Fujifilm goes medium format: What you need to know about the GFX 50S

Lol... I imagine it would... given that the phase one 100mp 645 format camera is just shy of 50 grand with one lens... I think $15000 is a pipe dream for the foreseeable future. I have read estimates for the cost of the 100mp sensor alone that approach the $15000 mark... and an even larger sensor would cost exponentially more.

But what would it cost to make a 25MP sensor in a 645 size? Spread those pixels! Bigger photosites. What would we get? Surely lower cost, but what else?

The cost of sensors as the cost of LCDs is a comparable process. You just make & cut the silicone larger, that's all it is. I remember when all LCD monitors for computers were all like 17" & 19". If you wanted a bigger one, needed be CRT. That's because no one was making them larger. Simple. The first larger ones were ridiculously expensive. Eiki & Fuji among the first. Then suddenly Sharp started making cheap large LCD panels. In a couple years we went from 19" to 30" & larger at low cost. Like that. Only a handful companies make imaging sensors. They have the monopoly of the industry & supply them to everyone. So they dictate the output, specs & price. One day, there will a newcomer or established company that decides to jump on it & the next day you have 5x9 & larger monstrous sensors for cheap. It's a matter of someone wanting to do it & getting it done. Most the cost is the design & research. Making tech stuff costs nearly nothing. It will only take someone to do it first..

the larger the sensor, the fewer sensors you get off of a single peice of silicon wafer... and the more likely there is to be a defect in an individual sensor. I think that the average sheet of silicon has something like 9 defects in it. When you are getting 20 chips from that wafer and having to throw away 6 or 7 of them you still have a lot of usable chips... when you are getting 4 chips off of a wafer you could have many instances where none of those chips pass inspection. If you go really big and have 1 or 2 chips on the wafer you would have a hell of a time getting a good one... one of the biggest limitations at the moment is in the manufacture of the "blanks" that they make the chips out of... continued below...

continued from above... this is why really large sensors are more than simply a decision away from reality. They need to invent a new way to make the silicon sheet with less defects before they can even start to look at something like that... I am not saying that it will never happen... I am saying that at this point it is a ways away, barring a game changing stroke of genius somewhere along the line... and that isn't something you can count on.

I also remember the days when it was almost IMPOSSIBLE to find an 13" laptop LCD panel without dead pixels somewhere & manufacturers like Fujitsu making big campaigns guaranteeing 99.9% defect free panels. Fast forward a few years & dead pixels on GIGANTIC 150" plus LCD TVs & displays are almost nowhere to be found anywhere. It only takes will to make more efficient wafers, better silicone & improved production lines to make 1 square foot large imaging sensors with as many gigapixels as you want in them. We put a man on the moon more than 40 years ago & we make mass produced 4K displays for phones today! If just a couple years ago someone dreamed of a 4K TV shrunk into a 5.5" screen area, they would have said would cost a million dollars to do! Now, costs around 100 bucks! Imagine if they made a 30" monitor using larger wafers of the same 5.5" 4K phone display.. a multi gigapixel resolution monitor? It's all here, now, today! All of it! It's a matter of will & want. Think about it..

Highly thought provoking guys!
I understand you can't use displays with dead pixels. However, for a sensor, nobody is looking at directly, couldn't software in the camera interpolate a fix for the very few isolated pixels that would be virtually undetectable?

NoMirror99, all sensors have dead pixels to some degree. They get zeroed out with pixel mapping and other tricks. Is very simple. Before this was done by sending the camera for service,after some cameras had it in their menu. Now most of them just do it automatically randomly every x cycles without any user interaction. Unless you have a larger area of several adjacent pixels all bad in a large group, this is a problem affecting less an less cameras. Also, dead pixels are easier to get seen under very high ISOs and because high ISO performance has radically improved in the last few years (so have high ISO Noise Reduction algorithms) the ratio at with faulty pixels are detectable is much lower than before.

It's not only megapixels you know? Sure my iPhone has 12 megapixels but it's no match to the 12 megapixels of a Nikon D700 with a 28mm prime lens despite being "old technology" Size does matters ...

notpc Would you ask me to explain the totality of cosmos too?

Thats logic in its strongest term. Bravo argument!

The only sourness apparent here are the people who don't seem to have adequate responses to the challenges people have made. So they take swipes at people.

Challenges? Fuji introduces a medium format (or FF+) camera and the forum is full of complaints: too large, too heavy, poor AF, not really MF, not Sony, etc. In many cases followed by "my -insert camera-" is fine for me. OK. That's OK. Fuji did not make this camera for users of other cameras to suffer. Or feel inadequate. Or to kill Canon/Nikon/Sony/... DSLR or mirrorless line. Why so many people need to take swipes at this camera, while 98% are unlikely to need it, use it or own it?

Maybe, just maybe (among other things) because at the expected sub 10k USD price it is a MF camera just within (and just out) of many people's price range? Only 24 replies under the Phase One 30k+ USD MF camera announcement.

It is a tool for specific jobs. It is not a swiss army knife (although if it was, there would probably be another set of people complaining about that).

