should be sure about committing to the platform. When I saw the IQ possible in landscapes taken by full-frame cameras I just caved in and rethought my whole direction.

        Guys were taking shots directly into the sun and still getting good detail in the shadows. Something impossible with the E-M1, which just turned everything black once you ETTR. Full-frame just gives a bit more to work with.

        Getting back on topic, I think I will consider the A7R as a system comprised of manual-focus Voigtlander lenses for street and the Canon zooms for landscape work. The Canon 135mm f2 USM L prime looks like it will find its way onto the A7R, too (which will give me something roughly equivalent to the Olympus 75mm f1.8 🙂 Although, there is an Olympus OM Auto-T 135mm f2.8 available for the A7R as well.

        I do like the option of using crop mode on the A7R if it proves to keep IQ high.

        • I’ve been reluctant to do much with the multi shot approach because you have serious issues with anything that moves – water, leaves, clouds, etc. Unfortunately there’s no way around this other than single capture…

          I wouldn’t use crop mode on the A7R – downsample instead…

    • Any of the 24-ish MP cameras will do the job, paired with a couple of fast primes…

    • John bresnen says:

      Hi. nick, how much for the em1..thanks

  37. Good post Ming. I wish Leica S had a 4×3 sensor crop! I really like the mobility of the camera, quality of the lenses and ergonomic of the body and I’m almost at the point to move from Hassy to Leica S, but I’m worry that lower megapixel and crop factor might make me regretted from the switch.
    I would like to hear your perspective on that.
    Take care

    • No reason why you can’t crop afterwards, but then you’re not much larger than a normal 35mm sensor…there’s always the Pentax 645Z – it has a better sensor anyway 🙂

      • Agreed! bigger sensor has its own benefits, also as you crop you loose more megapixel. I guess what make it is hard for me to move away from 4×3 is the look and feel of the photos as I shoot and how my brain processes the framing and composition. I was thinking to get the new S and use the Hassy (lenses) on Leica S, but to your point those lenses are not as magical as Leica lenses. I shoot with M rangefinder and the experience of shooting with M glass is magical 🙂

        I will try the Pentax for sure. All I need for the studio and location work is high sync speed, and two to three prime lenses. Currently I’m using Hassy HC 80mm and HC 150mm and that cover is well for fashion portraiture and beauty.

        How is the built and glass quality of Pentax MF lenses?

        Take care and keep up your awesome posts coming.

  38. A7R: I still have not seen any examples of shutter vibration handheld on the net. You responed in another post that Lloyd Chambers did extensive testing – but not handheld. Also; any real world examples of compression? I did try the OMD EM1, and I can not understand how you can talk badly about IQ from the A7R and live with the muddy files of the EM1…

    • Google for “sony a7r shutter vibration” and click on the images tab. You will see many examples.

      As for muddy E-M1 images, look in Ming’s Flickr pool for images by Matthew Stark or Gerner Christensen. There is no way their images could be considered muddy. Ming’s own images are also great examples of m43 quality.

      • They can be if you mess up exposure and then try to push it in post. ETTR is pretty critical with this camera, though frankly you need to be doing that with any camera to maximise image quality anyway, so I didn’t see it as a big issue personally.

      • I own an A7r and have to say that, for me at least, There’s is no denying that shutter shock does exist and it can be seen clearly as a double image in some hand held shots.

        That said I like the camera very much, it just needs care when taking critical shots.

      • Handheld? Please; give me a URL to such an image so I understand what to look for.

        • Sorry, I missed the handheld part. You’re right that most of the test examples out there are tripod-mounted using the lens’s tripod mount. Having said that, the developer of the bolt-on weight solution does say that there are problems handheld, too: http://www.josephholmes.com/news-sonya7rshuttershake.html He addresses this more directly in part 3, but as Ming says, sometimes it’s hard to tell hand shake apart from shutter shock.

