Fujifilm GFX 100 review

@Pixel8888

It is just that Fujis version of IBIS is bigger. They also created the X-H1 because they couldn't fit the IBIS in the X-T3 sized body. Something many other manufacturers have shown is possible.

I bet we will see a GFX50S sized version some day in the future

Probably a GFX100R without IBIS is already under development :)

"Probably because it easily surpassed film medium format when digital 35mm format was introduced."

Eh, actually that did not happen until pixel count for 135 digital sensors passed 16 MPix mark. That took almost a decade from the introduction of first serious professional digital cameras (which had less than 1 MPix sensors if my memory serves, and cost north of $100k)

I ran a bunch of tests around 5 years ago. My at the time Canon Eos 1DS Mark III (21mpx) vs. my Pentax 67II..

Conclusion:

Digital shot at ISO 100 vs using the ISO 100 film on 67 (both slide & negative) of the same subject in the same light, the 67 (drum-scanned) did not stand a chance for detail.

The term "medium format" vs "35mm" etc.. really only has to do with film vs film in my opinion.

@MyReality.. I think a couple factors could skew the results one way or another.
At the time I ran a bunch of tests before flying to Tokyo from Los Angeles for a planned shoot. For this project:

1. If one was printing direct from the negative/slide (not scanning) to the same sized print.. (but images were going to be scanned)
2. Choice of lens on the 67 vs the Eos 1ds Mark III. I did not want to bring both systems with me..and the resulting image was to be around 100cm x 100cm+. For the subject matter to be captured on the 67 a 165mm - 200mm was most likely ideal. On the Canon that would be around 80mm-135mm. Although the 67 165 is quite sharp, the 200mm not so much. For the Canon I ended up bringing my 85mm f1.2 & 135 f2 amongst others.

I think depending on lens, subject matter, focus speed, number of frames the Canon had the advantage in this case. If I was shooting portraits with a 67 105 & 165.. especially in B&W, the 67 would have been the choice.

@MyReality.. yes, if you had on each camera, glass with the same resolving power, I agree the 67 would have the advantage in image detail over the Eos DS Mark III. I currently shoot with 5dsr's..and it certainly has the 67 beat.

Edit: Again though.. Shooting negative vs slide. Scanning or direct print from neg/slide. Shutter speed/mirror vibration are all factors that could come into play.

I also do enjoy shooting portraits with the Pentax 67II more than with the Canon. The brightness of viewfinder, the medium format slowing you down, the light through an 81b or sepia filter etc...

@cbphoto123 - You still use a Pentax 67II?. Are you a masochist? I have not seen anyone use that camera since the 1980's when travel photographers used to use it overseas for print medium. The weight stops a lot of photographers.
I do agree with you last paragraph, since I sometimes use a Mamiya 7 with Velvia 50 for landscapes and architecture.

Ctein thought Micro 4/3rds had well passed film 6x7 in 2010.

https://theonlinephotographer.typepad.com/the_online_photographer/2016/03/when-will-micro-43-equal-medium-format-film-we-have-the-definitive-answer.html

Oh I kind of understand now, I just thought about math wise, 16bit was 4 times of 14bit, but back to brightness can be captured by each pixel, it does not make big differences in this camera. Thanks

Oups, I rounded .482 wrong! Sorry!

We both know that cameras are capable of more than 12bit now. But the fact that 14bit is a bit of an overkill didn't stop them from implementing it. Maybe they are simply afraid of the number 13 for superstitious reasons :). Or perhaps adding two more bits at a time is a more reasonable approach. Just like skipping the 15 and going for 16. Maybe we'll never know, but it is what it is. I'm just stating facts. It wasn't my choice to make. Anyways, like I said before, it matters for image editing more than anything else. And it makes sense in the end. So, it might be BS, but it kind of works. Have a nice life!

> “the fact that 14bit is a bit of an overkill didn't stop them from implementing it.”

Are you sure? On many previous generations (at least on Sony platforms) what was ACTUALLY implemented was 13bit! It was just (essentially) padded to 14bit in ARW.

