What do the Sigma FP and the Hasselblad CFV-50CII/907X have in common? Hint: it’s in the title. Of course, modularity is nothing new, but for whatever reasons it’s been restricted to very niche applications in the past – medium or large format, cinema, or strange mutations like the Ricoh GXR. We’ve seen the CFV backs before, of course – but this is the first one with an integrated battery, electronic shutter and full controls, plus electronic system support. It’s only in recent years with the growth of mirrorless cameras that we’ve seen the first tentative steps towards true universality – in the form of adaptors. Any lens with a longer flange distance can be used on any body with a shorter one, so long as the lens has mechanical controls and the camera has its own shutter. There are some cross-platform fully electronic adaptations, but they obviously don’t work as well as something native thanks to the protocol reverse engineering required. Still, it’s impressive that they work at all – moreso when you consider the mount mechanisms and the electronics are crammed into something as thin as a couple of millimetres, in the case of the Sony E to Nikon Z adaptor. Adaptation is now commonplace on pretty much every format – from 1″ to medium format; but read on for the reasons I think these two specific “cameras”* might be the start of something greater.
*I use the term camera very loosely, because whilst both are capable of capturing images on their own or with minimal accessories from within their own system, you really need to start attaching things beyond for them to reach their full potential.
The FP will take an L mount lens and work much like a Nikon 1-type thing or a M4/3 compact with an overdeveloped sensor; of course assuming that the electronic part, firmware part and UI/UX is developed to expectations. The CFV is a bit different because it needs either a V series body or the 907X to be fully functional, but when attached to either it’s fully integrated. Sigma and Hasselblad have take very different approaches to design here: The FP and L mount are both relatively new; the L mount is enormous for the sensor size that has to be covered, and has a relatively short flange distance. This means a lot of flexibility in wide angle optical design, mounting to other accessories that might entail movements etc. (and thus not having a deep or narrow mount physically shading the sensor) – but most of all, the ability to adapt a wider range of lenses than something with a longer flange. The CFV has no mount at all – being a digital back primarily – and can thus capture an image through anything with the right adaptor. Both ‘cameras’ have their own shutters, albeit electronic only.
We have some limitations here, still: electronic shutters don’t play nice with large sensors as there is a line-by-line readout time involved, resulting in a rolling shutter effect. Newer sensor designs continue to reduce this readout time and thus the risk of motion-induced distortion if camera or subject aren’t perfectly still (and of course picking a sensible resolution tradeoff also means fewer lines to read and thus a higher frame rate). Flash sync remains a problem, as you will need either a continuous source (impossible for some setups) or a physical shutter that controls the actual exposure duration (while the electronic shutter is switched on for longer). But it will only be another generation or two before we have global shutter capability on even larger sensors and at higher bit rates – and at that point, sync speed becomes a non-issue.
I am of course heavily biased and positively disposed towards the CFV/907X as this was one of my pet projects whilst at Hasselblad – whilst the FP represents something much closer to the cutting edge (and a lot like the X1D, actually) – the CFV manages to straddle both legacy systems (V, and anything V mount including technical cameras etc) and provide full functionality with new ones (X, and anything that can be adapted to an X mount). Theoretically, there’s no reason why any electronic protocol can’t be adapted to work with the back and provide full functionality – you could theoretically mount a Nikon Z lens on the CFV (and use as much of its image circle as possible, in aspect ratios beyond FF35) which you can’t do with the Sigma since that has a fixed and longer flange than the Z system. It is a) format independent; b) system independent; c) even somewhat open source, as the V mount specifications and interface protocol have been public domain for quite some time. I believe it’s a very interesting option for photographers with heavy investment in legacy systems as all of your old gear is usable – with the advantages of the better sensor, and in some cases, extended capability that was not possible before. It is a current strategy: you can use your old hardware, the current offerings, and future gear that hasn’t yet been developed. For example: most of the Otuses will cover 44×33; but with the flange distance of Nikon F, it’s possible to build in an adaptor with tilt. There’s no way this could have been possible on its original mount, but all it requires now is some careful machining work.
