Long term review: Canon 5DSR

_5R00808 copyWhen the stars (trees) line up

These will be my closing thoughts on the Canon 5DSR, first reviewed here. It turns out there won’t be a part two for a simple reason – I don’t see the point. My opinions that follow are going to appear initially conflicting and probably be misinterpreted by the fanboys, so I’m going to state this upfront: I really, really liked the camera. But in the end, it just isn’t for me. Allow me to explain why.

_5R03711 copyEncounters at the train station

There’s a lot to like about the 5DSR, even relative to the competing Nikon flagship D810:

  • A visible resolution advantage both on screen and in print, though admittedly only for very large sizes or Ultraprints
  • Much better color out of the box, especially for foliage and landscapes. This was also very visible in print, with a significant advantage over the Nikon; I’d have to do significant compensation work on files with the D810, but the 5DSR’s files just looked correct and pleasing after a single initial camera profile. I have no explanation for this other than native tonal response at the highlight and shadow ends of the green channel isn’t linear; this is probably related to the non-ISO-invariance of the sensor.
  • Preferable (to me) ergonomics and handling – the grip is perhaps the most comfortable I’ve held, and whilst I still get blisters with the D810 on my little finger second joint after extended use with heavy lenses or the grip because of the location of that one lower seam, I didn’t with the 5DSR
  • Lenses not in the Nikon system at all – the 17TSE, 70-300/4-5.6 L, all of that vintage Contax Zeiss goodness that isn’t F-mountable because of flange distances. Stuff that has no equivalent like the 35/2.8 PC-Distagon and the 100-300/4-5.6 Vario-Sonnar.
  • Custom shooting settings on the mode dial, easily accessible.
  • Better AF acquisition and tracking – only one of my lenses required fine tuning, and even then it was by a 1/20 amount at one end. Oh, and you can tune both ends of the zoom separately – which is impossible on a Nikon.
  • Better mirror lockup implementation. Choose the delay and trigger it with a cheap IR remote, or just hit the shutter once.
  • Electronic front curtain by default in live view.
  • A mirror/shutter mechanism with subjectively less mechanical vibration – or better damping, or both.
  • The ability to move the magnified point quickly in live view, hit 50 or 100%, and return to centre with single button presses – again, not possible on the D810. That is slow and you either have return to centre or magnify, but not both.
  • Preservation of the original file when running crop modes; only the crop selected is tagged but not permanently baked in. With the Nikons, you lose the rest of the file.
  • A slightly larger and slightly clearer viewfinder.
  • I personally thought the camera felt a bit more solid, and there was no popup flash to accidentally deploy and break.

_5R00608-10 copy

But there were things that I didn’t get along with, too:

  • Dynamic range is limited compared to the D810 – about 2 stops at base ISO, by my estimation. And that gap increases as you go up the sensitivity range.
  • Metering in evaluative mode is unpredictable.
  • At the pixel level, I’d stop at ISO 3200 with some NR afterwards – beyond that, it isn’t so much the noise but loss of acuity and color accuracy that become an issue. Files are larger but with no more perceived detail or clarity.
  • Some button function combinations are just impossible: you cannot have AE lock on the shutter button in anything other than evaluative metering, for instance, and then spot metering point doesn’t follow your AF point. You have no choice but to dance a jig with your fingers over the back button.
  • Battery life left a lot to be desired: I never saw more than 600 frames, and often more like 300. This is in comparison to 1,200+ under similar circumstances with the D810.
  • Long lockup times in live view ‘BUSY’ became a familiar screen.
  • That slightly larger and slightly clearer viewfinder was also not quite so snappy or easy to manually focus with.
  • You can’t swap the front and rear dial functions either, but you get used to that fairly quickly.

_5R01383 copy
Through the wormhole

By far my biggest headache with the camera was that it tended to be a ‘fair weather friend’: under ideal circumstances, on a tripod, or in very good light, with critical focus obtained with the very best lenses, the files are pretty incredible. But if any one of those stars fail to align, then you’re going to land up with something slightly soft-looking and just a tad disappointing: as though somebody slipped an AA filter back in the optical chain somewhere. Somehow, the jump from 36 to 50MP magnifies the visibility of a lot of things: operator shake and critical focus/focal plane being the main ones. It is both very rewarding and extremely brutal on those who are sloppy on technique or cheaper out on their support gear, perhaps the most unforgiving I’ve ever used.

I suspect contributing to the duality of results is the unpredictability of metering in evaluative mode, and the files not being very tolerant of underexposure; combined with the reduced dynamic range, this means your exposure has to be bang on all the time – it was a bit like shooting slide film. There is very little latitude for recovery, but at the same time, you don’t have to do very much work at all to the files if you get it right – the tones and colors somehow just look ‘right’.

_5R03611 copyOver Genting

My conundrum is that I already have a system that is capable of delivering excellent results under a wider range of circumstances. I give up something at the extremes of the performance envelope, but under those conditions I can probably stitch to close the gap. The rest of the time, I don’t see any clear advantage one way or another. Maintaining an additional system for vanishingly small returns – and carrying it on assignment or during travel – is another challenge entirely. It either calls for a significant financial commitment to field a complete Canon system (i.e. second body, full lens range) or a significant amount of weight and partial coverage of either system, neither of which is ideal.

_5R03957 copy
Eternal friend of the state (from The Idea of Man)

I think it’s important I comment on my initial hypothesis when buying the camera – it would be effectively a body-only cost since I would be adapting lenses. In practice, whilst this works, it’s very clunky and means manual focus with live view on sturdy tripods all the time. It also means having one adaptor for each lens since thin adaptors in the field are not so easy to change quickly. The usual suspects proved to hold their own resolution-wise – the Zeiss Otuses, Zeiss 2/135 APO, and surprisingly, some of the older aforementioned Contax designs. The 2.8/35 is a superb lens that has a really beautiful rendering and conveniently, a 28mm equivalent FOV when shift-stitched vertically. The 2/28 Distagon shows significant CA, fringing and all sorts of optical gremlins until f8 – after which we hit diffraction limits because of the camera’s pixel pitch anyway. All in all, the Zacuto-F mount-Otus-5DSR setup was capable of some impressive results: but not so consistently. And let’s just say it’s rather large, and 100% live view also means finishing at least three batteries a day.

_5R03637 copy

Since my assignments tend to be somewhat varied, I still needed run and gun versatility at least some of the time. I landed up buying a 24-70/4 IS L, 40/2.8 STM and 70-300L, and carrying these in addition to the manual focus lenses and a Nikon body for backup. The 40 STM was the surprise of the bunch, and very enjoyable to shoot with because of size and image quality. It made a great aerial lens, actually – the right focal length for a natural perspective, not so long that you run into shutter speed problems, and not something that would be so physically large as to be affected by wind buffeting. It’s also a steal at $149.

_5R05901 copy

In total, I shot three assignments, one workshop and north of 10,000 frames with it. I would say that with the exception of trying to pull critical focus at f1.4 on off-centre moving subjects in low light handheld at the Hanoi Cinematic Masterclass – a tall order under any circumstances, much less with 50MP – I enjoyed the experience very much. I was initially going to do a comparison between the D810, E-M5II in high resolution mode, and 5DSR – I don’t think there’s a lot of point, because they all each have some fundamental weaknesses, and some strengths – what is ‘ideal’ depends very much on your own needs. The E-M5II has the most accurate color, but the poorest acuity, slowest capture time and worst dynamic range. The 5DSR has the highest resolution but visibly less distance is in critically sharp focus because the increased resolution also differentiates planes of critical focus more distinctly. The D810 has the most dynamic range, and slightly better pixel-level acuity than the 5DSR.

_5R07249 copy

It is clear to me that neither Nikon nor Canon can claim to be decidedly superior at the moment, each with their own strengths and weaknesses – much like their equivalent lens lineups. Canon lacks Nikon’s new f1.8G primes, but Nikon lacks all of the special purpose stuff and some of the zooms – there is no 17 TSE, for starters. There’s no question that if I were a Canon shooter, I’d be very happy at the moment: you have a tool that’s miles ahead of anything else in the system, can hold its own against the competition most of the time, and can pull ahead if the conditions are right. If I shot only subjects under which I could satisfy those conditions – in a studio, or perhaps landscape or architecture – it would be very easy to justify a complete switch. Unfortunately, I don’t, and I tend to have multiple very different objectives within one assignment – and under those circumstances, I’m left with either a very heavy solution or one that doesn’t quite cover all the bases in the most optimum way. And with shrinking client budgets and increasingly militant airlines, I can only afford to keep and carry one system – I’ve already been forced once to buy another seat last minute at the airport at full price to avoid having to check in my equipment. Since my glass investment is already F mount, it is with great sadness, I have to bid the 5DSR goodbye. I’m missing it already. MT

Equipment in this post is available here from the site’s B&H and Amazon affiliates links – every bit helps!
Canon 5DSR B&H Amazon
Canon 24-70/4 L IS B&H Amazon
Canon 40/2.8 STM pancake B&H Amazon
Canon 70-300/4-5.6 L IS B&H Amazon
Nikon D810 B&H Amazon


Be inspired to take your photography further: Masterclass Chicago (27 Sep-2 Oct) and Masterclass Tokyo (9-14 Nov) now open for booking!