          • OK, I just find it a bit strange that no evidence is present on the net for this issue when the camera is handhold, considering just how much of a scandal this is supposed to be. And frankly; people interested in the A7R are looking for a small system; I do not think carrying along a tripod is their intention. I can see no issues with my own pictures, but I might not know what to look for, I guess.

    • Might well be because people shooting it handheld cannot decouple it from camera/hand shake. But the behaviour is there: shoot a 35mm lens at 1/350 and you get a double image that looks like poor print registration. That cannot be a camera shake issue at 1/10x.

      Compression happens in skies and areas with little tonal separation, especially once you start doing any post processing work with the files.

      Short answer: it was a big enough issue to prevent me from buying one.

      E-M1 files are only muddy if not properly exposed – you have to be extremely diligent about ETTR with that sensor, otherwise you get mud AND noise. IQ is not just resolution, it’s also tonal quality, PP latitude, dynamic range, acuity and a whole host of other things. In any case, at this point, I’m not happy with either the A7R or E-M1 for stills because my print requirements dictate otherwise. Neither cuts it at the pixel level, and ‘serious’ starts for me at D800E and goes into medium format territory.

      • Let’s face it Ming, and I say this with all due respect, if Sony rolled out the A7x 50MP next gen A series that resolved EVERYTHING you’ve complained about (lenses included), I highly doubt you’d be showing up at the job with it mounted to your tripod 🙂 and we both know why 😉 That said, I’m pretty sure Sony has implemented a solution that may not have completely alleviated the shock problem but has significantly improved the situation. I am now able to get a 99% shot rate handheld at 125 and higher on my new A7R. The first unit I bought when they were first introduced last year was a mess at 50% or less handheld shot rate. And when I don’t need to shoot buildings n such, my A7S is sublime. Very excited about the new Zeiss 35 1,4 M mount. I don’t share your aversion to adapters.

        • You assume too much. I’ve acquired systems that fill far less of the requirement because they offered some capability I didn’t have before – Pentax 645Z being a good example. If they did resolve everything – raw compression, shutter shock, diffraction at 50MP and now to a lesser degree, lenses, then I’d be the first to buy two.

          You won’t see much issue with an adapted lens on an A7S. Large pixels are far more forgiving of everything. Try the same on your A7R, and I doubt you’ll be seeing perfect pixels. I can get those perfect pixels now, so why change for something that isn’t going to deliver them?

  39. As far as system cameras are concerned, it really comes down to lenses available. Lack of ultra wide primes and tilt shift lenses for m4/3 has me kept me from ditching FF completely.. even though I prefer m4/3 for it’s ergonomics.

    For everything else there is Ricoh GR.

    • You can put pretty much any mechanical FF lens on M4/3 for tilt/shift via adaptors as the image circle is more than large enough, though of course planarity is an issue again. I don’t find myself wishing for tilt as much as shift with smaller sensors, though. By f8 you’ve got everything in focus and aren’t really challenging diffraction too much, either.

      That said, I don’t know about you, but the FF T/S options aren’t that fantastic either, optically. If only Zeiss would make a couple in the Otus line…

      • I see your point. I maintain a small collection of FF glass for UW and T/S options. Everything else I am more than happy with m4/3 provides.
        I am not really a fan of adapted lenses from the usability standpoint, besides the issue of planarity you mentioned. Size and lack of AF to some degree are the main concerns. That’s not to say that those lenses in native form wouldn’t be huge, still smaller than the FF equivalents though.

        Zeiss should totally make a FF T/S lens! But for me personally it would be another lens to drool over. 🙂

  40. No easy answers. I’ve gone round in circles on it. The drawbacks to each serious system are considerable, and you are right to show that there are no easy answers.

    Given Sony’s rapid product cycle I have a feeling they will get there. That 1.8/55 FE lens is very nice. If they properly fill out that system (say, a mirror of Olympus’ lens line up for M43) and resolve the shutter problems it might become a leading system. The A7/R is selling well in Hong Kong. The market scale should give them incentive to continue to roll out the lenses.