Sigma’s approach is heavily future-forward – you can use the current L mount lenses, but those are still being fleshed out. You can use older lenses with an adaptor, but those are also still being fleshed out. But the camera has been constructed (and clearly positioned in the marketing spiel) as being the centre of a modular system – there are even rigging points built in, masquerading as strap lugs. It is clear it’s meant to be cinema-biased, given the ergonomics are once again not ideal for stills (square grip, optional handgrips, no eye level finder other than the LCD magnifier) – when fully rigged it’s no longer small, but it is about as minimal as you can get whilst still offering full customisation possibilities. Kazuto-san has said much of the open source spec and protocols – it’s only a matter of time before an accessory maker offers whatever specialised widget it is you require. One can only hope the long-standing history of the V mount means the same thing for the CFV.
I think for the last year or two myself and many others were worried that we were seeing the beginning of the end of the photography market – I think that’s still true, insofar as the companies that choose to repeat their history will land up dying simply because they have run out of reasons for people to buy their products. There is only so far you can go down a certain evolutionary path, and the prevalent closed system philosophy has pretty much achieved that end. We have put up with the foibles and limitations of various systems because their strengths outweigh their weaknesses, or they offer one critical feature we can’t get elsewhere. There has been slow leakage already thanks to short flanges and adaptors – and I bet unexpected sales of certain products because of this – but we still need that big change to motivate people to start buying and wanting to shoot again. If it’s not a fundamentally different shooting experience or form factor, then being able to cherry pick the best of all worlds and put that together into a customised, workable solution is perhaps even better still. MT
__________________
Visit the Teaching Store to up your photographic game – including workshop videos, and the individual Email School of Photography. You can also support the site by purchasing from B&H and Amazon – thanks!
We are also on Facebook and there is a curated reader Flickr pool.
Images and content copyright Ming Thein | mingthein.com 2012 onwards unless otherwise stated. All rights reserved
Ming Thein | Photographer 
Hi Ming, I really like the simplicity and design of the Hasselblad CFV-50CII / 907X combination. It reminds me of that great quote from Obi Wan Kenobi in Star Wars, A New Hope: “An elegant weapon for a more civilized age.” when describing lightsabers verses blasters. I am not sure I would spring the current asking price for it – but the MF lens adaption options it provides, combined with small size/weight, and backwards compatibility to V-series cameras is technically excellent, top quality design and engineering.
My photographic pursuits (enthusiast not pro) are more lightsaber than blaster. Like everything, you choose the tools you use based on what you want to achieve, and what suits your circumstances at the time.
I definitely think there is a place for this Hasselblad combination in modern photography.
Put it this way: it’s way cheaper than MF used to be, and I know the construction is a lot more expensive than the usual mirrorless-style cameras because of the level of fit and finish required. Interesting they still managed to keep the price in the tool category rather than the jewellery one, though!
I’ve devolved to a riverboat special derringer. 😛
Thanks Ming for the pixel size answer. (Re: July 23RD answer) It’s as I suspected in that the smaller pixel size is more sensity to blur, beside diffraction.
That said, is 24mp the overall optimum useablity size on a FF camera? (Or 37mp, 47mp, 50mp?) where would you draw the line?
Thanks again.
P.S. I thought the pixel sites on the Hasselblad X1D were quit a bit larger then the Sonys’ and would make a difference!
For a given sensor size, yes, the smaller pixel sizes are more susceptible to blur since the density is higher. But not if the overall sensor size is scaled accordingly.
