Focus issues of Fuji GFX50

[Jim Kasson]

Great stuff, Jim.  Particularly happy to learn about the April 1 flooby dust issues    Also glad to read someone post about a reasonably good copy of the Fuji 63.

A long shot question -- are you contemplating comparing the X1D lenses anytime soon?   This kind of objective testing between the Fuji and X1D lenses I haven't seen yet and would be at least selfishly speaking unbelievably helpful.

I generally test only equipment that I intend to use for my own photography. I have no plans to use the X1D. The reasons are the lack of FP shutter and the fair to only-good performance of my HC lenses on the GFX, and my assumption that they would perform about as well on the X1D, leaving the X1D with a limited lens selection. I am happy with the performance of the Otus 55 and 85, plus the Zeiss 135 Apo, on the GFX for square images, and mostly happy for 4:5 images. I suppose that when the X1D become easy to rent I could do something.

Thanks for the kind words. I've got a lot more GFX testing ahead of me, and I'm still taking requests.

I think that this afternoon I'm going to do a head to head quantitative on-axis test of the a7RII/Otus 55 and the GFX/63.

I already did a qualitative test of the a7RII/Otus 55 and the GFX/Fuji 120 that was interesting, st least to me:

http://blog.kasson.com/the-last-word/otus-85-on-the-a7rii-fuji-120mm-macro-on-the-gfx-50s/

Jim

[Brad P]

I'll at least follow your A7RII/Otus 55 GFX 63 results with great interest.  I'm contemplating leaving the Sony platform for either the X1D or GFX as my main shooting platform in the next week or two, am leaning toward the X1D but totally despise spending money unnecessarily.  (I actually read about flooby dust a few days after you broke the story). 

[Michael Erlewine]

In today's blog, Lloyd Chambers compares the Fujifilm GFX vs Hasselblad X1D, Canon 5DS R, and Nikon D810. In detail.

https://diglloyd.com/

[Christopher]

No need to read his stuff, sure you can, but I would suggest GETTING the stuff and actually comparing it side by side.


Gesendet von iPhone mit Tapatalk

[Jim Kasson]

In today's blog, Lloyd Chambers compares the Fujifilm GFX vs Hasselblad X1D, Canon 5DS R, and Nikon D810. In detail.

https://diglloyd.com/

WRT imprecise AF, I have a test protocol, which you may or may not think is adequate:

http://blog.kasson.com/the-last-word/another-medium-tele-test-batis-af-vs-mf/

If a few of you like the protocol, and if there is sufficient interest, I'd be happy to run that test with the GFX and the 120 or 63.

Jim

[Brad P]

WRT imprecise AF, I have a test protocol, which you may or may not think is adequate:

http://blog.kasson.com/the-last-word/another-medium-tele-test-batis-af-vs-mf/

If a few of you like the protocol, and if there is sufficient interest, I'd be happy to run that test with the GFX and the 120 or 63.

Thanks Michael - I would likely have missed that page as I've already read way past it.  That's exactly why we subscribe!

Jim, you have one vote here.   I can see the misfocus in Lloyd's pics and words, so it could be very helpful to see something more measurable.

[Jim Kasson]

Here's the on-axis a7RII/Otus 55 vs GFX/Fuji 63 test:

http://blog.kasson.com/the-last-word/a7riiotus-55-vs-gfxfuji-63/

WRT "focus instability", there is no indication that the Fuji manual focus moved at all during the considerable time it took to collect the data. Of course, on the focus-by-wire lens, you can't expect focus position to persist through power cycling.

Jim

[Brad P]

Here's the on-axis a7RII/Otus 55 vs GFX/Fuji 63 test:

http://blog.kasson.com/the-last-word/a7riiotus-55-vs-gfxfuji-63/

Your whole GFX series, including this one, is fascinating and extremely helpful to me Jim.  It's scientifically allowing at least me to reach the conclusion (supplemented with a week's worth of other mostly less helpful reviews) that at least for non-sports, non super telephoto shooters who manually focus at each aperture (as I do), the GFX platform is a likely a step up from a more expensive Otus + A7RII platform.  Never mind the improvements when the GFX gets a 100MP back.  Of course there's a lot of qualifications in that statement that readers are advised to understand, but it's right up my personal alley anyway.  Bravo!   You could charge for this stuff!