> “it matters for image editing”

Absolutely not. “It does not matter” means exactly that!

No matter if your editing pipeline is 16bit, or 32bit, or 64bit: having input of 14bit vs. 16bit would not make any difference. (Of course, this assumes that the full well and read noise of the sensor are as they are now.)

"Are you sure? On many previous generations (at least on Sony platforms) what was ACTUALLY implemented was 13bit!"
- I'm talking about the 14-bit ADC. Not electrons.

“It does not matter” - Yes, it does. Every bit of extra data makes image editing more "flexible". 12-bit would not give you the same results.

And I’m talking about 13-bit data inside 14-bit RAW file. (I have no idea about where “electrons” came from.)

Moreover, as I said many times: NO, it does not make ANY sense to keep EXCESSIVE bits. Here “excessive” is, in fact, very flexible: even when read noise is 1/4 of a DN adding an extra bit is “almost excessive” when DN=3e.

To see, just model the whole “physics pipeline”, and calculate the resulting pass-through tonal curve. If you think that your blanket statements make sense — verify them!

"I have no idea about where “electrons” came from"
- Just like about many things you say.
It's never 12bit or 13bit exactly. You still need 14bit ADC for let's say 13.4bit or whatever overhead there can be on different cameras. And it makes no sense to develop an entirely separate color matrix for each of them. You are being unreasonable. Stop overreacting. There are no excessive bits unless you export them all into 16bit TIFF.

Sigh… I see that you have no clue what you talking about.

There are two ORTHOGONAL issue in this discussion.

1) Just dump out the content of a RAW file (you can use dcraw to do this). Then Inspect how many last reported bit are ALWAYS zeros. This is the number of bits REALLY produced by ADC. (And often, this is way less than what MAY HAVE BEEN stored in the RAW file. This is what leads to nominally-14bit-but-really-13bit RAW files.)

2) As a thought experiment, consider rounding the data BETWEEN the ADC and the RAW file to a smaller amount N of bits. Inspect how the result of processing such RAW file depends on the number N.

This would produce a certain “tonal curve”: the graph of how the brightness of the input (optical!) image is converted to the (AVERAGE!) brightness of the output image. (We need to average because of the noise in the pipeline.)

Up to some moment, increasing N decreases the “staircase” in this curve (this staircase is visible as “posterization” in the actual images). However, at some value of N the staircase disappears. Any increase of N after this value does not matter FOR ANYTHING: for these N⸣s, quantization is subsumed by the noise dithering (due to photon and read noise).

(Continued…)

(Unfortunately, the WikiPedia article on this topic is long-winged and mostly focused on not-that-important-in-practice subtopics — so I cannot recommend it. Myself, I needed to read a lot of original papers [many years ago…] and do a lot of modelling to understand the fine details of what happens…)

You are making no sense at all, sorry. It doesn't have to utilize the full tonal range on every single image that you shoot. If you expose "correctly", you use the middle part of the color matrix. If you over- or under-expose, you shift it towards either one of the extremes. Which is why image editing potential still benefits from a higher bit depth. I mean, regardless of the exposure value that you choose, it can record more data if/when it can.

Sorry, ecka84, I would love to write something making more sense to you, but it looks I’m running out of my allotment of time. (And: basically, what you say above is completely orthogonal to the issues of “how many bits are needed, and how many are excessive”.)

Maybe you just don't get it. Raw files do not contain "empty" bits.
Look, if you want no excessive bits, then shoot JPEG, it's only 8bit. Problem solved. At least then you will see why more bits is better for editing. Bye.

Yes without the IBIS.

Though not before a 60-70MP will be incorporated in a GFX 60S.

Only in terms of resolution

Price, speed, size, autofocus, video, lenses....not yet I think

More like the opposite, even 9 out of 10 pixel peepers who want 50+ MP will look at the price tag and pick FF and the much more affordable and varied E-mount lenses instead. I mean it's amazing detail but the biggest posters I can easily order is 60x40" and the a7r4 is already at 160 DPI, if you can use the half pixel shift it's in theory 320 DPI even though that mode seems incredibly fickle and best used for product photography and such. But hey, it's not MF...