    The breakthrough I’m looking for is a super hybrid OVF/EVF – somewhere around triple the resolution of current best in class EVFs, and the ability to switch back and forth between OVF and EVF for different purposes. When I want to work fast, I still really like the discrete, natural feel of the rangefinder system. I just hate all the mechanical mis-alignment issues – serious back and front focus. It wasn’t as big a deal with film, as we know why.

    I’m getting rid of most of my Leica M system now – just keeping a film body and two lenses – and returning to Nikon (D810), and the Zeiss line up for project work. Will I be too lazy to go back to lugging rather than carrying, having been spoiled by the M? Let’s see. I’m not convinced the F-mount / SLR combo will last many more years. It’s big for the sake of a mirror, and the mirror is an inferior focusing tool.

    My hedge is having just bought the Sony RX1R. I am persuaded by Lloyd Chambers’ arguments in favour of it – integrated lens/sensor design, quiet leaf shutter. If they make a 28mm version I suspect you might be tempted…

    • This is why we land up with multiple systems. My fear with Sony is that they will get there, but overshoot or miss the point and then land up going too far. Or they will veer off in a different direction (as they’ve done in the past).

      Integrated lens-sensor is the way to go for many reasons – optical synergies, size, no mounting/planarity/AF alignment issues etc., but it does effectively mean you’re limited to live view or EVF. I’m fine with that. Actually, the more I use the Zacuto hood on the D810, the more I’d like to see an iphone-style panel with an equivalent hood; flip it up to use the LCD normally, flip it down or detach it for eye-level shooting. We wouldn’t even need the EVF anymore. Sigma has it right with the accessory finder on the Quattros…

      • Tibor Kadar says:

        I like the notion that the right system is your system and what you make of it. On the basis of your thoughts I guess there is a sore hiatus of a Ricoh GXR like system, where you can match the sensor size and architecture with the lens design to serve the specific purpose. I agree with you that the module concept of Olympus can be a huge success (actually they already have a patent for a separate viewfinder similar to Google Glass). Camera makers will develop better cameras if the demand of the enthusiast crowd is strong enough about the specifics what they want. The enthusiasts can only rely on the knowledge and experience of the pros.

      • Does the D810 allow you to use Zacuto + live view when using the aperture ring of a Zeiss ZF.2 lens? (As previously discussed, there is a bug in the D700 firmware which prevents using both live view and the aperture ring if the lens is chipped, i.e., live view only turns on with an unchipped lens or with a chipped lens, when the camera, not the lens’ aperture ring, is set to determine the f-stop.)

        • No problem, that works – you just need to enable the aperture to be set via aperture ring and command dials (not enabled by default, no idea why).

  41. Kinda amusing you’re not jumping into Fuji X serieswagon Ming

    • Ming has talked about it in the past. He’s not happy with the workflow required to get the best out of Xtrans sensors.

    • No, why would I? I’ve been asked this about a dozen times in the Photokina comments: basically, it doesn’t do anything for me that I can’t already do; it doesn’t fill any gaps or do anything better or smaller or lighter, and the workflow is poor – I can’t easily get the output I want. All in all, buying into a system like that – Fuji or another brand – would not make sense at all economically…

  42. Hi,
    Excellent post as always. I too have quite a number of systems myself somehow :). Just wondering if you have given up shooting with m43 since you seem to use the GR a lot.

    Regards
    Sukhdev

    • A M4/3 system requires space in the bag, the GR does not. Given my current printing requirements, M4/3 is insufficient and doesn’t justify that space…we still use a pair of E-M1s to film all of the workshop videos 🙂

  43. I’d second that as well, I’ve always been of the opinion that folk buy far beyond what they need, driven by the must have the latest and shiniest. I’ve got the em5 and primes 12, 19, 25, 45 and 60 all of which fit my needs. Needless to say they perform beautifully, I also shot purely b and w with my blad and my car everywhere camera like you Ming is the Ricoh gr the rendering this gives is every bit as unique and stunning as Leica

    • I suspect if I ever gave up ‘serious’ photography, I’d probably land up with just the GR. It really does strike the right balance of usability/ portability/ quality…

      • Kristian Wannebo says:

        Let’s hope for a Ricoh “GR” with a collapsible short tele with an additional macro setting.
        One GR in each pocket …

        • That would be very nice indeed!