As an aside, now my only remaining task is to figure out whether the X1D is worth its premium over the GFX.   That has me stumped for now.  The driver in that last choice may be more my wife than the specs!

[Christopher]

It ones again shows how baaaaaaad the Fuji lenses are.


Gesendet von iPhone mit Tapatalk

[ErikKaffehr]

Hi,

Jim Kasson is quite positive about that lens. He also has the Otus 85 and the Otus 55.

Yesterday he posted a limited comparison of the GFX 63 with Otus 55 on the Sony A7rII:

http://blog.kasson.com/the-last-word/a7riiotus-55-vs-gfxfuji-63/

Best regards
Erik

I am currently also a subscriber of Lloyd chambers blog and generally I appreciate his very critical view on cameras and lenses.

But - as a user of the GFX w/ 63 - I have to disagree with many of his critics because I just don't see them. Of course I only can speak for the GF 63f2.8 lens, because that's the only one I got (yet).

AF focus accuracy
If I select a small focus point and twenty times focus at open aperture to objects (w/ sufficient contrast) in various distances I 19 times perfectly catch sharpness at the focus area and 1 time I'm slightly off.
Of course focusing is not super quick because of contract detection but everyone knows that phase detection is technically impossible with this generation of Sony sensor, for that feature one has to wait for a camera update in 2018/2019 using the upcoming 100 MP BSI sensor.

Manual focusing
Of course it's focus-by-wire but IMO well implemented with very little clearance when focusing back and forth close to the perfect focus point.
It's true that somehow the image in the EVF or back screen does not 'pop' so much into the focus than other cameras (I will do a comparison w/ Sony A7RII) but I see no problem to focus correct manually, as I have to do using the Zeiss ZE 135f2, Leica R 180f3.4, Canon TSE 17, 24 a.s.o..

GF 63f2.8 performance
I would rate my Fuji 63 sample as one of the best lenses that I know and have, summing up all characteristics:

Sharpness
Center sharpness is perfect at any aperture (apart from diffraction closing more than f8 of course), corner sharpness is already impressively good at open aperture, very good at f4 and excellent at f5.6. My sample is also very evenly sharp at all four corners.

Focus shift
I did tests using a tripod, manually focused at open aperture at f2.8 to f11 at about 2 m and 3 m. Can't find any focus shift that would diminish image sharpness.
And: In autofocus, the lens focuses at the chosen working aperture (like Sony A7), so even if a certain amount of focus shift would be apparent, it would not cause any trouble.

Distortion
Fuji incorporated automatic correction for distortion, light fall off and chromatic aberration into the raw file (and the jpg, but I don't use). ACR carries out the corrections accordingly so one can't see them - normally .
Only if opened in Iridient X Transformer, switched off the auto-correction, exported as DNG and opened in ACR one can see the effects. Distortion is very little, maybe 1%, light fall-off at open aperture also, maybe half a f-stop, CA too. I see no problem in the auto-correction because I anyhow would apply it in ACR (apart from the fact that ACR adds a slight concentric wave pattern interpolation artifacts towards the image corners correcting the distortion, visible especially in darker even areas, but that's an Adobe fault, not Fuji's).

LCA
Very little greenish/ reddish halos back/ in front of focus point.

Bokeh
IMO absolutely great, one of the best I am aware of: soft transitions in the out-of-focus areas and absolutely round out-of-focus lights (at open aperture).

Otus comparison
Chambers every 20 lines of his blog writing repeats that the Otus (55) is better than the GF 63. The Zeiss Otus lens line is close to perfection, no question, but I don't see the point to compare again and again a manual focus 135 3K EUR lens with an auto focus medium format 1.5K EUR lens.
I anyhow would rate the resulting image quality of the GF 63 to be very close to the Otus (at comparable apertures).
I don't have an Otus but maybe I do a comparison to the (IMO also excellent) Sony/Zeiss FE 55f1.8 next weekend.