24MP FF is a pretty good sweet spot – diffraction doesn’t kick in until f16 or so; it’s very tolerant of lens resolving power, and the pixels are sufficiently large that at 24MP sensor seems to have several stops of high ISO advantage (more than you’d expect). Low ISO resolving power aside, I’d say the Z6 for instance pulls away from the Z7 in overall image quality for a given output size (a balance of resolution, color accuracy, DR, noise, general look etc.) at ISO 3200 and above; by 25,600 it’s still usable, but I’d stop at 6400 on the Z7. That’s a two stop advantage, perhaps a little more. You could realistically have the 24-70/4 as your only lens on that camera and not feel too pushed in low light; impossible on the Z7. Absent of commercial client considerations, it would probably have been my pick. Cheaper, too. Somehow, I doubt the 61MP sensors are going to be better…
This is interesting to me. I’ve found that of the cameras I’ve owned, those with 35mm sensors somewhere around 24mp seem to produce images with the least amount of visible artifacts when viewed on my NEC display at 200%. (APSC seems best at around 18mp.)
So I did a little digging and find when using something called a pixel calculator (http://lcdtech.info/en/data/pixel.size.htm) a 35mm sensor with a 6000×4000 image size (24mp) yields a pixel size of .126mm. The new Sony A7R4 61mp sensor has an image size of 9504×6336 (according to a Sony web site (https://www.sony.com/electronics/interchangeable-lens-cameras/ilce-7rm4/specifications). This sensor then has a pixel size of .08mm. I’m terrible at math but it looks like about 55% smaller size pixel.
The computer processing engine in the new Sony camera (Bionz X; new generation) must be doing a significant amount of work to “tidy up” those very tiny pixels.
So while one is indeed getting a lot more pixels being recorded in image capture, these are very small pixels maybe needing a lot of computing power to sort these pixels out and assemble them into a useable image. I wonder what kinds of interpolation and compromise is going on to sort those pixels out. There’s got to be some compromise to image recording in this process.
Finally, please Sony supporters, realize I’m not knocking Sony, just thinking out loud about the behind the sensor activity needed to render a final digital file. I like Sony. It’s a nice short name; easy to remember and spell.
The smaller pixels definitely start looking a little gritty/grainy/crunchy – I see this manifested as the tonal smoothness or clean edges that larger sensor/larger photosite cameras can produce that’s increasingly absent at the pixel level, especially on smaller sensors. The samples from the A7R4 don’t seem to deviate from expectations here; the edge acuity doesn’t seem to be as consistently high as the lower density cameras.
I have completely mixed feelings about this camera. I like the idea of a modular camera, but I seriously wonder about the usability of it. The photos of the body/button layout out there don’t look like something I would want use, even with a huge amount of customizability. I would think the idea of modular camera would require a lot of thought and careful design to pull off gracefully, but Sigma seems more intent on giving us a box of the smallest size and hoping you attach whatever you think is missing to it. It also seems like the things that you would be most likely to want to swap out are fixed (sensor and E-shutter). The parts in which we expect the most advance in technology in the coming years are totally fixed.
I wonder if maintaining multiple purpose-built systems, as complicated it may be, would be a better solution. Somehow I just don’t see users building up a camera for, say, studio use, then video use, then stripping it down for travel and being happy with the package in all situations.
Maybe I just don’t trust Sigma to pull it off (their past cameras are ergonomically strange or just plain bad). The irony is it seems they are the only manufacturer that would even attempt such an unorthodox product. I remain pessimistic, but we shall see.
There are definitely ergonomic questions and a somewhat patchy history in their previous designs. My biggest question remains whether you can use something like this handheld with anything but wider lenses before hand shake becomes prohibitive.
You seem to be looking at it in an odd way. Why do you think the sensor and eshutter are fixed?
Maybe the addon flash, grip, microphones are fixed and ports will transfer to the next model. Just swip out the old sensor with new core box.