        • Kristian.. what about 3 pockets with GR’s. One pocket for each focal eqv. 28, 50 and around 80 w. macro 🙂

          • Kristian Wannebo says:

            Right ! 🙂

            Of course, you can’t always zoom with your feet to overcome a gap between a 28 and a short tele, so I have considered that (as have Sigma with the Merrils), although I can’t see myself pocketing all three … 😦

            The more cargo pants friendly solution would be an MFT plus three primes?

            Personally, I’m prepared to carry two, easier on the pockets than the thicker LX100 and probably with better image quality – except for that lens gap, 🙂 .

            Without cargo pants or jacket pockets I guess we will have to wait for the next generation of sensors.
            Perhaps http://www.fujifilm.com/news/n130611.html
            will give us a larger shooting envelope in pocket size (except, of course, for DOF).

            • Or Ricoh could just make us a 2x TC to accompany the 0.78x. They actually used to do one for the GRD1, but for some reason stopped…

              • Kristian Wannebo says:

                How does the wide converter for the GR perform?
                Does it give similar IQ as the GR without it?
                ( Until now I thought wide or tele converters were to hard to make affordable to the standard the GR sets?)

                • Pretty good, actually. I shot Forest I with it in Prague, wide open. It needs 4-5.6 for the corners to match the center, but as far as converters go – it’s impressive. And would certainly do as your wide solution in a pinch if you don’t use it that often.

                  • Kristian Wannebo says:

                    Thanks!
                    I checked your photoessay The Magic Forest – I’d forgotten you used the wide converter then.
                    Considering your demand for detail for printing, I realise it must be a good one.
                    So I’ll add your suggestion for a tele converter to the Ricoh wish list.

                    ( F:4 for corner sharpness seems good enough, a 21 mm eq. lens for APS-C faster than 2.8 up to your standard would hardly be pocketable ..)

          • Now if only they’d actually make them!

  44. Ming, I am wondering, that you don’t mentioned the focusing problems, with your excellent manual Prime lenses!

  45. It still seems like there is a fairly large gap in image quality between Micro 4/3 (at least the EM-1 I tried still had substantial noise at base ISO) and DSLRs. And once one commits to carrying something as heavy as a DSLR, it is hard to give up the potential quality of the D800E/D810 sensor.

    I suspect your comment about the Canon 24-70 f/2.8 was in reference to the Mark I variety, as the Mark II is widely acknowledged to be perhaps the sharpest production zoom lens and certainly far superior to the Nikon offering.

    • Perfect exposure becomes more critical with smaller sensors, but also easier thanks to live view/EVFs and highlight peaking. There’s a big difference between M4/3 and D810, but not so much between M4/3 and APS-C, especially under sunny conditions. Also worth noting that the stabiliser tends to be able to claw back a stop or two, and the lenses are faster, so you may well see a net advantage to M4/3 in practice…

      Yes, the Canon 24-70 I was a dog. I’ve not used the II, but it must be pretty spectacular to be better than the Nikon one – or sample variation at play again…

  46. Hi, Ming,

    Thank you very much for the nice article. I completely agree with everything you have said. I used the same reasoning and ended up with Olympus OMD EM1 with three primes: 15 mm, 25 mm and 75 mm. Right now I am all set. If 150 or 300 mm prime lens becomes available I might consider it but not sure as of right now.

    I always enjoy reading your posts. Thank you.

    Val

Trackbacks

  1. […] into any camera system is a big deal – not just because of the financial investment involved, but because […]

  2. […] even other photographers) which camera you should probably buy, Ming’s latest article “System thinking” has some good […]