Ultraprints from this series are available on request here


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!


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


  1. Hi Ming …. This is the first time I have read a review by you. I must say I liked it much. You seem to be very honest about what you think & feel about things …… You tell it how it is …. and I thank you for that. I do have a question. After reading this review about the 5dsr…..well….It almost sounds a little like you were reviewing a medium format camera in some ways….. How does this camera hold up to medium format?

    • Honesty is necessary but apparently very unpopular with the camera companies. I had my CPS membership rejected after publishing this, and got the cold shoulder from Canon Malaysia. Oh well.

      Medium format: resolution is there, dynamic range is not, color accuracy is not (but it’s pleasing, at least). For the price, you can’t really complain though…

  2. Hi Ming,

    Excellent articles, although I’m a little sad that a comprehensive part two never came about. I’m a Nikon shooter (vanilla D800), with a bunch of Nikon lenses, but no Otuses. I’ve been wanting to upgrade my D800 for a while now (I print my images with an Epson 7900, with the bulk of the prints at 24×36 inches (single image), or larger (stitched panorama)), and have seriously considered switching to Canon and even Sony the past year (the D810 is tempting, but I feel like I’m so far into its life cycle that I might be better off waiting for its successor). I tend to shoot mostly landscapes, and bracket quite a bit as well. Given the use of a tripod and bracketing, is this good enough to overcome the DR limitation of the 5DS-R? Will HDR software in any way counter the benefits I’d get from using a 5DS-R? (I use HDR software as a starting point, then fix all the problem areas in photoshop.) I’d obviously pick up an Otus (55 for sure, probably 28 as well when its available) if I made the jump. I’m not a pro photographer yet, but I want to make the switch, and while I know gear can only contribute so much, I want to push the image quality envelope as far as I can.

    I’ve also read a few reviews mentioning the higher than average noise in the shadow areas of a 5DS-R photo. Did you ever find this to be an issue? I tend to do quite a bit of night landscape photography (always on a tripod, and bracketing again if necessary), and am planning to do a pano of the milky way during the summer of 2016. Would a 5DS-R + Otus 55 at 1.4 be able to do this without compromises in image quality? Or am I better of sticking with Nikon and just getting an F mount Otus? Thanks!

    • Good question: you’ll see the resolution gain if you’re working off a tripod at base ISO (stability and focus are critical) with the right lenses. But HDR doesn’t always work, for instance with long exposures or moving subjects (if it’s windy, for instance). You can’t push the shadows that much without noise – that I agree with.

      I guess the concise answer is you’d gain from a switch, but questionable if it’s enough to make it worthwhile. I’d try to rent one first if possible.

  3. Juan Pablo Bialade says:

    Ming, I have just read this entry blog. I really like your site, it’s very helpful. Just once question: i have three zeiss ZE lenses a 21/2.8 distagon, a 50/2 makro planar and a 100/2 makro planar. All or them working fine on my Canon 5d3. How do you think they will perform on the 50MP sensor fo the 5ds/r ? Should I upgrade to milvus? (Optical formula seems to be identical on this lenses) thanks in advance and sorry for my english, it is nos my native lenguaje.
    Juan P. Bialade

  4. Hi Ming…very interesting two articles: I did a double take when I saw you reviewing a Canon. So sorry you’re out of the system so quickly: I had high hopes for your review of some Canon EF lenses (and perhaps an FD or two). C’est la vie; still will comb your reviews as I find them more holistic than many other sites. Still, you’ve made me reconsider the 5DSR which I initially looked askance at — seemed to me to be a marketing ploy by Canon. Nice to hear there is actually some substance to it, although I may still pass on it.

    • Thanks. There definitely is a good reason to get one if you’re heavily entrenched in the Canon system, but it duplicates too much of the Nikon for me and only extends performance under a very limited set of situations – I just don’t encounter those enough to justify it.

  5. Ming, you give some real interesting explainations here.
    BTW, I was wondering, in the case of the most “prolific” posters, what would be their ratio of “hours spent shooting” versus “hours spent posting on the net”…
    I suspect very low… 🙂

  6. Great review Ming, thanks a lot!
    From everything I could gather on the net, the only advantage of Nikon (and Sony) over Canon would be the (Sony) excellent sensor and its wide dynamic range. People (and you, too, in this review) believe it is roughly a difference of 2 stops. Yet, If I read the graphics, the differences of around 1 to 1.5 stops are until ISO 320 or so. Beyond that, not only the difference does not increase (as per your statement), but they seem to become practically the same. Please see the DXO graphics:


    Now, it is not uncommon with the current (full frame) digital sensors to shoot at ISO 400 or 800, with no visible loss of quality or significant noise issues, in case the shutter opening duration cannot increase. So, in other words, if there is enough light or you can shoot wide open (sharpness all over the frame not being important, or on wide angle lens), so that you can shoot under ISO 400, then it might be easier to recover some highlights and/or shadows with D810 or A7IIR. They say on DXO that a difference of half of stop is not really noticeable. But how often is that a real challenge? How many professional photographers could not deliver a quality product, due to the “limited” DR of their Canon sensor?

    Usually, a quality image is made of quality light. A high quality light means not a huge contrast, not harsh highlights and deep shadows, as we experience in a bright summer day at noon, right? Therefore, the “natural” dynamic range is not a huge challenge, is it? When the light can be fully controlled (studio, reflectors, strobes etc.), there is no issue, whatsoever.

    Then, the DXO definition of the landscape dynamic range states that 12 stops of DR means an excellent one. Well, Canon 5D sensor has been measured for 12.4, at ISO100.

    Yes, I would also prefer a wider dynamic range on my Canon cameras, but I don’t see it as a real issue. Not a deal-breaker or a supplier switch trigger, anyways…

    I was looking for arguments not to buy me the 5DsR…Well, except for the price and the “unforgiving” resolution (which I am still wondering if I really need), I still couldn’t find any. 🙂 What am I missing?

    Oh, yes, I confirm that the evaluative metering is not stellar at Canon, but why is that an issue? I am always shooting manually, using spot metering, bracket and/or test the exposure, since shooting digital allows all these. Now, I am sure that evaluative metering combine with one of the semi-automatic exposure mode can fail you every now and then. Hence the reason for me to always shoot manually and selecting my own exposure, as being part of the picture making process. My two cents.

    • I buy, use and evaluate cameras based on results: images. Not DXO measurements, which don’t always agree with the real world. And if you want to compare DXO numbers only, even then it’s clear the 5DSR still lags.

      • 5DsR still lags, true. Between 0.5 and 1.5 stops, at ISO values below 400. Please note I also don’t fancy the DXO measurements…For instance, the overall IQ scores are kind of strange, when compared Nikon to Canon to Sony. Must admit I have no direct shooting and processing experience with Nikon. I can only tell what I saw at my fellow photographers: Nikon has a real issue with WB, colors rendering etc. If that is true, even it can be easily corrected in post processing, isn’t it a parameter in the IQ overall equation?

        Anyways, I apologize if my intervention has been perceived as defending Canon against Nikon. Why should I do that when your conclusion is quite clear: if it wasn’t for logistics and the lens lineup, you would have switched to Canon, right? All I wanted to say is that the long held dear argument of Nikon (and Sony) fans claiming superiority over Canon sensors of a wider dynamic range is rather a minor aspect, when we look at the complete and greater scheme of things. And so was the number of pixels. And I say this now, when Canon is well above 36 megapixels. I actually have preferred something in between 36 and 50. I am truly scared that I could miss shots, due to this insane number of pixels over crowded on the same full frame sensor…Maybe, Sony made the right decision with their last 42 megapixels sensor, have chosen the correct compromise at this stage of technological progress.

        • There is a difference in color rendition between N and C – whether it is an issue or not depends what you shoot. I prefer the Nikon for product and architecture, because it seems to be more accurate under controlled light circumstances; the Canon does much better with landscapes. I could go either way for skin tones. They can’t easily be made consistent in post, because I suspect this partially has to do with the tonal response of each channel at different luminance levels AND the dynamic range differences.

          I’ve shot both side by side under identical conditions and can confirm that in general, under the conditions which you would use these cameras – i.e. optimal exposure/ISO/tripod etc. – the Nikon has superior dynamic range by quite a noticeable amount, but the Canon has superior resolution and sometimes more pleasing color. At high ISOs I’d rather use a different camera to either of these because you lose a lot of acuity and land up with big but ‘loose’ files. That is not their purpose.