So far I'm perfectly happy with the camera and lens (just waiting for an one f-stop faster version ;-).

[Juanito]

Anyone care to summarize Lloyd Chamber's opinions on the comparison between the GFX, X1D and D810?

John

[hcubell]

Jim, to be completely candid,  I have no idea at all what all of those charts and graphs you are producing demonstrate in terms of real world results. The analysis that Lloyd Chambers provides is derived from actual PHOTOGRAPHS that we can look at with our eyes and judge for ourselves. That's the way I like to evaluate things. When Lloyd raised questions about the accuracy of the AF performance of the X1D and the 90mm lens, I looked at the photographs he produced as evidence. Then, I went out to see if I could see the same things he was seeing by shooting photographs myself. And, guess what, I saw the same things Lloyd saw. Same thing with the sloppiness of the MF focusing in magnified live view that Lloyd is complaining about with the GFX. I saw the same thing with my eyes.

[SrMi]

.....
Focus shift
I did tests using a tripod, manually focused at open aperture at f2.8 to f11 at about 2 m and 3 m. Can't find any focus shift that would diminish image sharpness.
And: In autofocus, the lens focuses at the chosen working aperture (like Sony A7), so even if a certain amount of focus shift would be apparent, it would not cause any trouble.
...

Is it really true that GFX auto-focuses at working aperture? I know that my A7rII does and it causes me headaches as even the GM lens occasionally does focus-hunting if the aperture is f/8 or smaller.

I read that you can focus manually at working aperture by using M mode and pressing DOF button.

[Jim Kasson]

Jim, to be completely candid,  I have no idea at all what all of those charts and graphs you are producing demonstrate in terms of real world results. The analysis that Lloyd Chambers provides is derived from actual PHOTOGRAPHS that we can look at with our eyes and judge for ourselves. That's the way I like to evaluate things. When Lloyd raised questions about the accuracy of the AF performance of the X1D and the 90mm lens, I looked at the photographs he produced as evidence. Then, I went out to see if I could see the same things he was seeing by shooting photographs myself. And, guess what, I saw the same things Lloyd saw. Same thing with the sloppiness of the MF focusing in magnified live view that Lloyd is complaining about with the GFX. I saw the same thing with my eyes.

I appreciate that there are people who understand the world in different ways, and respect your point of view. I like quantitative results where possible, so our perspectives are different. I'm sorry that you don't see the relevance of what I have posted to your photography. I suggest you not bother to click on the links that I will continue to provide if you find that my approach is not helpful to you.

I have so far posted no measurements of AF or MF accuracy with the GFX. I have made a set of test exposures for AF performance this afternoon, and I will be working on analysing them. However, I don't think that you'll find the results satisfying, as they will be quantitative.

Jim

[ErikKaffehr]

Hi,

At this time I sort of feel that Diglloyd's reporting is a bit unprecise. I don't think he mentions focusing method or focusing point selection.

If you don't know the circumstances it would be difficult to draw conclusions.

Let's take an example. He has an example with a mountain landscape being in perfect focus with some boulders in front out of focus. If he put the AF-point/points on the boulder the image is clearly mis-focused. If the camera was allowed to choose AF point, it would be a different question. Many cameras would focus on the object closest to the image, but it could also be a good focusing strategy to focus on the mountains. That is what I would do.

Shooting medium aperture covers up the focusing difference pretty well.

It seems that corners/borders are not so great on that 63 mm. It is quiet possible that the 63 mm lens on the X1D would deliver better corners would it exists. But fact is that Hasselblad doesn't have a lens round 63 mm. Standard lenses are often a compromise.  The Fuji 120 macro seems to be a truly great lens with almost no focus shift at close distances and very little axial chroma as measured by Jim Kasson. Jim mostly measures near axis data, so his findings don't tell about borders or corners, but that lens probably has little focus shift.