Yeah, 60x40 or whatever by 40 at 150+DPI is good enough for most photography nuts like me :). But there are "by 80" options too, so we have room to grow.

Maybe it only shoots children without airbrushed makeup.

@thegrammarfairy
While I imagine someone will buy this and use it for family pics, I doubt that the majority will use it this way. This is a professional tool for professional applications so I don't think your problem is actually a real problem.

I have a hunch that @TheGrammarFairy also thinks this is not a real problem.

PLShutterbug, TheGrammarFairy is on the fence. Seems like a situation where you won’t be able to show clients unretouched pictures on a big screen, the positive side is that they’ll be easy to retouch.
The negative is that you won’t be able to have even a speck of dust on any product shot, you will definitely need to raise your housekeeping game and carry a jumbo size can of DustOff with you wherever you go.

That is the way it was “back in the day” with 2-1/4. 4x5, and 8x10 were the same. Nothing really has changed. You need spotless models for commercial work.

Unfortunately it’s out of my reach, so all I can do is mess up the model with drool.

" TheGrammarFairy is on the fence"

There is something about the use of illeism on social media that I find to be so contrived. It's like trolling but done by old people.

Fairies are a very old race.

Not smaller price gfx 50r is $5k

@John Clinch... I think the question that should be asked is whether or not a larger sensor makes a difference in any visually detectable performance and whether or not the larger sensor can make a practical difference in aesthetics, e.g. could a larger sensor affect the wide open performance of a f/4 wide angle lens? *What about practical differences such as does a larger sensor allow a photographer more latitude when shooting in small confines when compared to the same focal length mated to a 33x44 system? So much more to think about than just sheer image quality :)

john Clinch:

The 100MP raws I've seen from the Phase (with Schneider lenses) are staggering. (My sample attempt didn't work.)

Even the best 16 bit raw images from this Fuji aren't.

I know that Fuji's best G lenses are very few, so that's not a surprise. What I can't tell you is if should you try Fuji's best (the 63mm or 250mm) against the Phase with a 100MP back (and remember Phase has the even better 100MP achromatic back) with a Schneider lens that the Fuji GFX 100 samples would be as good as the results from the Phase-Schneider combination.

DPR and Imaging Resource both have Phase 100MP raw samples you can download. I believe IR has achromatic back samples too. The only raws you'll find that compete with these (and such raws are very hard to find) will be from the various Leica S2/S bodies. Right "only" 37MP. No, Hasselblad's 100MP back ain't in this category--the lenses are Fuji.

MirrorLessHater:

I'm not clear why you mention the GFX-50r.

Same dimensions for the sensor, but very differed AF, and very different body.

Now could Fuji put this 100MP sensor in the GFX-50r with a contrast detect AF, and a slower buffer, probably. But in 2019 that wouldn't be cheap.

Our Aibotix X6 can lift 2kg so not much for the lens it seems.

I always wondered about the insurance issue when a drone of this weight falls down.

Thank you Richard.

Well yeah, it's dying to low end phones, there has always been the low end market but photographers have always wanted more. This camera would have come out anyway even without the loss of the low end market. The market will revert to where it was pre-digital. Some low end sales and high end sales but not the crazy digital bubble of the last 18 years.

This camera will sell to people who will see it as a cheap way of achieving this resolution

Elaborate please.

When a wall is limiting your placement, fov/format matters. A GF23 can see as wide as 18 without the 18mm distortion.

What is "glass?"

It doesn’t matter what you call it. The creative effect is much more important then the nomenclature. Longer lenses show less background around your subject—they help isolate. Shorter lenses show more of the environment, they help tell a story about the subject.

@panther fan

no its not. Im standing in a spot and I want to take a picture. Put a lens on that medium format camera what will give me exactly the normal perspective. Take the picture. Look at the edges of the scene and how far left and right you captured it with your normal lens. Now stand in that spot and put a lens on your micro 4/3 that gives you a normal perspective...and see how much of the scene you caputred. It will be a lot less. A small square in the middle compared to the medium format.