          The answer to your question is no, I wouldn’t switch even if both lens/system considerations were identical. I do enough work in challenging light that I can actually make use of all the dynamic range the camera has. In any case, it’s an academic question as such a situation can and will never arise. In the equally unlikely situation that cost were not an issue, I’d run them in parallel, fly first class and pay a sherpa to carry what I might need ‘just in case’. 🙂

          • Right, that pretty much summarizes it, as you’ve said it already in the beginning: none of the two systems can claim overall superiority. As for the dynamic range, I definitely don’t have your work span and experience in lots of different and challenging situations, so I will only take note of your findings. The only thing I can testify for is that I bought me a Sony RX1R precisely for portability, the great Zeiss F2 lens and the famous wide dynamic range sensor. Well, after a few weeks of shooting and testing, I can honestly testify that it is not a huge difference when compared with what I can recover in post processing from a “correct” 5D Mark III (or 6D) exposure. It is, but nothing jaw dropping kind of thing. Moreover, I am entering a new world and even a bit strange for me: when am able to light up the shadows to that extent, with no visible noise increase, I am getting more of a “plastic”, “unreal” or “CGI” feeling…Therefore, I am sliding back the shadows, with no real gain from that great dynamic range. What I am trying to say here, I guess, is that there is a sort of “poetry” and an “old fashion” of doing emotionally filled photography, when producing images that are not perfect technically. Think scanned film, for instance…There are even some opinions out there saying that sometimes blowing out the highlights or clipping the shadows is exactly what you want to do, for a greater contrast and more compelling storytelling.
            I don’t know, just saying… 🙂

            • What you’re seeing is the great Sony 11+7 raw compression. There IS more dynamic range at the sensor level, but it’s being lost in noise and posterisation. It’s just like a JPEG in some ways, but larger: you have 14 bits of information to stuff into an 11 or 8 bit container; some of it is going to overlap. I suggest downloading a good D810 raw file then moving the sliders and seeing the difference – it’s really quite amazing just how much information is recoverable, and without noise or strange plasticity.

              Poetry aside and narrative device aside, there is a) no excuse if the other company using the same sensor can make it work, and b) I want to have the ability to make that choice myself, not have the camera make it for me. On top of that, negative film has enormous dynamic range if used correctly…

              • Well, I was not aware that would be the reason behind. I guess will just sell the camera, then…:)

  7. Ming, considering your D800/E, D810 and this review, it seems to me that it is not advisable at all to use this camera hand-held despite its better mirror/shutter mechanism leading to less mechanical vibration, but, one can -at least occasionally- use D810 and with more care (due to its inferior mirror/shutter mechanism) D800/E hand-held.

    Do you agree with this statement and if so would you mind elaborating on those occasional instances?

    • No, because if you have enough light/shutter speed for a D810, you’ll have enough for the 5DSR. The problem is when you have to raise the ISO significantly and dynamic range suffers as a result. This is before we even consider metering inaccuracy losses…

      • Ming,

        1) The reason for my question was your statement “under ideal circumstances, on a tripod, or in very good light, with critical focus obtained with the very best lenses, the files are pretty incredible. But if any one of those stars fail to align, then you’re going to land up with something slightly soft-looking and just a tad disappointing: as though somebody slipped an AA filter back in the optical chain somewhere”.

        There one of the factors you mention is the use of a tripod (and hence not using the camera handheld), regardless of the shutter speed. In fact one of the factors is the availability of “very good light”, in which case the shutter speed would likely be not a problem especially on short focal lengths.

        As a matter of fact from your statement I have deduced that unless used “in the ideal circumstances, on a tripod, or in very good light, with critical focus obtained with the very best lenses”, one should not use this camera handheld regardless of the focal length, shutter speed and the intensity of the available light. If my understanding is wrong, I will appreciate your comments.

        2) Leaving all the constraints due to ISO, dynamic range and metering, in terms of the reciprocity rule, i.e. one needs to use shutter speed equivalent to the focal length, how does this camera perform compared to D810? (In your D810 review you stated that for D800E you “could use down to perhaps 1/focal length at best; everything else was 1/2x (borderline) or 1/3x + (consistently sharp)”, and that for D810 “without changing anything in the way I shoot, I’m finding myself with an extra half stop to full stop of shutter speed before I see camera shake creeping in”.

        How does this camera perform in this regard?



        • 1) Image quality degradation is not binary – but the higher the resolution you go, the easier it is to not achieve the full potential because increasingly tight tolerances are required. It obviously depends on what you shoot and what your personal threshold of acceptability is – fair weather vacation snaps are going to have a far higher hit rate than cinematic still in low light.

          2) Still 1/2x-1/3x. The mirror/shutter has less vibration than the Nikon but focusing requirements are more critical.

  8. Nice review, and thanks! Your last comment on the Pentax 645Z is interesting. Did you get rid of yours or have you stopped using it? How does this Canon compare to the CMOS medium format sensor? (apart from ratio, etc) . As someone with experience with the Pentx 645Z and a Hasselblad CCD medium format system, Do you consider the Canon close to medium format quality, apart from the issues you mention with dynamic range?

    I am a fashion photographer working with the Pentax 645Z and Phase One IQ250, and hope for a smaller, lighter, and cheaper DSLR that has comparable image quality. Pentax 645Z is amazing but the overall lack of lens choice (esp a high-quality 80mm) and slow 1/125 flash synch speed make it less than ideal in studio or location settings using flash, so I do not use it much and may sell it. The Phase One is truly remarkable but leaves a lot to be desired, especially in terms of autofocus and overall reliability, and size/weight and of course cost.

    • I sold it because I could not justify keeping and using a system of that cost/bulk that couldn’t serve as primary (cost of backups, lack of specific lenses, flash sync limits, lack of pro support).

      Under ideal conditions, the Canon has about the same DR as the Hassy MF system, but slightly less acuity and color accuracy. The Pentax is superior in all ways except resolution (for which it would be similar). In the field, if you’re using flash, you’ll get pretty close. But your lenses are likely to leave something to be desired, and you’ll probably have to go Zeiss Otus – which aren’t AF.

      • Thank you Ming. I really enjoyed reading your reviews which are refreshing compared to the vast majority of geeky/bird watcher pixel counters amateurish reviews. Your site reminds me of a restaurant review book called “Where the Chefs eat” as it is dedicated to pros. Not only that but your ‘test’ imagery is beautiful and the design of your site is simple, clear and sophisticated.

        I have suffered a bit of ‘grass is greener’ syndrom lately, I shoot interiors and still life commercially and architecture and landscape personally. The commercial side of my photography requires both studio and location work and whilst I currently use my 5DIII for natural light/editorial content and the 5DsR for studio or on location artificial lighting situation, I feel like am condemned having two different systems for those applications but perhaps I am mistaken. Having both with Canon offers the obvious advantage of glass compatibility but am I getting the best result in both situation?

        I have tried the 645Z and really liked it although the lens offering isn’t strong enough, yet, compared to the Canon for my type of use in the prime under 50mm lenses. I haven’t shot with the D810 yet but the superior DR is very tempting, particularly in those interior situations where I usually need to plate windows and shadows.

        Do you think that the 645Z and D810 would be a good combination and would it be superior to a 5DsR/D810 combination? I can’t ignore the financial aspect but I am also conscious that I would like to get the best quality files to my clients and also that unfortunately sometimes client paying you a good advertising rate expect to see an expensive camera on set. I am conscious that the only thing/person responsible for the quality of the photograph is me. I have used phase one systems but never felt anything special using them, sure some of the Schneider lenses are beautiful but they lack soul which is very important to me. The problem is that I would love to fall in love with a system again but that system would probably not offer me sharpness from corner to corner which is an absolute necessity in a good majority of my jobs.

        Maybe there is another single camera body option or a better combination option, I could dust off my RZ67 and get a back for it or would a Cambo, Hasselblad or Alpa be more suitable? There seems to be so many options and unfortunately not many of them can be tested where I am currently located so I would much appreciate any input if you have any.

        Thank you very much,

        • Thanks Felix. I have that book too – it hasn’t pointed us wrong yet 🙂

          Two 5Ds is really the same system, no? It’s more like two different tools. I don’t see anything wrong with that – best tool for the job.

          Moving to Nikon will only gain you dynamic range. You need to decide how important that is. Having run a 645Z/D810 combination for a while, I’ll say it’s worse than two same-manufacturer bodies – you can’t carry ALL lenses for both, and the weight penalty is significant, plus having a field backup becomes costly and heavy. I didn’t find it to be a good solution at all and dropped the 645Z.

          Adapted backs for older MF systems work pretty well IF lenses can hold up, and remember you’re going to need something below 35mm for interiors since those sensors are at best 54x40mm.

  9. Hello, Ming. I have a question for you: is it true that you can upsize D810 files to 50mp and obtain at least the same level of detail and resolution as 5DsR, if not better? I do find 5DsR files are surprisingly a bit noisy even at base ISO, however D810 files are clean as a whistle and hold up upsizing remarkably well.

    • Err…no. If you are using lenses that out resolve the sensor and do not have shake/focus/exposure issues, the 5DSR will out resolve the D810.

  10. Hello,

    Great, thorough review of the camera! But a couple of minor notes about the controls – you say that:

    “Some button function combinations are just impossible: you cannot have AE lock on the shutter button in anything other than evaluative metering, for instance, and then spot metering point doesn’t follow your AF point. You have no choice but to dance a jig with your fingers over the back button.”

    -If you use the Custom Controls to set the Shutter Button Half-Press for “AE Lock (while button pressed)” you can then use the Shutter Button for AE lock in all of the exposure modes.

    Also, you say:

    “You can’t swap the front and rear dial functions either, but you get used to that fairly quickly.”