It is quite possible that Hasselblad achieves better AF, although that would come as a surprise in a sense. Fuji has doing CDAF for a long time, the X1D is Hasselblad's first try.

Twenty times view finder magnification may be a bit too much. You want to focus at axtual pixels and 20X is probably beyond that.

Gearing of focusing is extremely important with magnified live view.

Best regards
Erik

Lloyd Chambers has criticized a lot on this model's focus issues and quite frankly by looking at the images he put on I do not see what the problems are. I may not have trained eyes and I would appreciate it if anybody would help me to understand the points. Thank you in advance.

[Michael Erlewine]

At this time I sort of feel that Diglloyd's reporting is a bit unprecise. I don't think he mentions focusing method or focusing point selection.

If you don't know the circumstances it would be difficult to draw conclusions.

I have followed Lloyd Chamber's Medium-Format subscription for some time, also vetted the same cameras (Pentax K1, GFX, X1D, etc.) and had plenty of personal correspondence with Chambers over these issues. I don't find him "imprecise" as you mention it, but very careful and thorough all the way as regards these AF issues. He certainly is well aware of focus points and their selection.

Personally, I don't use AF on these cameras, but I do very much use manual focus. I don't care for focus-by-wire (in general), but having used both the GFX and X1D, for the kind of still-life manual-focus I do, I find the X1D to be very usable and accurate. I had a lot more trouble with magnifying and focusing the GFX. Could be just different strokes for different folks.

[ErikKaffehr]

Hi,

What I find imprecise is his reporting, not necessarily his testing.

Best regards
Erik

I have followed Lloyd Chamber's Medium-Format subscription for some time, also vetted the same cameras (Pentax K1, GFX, X1D, etc.) and had plenty of personal correspondence with Chambers over these issues. I don't find him "imprecise" as you mention it, but very careful and thorough all the way as regards these AF issues. He certainly is well aware of focus points and their selection.

Personally, I don't use AF on these cameras, but I do very much use manual focus. I don't care for focus-by-wire (in general), but having used both the GFX and X1D, for the kind of still-life manual-focus I do, I find the X1D to be very usable and accurate. I had a lot more trouble with magnifying and focusing the GFX. Could be just different strokes for different folks.

[Michael Erlewine]

Hi,

What I find imprecise is his reporting, not necessarily his testing.

Best regards
Erik

I hear that, but then you go on to give a hypothetical involving him (with a boulder), with a lot of "If he did this, etc.)  and to me we should be thankful that he (one of a few) is doing this kind of testing. Just my two cents. It comes off as disparaging of him, rather than trying to get to the bottom of all this AF stuff. I don't mean to be aggressive, just I feel it is a little unfair.

[cgarnerhome]

I guess the way I look at Lloyd's review is it's just a data point - a very helpful one at that.  At the end of the day, all I care about is whether the camera does what I want it to do in my hands.  For my kind of photography, landscape/seascape, I have found the GFX to be an excellent performer.  Most cameras have some limitations/quirks and understanding those limitations/quirks is helpful.  I have also been around long enough to know that testing one camera and one lens is helpful but we have all seen the variability that can exist from lens to lens. Reviews are helpful but draw your own conclusions! Sometimes I get the feeling that people like to use reviews to support their decision or to point out why their choice is better than someone else's choice.

[Michael Erlewine]

I guess the way I look at Lloyd's review is it's just a data point - a very helpful one at that.  At the end of the day, all I care about is whether the camera does what I want it to do in my hands.  For my kind of photography, landscape/seascape, I have found the GFX to be an excellent performer.  Most cameras have some limitations/quirks and understanding those limitations/quirks is helpful.  I have also been around long enough to know that testing one camera and one lens is helpful but we have all seen the variability that can exist from lens to lens. Reviews are helpful but draw your own conclusions! Sometimes I get the feeling that people like to use reviews to support their decision or to point out why their choice is better than someone else's choice.

The problem here is that few of us can afford to buy both systems and compare. I had to do it one system at a time, sending back and trying another. That is why I read these various reviews, aside from incidental interest in all things photographical.