#844. Monday Post (15 April 2019) – 100Mp mirrorless cameras, at long last …

#844. Monday Post (15 April 2019) – 100Mp mirrorless cameras, at long last …

… said no one, ever.

 

Fuji says โ€œYou will be amazed by the image quality of GFX100 coupled with GF lensesโ€

say mirrorless rumours.
 

Not only will that image quality be entirely believeable, my guess is you won’t tell the difference with the current GFX, in similar photographic scenarios. And I find the claim by Fuji worrying.

 
 

Format for format, we’re long past peak IQ. And every new sensor generation in the past few years has “only” expanded envelope, bringing the goodness of previous iterations to more varied shooting conditions, sometimes even at a perceivable loss of visual magic. If the new 100Mp GFX is well designed and built, it will bring to the table a host of very valuable envelope-expanding benefits : faster AF and possibly IBIS, for example.

 

Both would be major boons on my X1D. Although I have pulled off a couple of 1/5s exposure with a 30mm lens, most of my shots cease to be critically sharp at speeds below 2/f (1/25s for a 50mm lens, eg). That’s already pretty good, but another 2 or 3 stops of added reach would be lovely to have.

 

As for faster AF, well … no need to comment on that. I’ve never used a GFX, but the X1D is one slow focuser.

 
 

That being said, neither would be a life-altering benefit. For my style of photography and – I believe – for most who lust for medium format, neither of these is a priority. Big fat pixels and a beautiful image is all that matters. I’ll even grab a tripod for the X1D, if that can help ! The camera has been so good at doing what I have longed for so long, it could be 3 times slower and still not bug me.

 

So, you’d expect the 100Mp sensor to fill me with joy. It doesn’t. I don’t believe for a second that the X1D’s resolution has much to do with the beauty of its images. In fact, it’s probably the low-ish pixel-density that makes it shine. Much like the Nikon D700 in its time, among many others. But serious calibration is likely responsible for 90% of the camera’s goodness.

 
C-Sonnar 1.5/50 ZM on Hasselblad X1D – The Hassy’s large pixels are extremely tolerant of old lens designs
 

More pixels equals smaller pixels. This means a lot more stress on the glass and the shooting technique, and more noise. And more processing power, and more storage. And all for a resolution increase that will only truly benefit a few % of the target audience of the GFX line, already a pretty small population.

 

Why does this bother me? Because Fuji know better. They created one of those legendary cameras in which sensor pitch is a sweet spot and that drew the crowds (and continues to elicit teary-eyed comments) because of quality, not quantity : the X-Pro 1 / XT1.

 
 

So, what’s the nonsense about ? Either Fuji mucked up something in the current GFX and are correcting it in the new version (I doubt it) or they are giving into that most populist of marketing tactics : quantification of performance. Fuji, you’re better than that smoke.

 

When the X1D was first rumoured, it felt like an electric shock on a dying heart. To me, at least. We were being bombarded by stats, higher ISO, higher AF speed, higher frame rates … expanding shooting envelope while pleasure of use and image beauty stagnated. The main “culprits” (others will regard them as heroes ๐Ÿ˜‰ ) were Sony, who were milking their technological superiority to the max.

 
 

Fuji, in that era, appeared to be the sane ones, pushing reasonable pixel count, fun shooting and inspiring ergonomics. And Hasselblad appeared to offer an alternate road altogether. One on which engineers do their utmost best to make the image great and don’t give a number 2 about specifications. Medium format digital seemed like that quiet little corner of the world for haiku lovers, one in which fps, and all the other 3 letter signifiers of technical performance, had no dominion. A sheesh niche.

 

And now, here comes this new megapixel race, like it’s 2008 all over again. Please, Hasselblad, please, be the better company, here. Keep the moral high ground. The X1D is hands down the best digital camera ever made for photographic sybarites. Give us an X2D with a better EVF, fewer bugs and faster glass for shallower depth of field. Heck, why not throw in your exceptional pixel-shift tech if you must satisfy pixel-peeper. Even make it 100Mp if sensor markets really dictate it but please don’t insult our intelligence by claiming we won’t believe how much better the IQ will be. It won’t.

 
 

Philippe adds

Hahaha! Pascal has brought out the heavy artillery against quantification, the pixel race, etc… He waxes lyrical about the Nikon D 700 and its fat pixels. Fat pixels, ha! I thought he was talking about me…:-(

But is he using one? Of course not! He says that the megapixel race trashes the soul of photography, but what resolution is his present toy? 50Mp!(nobody has more, yet). What size his sensor? 44×33! Nobody has more yet, this side of a Hassy H6, a Phase One, and 30K$. How many stops of DR? You guessed it, nobody has more, yet.

So there he goes, ranting against exactly what he is using, loving, and flaunting.

 
 

Of course, being who he is, he will very cleverly explain that it is not because of the high resolution, high DR and large sensor that his pics shine. But it is rather hard to unbundle why something works, methinks, when you can’t compare like with like and change just one variable.

 

Fact is, a Nikon D 850 will produce much better output than a D 700. To date, the best IQ of any brand of cameras is delivered by their most recent models. I have yet to see a brand introduce a “worse” camera than before, though it could of course happen. Except that some people wax lyrical about how good some “golden oldie” camera was, while using a newer one, of course…

 


What I see is that, in some cameras, technology and quantitative peformance are not in synch/step with “real IQ”, and certainly not always with the shooting experience. But trashing them, Pascal? Well, fact is, much as it chagrines/pains me to acknowledge it, especially publicly, his images with the X1D are in fact very clearly better than any output I have seen from any 35mm camera, as well as from the Fuji GFX. Maybe the man has a point after all… Except that, now when the Fuji is out, I can tease him about his “mediocre” resolution. And even more so when Hasselblad, eventually release a 100Mp camera with IBIS….

 
Email: subscribed: 4

25 Comments

  1. Avatar
    Kristian Wannebo April 15, 2019

    Pascal,
    you “find the claim by Fuji worrying”.
    Right!
    But just think of an angler describing his latest catch…

    ( I note that Fuji says that *very* carefully guarded, sensor + lenses, and they don’t (explicitly) compare to the previous sensor. And so it comes out as a nice catchy (angler’s?) phrase.)
    – – –

    With the same pixelsize as the X1D
    a FF would have 30 Mp,
    and an APS-C 12 MP.

    So one might compare files from the X1D with later FFs and not so late crop cameras (though the latter have older generation sensors).

    Except that Hasselblad calibrates each sensor, and probably has invested a lot more in their colour science software.

    Philippe says:
    “To date, the best IQ of any brand of cameras is delivered by their most recent models”

    Certainly, if one needs the resolution.
    If one doesn’t need more than 10-15 MP, the difference to those (not so recent) models seems to be rather smaller.

    E.g.:
    Kirk Tuck has in a couple of posts on
    https://visualsciencelab.blogspot.com
    praised the *file quality* of older 10-15 MP FF(?) Nikons when processed with *modern* software – rather better than his original results, and closer than expected to his modern higher resolution cameras.

    • Avatar
      pascaljappy April 15, 2019

      I like the comparison with the angler ๐Ÿ˜€

      Also, โ€œTo date, the best IQ of any brand of cameras is delivered by their most recent modelsโ€ is not at all how I feel. Having compared my very old Nikon D80 with my Sony A7r2 with identical lenses on identical scenes, I can tell you the old timer can produce images with a lot more soul. What modern cameras do is carry most of the goodness of older models to many, many more shooting scenarios. I don’t think many Leica fans would dispute that the M9 created a look that hasn’t been matched since. But more recent models are more convenient. Ditto XT1 and successors.

      As you say, the X1D has the pixel density of a 12Mp APS-C. Not a lot of pixels for the surface. All other things being equal, that means much better signal/noise ratios. Coupled with great calibration, you have a fantastic camera. Putting a 100Mp sensor in that would let us print larger but we’d lose some of the magic, unless sensor tech improved in parallel (which it hasn’t for at least 5 years).

      Cheers

      • Avatar
        Kristian Wannebo April 15, 2019

        Pascal,
        I can’t but agree.

        Yes, the Nikon D80 has slightly larger pixels than the X1D.
        ( The Sony A7 (FF 24 Mp) might be fairer to compare with than the A7r.)

        The Leica M9 has a CCD sensor, and I’ve read many praises of those compared to CMOS sensors (except in low light).

        And then there is the in-camera-ware handling of colours and tones…
        ( whether it is the choice of colours and density of the colour filters or different readout hard- and software)
        … which gave you all those Sony A7r problems!

        Perhaps Leica prices gave them the economy to invest more in their colour science.
        – – –

        I’d be curious to see a comparison between the X1D and the Phase One Trichromatic sensor, *just for fun*. Phase One claims it has truer colours.

        Now art (also photographic) isn’t about *true* colours! (Unless it’s documentary.)

        How far would the X1D calibration suffice compared to the results from the different and denser Trichromatic filter? Would the probably lower noise from the X1D (also from the ~25% larger pixels) compensate?
        – – –

        > “.. 100Mp sensor in that would let us print larger ..”
        Keith Cooper has written two good articles on printing large
        http://www.northlight-images.co.uk/printing-low-megapixel-images/

        and huge
        http://www.northlight-images.co.uk/making-a-giant-print-from-an-11mp-image/

        from small(er) files.

        Cheers!

        • Avatar
          Brian Nicol April 16, 2019

          I am so delighted I got my ancient new X1D at a dramatic lower price than new release. A X2D of any sensor density will be beyond what I want to pay. The X1D pixel density is perfect for me and is the king of low noise. The sensor did not have great AF speed ability and the Fuji was stuck with the same sensor AF abilities. What I do care about is that the X1D nails the correct focus every time which is a first for me. I do not need ultrafast AF for a medium format methodical camera. Even if it had superfast AF, I would choose a different camera for that application anyway. The key difference for me is the Natural look and colour of the files which Fuji does not have calibration of sensors. Anyway, I will not even think of โ€œupgradingโ€ Max, my X1D, as he already exceeds my skill envelope and a denser sensor will require even greater care shooting care such as higher min shutter speed and so on

          I disagree that the latest sensor is the best. A higher density sensor means you have to get expensive new glass as a lot of cheap glass does not work well with 50MP, usually has more noise than less pixels, file sizes start slowing software down, … Sensor rendering can change. How many people are printing large or just feel great that they are using a bazooka to shoot birds. Sensor tech made great strides up to 5 years ago but the tech has plateaued and new sensors are minimally incremental.
          Camera sales have dropped 35 percent over the past year. I think it will continue to decline as people realize that the difference in 425 Horsepower and 410 horsepower is not significant day to day but the cost in redevelopment of muscle memory to capture fleeting moments is the biggest real cost.
          Camera release cycles will have to slow down and I would be real happy if they get rid of the video crap in my serious picture capturing tool as well as anus AF, animal eyeAF

        • Avatar
          pascaljappy April 16, 2019

          Very interesting Kristian, thank you.

          In the Phase comparisons between their ordinary sensor and the trichromatic version, I was actually shocked by how way off the initial colours really were. Not that I’d have noticed without the comparison, but it just shows how good that technology can be. I’d much rather see that come to the X1D than more pixels.

          As you say, art is not about colour fidelity. But I think neutrality is a better base for interpretation than a biased camera. Unless the bias itself is part of your process, as some film were, obviously.

          Cheers

          • Avatar
            Kristian Wannebo April 16, 2019

            > “.. neutrality is a better base for interpretation..”

            Agreed, Pascal !
            I just thought the X1D was perhaps good enough to be practically neutral.

            ( And the more neutral – when one takes advantage of it, the more requirements on screen and print colours grow, and the more precision is needed in the chain RAW > screen > RIP > print.)

            Cheers

  2. Avatar
    jean pierre (pete) guaron April 15, 2019

    This is going to end up like “Murder on the Orient Express”, where every passenger on the train shoved the knife in!
    Pascal, you invited this onto your own head. Years of extolling the virtues of Sony – then out of the blue, abandoning full frame for medium format and lashing out to buy a Hasselblad X1D.
    To make matters worse, confirming some of the intuitive remarks I’ve made in my somewhat briefer term as a member of the DS confraternity, about the idiocy of assuming that more pixels EQUALS bettter image quality, when it really simply means smaller ones – capable of leaking into each other. And unless they are staggered, from one row to the next – ie, IF the are all lined up both vertically & horizontally, instead of using the diamond shaped spaces between them to slide alternate rows sideways and down – they will have the same total surface area anyway – that is, unless there’s a gap between adjoining pixels.
    But hey. who’s pixel counting? Besides, it’s not the pixel numbers you’re going on about – scarcely any more than Dallas & I have on our D850’s, or Dallas’s Z7. It’s the quality. Ah! But that takes us back to Sony, and all the virtues their sensors had. Until recently, apparently.

    I’ll stop at this point. Too many hands waving, too many voices crying “My turn, my turn!”
    (BTW – loved the shot of the bicycle. Very reminiscent of the way Arthur Hitchcock used to make cameo appearances in all his movies, and you had to spot where he turned up. As he gained in experience, he gained in pure devilry, and made it harder & harder to spot him, in each successive film. HINT – enlarge the lead photo in a separate window on your screens, if you want to look for it! ๐Ÿ™‚ )

    • Avatar
      pascaljappy April 15, 2019

      Ah, nothing like a good murder mystery ๐Ÿ˜‰

      The virtues were real but I seem to remember bashing the ergonomics and flaky colours quite a bit, to be honest. The X1D had been at the back of my mind since one had been sent to me for a review, but I was just waiting for the right offer. Which, luckily, happened a few months ago.

      Also, I’ve always been an advocate of larger pixels; you must be confusing me with someone else.

      What bicycle picture is that?? ๐Ÿ˜‰

      • Avatar
        jean pierre (pete) guaron April 16, 2019

        Dunno how my reply to you has ended up as a reply to Kristian’s post! But if you’re seeking answers, read on (below).

  3. Avatar
    Dallas April 15, 2019

    Pascal, well done on another excellent article. The FF MP wars I thought were over, now the MF boys are at it hammer and tongs if the rumour mill can be believed. Let them play I say as I will stick to my 46.??MP sensor which is more than enough for me and the MacBook Pro and storage capacity I have.

    • Avatar
      pascaljappy April 16, 2019

      Thank you Dallas. Not that you mention the 46Mp of the Z7, I wonder whether the 100Mp thing is about beating FF commercially rather than beating Hasselblad … It’s a poor argument, whatever the target. Enjoy your lovely camera. I was stunned at how nice the viewfinder is. That matters more than pixels. Cheers.

    • Avatar
      jean pierre (pete) guaron April 16, 2019

      OK – win some, lose some – can’t remember who else. ๐Ÿ™‚ Another one of my “oops”!
      At the time, it struck me as nuts to suggest that more pixels automatically meant a sharper image, when alongside the trend I was finding pros (with their knowledge & experience – AND gear) telling us all they could get perfectly sharp images from 6 MP cams, or 12 MP cams, and that unless you were printing shots the size of a double bed sheet, you’d never be able to tell the difference anyway.
      And techos telling us all that smaller pixels capture less information and leak into one another, blurring the image. I did eventually come across evidence of that, in the relentless competition between the majors in the DSLR FF market – enlarging images with (I think, from memory) a 50 MP or thereabouts sensor being decidedly blurrier and less contrasty than others with a 36 MP sensor, or thereabouts.
      Of course that’s nothing to do with pixels is it? – it’s all the processor’s fault! Or maybe the camera should have been on a tripod. Or the lens shouldn’t have been on MF.
      But the bicycle? – quelle honte! The very first one – actually, I think there are three of them – velib bicycles, I imagine.
      FYI – I have enough set aside for one more cam – not, perhaps, a 100 MP Hassy – but a jump into the major league, nevertheless. Truth to tell – although I was always happy with my Contarex, in the days of 35mm, I grew up on 620 and (later) 120 roll film, producing 6×9 images. So by the 1970s I had already experimented with the Linhof 4×5, and later settled on a Bronica (mainly because it did everything the corresponding Hasselblads did at the time, at a fraction of the cost).
      Truth to tell – I have an affection for larger format cams. And it’s a bit like some of the other social diseases – pretty much incurable. Best examples include a great-great-uncle of mine, who obsessed over it – left his brothers to run the family farm, while he built a darkroom on the side of the family home on the farm, and travelled all over in a pony cart, laden with all his gear – making his own 10x and 12×16 collodion wet plates, on the run, and processing them in a tent, on site, before heading home to make prints. Thousands of them! When the State Archives fell upon them, they nearly wet their pants – they had no idea there was any such treasured memory of rural life here, in the 19th century.
      And a friend from NY, whose father was a wealthy Wall Street lawyer, so my friend could afford whatever. He chose a 4×5 Graflex Speed Graphic, and spent a fortune going around the world taking photos, instead of going home and starting to work, after he graduated from college. Where others would run around with their 35mm cameras, blazing away at everything in sight, he’d scout a location, set up for the shot, and trip the shutter once – for a perfect photograph. Quality, not quantity. It made an interesting basis for comparison with the rest of my (then) photographic friends!
      BTW – I love immodoc’s response to your post! It’s not a very good idea, though, to run high performance cars in traffic – it doesn’t do their engines much good – better to wait till Lamborghini et al provide them with EV high performance cars, when being in traffic will probably be better for them than blasting away with the pedal that releases the full fury of cars that can do 0-100 in 2 seconds flat! Seems that if you do it too frequently, you are automatically tamed by the gear in those cars, till the battery and engines recover from your antics. ๐Ÿ™‚ They should have thought of something like that years ago, and fitted it to all the ICE vehicles on our roadways!