    -Again, use the Custom Controls to swap the function of the top Main Dial and the rear Quick Control Dial for shooting in Manual Mode. For example, you can set the top Main Dial to control the aperture setting (set for Av) and the rear Quick Control Dial to control shutter speed (set for Tv).

    Doug Klostermann
    author of Canon 5DS / 5DS R Experience

    • Doug, I tried both of those things. It won’t AE lock in spot meter regardless. The manual states that AE lock only works with evaluative metering.

      You can swap main/rear dials in manual only but what would be the point of having confusing controls that are reversed for one mode only? You would then have to remember things go back to front again in Av or Tv, which will eventually cost you a fumble and a shot.

      • Sorry, my mistake! I was thinking of the rear AE Lock (*) Button, which should lock exposure in any metering mode. I will have a look at it to confirm this, and see if I can find a Custom Control setting to do this with the Shutter Button.

        Yes, the swap main/rear customization works best if you typically work in M and Av, OR if you typically work in M and Tv. Then you can set the buttons to be consistent for two modes. If you go between M, Av, and Tv, you are going to have to deal with the inconsistency for one of the modes!

  11. Hi Ming,

    I’m a little puzzled by your comment “Dynamic range is limited compared to the D810 – about 2 stops at base ISO, by my estimation. And that gap increases as you go up the sensitivity range.”

    More specifically the part “go up the sensitivity range”. This would imply that the sensor changes it’s sensitivity? I may be interpreting this incorrectly.

    Thanks for this very informative review. Always look forward

    • Something in the processing of the 5DSR means that as you go up one stop in sensitivity, you don’t lose one stop in dynamic range – it’s more. The D810 is linear.

      • Maybe its just semantics but I’m not sure what you mean by going up one stop in sensitivity?


        • Increasing the ISO. ISO is a (supposedly) linear and standardised measure of light sensitivity.

          • OK. I was under the impression that ISO settings do not change the sensors sensitivity.

            Thanks Ming

            • Signal to noise ratio decreases as the signal is amplified above base ISO on any digital camera, whether the amplification is in post process pushing or by changing the camera’s ISO setting. While the noise of the sensor should not really increase if two exposures were the same duration and the same temperature, if one exposure has the ISO set higher you will see the noise getting more obvious in that exposure, because the amplification spreads the tonal values of the noise over a wider swath of the shadows, and more of the higher luminance value pixels produced by noise will become visible as their contrast with lower luminance value pixels will increase due to the signal spread. At the same time this high ISO noise increase will effectively decrease the overall perceived contrast as the shadows become rather foggy looking. High ISO noise may also reveal some non random pattern noise, which on Canon cameras and the Leica M Typ 240 may look like cloth texture or “tartan” patterns. Nikon and Leica M9 series cameras have a more random, film-like, graininess to their noise.

  12. Dear Ming,

    I’m certainly not a fanboy as in this times of industry I can’t be. But I make decisions which have to be merely compromises. The old scientist Pareto already found out that on the way to 100% results one have to put approx.. 20% of the effort it to reach the first 80% but 80% effort to reach the last 20% of 100%. This counts for every ting we undertake and thus also bears down to the article you recently published about what to take with you.

    Your article about the Canon % DSR certainly caught my interest. But then, NO, I won’t go for it. First of all I decided, to hold only one 35mm DSLR system, as I don’t like to adjust and readjust on a frequent basis. It’s either or, Canon or Nikon.
    And actually it’s a love-hate relationship to both of the big makers. Because they never implement into their gear, what they are able to, but do their tactical maneuvering of what they do now and what they reserve for the next model to buy respectively for the model one step up in their lineup. Nikon will never accomplish their profit targets because, more and more customers realize that they only serve as a peg in the profit game.
    The Fuji X T1 shows what the Nikon Df could have been, besides of course that the X T1 is a mirrorless and thus the Nikon concept of a camera for old lenses without the possibility to at least change to a split field screen is a fail!!! As is Nikon’s preference of certain customer groups as wedding Photographers.
    I wouldn’t even trade the 1.8/50 for the 1.4/58 if I have to actually use it. This lens is a total fail in my opinion cause instead of sharpness Nikon opted to maximize bokeh!
    For my photography I want to maximize sharpness and NOT bokeh! Thus I also dislike the entire series of 1.8 lenses that Nikon make so much fuss about – I’d prefer a 2.8 series much better. (cause their max. resolution is around f 4) Besides 2.8 and where necessary 3.5 would give more headroom for corner sharpness, flat field and microcontrast.
    Neither Nikon nor Canon implements the UHS II standard for SD cards to name just one of many, many things, where both are not up to the possible standard. No touch screen and no high resolution OLED.

    As I don’t do billboards, I bet until I can master ultraprints, Nikon will have a camera above 36 MP. Nikon having a 24 MP sensor in an APS-C advocates for the possibility of 54 MP FX. And Sony has already a 42 MP FX sensor.
    The second question is, if would I be ready to accept the side effects of a resolution that high or the 50 MP of the Canon for that matter. I doubt.
    Great, Canon offers 50 MP, so does the 645Z. If I desperately wanted that resolution, which one gives me more advantages?
    Besides, the only things I envy the Canon system for, are the 17mm and 24mm TS lenses and the radio trigger of their flash guns. (and again Nikon is too ignorant to come up with similar products!!!)
    For many other things, I like the Nikon system better, e.g. the much better dynamic range.
    The Zeiss Otusses are the best lenses money can buy. But then, straining my nerves and the ones of who travel with me with numerous shots to get the perfect sharpness with manual focus more or less by try and error is not worth it. Thus I decide: NO manual focus only, above 28mm. (WW can make up for the risks with higher dof) And the Sigma Art 1.4/50 – if not reaching the IQ of the Otus- is still far better than anything Nikon or Canon has to offer.
    And yes, due to the air travel hand carry limitations I also integrate Zooms. For the D750 good Zooms are sufficient in most of the cases.
    (2.8/14-24= 1000g, the new 2.8/24-70 VR = 1070g, 4.0/70-200 VR = 850g and the D750 = 840g adding up to 3760g providing enough leeway to add a Gitzo serie 1 Traveler, some batteries and filters.)

    Decision making can be tough at times. Striving for perfection and having fun are not always friends. Especially what to carry around and not becoming a mule requires limitations.
    Again the Paretto-principle is a good advisor: why must everything be always perfect while nobody pays for it and less and less people in the broader culture care at all?
    Furthermore the less photography can serve as a frequent source of income, the higher weighs the amount of investment in gear. Thus, if quality matters, one cannot have (or carry around) all. Some things simply rule out.
    In my case this starts with this crazy Photoshop subscription-mode. I’ll never make myself a servant to Cloud and arbitrary and constant changes. My computer for post processing is a standalone solution based on windows 7 without any Internet connection after the installation.
    I decide for the D810, the D 750 and the D7200. If JPGs are sufficient, I’ll take the Fuji X T1 (meaning no post processing by computer). I take some Zooms and some primes into my arsenal and the select my 7 kg Hand carry. Only if I can go by car straight to the location I’ll take more.

  13. Samuel Jessop says:

    Again, a great read. You have a need for coverage of focal lengths that far exceeds my own, but the themes of consolidation and rationalisation have been on my mind as I work to regroup with one system and the right native mount lenses. A world apart, but your observations about dynamic range and shooting envelope are what put me off of moving from Fuji to Sony A7, even before I can consider the cost.

    All of my favourites of the shots in this review and the first are those from the 40mm STM. The field of view and rendering to my eyes are a lovely combination. What kind of proportion of your keepers were from the Otuses?

    As a reader it will be fascinating to see how the consolidation pans out, and I can’t complain with the output I’ve seen from your D810 and Zeiss lenses.

  14. Out of the thousands of frames you shot, how many times did you see moire? Is moire an issue with the 5DSR?

    • A few times; no more frequently than with the D810 or D800E – if anything, less, because slightly more resolution means you won’t see aliasing til higher frequencies.

      • As this is one of, if not the main potential issue with the 5DSR, certainly compared with the 5DS and the Nikons I think perhaps a bit more work and testing would have been useful. We still don’t know how much of an issue it really is. You haven’t really told us anything we didn;t already know.

        • I have done enough work to convince myself it isn’t an issue, and I probably work under a far wider variety of conditions/subjects than most. My job isn’t testing; I derive no benefit whatsoever from anything beyond what is necessary.

  15. Hi Ming comprehensive review as always. Obviously I don’t know his testing methodology, but I just read a review [part one] by Ron Martinsen, a Canon shooter, and he says that the 5DS appears to be sharper than the 5DSR. Interesting!

    • Hmmm…that does not make any sense – the removed AA filter will always be sharper; the AA filter is effectively blurring things.

      • Anton Kerek says:

        But the 5DSR does not have a removed AA filter. It has a cancelling filter (basically, it’s a D800e, not a D810). It’s conceivable that having the traditional AA filter and sharpening could be sharper than a cancelling filter with a sharpening algorithm that doesn’t know how to cope.