None of this is going to save camera companies. They have lost the war and it is the smart phone that has won. Pros and enthusiasts cannot sustain a market. It takes mass market success to provide the revenue and needed volume to keep factories busy. The mass market has gone completely over to smart phones. It is beyond rare to see a traditional camera these days. At my son’s high school, it’s nothing but smart phones at events where kids perform. No SLR’s, no ILC at all. Traditional cameras have been utterly obliterated. They are not coming back.
I guess we’ll never know since none of them will try it! But look at it this way: Fuji offers only a very slightly different shooting experience and they managed to carve out a significant chunk of market share and be profitable, so anything’s possible.
Fujifilm’s market share is barely in the upper single digits. The highest I have ever seen them reported as is 8%.
Fujifilm is only recently profitable with cameras. I have been reading their quarterly reports since the days of the S5 Pro and they have been losing money hand over first for many years. They crossed over into profitability about half way into the X camera line’s life. Regardless, they make so little money from cameras that it is hardly worth their effort. Only the Japanese sense of legacy keeps them at it. Fujifilm could easily shut down their entire camera division and overall profits and revenue would hardly drop.
“They crossed over into profitability about half way into the X camera line’s life. “
I should have been more specific – the X cameras are profitable. Not immediately, given R&D costs etc to recoup, but more profitable than the others…
Sigma is another example, though with their lenses rather than cameras. I suspect their foveon efforts are wildly unprofitable, if anything.
I like the Sigma FP. Great size and weight. Reminds me of the first Canon M. That was a terrific landscape camera. I had one on which I put the Canon 17-40L lens. It was perfect for mountains and forests, etc. Stuff that doesn’t typically move much. I like that the Sigma will take lenses from multiple sources without needing an adapter. Fun times.
The one thing that remains to be seen is usability handheld with no finder or stabiliser – is shake going to limit envelope, or is the sensor’s high ISO clean enough and the e-shutter smooth enough to negate it?
“Adaptation is now commonplace on pretty much every format – from 1″ to medium format… ”
Don’t forget the Pentax Q 1 2/3″ (or 1/1.7″ on the Q7) which can even adapt any lens down to a d-mount—and all with IBIS!
The depth of field when using one these little marvels is amazing, especially when using a long telephoto or, on the other end, a 2.7mm fish-eye. The biggest problem with using a Q is finding a lens good enough for its astrophotography-derived sensor.
Ah yes, good point – too bad about the sensor, though. And it seems the line itself is pretty much dead…?
My IQ is more important than camera/lens/sensor IQ.
Have you seen the distortion caused by slow focal-plane shutters in early twentieth century photos? Now you can get it SOOC.
Sharpness is a bourgeois concept. Now I can easily use pinhole lenses, DiY magnifying-glass lenses, what-ever.
As a Dada/Surrealist, the Sigma fp sounds like the purrfect camera.
If you want blurry-lo-fi images, you don’t even need a dedicated camera…think of the options! But not everybody wants that look, nor does it work for every image/idea.
Not lo-fi, dreamy/nightmarish . Distorted like this https://people.rit.edu/andpph/photofile-b/lartigue-1.jpg SOOC from a slow rolling shutter. I have Leica, Canon and Schneider lenses i can adapt, if I want sharp. I have access to a machine shop to build dreamy https://petapixel.com/2017/01/26/build-soft-focus-dslr-lens-15/ There are lots of inexpensive lenses, both glass and plastic available online. I spent over twenty years in Hollywood, maybe I’ll shot some distorted motion …
I’m sure there’s an art idea in there, actually…it’s very Munch.
The fp is the first new camera I’ve been seriously interested in purchasing, because of the extreme simplicity and the modularity.
I want the ability to add a nice grip, but I also want the ability to pare the camera down to the bare minimum film-camera-size if I want. That plus the promise of reliability from a complete lack of moving parts is appealing, as well as a modern full-frame sensor I can use all my lenses on.