With medium format, you can shoot a wider scene with a normal perspective lens, than you can from that same spot with a normal perspective lens with a smaller sensor

Medium format is amazing for this reason. Its the only reason to go medium format. Not MP. Not anything else. To take images that are without skew, in the normal perspective...but get more of what you actually see with your eyeballs in your final image.

@panther fan we are talking about perspective distortion....

no matter how great your 18mm lens is at making things not look distorted in the image....it will still have perspective distortion which will affect the distance between objects in your image in relation to reality.

@otto k

You are right. They don't understand distortion or perspective. The problem is they make a claim, which they could easily prove with some sample images. But sadly they didn't put in the effort. Otherwise they would have probably noticed they disproved themself

@otto k @panther fan -which competing camera company do you two work for exactly?

"The key point as a result of that difference is that even though the field of view is wider, the geometry, or "look," of the 50mm focal length remains. You don't get the exaggerated perspective that wide angle lenses usually produce on 35mm cameras. In that regard, medium format mimics how your eyes actually see the world, at least more so than the smaller 35mm size. The effect is even more apparent with large format 4x5, 5x8, or 8x10 film. You can shoot an extremely wide scene but it will have the "real-world" look of a lens with a longer focal length. The subject remains flat and not 'stretched' out."

https://gizmodo.com/why-medium-format-is-so-gorgeous-its-about-more-than-r-1601938278

at the same distance and 50mm lens, you absolutely do not get the same scene. Absolutely not. This is a fundamental misunderstanding of optics. You use a 50mm lens on a medium format camera you ABSOLUTELY WILL NOT get the same scene as a 50mm lens on a micro 4/3 camera. This is just basic photography knowledge. Why are you trying to spread misinformation and throw shade on this camera for no reason with false statements.. I suspect its because you are paid to do so

the same focal lenght of 50mm means that you are at a normal perspective. The perspective of things look normal. That 50mm lens on a medium format will give you a wider FOV without needing to distort perspective to get there

https://www.dpreview.com/forums/thread/4141091

...."the larger the "sensor" area, the easier it is to get a wider view (angle) image without a wide angle (distorted, extreme depth) look. this is most important with portraits IMO.

the fuji gx 680 and 6x9 used to be my favorite cameras for the simple reason that you could stand a couple of feet from a person, get almost their whole body in the shot but never have their features distorted the way the same shot (of angle of view) would have looked with a 24 or 28mm on a FF 35mm camera...."

http://www.bokehblog.com/photography/post/Medium-Format-vs-Full-Frame-vs-Crop-Sensor.html

@panther fan lol, read the link. its right there. literally RIGHT THERE. To anyone that is reading this, it is right there in the link. There will never be a better test. This has nothing to do with marketing, and no I will never buy this camera. In fact, I dont even have a digital. My film camera is fine, and if I had to buy a digital I would buy a used k10d. But that does not mean I think its the same as this camera, its not...but dont care. But there are differences and its just a different tool. And its a different tool that cant be replicated with a different lens or a different distance.

@panther fan. Literally nothing you have said makes sense to me.

Which Holy Grail obviously is the D850, but.... oh well...

@panter fan its RIGHT IN THE ARTICLE lol

@gravis

1. Watch the video
2. As I explained there aren't actually images in the article where the tripod hasn't moved

No one moves...

We can only know once the images can be compared.

To an extent, yes the print is going away, but not completely. The thing a print gives you (regardless of it's size) versus something on the screen is something you can hold. A tangible thing. A file on a computer doesn't really do that. Yes the days of people printing everything and putting it on the wall has severely decreased, but there will always be value in the print, and people will continue to print. Why do you think the instant cameras are popular? Maybe that's just a fad that will go way too, but it's the idea of it produces a physical print and that is something that can be shared and given to people (without or without a computer).

Now, that being said, majority of my work is digital, but I"d say maybe 1/5 of it gets printed and either given to people or I hang it on my wall.

Printing at home, yes is probably not much of a thing, except for either the old-school or serious photographer. But people are still printing (some are using labs obviously).