        I have a D810 and a 5Ds. I’m remembering how much I like certain Canon lenses (the 135/2, for example). Even so, I can confirm much of what you have observed, including the lesser DR of the Canon. But when you nail it on the Canon, it can be glorious…

        • I am aware of that. Having gone through 120,000+ exposures with two 800Es, 10,000 or so with an 800, 50,000+ with an 810 and 10,000+ with the 5DSR, I find the Canon to behave more like the 800E than the 800. The 810 is slightly crisper than the 800E, but it isn’t always noticeable (dependent on subject and lens). I can’t see how the 5DS would be sharper since the 5DSR is already pretty close to the best pixel-level acuity we can get from a Bayer sensor; I suspect incorrect sharpening may play a part though.

  16. Thanks for the article. Covered a bunch of interesting points.
    But one thing caught my eye, when you said about the Canon: “You can’t swap the front and rear dial functions either”.

    Really? Using the Custom Controls screen you can swap them for Manual mode at least (and the same functionality is in the 1D X, 5D3, and 7D2 at least). Or were you thinking of somehow swapping them for Av mode/etc as well?

    • It wouldn’t make sense to swap them just for manual and then have confusingly reversed controls in Av…

      • davidburren says:

        Ok. Personally my fingers go the default Canon way, but I do know people who do the switcharound so that the front dial remains the one affecting aperture in both M and Av (they almost exclusively use those modes).
        So at this point I’ll throw my hands up and agree with you. Like I said my fingers have been trained to go the Canon way automatically: to me the Nikons sometimes feel backwards.

  17. Encounters at the train station is such a terrific image! Very cinematic with lovely blue tones.

    Ming, would you mind quickly describing how you go about to upload/link to your flickr images for optimal viewing experience with a thumbnail here on wordpress? I can’t seem to find a smooth way of doing this.

    • Thanks!

      Flickr never really gives you an ‘optimal’ web image – you have to make it yourself. There is a little arrow button in the bottom right of the image container that you can click to extract a link to different sizes; use that.

  18. Ming…as a D800e owner who still feels unsure to upgrade to the D810…and read your past D810/D800e reviews…was just curious if there was anything you like or like about the D800e over the newer D810??

  19. Martin Fritter says:

    Both the 5DSR and the D810 present focusing problems for the Otus lenses. As I understand it, neither Zeiss nor Canon will make their auto-focus interface available to Zeiss. (But Sigma and others?) In any case, I’m curious about what you think of EVF technology and if is now a viable way to address manual focusing issues. I just got a used A7 for use with Zeiss ZM lenses and find the EVF quite natural and easy to use. (I find the ergonomics weird and nasty, but my favorite camera is a Rolleiflex 3.5f.)

    Also, it seems that you’re now using lens adapters in specific circumstances. I remember you’ve had reservations in the past. Curious about your views now.

    Finally, your Hanoi series on Flickr is great.

    • I’m glad I’m not the only one having issues with Sony ergonomics!

      I still have reservations for critical work where cross-frame consistency is required; but for cinematic photography like Hanoi, very little of the frame is actually in focus, so it isn’t a problem 🙂

      • Martin Fritter says:

        You didn’t comment on EVF! (In general.) It’s really quite good on the Alpha 7, although I feel kind of like I’m wearing virtual reality goggles.

        The Sony is over engineered and under designed, but there’s a nice little camera under all the cruft. Not a professional camera, I think. The only reason for you to get involved to review it as an alternative to the various Leica digital M’s for the use of M-Mount lenses. Not exactly a vital concern for you, I suspect.

        • I found it okay but not great – lacking brightness especially for tropical weather, and lacking eye relief for spectacles wearers. It’s passable, I guess.

          M mount on the A7 series is pointless because the lenses need offset microlenses on the sensor to work without corner shading and hue shift. And that rather defeats the point of a high potential sensor by a negating it with poor optical synergy…

        • “It’s really quite good on the Alpha 7, although I feel kind of like I’m wearing virtual reality goggles.”
          The things is, with OVF you see what the naked eye sees, but not what you will get. With EVF, you see what you will get, but not what the naked eye sees.
          So it’s more of a choice of what do you prefer – you can’t have both for now… 🙂

      • exposuredl says:

        I had the same issues with ergonomics. Didn’t try the II series but went to have a look at the A7r when it came out and after testing both it and Sony A99 in a Sony store I was just unbelievable how Sony had a nicely thought out control layout in A99 and completely abandoned it in the A series. The A99 was to me a better camera in every way, so I kept my Nikons and left it at that 🙂

        Great review, Ming.



        • Agreed – right down to the fully</em duplicated controls on the vertical grip. I suspect that size has a lot to do with it – there is such a thing as too small…especially when the lenses do not (and cannot) shrink to match.

  20. I’m a bit surprised to read that the colors of the Canon are that much better than those of the D810. I think in your D810 review I read that it was the camera that needed the least color corrections from all the cameras you tested so far.

    • I hadn’t tested the 5DSR at the time. Note that I said pleasing not necessarily accurate. Neither of them is perfect over the entire color spectrum.

      • Thanks for the clarification. I guess colors will always be a mystery to me. 😉 But so far I can live with that. Photography is just a hobby for me after all.

        • They shouldn’t be; color hugely affects perception even in small shifts. So long as you are consistent and the final output looks the way you expect on multiple devices, and ideally, print, then you’re there.

  21. Ming, you wrote: “I’d have to do significant compensation work on files with the D810, but the 5DSR’s files just looked correct and pleasing after a single initial camera profile.”

    I wonder if your workflow for generating a camera profile wouldn’t make a good subject to talk about? : )

    I’ve never been too happy with my D800E’s color rendering, greens especially (which is disconcerting when one prefers to photograph landscapes) nor have I succeeded in making a viable profile in my attempts.

    • It’s not so easy to show in a post – but it is in PS Workflow II 🙂

      • Would that be Adobe Camera Raw workflow only?
        Or would the profiling process in the video result in one that is available to other programs?

        Is it me or are the comments here decidedly more rational of late. Great job people.

  22. Do you feel the 645Z still has an edge in terms of resolution under ideal conditions?

    • Overall, yes – not very much, though. Mainly because pixel acuity is slightly better leading to higher micro contrast and perceptually better edge transitions.

  23. ming…all i can say is “it’s all good”.
    your work will stand up either way. even in an art form that is more mechanically oriented than, for example painting…it is STILL all about the brain/eyes behind the camera that ultimately matters, not the sensor inside it. 9especially when you are vacillating between 2 extremely high performances examples of such)
    keep on.

  24. Great honest review Ming!
    I guess i will be the only one with both D800E and 5DsR and Zeiss lenses which fit both!
    Interesting about the Nikon sensor being better for B&W. Since I have a lot of MF lenses, the LV implementation is so much better in the Canon.
    Battery life is the worst thing for me so far, but i will just bring more batteries. But there will be fun late night battery charging to do when on multi day travel trips. May need another charger to charge two at a time.
    My Conurus converted Contax N 85/1.4 and 70-200 work well on the 5DsR when I need AF but want to keep the Zeiss microcontrast, colors, and rendering. Still need to try out my 500L or TS-E 17 on it. Should be incredible.
    The extra resolution really comes through in your above landscape shots and must be really good for your Ultra prints. I’m excited to be getting my set of Piezography Special Edition Gloss and Matte inks to print some B&W images with. The K7 inks and QuadToneRIP lay down 45% more ink in the highlights and thus should hold better detail there. Will be interested in how you and Wesley think about it if you try them.

    • I’d like to join you, Wayne, but the wallet simply does not allow. And it only makes sense to shift it now whilst depreciation will be at a minimum. Agreed LV implementation is better in the Canon – faster to move around and the ability to single press to go back to centre makes things so much faster. But the ‘BUSY’ lockup is not so good.

      The 17TSE might be a bit disappointing, to be honest. It isn’t that great in the corners even stopped down and unshifted.

      K7 – we’re still looking for a used printer to sacrifice. Once they go into the system, kiss color goodbye.

  25. This is an interesting closing review because I am very interested in the 5SDR, especially for the 24/17TSE’s. The ability to employ Zeiss optics and share them with µ4:3 bodies is extra gravy. I am very concerned about the focusing issues with manual lenses & focusing screens/Zacuto viewer however—enough so that I actually asked Zeiss about their intentions of producing their own focusing screens to allow easy focusing of (especially) their Otus optics. They replied they have received similar questions/requests before and have no plans to produce focusing screens or work with any third party to do so. That is very frustrating to accept. I have no problems with manual focusing per se as I used the older Olympus manual system in the past and loved it. I just want a modern, digital equivalent camera system for usage with real estate, landscape, & macro photography. As you said, Nikon has no 17TSE, and I wonder if their 24TSE-equivalent is as flexible. For most other types of photography, especially those genres requiring fast focusing, I thought I could use an µ4:3 body with carefully chosen zooms and a few primes.

    • Sadly, what you want doesn’t exist. We all thought the DF was going to be a proper digital F2/3…instead it turned out to be a retro-geek exercise with bad function and poor form. The chances of something like that ever being properly executed are nil – the closest you’re likely to get is a CFV back on either the FlexBody or V series; the only problem there is the crop factor and the designed-for-square camera.