Theoretically, the ability to rig up or strip down is useful – but the difference in handling is often such that you won’t use it without since the grip doesn’t make that much of a size difference anyway (I’m thinking every RF-style camera I’ve used that’s offered a grip has always lived with the grip on; the Ms, the Q, the Pen-F…)
You’re right about that, but when you’re counting the grams for a long hiking trip, you can at least downsize to the minimal accessory grip.
I would probably be looking hard at the D3500 again in that case 🙂
Ming, would you be so kind as to say something about the advantages/disadvantages of the physical size of the pixel, rather than the number (density) on the sensor. (That is if there is any!) Specifically is it easier to take sharp photos with the 50 mp Hasselblad then the Sony 61 mp? (Or am I confusing the issue?) I know, all things equal, bigger sites usually have better low light sensitivity but other than that, what else? (Forgive my stupidity on this issue but I have never seen anything written about it.)
Also I know you have mentioned before about the precise color palette that Hasselblad provides but there has been a back and forth discussions claiming one can actually customize any camera sensor the same way! Can one?
Thanks again for all the great insights.
Sharpness/camera shake – all things being equal (resolving power, AF accuracy, stabilisers, system weight etc) then higher resolution is more difficult as it permits less motion before being visible as ‘shake’ (i.e. one pixel’s information getting smeared into the next). In this case it’s probably easier with the Sony because the linear difference between 50 and 61MP is very little – barely 10% – and the Sony has both IBIS and faster native lenses.
Pixel level quality – newer sensors have higher quantum efficiency and light gathering ability, which is why we’ve been seeing more recent sensors improving on DR and noise even with smaller pixels and less collection area. However, there’s a point at which the technological improvements cannot cheat physics. I think we’re already seeing the ceiling at 15 stops or so. Going to smaller pixels isn’t going to help that, and it’s also going to bring in diffraction issues earlier on – requiring even better lenses.
“Also I know you have mentioned before about the precise color palette that Hasselblad provides but there has been a back and forth discussions claiming one can actually customize any camera sensor the same way! Can one?”
Yes, but not in-camera, as those calibration files are not user accessible. This is what I’ve been doing for years with the PS Workflow series and most recently with the ACR profiles provided in PS Workflow III.
Ming, I would love to see your objective take on the two new offerings from Hasselblad. I’ve yet to see anyone assess these tools and conclude their cost is justifiable to an advanced prosumer. Most people who want them are commenting on how beautiful they are. I agree, but I want to know why I, as a photographer, would want one or the other of these beauties over and above what I believe Sony and Nikon have and are planning for 2020.
“I am of course heavily biased and positively disposed towards the CFV/907X as this was one of my pet projects whilst at Hasselblad”
This should pretty much sum it up for you.
The CFV is modular and offers access to a wide variety of different shooting experiences and working methods – which should produce different results because the process is fundamentally different to shooting with a Nikon/Sony.
All of the Hasselblad cameras have individually calibrated sensors, which means much better color – and visibly so. More pixels aren’t better; but more robust pixels are (and what I’ve seen so far from the new cameras suggest that we have hit linearity in the underlying physics.
It’s a beautiful camera (and a bridge that connects V, H,and X series). Nice job, Ming!
Thanks!
I concur. It’s a beauty! Nice work! If I was to weigh up the X1Dii or this combo as a potential purchase, in your informed judgement, Ming, could the 907X/cfvii do pretty much everything the X1Dii does? Most of my shooting is tripod-based landscape work, especially long exposure photography, so I’m guessing that if there are any limitations compared to the X1Dii they might be more on the versatility/robustness side of things? I actually did have an X1D but sold it earlier in the year out of frustration with the buggy interface and dial freezes, which happened a bit too often for my liking while on a workshop in Venice, but I do kind of regret it-the output files from and aesthetics of the camera are just beautiful. Appreciate any thoughts you might have since I might take the plunge again with either the combo or the X1Dii. Thanks!
Short answer – other than have an EVF, which you probably won’t use for tripod use anyway, yes. The only thing you’re going to give up is weather sealing.