What's wrong with pointing out deficiencies in a camera? Or pointing out that its appeal is hard to pin down? With your admission that 98% won't want it, what's wrong with those people explaining why? This is, you know, a digital camera site forum where people give their opinions. I think some of them are just trying to elicit a real answer to what about the performance/output justifies the price of this one (and yes, you could apply that to other MF options).

I thought DOF and quantity of light depends also by the lens, not only by the sensor size.
I am wondering why DOF & light calculations is used only for small sensors, but not used for the big sensors. :)

Ok, nevermind, I guess my previous question was retoric.

A larger sensor gaters more light ONLY at the same aperture compared with smaller sensors.
An MF + F 4.0 lens gather less light then an FF + F/2.0 lens and has a larger DOF.
This is the logic which is polutting the small sensors areas.
When it is about an MF camera, this logic is switched into a another direction.
We all know, that they will not build brighter lenses for MF cameras because it is dificult and because it would be extremely expensive.

When the discussion is about an crop camera, is extremely important to have the tinyiest DOF. When it is an MF camera, the DOF is not that important anymore. Can I know why ?
An MF + F/4.0 lens has larger DOF than an APS-C + F/1.4 lens.
So if you want tiny DOF, an APS-C could serve better because you can find even F/1.2 lenses.

A couple years back I ran a similar test whilst trying to figure which system (in my kit at the time) would give me the best results possible for making larger prints. I shot with a Pentax 67II using Kodak Ektar 100 vs Canon Eos 1DS Mark III at Iso 100. The digital 35mm had more detail, better IQ.

@cbphoto123 My 5mp Olympus E-1 out resolves my Leica even when using tabular grain films and high acutance developers such as Tanol, Rodinoland Perceptol. But I do have a preference to the "feel" of film.

@fotopizza Why not connect the term "medium format" to specific hardware? The very format is irrevicably linked to specific hardware. Technically you are right, anything larger than full frame 35mm is medium format. I wasn't talking "technically" as in the minds of most MF photographers (that I have met) 645 is considered the smallest MF format.

Even my venerable Olympus E-1 produces better sharpness and resolution than BW 35mm film (see above). I really love the colour I get from Kodak sensors (E-1, E-300 and Leica M8) and my fuji cameras (X-)ro 1 and X100T) but prefer the "fell" of film either BW or colour.

Ultra bright apertures are always a compromise, having lens with a smaller aperture but compensated by a larger format leads to better sharpness (and uniformity of sharpness on the whole frame), contrast and bokek, and makes easyer to correct aberrations, coma, astigmatism and so on. All this without losing control on the depth of field (due to bigger format). Anyway also on 35mm it's really hard to say that f/1.4 aperture is really needed!

Dudes. Please. F4 is f4, if you use a hand light meter you don't need to set the film/sensor size. It will expose like f4.
Only equivalence of note is for depth of field.

Picture quality.

@Tom - it is not about metering light, but about DOF & noise. They are the same on a FF f/2.8 or fuji's f/4.

@padang you really don't need a MF camera, right?

Padang is geometrically right, as long as you have the same pupil diameter and the same angle of view you have the same performances, ON PAPER. The real world difference is that narrower max aperture lenses have less design compromises, leading simply to a better image quality. A bigger sensor means also more glass used, making production tolerances (which are the same regardless the size) smaller in proportion, allowing also more resolution.

FF killer for under 10000usd?

APS-C and FF are too similar? Oh really? In that way this MF and 135 FF are too similar as well. Please don't write silly things.

@Azimuth46 The word "similar" is found in your last sentence.

If Pentax go mirrorless than the legacy lenses will have to be used with an adapter... But the next Pentax medium format camera very likely will have a larger 6x4.5cm sensor, with 100MPix or more... Fujifilm and the Hasselblad mirrorless systems won't be able to use larger sensors.

That would be my preferred option as well, since I don't mind adapters at all. But many people seem to be averse to that solution for some reason.

It would certainly allow them to save a lot of size and weight and cost on the wide-angle lens options, as the Pentax 645 has a monster 70.9mm flange distance, which is just about the length of a normal lens on that system...

Correct me if I am wrong but the D750 weights 755g with battery. How is that a fraction of a 800g Fuji body ? :P

Though considering the Hasselblad X1D-50c is 725g with battery 800g for the Fuji isn't far fetched.

I'll be not surprised. It is a mirrorless body anyway. Well done fuji.

"What will this thing do that the Canon 5Ds can't? More dynamic range? Better high ISO? Something else?"

If you can't see the difference between FF and MF sensor sizes you must have sight problems, bigger is better, look at the richness and range of colours and the better contrast. Fuji lenses are exceptional at that anyway, this could be a really special camera at a cheaper price than the Leica equivalent.

Really? Better "richness and range of colors"? So you're saying that FF has better range of colors than let say APS-C?

kbryd - Yes of course it does. Its a minimal difference and one that most probably cant ever see, its its there. It has bigger pixels, meaning more light, more accurately, can hit the sensor. in great conditions you would be extremely hard pushed to tell the difference between this and the Sony FF sensor, but as conditions get more testing the difference will become more visible. You must have seen the horrible colour casts in dim light from small sensored cameras?