      The Nikon 24 PCE is fine if you get a good copy. I didn’t find the 24 TSE II to be markedly better or worse. The 17 is a bit compromised, but then again it is the only 17 TSE…so if you need it, there’s no choice anyway.

      • Barry Reid says:

        I don’t see too many problems, resolution-wise with the 17mm on my Sony A7r. OTOH…

        If there is a bright light source anywhere nearby then the lens absolutely MUST be flagged. So mine never goes on the camera without a ‘FlareDinkum’ show mounted lens shade these days. Incidentally, per Ming’s earlier comment on the first 5Dsr I believe the same may apply to the Zeiss PC Distagon too.

        • Could have had a bad sample; being a not-so-popular bit of glass they only had one in the shop for me to try. It didn’t look astigmatic though, which makes me think it wasn’t necessarily misaligned. In practice 36MP is noticeably more forgiving on glass than 50MP though.

  26. Lichtbild says:

    A interesting post, enjoyed reading it, as always. Have you ever considered to transport all of your photo gear via Fedex or UPS since the Airlines have excluded all risk of lost camera gear?

    • Same risk problem – then compound it with taxes at destination and the risk of somebody at the receiving end pilfering it. Plus you don’t know how it will be handled in transit, either.

      • Lichtbild says:

        Ok I can´t argue with that so the only solution is the need to plan ahead when flying with lots of gear and with a camera equipment insurance, which will give some sort of peace of mind 😀 hopefully

        • I’ve been looking for insurance in this part of the world for the last 10-odd years – nothing that doesn’t require a 25% premium, which is crazy. 😦 In any case, insurance is not going to fix the problem if your alignment gets out on a critical part just before a job…simply too risky.

  27. Willian Wives says:

    I really enjoy that you mix your reviews with serious art! They are all very good photographs, and I really like your consistency in the colors.

    • Thanks. Unlike other reviewers, I have a professional reputation as a photographer to maintain. So even if it’s ‘just for a review’, images that get associated with you have to be of a sufficient standard, too. And more importantly: if I can’t use a camera to do my normal work, anything I write is rather meaningless, isn’t it?

      • You are head and shoulders above the regular bloggers Ming. Great artistry together with excellent technical understanding and a willingness to teach. I have never thought you influenced by the companies in your reviews. Thanks for your hard work in putting these free posts together.

  28. Hey Ming,

    Great review and balanced thoughts as always. Are you planning to test new Sony A7R II?


    • Probably my most frequently received question of the last few months…

      • and least frequently answered? 😉

        • Well… it is listed on the recommended gear list as added on 13 August 2015 under “current equipment”, so I suspect some thoughts will be shared down the road.

        • I think Ming respects what Sony’s doing, but is unimpressed by lens lineup and he’s sophisticated enough that the NOT 14 bit RAWs are crippling. And it’s expensive.

          • The lens lineup is only a problem for any system if it doesn’t contain what you need now – the speculation of what might come is a bad reason to buy any hardware. What we have with the lossy RAW files are tonal and then color compromises, which manifest themselves at worst as banding and artefact, and at best as a slight loss of transparency that the viewer can’t quite put their finger on (or might not notice) until you see say a D810 image side by side. And yes, it’s obvious in print – we tried with the A7II and D750.

  29. A little bit out of context: since you talked about the Contax Zeiss glass, have you ever used the 135mm f/2.8? Thinking about buying one to use with adapters.

    Good review, as usual.

    • Nope, sorry. Didn’t investigate that one as I have the 2/135 APO.

    • I have the Contax 135/2.8 and while it has a great ‘look’ (ultra smooth bokeh) it’s noticeably soft wide open on my 60D due to strong longitudinal CA. Not bad for portraits and shallow depth of field, but it mostly likes f/5.6 and smaller. On the other hand, it’s fairly cheap (I bought mine for $220 including an adapter) so that can be forgiven. It takes extension tubes well for macro use, but the focus range is fairly small so you may be swapping lengths frequently.

      Although it’s a different focal length, I personally recommend the Contax 85/2.8 if you can find one. It is spectacular at any aperture with similarly smooth background blur and a great transition between the plane of focus and the background, and it’s extremely tiny. It’s the sort of lens that makes me want to go out and shoot.

    • Barry Reid says:

      It’s one of the cheapest lenses in the Contax range and the price is entirely justified by it’s relative performance – I have the later, optically improved MM version and it’s possibly the worst of the 15 or so C/Y fit Zeiss lenses i’ve tried the 80-200 zoom has less CA and may be sharper. I’d second what vcarlo says – the 85/2.8 is a great little lens but the 100/3.5 is also a very sharp short tele.

      If you are on mirrorless, the Contax G fit 90/2.8 is also reasonably priced and sharp.

  30. Hi Ming,

    Can you share your thoughts on the 24-70 f4 and 70-300L?

    • Sure – the 24-70 is good but not spectacular; it is in a similar performance class to the Nikon 24-120/4 VR. It does fine stopped down a bit and with the benefits of software correction. I wouldn’t use it for super critical applications. Wide open there is some variance in performance especially at close distances, though this appears to be a focusing issue rather than an optical one. Weakest performance is around 50mm.

      The 70-300L is a surprisingly good performer even wide open and even to the edges of the frame. It weakens a bit above 200mm, but not that much. The only thing I’d watch out for is some loss in performance when the lens is in portrait orientation – I’ve seen this behaviour on many IS/VR lenses and can only surmise it has something to do with the suspension of the stabilising elements and the way this is done to allow for compensation in one axis and not another when panning. It probably also has something to do with the lenses being calibrated in the landscape orientation.

      • Thanks for sharing, I’ve enjoyed using mine with the 6D, but I shoot personal landscapes and family photos, vacation, etc. and don’t normally print above 13×19. I can see how your needs would differ for sure.

      • That’s interesting to hear — the vertical orientation and IS, that is. I’ve always thought my Sony 24-70 was a little bit wonky in vertical on the A7r but this is the first time I’m hearing that it can be an IS-related issue. I’m more surprised that there isn’t much talk about this (or perhaps just not talked about at the “usual” places).

      • Mirror/shutter vibration is worse in portrait orientation because the movement of those elements becomes perpendicular to the axis of where the camera is held (hand-holding) or mounted (tripod) – rather than the vibration being absorbed at the bottom of the camera as it would in landscape mode, in portrait mode it resonates back into the camera. Here is a comparison I did of portrait vs landscape on the 5DSR, tripod-mounted:


        And a full 5DSR mirror/shutter vibration test:


        • Thanks – the direction of travel of the shutter would explain why we see that when using the optical finder, but it doesn’t explain why it still happens in live view with EFC…

  31. Great review and detailed explanation. I’m glad that this camera interested you enough to give it a 10K spin. Hopefully Canon will be able to match Nikon DR in the near future… that is the main thing that bothers me with my 5D3. I like most everything else about it.

  32. I was expecting part II; so thanks Ming.
    Any black & white tests?

  33. “Encounter” and “Wormhole” are great. Love your work.

  34. What I and maybe also you need is a body with a mixture of Sigma DPM Foveon ISO 100 IQ in all ISO levels with noise & dynamic range on all ISO levels like a Nikon D800 and D810, with an AF-system like a Nikon D4s and all that in a body size like a Sony A7 and of course small & light native Zeiss lenses with AF in all focal ranges (FFL and zooms) for this available.

    And this body please all for less than 1.500 USD. We are not there yet. But it will come. Sooner or later.

    • $1500?! That’s an outrage. :p

      But I agree: sooner or later. Look how far we’ve come even from the D3 generation.

    • As a long-time Sigma Foveon user and fan, it’s great to see a nice comment about their image quality, even though it is severely “ISO challenged”

  35. I saw some sample images for this camera months ago on Canon’s Japanese website and must say that I was not impressed with the level of detail. I have taken night-time shots with the Sony A7R that I am currently usign at the moment and found the camera’s ability to reproduce the smallest details to be impressive (I was using the 55mm Zeiss 1.8). What is the point of high resolution if the level of detail / sharpness in the shots leaves a lot to be desired???

    • I have never seen a single well-executed sample image from any camera maker, and I agree, their samples are rubbish – and not at all representative of the performance that is possible. The images shown and printed at the local Malaysian launch were equally disappointing; diffraction, CA and camera shake everywhere. Perhaps it was the photographers the organizers hired. Why they don’t make a better effort at presenting their flagship, I don’t know…

      • i suspect the reason questionable images get used is that (sadly) what you have noticed already is true: images don’t matter (at least in many cases). specs and selling gear are what it’s all about. i would be scared to guess the ratio of cameras (including the ever more capable camera phones) sold per year as vs high res fine art prints. i’m sure it’s a crazy ratio.
        everyone is looking to cut costs where possible and if the marketing department realizes they can get by using inexpensive/technically questionable shots they just…will.

        • eli may have found the problem behind the (way) less than stellar images used by almost all manufacturers in their ads; … the images are most likely chosen by the marketing department rather than photographers. As Ming mentioned, great shots in camera ads are rarely seen. They’re so consistently bad that the Evil Conspiracy side of my brain suggests that choosing bad sample photos for camera ads is done purposely, to entice us to look and say to ourselves; “This is junk! Even I can do better with that camera! Where’s my wallet?!”

