Need Advice: H4D-40 or Fuji GFX 50

[Bo_Dez]

If you ask me, the way an image looks, and the testable quality in terms of MTF etc of the lens used are antithetical. The more one introduces a battery of normative tests into lens evaluation, the less one should expect a lens to look "special".

I own two 85mm 1.4 Nikon lenses, the old and the AF-S design. When I went to visit Norman Koren in Boulder some years ago, I took the new one and my D3s,  and we ran the lens through Imatest, which Norman probably knows how to use as well as most people on this forum, and as he said it was then the best lens he'd ever tested. It basically outresolved the sensor, edge to edge, mounted on what was at the time a bleeding edge dSLR. And yet, another image geek I know, Iliah Borg, had the same impression as me about the "new lens, if I remember rightly  he called it politely "sterile".

Each of the above lenses has its uses, I guess. For repro use or for night photography *against* lights, the new lens with its flat rendering, incredibe sharpness and zero flare is perfect. I actually used it all the time because the images are ok, it focuses super-quickly, and it works perfectly in any context. But if what you want is painterly rendering the old lens is better, and it even has very good sharpness.

People who critique a lens should mention the intended usage to provide context for their opinions.

Edmund

I agree that just because a lens is sharp, corrected etc doesn't make it a nice lens or a lens that is suitable for certain instances. That is contrary to what I'm speaking of. I'm talking about lenses which are designed to be well corrected and yet still have gone that extra measure of creating pleasing aesthetics. Many older lenses are pleasing, not by design, but by mix of luck in design, and also interesting aberration. Many modern lenses are corrected for aberrations but designing aesthetics into them as well comes at extra R+D and manufacturing cost that is often beyond what their customer will buy or have the budget for.

If you need to keep medium format affordable, because that is the design objective, then there is only so much you can do before budget caps things. The focus is on other things. It's easy to fly under the radar with these designs by optimising the things that are tested in mainstream testing such as sharpness, CA etc. But "the lenses are really sharp" only means the lenses are sharp and often nothing more.

It seems the ideal to design to is a lens that is free from aberration but aesthetically pleasing. A lens that doesn't get in the way but still heightens reality. Some of course want aberrations too which is fine. A design that heightens dimensionality through tonality does generally carry widespread praise. 99% of people don't even know what they are seeing and say things like "it renders the light nicely" but can't say why.

I go for aesthetically pleasing designs but there are no absolutes either as taste is subjective. What I find pleasing others might dislike.

[Bo_Dez]

I agree, except for the Heisenberg-ish way you phrased it (the testing of the lens doesn't affect the "look" of the lens except for atomic-scale lenses). I have long been fond of saying that the "look" of a well-loved lens is attributable to what I call "adorable flaws". A diffraction-limited lens with zero SA, LoCA, LaCA, coma, etc wouldn't have a "look" at all.

Jim

Well that is a broad and far reaching statement but with plenty of caveats I could disagree and say yes it would if you consider other factors. It's just that your testing parameters are all you are basing things on.

[eronald]

I agree, except for the Heisenberg-ish way you phrased it (the testing of the lens doesn't affect the "look" of the lens except for atomic-scale lenses). I have long been fond of saying that the "look" of a well-loved lens is attributable to what I call "adorable flaws". A diffraction-limited lens with zero SA, LoCA, LaCA, coma, etc wouldn't have a "look" at all.

Jim

Jim, please read my comment as Darwinian rather than Heisenbergian. Relentlessly added and publicised laundry-lists of lens tests operate as selective pressure breeding the "adorable flaws" out of the lens designs even before they are fabricated and enter the market.

Although of course, one might consider it demonstrated that in some cases observation retroactively affects a facet of a system, I believe, but that point of view would rather be called "Einsteinian" (Einstein, Podolsky Rosen) or entanglement or something like that.

https://blogs.scientificamerican.com/critical-opalescence/how-to-build-your-own-quantum-entanglement-experiment-part-1-of-2/


Edmund

[Jim Kasson]

Jim, please read my comment as Darwinian rather than Heisenbergian. Relentlessly added and publicised laundry-lists of lens tests operate as selective pressure breeding the "adorable flaws" out of the lens designs even before they are fabricated and enter the market.

Gotcha, Edmund. I was pretty sure that's what you meant. Always fun seeing your perspective on things.

Jim

[Jim Kasson]

Well that is a broad and far reaching statement but with plenty of caveats I could disagree and say yes it would if you consider other factors. It's just that your testing parameters are all you are basing things on.

What I'm saying is independent of testing. It's about the inherent characteristics of the lens. I have written a lens simulator, though I haven't used it in a while and it would take me some time to dust it off. If I feed it a ray-traced image and set all the knobs but diffraction to zero, there's no "look" at all, at least, no look that I can see.

Jim

[Doug Peterson]

Jim, please read my comment as Darwinian rather than Heisenbergian. Relentlessly added and publicised laundry-lists of lens tests operate as selective pressure breeding the "adorable flaws" out of the lens designs even before they are fabricated and enter the market.

Although of course, one might consider it demonstrated that in some cases observation retroactively affects a facet of a system, I believe, but that point of view would rather be called "Einsteinian" (Einstein, Podolsky Rosen) or entanglement or something like that.

https://blogs.scientificamerican.com/critical-opalescence/how-to-build-your-own-quantum-entanglement-experiment-part-1-of-2/


Edmund

Bingo.

More broadly if you build a camera system or camera system component based on specifications (“teach to the test” “paint by numbers” or “built for bragging rights” rather than wholistic utility you tend to get garbage.

It takes a first product manager motivated by long-term thinking and wholistic real-world use, r+d folks who are shooters and want to use the product (not just engineers looking to collect a paycheck), and a lot of guys to build a system that is best for actual use, not just best on paper or according to the specs and numbers.

In my (heavily biased) opinion Leica and Phase One do the best job of this.

[Bo_Dez]

at least, no look that I can see

Because you can only see what your simulator has been programmed to simulate.

A lens with zero SA, LoCA, LaCA, coma, etc can still be designed to manipulate tonality.

[Jim Kasson]

Because you can only see what your simulator has been programmed to simulate.

A lens with zero SA, LoCA, LaCA, coma, etc can still be designed to manipulate tonality.

Please elaborate, quantitatively if you can. When the sim is set up with the knobs at zero, the spectral transfer function is flat.

Jim

[pschefz]

Thanks for this, is there any way you can elaborate on this process and what I need for it?  Am pretty new to tethering in general...

It is only the jpeg's that are viewed this way though right?  Is there any tethering solution that uses the GFX Raw files?

Thanks in advance!

when you tether the GFX to LR classic, you will need the (now free) fuji plug in, once that is installed, one simply activates tethering within LR and sets the parameters.....you can shoot raw, jpeg,...whatever you like....files are transferred over to the computer as well as stored on the internal card (which is great, great, great...instant back up)...works really well.....
if you are happy with just jpegs, there is also the fuji iOs app which lets you control the camera as well as pull images over a direct to camera wifi connection....it might work for raws (?) but in reality i dont see the point, even large jpegs take a few seconds each to pop up....you can shoot raw to one card, jpeg to the other and use the fuji app in this case to get a quick idea of the shots on the iPad (or iPhone) and still have the raws on the card.....this works well but i dont really care much about jpegs, it adds another layer to my workflow that i dont need.....
hope that helps

[pschefz]

This is the very sort of comment and attitude that keeps forum boards the domain of the uninformed sock photographer.

i think you are taking this way to serious....Jim's comment just made laugh out loud at my screen, which does not happen that much....

[Bo_Dez]

Please elaborate, quantitatively if you can. When the sim is set up with the knobs at zero, the spectral transfer function is flat.

Jim

Your simulator isn't accounting for everything in photography and lens design, it's just removing some things. It can't account for modern design processes, modern coatings, modern manufacturing processes and materials.

[pschefz]

A shoulder doesn't have a blade it has a scapula. Why didn't they call it a shoulder lever since that's what it is? It used to be called a wing bone of all things. Since we don't fly, that was confusing for people too. How we settled on blade is still a mystery since the idea of cutting anything is just as stupid as flying. It's been named as several things such as shovel bone as well. Why the hell don't we just call it a Scapula? We can waste time scoffing at lay terminology or we can just get on with the more important things in our life. 3D, similarly, is a lay term and also an umbrella term because there are a number of distinctly measurable traits that model the dimensionality of an image. Since it's a collection of things, but produces an effect that people can actually recognise, it needs to be given a name so people can argue about it with sock photographers on forum boards.

It's a combination of lens design characteristics, some of which are either neglected through poor design, eliminated for budget, or not important or relevant to the lens designer and/or their market... For some, it's a contribution of design aesthetic. It's considered a luxury in terms of optics and has only gained relevance in design in more recent times. Why does a Ferrari look so much better than a Toyota? There's lots of reason, but mostly because it's more expensive.

Scapula is just a latin word scientists agreed on using for a very specific part of the body so everybody regardless native language knows what everybody is talking about.....
if you ask 1000 photographers what they consider the 3D look, most would probably bring bokeh into this discussion....
i disagree that there is a quantifiable 3D look that lenses can be designed for.....
i agree 100% that there is much more to lenses then numbers and to me the best lenses have character.....if "3D" is part of that character for you, i totally get it....

[pschefz]

3D is not a feeling. It's a quantifiable thing and part of lens design.

i am not sure how you make this statement but then dismiss quantifiable testing.....
i agree with character, i am not sure how much of it is part of lens design (more likely to be a flaw in design)

and if 3D is (like you say) a quantifiable thing, we are doing a full circle:
what are the units?

[Jim Kasson]

Your simulator isn't accounting for everything in photography and lens design, it's just removing some things. It can't account for modern design processes, modern coatings, modern manufacturing processes and materials.

It can model the effects of all those things, although setting it up to do so can be very laborious. If you can quantify the effects (at least the ones that you consider relevant) on the image, please do so.

Jim

[Bo_Dez]

i am not sure how you make this statement but then dismiss quantifiable testing.....
i agree with character, i am not sure how much of it is part of lens design (more likely to be a flaw in design)

and if 3D is (like you say) a quantifiable thing, we are doing a full circle:
what are the units?

I'm not sure how you can be so continually snarky and expect people to share info with you.

Just because you haven't learned about something doesn't mean that those who have are having "religious belief experiences" or what ever you called it.

[pschefz]

I'm not sure how you can be so continually snarky and expect people to share info with you.

Just because you haven't learned about something doesn't mean that those who have are having "religious belief experiences" or what ever you called it.

i would like to see a snarky comment i made?
i wont get into this further since i already stated several times that i absolutely believe in "character" or what ever you want to call it....i personally have and use technically crappy old lenses for that reason....

if that personal opinion reaches the realm of "religious beliefs"...as in brand X or Y is better because of non measurable qualities like..."3D"....and at the same time brand Z is being criticized for lacking some non measurable qualities.....

but if i got it right you are saying that 3D is measurable and a design feature, but then jump on Jim for trying to put design and performance into absolute numbers?

lets not forget that this is a discussion about hasselblad H4D vs fuji GFX....my "snarky" remarks refer to the phenomenon that some "true believers" always praise the "3D" in the hasselblad H glass (when it was made by fuji)....and now only see the incredible "3D" in files from the X1D (with lenses not made by fuji) and comment on the "lack of 3D" by fuji lenses on the GFX.....
and it is usually the same people who will put leica glass (and even bodies?!) over everything else....even if they are not, because of....3D...
also: good design is never expensive...never...good design is maximizing available resources in order to maximize functionality....good design can often cost a lot of money but it does not have to and good designers are always trying to keep it simple.....

[Jim Kasson]

I own two 85mm 1.4 Nikon lenses, the old and the AF-S design. When I went to visit Norman Koren in Boulder some years ago, I took the new one and my D3s,  and we ran the lens through Imatest, which Norman probably knows how to use as well as most people on this forum, and as he said it was then the best lens he'd ever tested. It basically outresolved the sensor, edge to edge, mounted on what was at the time a bleeding edge dSLR. And yet, another image geek I know, Iliah Borg, had the same impression as me about the "new lens, if I remember rightly he called it politely "sterile".

I tested the G version of the Nikon 85/1.4 and found it to measure nowhere near as well as the Otus 85/1.4:

https://blog.kasson.com/the-last-word/another-medium-tele-test-otus-nikon-loca/

Doesn't look as good, either:

https://blog.kasson.com/the-last-word/another-medium-tele-test-f1-4/

But the Leica 90mm f/2 Apo-'cron ASPH was also in that test; it measured so-so, but looked almost as good as the Otus overall, and better for some things.

I have heard some people complain about the look of the CO 60/4, and I think that's a commentary on how good the LoCA is (although it has other foibles). We're not used to seeing lenses that good wrt LoCA.

So I'm agreeing with you about the use case being important.

And, to bring this back around to MF, the CO 60 works very well on the GFX, except in the distance range which has the sensor reflection issue.

Jim

[hubell]

Let’s not forget that the back and forth here was sparked by someone who failed to genuflect to the insistence of the GFX Shooters that everything about the GFX lenses is PERFECT. If anyone dares to dissent even slightly, the same prime defenders of the faith descend. Not just here. It’s everywhere, though the DP Review MF Forum is the worst. The very same people are trying to enforce the faith here.

Oh, no. The latest blog entry from DIGLLOYD. Time to call in reinforcements.

[eronald]

Let’s not forget that the back and forth here was sparked by someone who failed to genuflect to the insistence of the GFX Shooters that everything about the GFX lenses is PERFECT. If anyone dares to dissent even slightly, the same prime defenders of the faith descend. Not just here. It’s everywhere, though the DP Review MF Forum is the worst. The very same people are trying to enforce the faith here.

Oh, no. The latest blog entry from DIGLLOYD. Time to call in reinforcements.

That terrible Digilloyd ... "in my experience with both, the Fufjilm GF lenses range from grades B+ to 'F', but all the Hasselblad lenses get an 'A'. So I see trouble ahead for Fujifilm GFX 100-S users—see for example the worst lens I ever have come across with one sample of the 110mm f/2. How that one made it out of the factory seems a bit ominous in terms of quality control. The 2nd sample was excellent, but that’s the point."

I think Digilloyd must be a cousin of mine or something, I always get the lemon bodies, he gets the bad apple lenses.

Edmund

[fredjeang2]

good design is never expensive...never...good design is maximizing available resources in order to maximize functionality....good design can often cost a lot of money but it does not have to and good designers are always trying to keep it simple.....

Correct