Thanks Ming. Much appreciated! Sorry, just one other thing: I’m assuming that were a cfviii back ever to be produced, let’s say with 100mpx, that there won’t be issues there?
The larger sensor is physically impossible because of the size of the board and certain connections required within the mount spec/ housing. 100MP 44×33 – sure; I don’t see why not…
I’m about as disinterested in buying new gear as I’ve ever been, and in the past I was quite keen on gear acquisition! I just don’t see any reason to move on from what I’ve got now (D810 and D850 with Sigma Art glass and an X-T3 for holidays). The new Sigma camera is interesting but I don’t want one; ditto the new version of the A7R. I like the look of the Panasonic S1R but don’t see myself buying one and changing all my lenses, and really, the only camera with MUCH better image quality than my Nikons is the new 100MP Fuji, which is expensive and looks vile ergonomically. Furthermore, I wonder who really needs all those pixels. It’s getting like the audio industry where the electronics record and reproduce sounds no one can hear and manufacturers improve on already inaudible distortion and nonlinearities; that 100MP Fuji is surely capturing detail no one would see with the naked eye!
I dunno. Certainly the industry cannot survive as it is.
I’m not 100% convinced on the Fuji: with that pixel density, we’re into the realm of diffraction limited performance even before your whole subject is in focus. No doubt the envelope is larger than normal MF thanks to IBIS, but I think the pixel density might be too high to make that much of a visible difference under most conditions…
You can easily find yourself in diffraction territory with the Nikons with a lot of subjects. Although the Sigma Art lenses seem to perform surpassingly well at the smaller end of the aperture range; I’m perfectly happy to shoot at f/11-f/14 when necessary, and they make gorgeous sunstars in that range! They aren’t quite as sharp as they are at f/5.6, but the difference isn’t really relevant in actual photography.
I won’t be splashing out on the Fuji but I’m keeping an eye on Panasonic. I like Sigma lenses and it’d be nice to have accurate AF without messing around calibrating them, which is a total pain!
I think it’s an overall balance thing – I find myself having to compose for effectively infinite DOF with say M4/3, regardless of aperture; or anticipating some separation so long as there’s no tilt with MF. With FF, I can still choose, and at least so far, for normal subjects/distances – still have mostly critically sharp pixels. But I agree, there’s definitely a drop-off as overall pixel density increases.
Hello again from Germany! Do you think, there could be small L-Mount Lenses in the future? The fp with high quality Pancake-Lenses (smile) could be nice to always have with you.
I think that’s a question better asked of Leica, Sigma etc…
The CFV/907X is all the camera I know I need and all the camera I won’t be able to afford 😂
It’s funny how it’s always the next thing that’s all you’ll ever need, right? 😛
That can certainly be true 🤣
However, this combo really ticks my boxes. The cost will prohibit me though, like the X1D.
And like every other camera, wait a couple of years – if you still think that’s ‘the one’, you’d be surprised how much the value falls…and with DJI’s history of handling Hasselblad pricing, I’m sure there’ll be a massive end of life price cut, too. Way to preserve value… 😦
I like it, however Iam not sure general users care about modularity. They care about end to end ease of use (Apple approach). IMHO.
Also I noticed first time my comment is first in line. 😊
Majority of general consumers, yes. But those who are still spending money – the hobbyists etc – want as many gadgets and bits as they can get their hands on…
modularity is great! i still have my GXR (albeit with just the M-module now) and the illusion of convenience a la Apple was what made me move to PC: was looking to buy a new iMac but then realized I have to decide on my computing power at the time of purchase whereas a PC canalways be upgraded down the road or self-serviced if anything goes wrong.
Some of the Macs can be upgraded, but sadly it seems they’re moving away from this unless you buy the horribly expensive Mac Pro…e.g. the 2012 iMacs were very upgradeable – you could add/swap in a second drive, RAM etc.