          Yeah, I know. It’s a stretch.

          • You may be on to something, at least to a certain extent and in certain circumstances.
            Keeping in mind the base line reality: camera ads are about selling gear. They are not really about displaying the artistry and skill of some great shooter. Showing off the talent/skill of a given photographer does not really make a casual shooter/camera consumer buy a camera.
            However the idea that a given camera is “good enough” for a pro (but is financially attainable by almost anyone with disposable income) definitely helps sell cameras. And an attractive/pleasant (if technically unremarkable) image viewed at medium red can be produced by almost any camera (and many or even most photogs)….so other things have to be focused on (both literally and figuratively)….like specs and fun subjects…and saving some marketing cash along the way.

            • I’m not sure I agree, Eli. Here’s a typical consumer (and frequently enthusiast too) reaction:

              “Wow, what a shot! What camera did you use?”

              Nothing about any of the conceptualisation or execution.

              And the reason pro endorsement is useful is simply because that person is known for a certain level of work; to imply that they use camera X for that work is basically the same thing by association.

              • sure thing. i’ve heard plenty of this…and marketing certainly does happen in all scenarios.
                the thing is…under the right circumstances i’ve heard this about iphone and point and shoot photos….which kind of turns the whole conversation on it’s head from a camera marketing POV.
                you also run into something which is central to much of photography: how much of the response to an image is due to the subject of the photo (an attractive and/or famous woman for example) /stunning tourist-y vista at magic hour vs the skill, talent of the shooter OR the brand/type of camera.
                it’s a bit a a maze…and i honestly think if camera makers were able to determine that the technically best shots indeed helped sell the most camera they would use only those. compared to product development investing in technically air tight shooters seems like a small investment.

                • I don’t even think it’s an investment as such – most would be happy to do it in exchange for a camera which they’ll then go on to use and indirectly market for you…

                  • which still begs the question: if getting better shots for promo could be so easy (find a skilled shooter and give them free gear) why is it not done more consistently?

                    • Internal politics. I’ve shot samples for Zeiss and Leica, but remember Nikon and Canon refuse to work outside a certain small pool…(though I believe McNally was one of those for Nikon a while back).

          • It’s far more likely to be ‘we have no budget to hire a photographer. Just get one of the engineers to take some snapshots’.

            • The shot I took with the Sony A7R was one that does not require any technical skill or know-how. I placed the camera on a tripod in the dead of night, pointed it at a famous hotel in Dubai and took the shot with the timer (this camera is notorious for all sorts of image-blur problems though I am not entirely sold on the “shutter shock” theory that has made its way around the net (equally, I am not denying that there is a problem – perhaps the latest version has solved it???). What I’m saying is that perhaps Canon and other manufacturers do not need skilled photographers to demonstrate whether or not their gear can deliver tack sharp, detailed images – the trouble with using technically adept photographers is that their results / findings may show up the camera’s shortcomings. I am not necessarily agreeing or disagreeing with any of the comments in this thread – just reiterating Ming’s earlier question in the thread – “Why they don’t make a better effort at presenting their flagship, I don’t know…

              • The shot I took with the Sony A7R was one that does not require any technical skill or know-how. I placed the camera on a tripod in the dead of night, pointed it at a famous hotel in Dubai and took the shot with the timer (this camera is notorious for all sorts of image-blur problems though I am not entirely sold on the “shutter shock” theory that has made its way around the net (equally, I am not denying that there is a problem – perhaps the latest version has solved it???). What I’m saying is that perhaps Canon and other manufacturers do not need skilled photographers to demonstrate whether or not their gear can deliver tack sharp, detailed images – the trouble with using technically adept photographers is that their results / findings may show up the camera’s shortcomings. I am not necessarily agreeing or disagreeing with any of the comments in this thread – just reiterating Ming’s earlier question in the thread – “Why they don’t make a better effort at presenting their flagship, I don’t know”. In other words, if you’ve got it, flaunt it. Given that it may not be such a challenge to do so.

                • Barry Reid says:

                  Really, shutter shock exists – but you’ll rarely see it on an exposure long enough to warrant a tripod – due to the cumulative nature of the exposure. Try a something like short tele at 1/100th of a second and mid distances…

                  • Based on my experience with the camera to date, I think you might be right. The first and LAST time I sent the camera to Sony, it took them 4 months to acknowledge the problem and then repair it. Although, I can’t say that it’s now perfect, I don’t experience it quite as often and nowhere near as severely as I did before (it was really bad with mid and very distant buildings with repeat patterns). Thanks for clearing that up. I am curious to see whether the new 42 MP model with the 5-axis image stabilization feature will resolve this issue. I think I’ll wait until a few reviewers experience the pain first (assuming there’s still a problem). Lesson learnt – if you’re in the market for a Sony mirrorless, be afraid; be very afraid!!!

        • Even if you change that to ‘cameras sold : number of people who print’ the ratio is going to be extremely scary. Even amongst this reader pool, undoubtedly representing a more serious bunch – I suspect the number of printers is perhaps less than 10%.

          • I polled the rangefinderforum folks about the number of *framed* prints amateurs had done the previous year. (Quite a few pros couldn’t help themselves and responded too, skewing the results significantly.

            Framed print poll: http://www.rangefinderforum.com/forums/showthread.php?t=113890

            • I think part of the reason for none for most people is a) access to good printers and b) once you decide to print, that’s a level of commitment to an image that most people simply do not have – deciding what to print is actually a very good way of curating 🙂

              I guess it must be people like me keeping the printers in business – I probably made close to a thousand prints last year…

          • absolutely. It is one of the paradoxes of the digital age and it occurs to me daily.

  36. If only they actually hire a professional photographer/engineer as the project manager…

    • I suspect that finding somebody with skills at the necessary level in both disciplines might be tough. But all the same…yes, I’d like to see a camera that was actually designed by a serious photographer, not a marketing team.

  37. Excellent, informative, and unbiased review.

    Not surprised about the DR. Also not surprised that the Canon’s evaluative metering is less predictable or precise as Nikon’s. As you probably know, there’s quite a substantive difference between Nikon’s RGB metering system vs that of competitor’s evaluative systems, one of them being that the Nikon system does an instantaneous cross-comparison with tens of thousands of image types stored in the camera’s database before making its final exposure calculation. This is something that’s been baked into the Nikon “matrix” metering system since the days of the F4.

    Further to that, your review does help to clarify why Canon remains a popular choice among so many pros, however, despite their sensor architecture lacking the high ISO performance or DR of Nikon’s.

    • Nikon must have changed the matrix formula at some point post D3, though. That camera – and the D700 etc. of the same generation – required almost no exposure compensation; the newer ones tend to be a lot more variable. Perhaps too much bias to the AF point in use? From the launch locally Canon were claiming to use something supposedly similar with the 5DSR/1DX, but I’m not seeing it.

      What I am seeing is that if you’re heavily invested one way or another, there still aren’t really any compelling reasons to switch unless there are very specific things you need (5DSR: resolution under controlled lighting, ergonomics, landscape color, some specific lenses, AF performance; D810: dynamic range, wider shooting envelope, some specific lenses, flash system, metering reliability, and I suppose also arguably ergonomics). Bottom line: the grass is always greener and all that…

  38. The third image from the top puzzles me a bit. The tone of all the greens appear to be unnaturally similar. I’ve seen the same sort of “hue/tone flattening” when comparing some Bayer and Foveon sensor foliage shots. For some reason unknown to a technically challenged person like myself, this is most noticeable when shooting green vegetation.

    Are my eyes playing tricks on me, or was all that green stuff really that similar in tone and hue?

    By the way, I just discovered your site, and in a very short time I’ve grown to respect both your knowledge and your refreshingly honest review methods. Great stuff!

    • Thank you. I think there are a few things going on here – one, the scene was pretty uniform; two, JPEG compression takes out some of the tonal differentiation and downsampling doesn’t help either; three, it may be a monitor calibration issue.

  39. Excellent unbiased report. I shoot D810 with a D800 as a back up. I like the Nikon ergonomics, the lens range suits my needs and the Nikon service, if ever needed, is first class. I also like the exclusiveness that comes with Nikon possession. Nikon seems to only push their product with discretion. You never see a Nikon add on the back of a toilet door.

    • Martin Fritter says:

      Thom Hogan assiduously covers the Nikon beat over at his blog and has been very critical of Nikon’s quality control and customer service in the US. Are things different in other markets?

      • Sadly, no. I get decent service because I know the guy who run the service centre here, but that’s about it.

      • Here in there State of New South Wales, Australia where I live I’ve been a Nikon user for 30 years+ and have never had an issue with any Nikon I’ve had, from my old FG, Nikons 5, D100 right through to the D810.Service first class, quality 1st class. The Nikons all provide more technology than use and the quality or their colour renditions, auto focus capability and all of their manual actuations are outstanding. In my old work we used first Pentax then changed to Nikon back in the early 70’s , gave the cameras a real hiding and never had a problem. I can’t speak more highly of them. I also use Nikon lenses on some of my enlargers without any issues.

