Fujifilm Quality Control... Decided to Be Picky: Returned the Fujifilm GF 35-70mm f/4.5-5.6 WR + Reader Comments
Shooting a 100MP camera and getting maybe ~75 megapixels feels like buying one of those cereal boxes which is the same dimensions, but now with 13 oz of product instead of 16 oz.
My field work with the Fujifilm GF 35-70mm f/4.5-5.6 WR impressed me in certain ways as to the merits of a nice midrange 2X zoom. So convenient, and really a lot of capture there in a package no larger than a Sony A1 and its lenses, smaller and lighter actually!
But the trip also disappointed me in showing me that my loaner test sample was consistently weak on the left side. That’s something I find hard to abide, yet I suspect it is middle of the road and perhaps even one of the better samples. My estimate based on using two dozen or so Fujifilm GF lenses over the past 3-4 years is that at best 1 in 6 samples performs as it ought. That is, without obvious lens skew or similar issues. Which means in 5/6 samples, you never are rewarded with what you potentially could get on the 100MP sensor. And even 50MP sensors show the issue.
So I’ve returned the loaner 35-70mm, having initially intended to buy it.
I want to own the 35-70mm, but not a sample that reminds me that I’m being cheated in image quality. I’m going to take my time and try another sample in June or so. And another and another if I have to. It doesn’t have to be perfect, but I’m not going to accept a lens with an obvious weakness on one side or the other (or top vs bottom).
Roy P writes:
On a different note, I saw your comment on returning the Fuji 35-70. A guy at Phase One explained to me a few years ago that the Schneider blue ring lenses were optically identical to their non-blue ring counterparts, but performed much better on the higher resolution backs (80 MP and greater) solely from much tighter manufacturing tolerances. But it took up the price by some 50% or so.
I think that is the issue with the Fujifilm lenses. Even with the existing lenses, Fujifilm should be able to extract much higher consistency and performance, but at higher prices, probably comparable to the Hasselblad lenses. And any more advanced / better corrected lens designs would also drive up the prices. So the Fujifilm 35-70 is probably built with about the same diligence and manufacturing rigor as a Nikon or Canon kit lens sold through Costco.
Fujifilm’s entire game plan from day one seems to have been to undercut Hasselblad and muscle their way into this segment of the market, so they had to keep the camera and the lenses at a significantly lower price point. Their shock value strategy seems to have worked – the GFX has seriously crippled Hasselblad. But the problem is, of course, in absolute IQ terms, you have these inconsistencies and problems. All the .internal parts are probably cheap and sloppily put together
At this point, I don’t know how serious a player Hasselblad is, so Fujifilm has quite a bit of head room to put out much higher quality lenses at slightly higher prices. Apart from Hasselblad, the nearest competitors for Fujifilm are the Leica S and Phase One systems, both far more expensive than the Fuji at even double the prices. What they have is somewhat in a no man’s land, which is why comparisons with 35mm systems is even a question (e.g., the question you asked, Sony A1 + 24-70 GM II vs. Fuji GFX 100S + 35-70). The differentiation for the Fuji should be much more obvious: higher price for higher quality images.
DIGLLOYD: we cannot really know if Roy’s hypothesis is correct without inside knowledge (other than the physical parts quality thing), but I concur, based on all my experience with GF lenses, including one case which makes we wonder if there is any optical quality control at all at Fujifilm.
To be clear, I am not singling out Fujifilm on lens quality. The same issue exists for all manufacturers, and it is only a question of how loose the quality standards are.
As for optical performance, it is my impression that Fujifilm has, with most of their optical designs, gone for “good enough” performance that falls far short of what Hasselblad XCD lenses can deliver (but see my concerns about the control/management of Hasselblad).
Higher costs from QC are large and real: the reason Zeiss Otus lenses are so expensive is partly because of higher optical element costs, but mainly because of (early on) up to a 50% rejection rate from optical quality control—a massive cost.
Fujifilm GFX100S + Fujifilm GF 35-70mm f/4.5-5.6 WR @ 57.3mm equiv (70mm)
ENV: White Mountains, altitude 7800 ft / 2377 m, 32°F / 0°C
RAW: LACA corrected, vignetting corrected, +20 Whites, +10 Clarity, diffraction mitigating sharpening
[low-res image for bot]
Claude F writes:
Mine is off at 70. The bottom across the frame is tilted or skewed, not sure which but my concession is that I will crop in at that focal length.
DIGLLOYD: we landscape photographers often have to accept some level of inferior lens performance. But we can also try multiple samples. Claude picked the best of three (3) samples!