    • Avatar
      jean pierre (pete) guaron April 16, 2019

      ROTFLMHAO
      Kristian, that’s naughty – that was a “poisson d’avril” – admittedly written BY a photographer . . . . It could shred some people’s brains, if they didn’t notice the faded footnote – spending 5 grand on a camera AND another 10 grand on a ticket to Aspen, only to find their precious camera fired blanks after the first 10 frames!

      • Avatar
        Kristian Wannebo April 16, 2019

        Aye, ๐Ÿ™‚ .

        But do check his photography!
        Some of it is very different!

    • Avatar
      pascaljappy April 16, 2019

      Brilliant. Actually had me worried for a minute ๐Ÿ˜€

      • Avatar
        Kristian Wannebo April 16, 2019

        Me too…
        ๐Ÿ™‚

  4. Avatar
    immodoc April 15, 2019

    Hi, why not 100 MP cameras? Those who buy them will enjoy the benefits – and the side effects of large file sizes.

    Some people like, and own 500 hp cars, which do not make travel in everyday traffic significantly faster.
    But they have fun with their cars, which is the main thing.

    Photography is a hobby. So, rationality should be applied flexibly, and sometimes sparingly,

    Have fun!!! If you need 100 MP, just do it!

    • Avatar
      pascaljappy April 16, 2019

      Oh absolutely. My beef is not with the camera itself, which will probably be brilliant. I just don’t like the marketing argument that more pixels make for better IQ. And yes, having fun is all that matters ๐Ÿ™‚

  5. Avatar
    Cliff Whittaker April 16, 2019

    Discussion of equipment details bore the #2 out of me. Good photographic art and discussions of photographic art excite me. I don’t care how digital cameras work, I only care that they do.
    Every new digital camera I have bought has been so much better than the one I had before that it inspired me to go out and make new images and try new techniques. So I did so I did. And I made better images. At first technically better. But once I got over learning the “wow” technical details I was able to concentrate on using the new technology to make better art.
    Camera details are simply a means to an end. Either they work or they don’t and I don’t care why.
    Pascal, your X1D is inspiring you to keep testing its capabilities and I, and other subscribers, are the beneficiaries of your efforts. I’m really look forward to your images posted here. But, honestly, unless you told me, I wouldn’t know from my computer screen whether you shot them with the X1D or your Sony. I just know I like the subject matter, the compositions, and the smooth, creamy contrast. But, the real proof is in the print and I know you must feel really good inside every time you make a new one. That’s the incentive that keeps fine art photographers going.
    BTW, the bicycle is on the left side of the first picture. Always gotta have at least one bicycle…. right? :))

    • Avatar
      pascaljappy April 16, 2019

      Thank you Cliff ๐Ÿ™‚

      The reality is exactly what you wrote : you cannot tell which is made with a Sony and which is made with the X1D. Pixels don’t matter near as much as ergonomics and how the camera inspires you to make photographs. And, in prints, the difference is even more difficult to spot.

      What the Hassy gives me is more lattitude to correct files and create more, better prints. At my print size (A2 max), resolution has no impact.

      I hate to admit it but, ahem, I discovered the bikes in the frame when they were pointed out by Pete in a prior comment. Ahem …

  6. Avatar
    NMc April 16, 2019

    Pascal and Philippe
    I donโ€™t want to take sides but in this case I think I will favour Phillippe. :O
    The simple fact is that if you want improved live view you need the latest generation of sensors, and that is as relevant to manual focus as it is for phase or contrast detect autofocus. I guess that Fuji have to take what Sony thinks is best or at least the highest common denominator of all its MF customers, and letโ€™s face it that the main advantage of MF over FF is resolving power (with tonal smoothness), not sensitivity or dynamic range.

    Having said all that, if I was in the market I would prefer a lower resolution fully up to date sensor, or maybe I am just blessed that the 16Mpix APSC that I have is able to show up my lack of technique so I do not need to go higher. ๐Ÿ˜‰
    Regards Noel

    • Avatar
      pascaljappy April 16, 2019

      Please don’t do that, Noel, I’ll never hear the end of it ๐Ÿ˜€ ๐Ÿ˜€ ๐Ÿ˜€

      Just to clear things: I’m not saying more recent cameras aren’t better cameras. They usually are. But what is improved is usually functionnal, not always IQ. And, to anyone printing less than 50 inches wide, the difference between 50Mp and 100Mp is just negative (better technique, better glass, more storage …) The good side of that sensor is that it provides faster AF and IBIS … I just don’t understand why Fuji wouldn’t insist on those very important aspects rather than focus on meaningless resolution.

  7. Avatar
    jean pierre (pete) guaron April 16, 2019

    Philippe – to relieve some of the pain of today’s tragedy, can I revisit some of your comments?

    You mention that the D850 produces much better IQ than the D700, and I’ve no doubt that would be true. What I found “interesting” was that shortly before the release of the D850, with its 45MP sensor, Canon released its EOS 5DS, with 50MP. So when the D850 was released, with only 45MP, I was startled.

    Yet when I later started seeing straight comparative shots – taken side by side, under identical conditions, with identical settings – I found the IQ from the D850 was sharper and clearer than the IQ of the comparative shots from the Canon. It seems absurd to mention it – the actual difference between a 45MP sensor and a 50MP sensor is hardly worth worrying about. And yet . . . Well the most obvious differences were two fold – blur and a lack of contrast. Of course you can always punch up contrast in post – but that’s scarcely the point, is it?

    And for some years I have been staring at – and lusting after the image quality of – Ming Thein’s work. Mostly done with Hasselblads. I’m not going to suggest that the reason why his photos are so great is because of the gear he uses – quite the contrary – I’ve seen plenty of his photos that have comparable IQ, taken with far more mundane and family oriented gear. He’s an arch-typical, modern day example for the maxim “it isn’t the camera that takes the photograph – it’s the person holding it”. And at the risk of giving offence when none is intended, few people these days could outgun & outshoot Marshall Ming, from the Wild East!

    But there is a special quality to his work from ‘blads. And it must, in part, be due to the size of the sensors. Not all that much larger, on the X1D – neither in physical dimensions nor in pixel count – but enough to make a difference. And with bigger pixels, which boosts its performance – which is where I suspect the EOS 5DS failed, in straight comparisons with the D850. (If I’m wrong, it must be the processor – either way, the one with the lower pixel count wins).

    Alongside all of that, I came across an interesting article about MF vs FF or HF. MF sensors give slightly better detail in the shadows, and they are the ONLY sensors that give any real detail in the highlights. No matter WHAT you do in post. Because post can only enhance what’s there – it can’t add detail that was never captured in cam, in the first place. And if you find any comparative shots illustrating just how different they are, in that respect, you will probably be completely blown away by the difference. In fact, you might even succumb, and buy a ‘blad, like Pascal. ๐Ÿ™‚

    PS – Pascal – i did this late addition to my earlier comments for a reason. I’ve just checked this post from DS on three web browsers, with NO firewall interfering with outcomes at this end, and there’s no sign of ANYONE’s comments on this post, on ANY of them – Google Chrome, Firefox OR Safari.

    • Avatar
      pascaljappy April 16, 2019

      Hi Pete, I’m sorry about this and don’t know what to say. There are 24 comments on this post, not counting this one. Why they should be hidden from you is beyond my comprehension.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.