        • Martin Fritter says:

          I wonder if this is a NSW thing. Maybe you have a great Nikon support team. Maybe the general level of technology service in Australia is better? Nothing gets serviced well in the ‘States, except maybe very high-end luxury goods (not including Leica).

  40. I like the mood in ‘Encounters at the train station’ and ‘Curiosity’ a lot 🙂

    I don’t think we are many who haven’t had an affair with a system that is different and better in some fields than what our present system may muster. For many and unfortunately we have to buy the things to know if the overall needs are met and the camera actually delivers where we felt we missed a bit with the present.
    In my native country renting and leasing has become more common and I know photogs who doesn’t even own the cameras and lenses they use for specific assignments. It is a luxury beyond imagination that we may buy whatever we might think of would be better for this and that.
    I myself have been longing for a more compact and versatile FF system that would be the answer to lesser the burden carrying a 810 + fast prime focal’s that reaches from wide to long, would have means of annulling my handshake and being as stealth as possible. Would sport a high end EVF to ensure better focus ability in low light and have super precise AF. I was close to jump on the Sony A7R2 system until I learned that Sony still didn’t address their crippled and awful lossy RAW and neither have dialed in their color algoritm so an easy profiling would be possible.
    My point is that if I just could have rented the camera and the lenses, it would have been far more economical than making a big mistake by buying the damn things.

    • Thanks Gerner. It’s just confusing: having had two of these affairs now, first with the 645Z and then with the 5DSR, I’m left more uncertain than ever. There were images I shot with both cameras that I could not have made ‘work’ with the D810; on the other hand, there were others I let go even though the D810 would have had no problems. I suppose if one never tried the other side, you wouldn’t know what you’re missing. I’m now in the position of knowing what I’m missing in every case, and coming to the only sensible conclusion that from an opportunity and ROI maximisation standpoint, the Nikon still remains the sensible choice. Am I troubled about knowing what I’m leaving on the table? Yes, but grass is greener and all that.

      In Malaysia there’s no way to rent. And none of the brand principals are interested in giving you a loaner unless you are willing to prostitute your ethics to swear on your eternal soul that theirs is the best product ever.

    • “Sony still didn’t address their crippled and awful lossy RAW”.
      Any first hand experience or just reporting the usual web myths? (BTW I don’t lke Sony, so no fanboy here).

      • He’s getting his information from me, and I’m getting it through my own examination of 7RII test files captured under controlled/known conditions and cross-referencing with Lloyd Chambers. The information loss in RAW is still there, but causes different perceptual issues to the 24MP 7II because resolution comes into play. Banding is reduced because are more spatial steps possible for an edge transition, but there’s now some strange color response/tonality instead. And the shadow texture/sine compression noise is still there.

        • Yes, and thanks you for clarifying Ming. I of course also downloaded many RAW images from the net, but those didn’t lead to new conclusions either. Perhaps next time Sony.

        • K – understood. As usual, is refreshing to see that there are still people like you that likes to test things IRL instead that repeating heard assumptions 🙂

        • Yesterday Lloyd were flabbergasted when he found the IQ of a correctly exposed ISO 100 shot of a mountain lake strongly compromised. I have seen 7R2 RAWs that were even worse. There should be no reason for posterizing due to lossy compression except the water in reality showed some very harsh contrast sun reflections, so perhaps it is something else? I can’t think of anything else except Camera RAW would have difficulties de-mosaicing the blue channel? I suspect Sony to do more than throwing away valuable RAW data and heavily cook their files. But I can’t know it, can I?
          I would love to shoot the camera, but what I saw I cannot unsee.

  41. Hello Ming, just a hint/question: Klicking on these (beautiful as always) pictures, doesn’t lead me to flickr, but to yahoo, where I have to register myself? Greetings from germany, Philipp

    • Appears to be yet another flickr issue, sorry. Everything appears normal from this side.

      • Wow, talking about the issue seems to make it go away! Everything is fine now, thx 🙂

        • Flickr has on and off issues quite often, it seems – depends on whether their server in your part of the world is playing nice on a given day or not. Sometimes refresh is all that’s needed…

          • I also have problems with your Flickr links quite often. Not today though. (Also from Germany)

  42. An very good report, but the results of this review were forseeable! The Nikon , nor the Canon are keeping the peak of the lastest digital technologies!
    Each camera has his advantages and disadvantages. 100% are not possible, but the better, is the good enemy!

    • Everything is a compromise. The only questions are, which ones really matter?

      • Would it help, to be on Nikon’s or Canon’s payroll? You get anything loaned for free, and being able to play with all new and expensive toys! Become an Nikon or Canon Ambassador! Apple for it, they need it !!!

        • No, because you can also leave your integrity at the door. No amount of money can buy honesty. And on top of that, Nikon has made it clear to me they are not interested in working with anybody who isn’t Japanese or Caucasian. Canon is only interested if they can dictate 100% of what you say and write, even if it makes you look like an idiot. Ask yourself: would anybody still visit this site if that was the case? Doesn’t that defeat the point entirely of having access to my audience? Personally I much prefer the Zeiss approach…

          • Ming – “No amount of money can buy honesty. And on top of that, Nikon has made it clear to me they are not interested in working with anybody who isn’t Japanese or Caucasian. Canon is only interested if they can dictate 100% of what you say and write, even if it makes you look like an idiot. Ask yourself: would anybody still visit this site if that was the case? Doesn’t that defeat the point entirely of having access to my audience?”

            There’s an entire article, boiled-down to just five sentences. Nicely – and concisely – put.

            • It is also in a nutshell why those two companies will be left behind in the innovation race. Their corporate culture represents the ‘old guard’ in every way.

              • It is a great point toward your thesis that no taking risks is a path that ultimately leads to stagnation, then loss. In buddhism there is a concept that explains that in life, not moving forward means that you not simply stay still, but in reality you are slowing falling back…

        • Well, it is interesting that speaking of corporation malpractice you used the term “APPLE for it”… You, dear Sir, understand very well the game! 😉

  43. “The fear of loss is a path to the Dark Side.”
    (Master Yoda Dixit)


    • But…but…I just came back from the dark side!

      • “Feelings of loss of 50MP, could back to it lead you.
        Fear you not of Nikon future releases.”

        God, trying to write like Yoda gives me a headache.

        In the end Canon, Nikon, Pentax, Ricoh, Hasselblad, etc. do not matter. What matters is the guy behind the viewfinder!

  44. Excellent write up Ming. Did you warm up a bit to the 40mm perspective?

  45. Gary Morris says:

    A very pragmatic statement and decision. In the first photo… Stars Trees Line Up, the highlights in the sky seem somewhat over exposed. You note the lack of dynamic range with this newest Canon. Do you think the D810 would have handled this intense light better?

    • It would probably have handled the transition at the edge of the clouds a bit better, but I doubt it would have been able to hold everything – I was basically shooting into the sun at that point. I do know it wouldn’t have rendered the warmth of the backlit grass quite right though.

  46. Richard P. says:

    Hi Ming,
    Here is a different take on the subject 😉… On the tail of your latest “rethink” article, it always seemed to me that the Canon was a bit like a mistress during a midlife crisis. Sure you were seduced by her curves and how she felt in your hands, not to mention bigger … But you realized it was a mistake. So the Nikon gives you blisters once in a while, but don’t all marriages – even the good ones? They require patience, effort and commitment to reap the rewards. Now you are parting with the classic, “it’s not you, it’s me” speech. You’ll fondly look back at the good times once in a while, maybe even with a pang of regret at what might have been. But know in your heart you made the right choice.
    Cheers, Richard.

  47. Well thought out and explained. I knew from the start that the 5DS/5DSR wasn’t for me. I was interested to see this camera from the perspective from someone who would actually use the camera to its potential.

    • Like the 645Z – I really loved using it and the results, but the economics just don’t work because the cost is high and the envelope over which it can be deployed is just too small relative to my greater work. The commercial justification doesn’t exist, and ultimately, I’m still running a business.


  1. […] series was shot with a mix of the Canon 5DSR and 40/2.8 STM Nikon D5500 and Zeiss 1.4/55 Otus, and post processed with Intro to Photoshop […]

  2. […] //embedr.flickr.com/assets/client-code.jsThe Garden of Cosmic Speculation. The color palette of an English garden in an English spring is a tricky one to reproduce. Light is hard, greens are subtle but varied and never really electric – there’s a warmth to the palette which I’ve only managed to print right with the 5DSR. […]

  3. […] ensure there are no strange effects. I had one such problem with the 2.8/35 PC: it sang on the 50MP 5DSR, but was very soft on the A7RII – almost as though I was getting free tilt in addition to […]

  4. […] pro or the reasonably affluent amateur. There is the all-rounder Nikon D810, the resolution-champ Canon 5DSR, the tech-heavy Sony A7RII, medium-format-on-a-budget Pentax 645Z and now the Leica SL. […]

  5. […] Source: Long term review: Canon 5DSR […]

  6. […] entirely. I shoot under conditions that are far enough outside the ‘ideal’ envelope to render the 5DSR too niche a tool to justify itself for my business (and the camera itself has a narrow envelope too), so it had to make way – as much as I love the […]

%d bloggers like this: