This study evaluates the Fujifilm GF 80mm f/1.7 R WR on a distant landscape scene, showing and confirming the findings in Pine Creek Valley view to Freshly-Dusted High Peaks. Sharpness across the field is the focus, as well as which aperture is required for maximum performance over the frame given the very troublesome field curvature.
Fujifilm GF 80mm f/1.7 R WR Aperture Series: View over Patriarch Grove to Southwest
Includes images from f/1.7 through f/8 at up to full camera resolution, plus crops.
This analysis should be very helpful when using the 80/1.7 for landscape.
CLICK TO VIEW: Fujifilm Medium Format System
View over Patriarch Grove to Southwest
f5.6 @ 1/55 sec electronic shutter, ISO 100; 2021-05-31 19:15:27
Fujifilm GFX100S + GF80mmF1.7 R WR @ 65.5mm equiv (80mm) + polarizer Breakthrough Photography X4
ENV: White Mountains, altitude 11800 ft / 3597 m, 55°F / 12°C
RAW: vignetting corrected, +81 Shadows, -100 Highlights, +40 Whites, +20 Contrast, +20 Dehaze, +15 Clarity, USM {9,50,0}
[low-res image for bot]
2021-06-01 - SEND FEEDBACK
Related:
black and white, computer display, Contrast and Phase Detect Autofocus, digital artifacts, Electronic First Curtain Shutter, Electronic Shutter, focusing, Fujifilm, Fujifilm GFX100S, Fujifilm medium format, Fujifilm mirrorless, Inyo National Forest, medium format, noise, Patriarch Grove, pattern noise, PDAF banding and striping noise, shutter, striping pattern noise, White MountainsAt least for making monochrome landscape images, the Fujifilm GFX100S is a total 'FAIL'. No one in their right mind should consider using it for monochrome imagery for landscapes.
Below, the entire image as shown is an unusable disaster because of prominent white stripes everywhere (plain to see at at a bit larger size than shown).
In diglloyd Medium Format:
Fujifilm GFX100S: Horizontal White Stripes — the Futility of Using GFX100S for Monochrome Landscape Images
Includes images up to full resolution, plus crops and ACR settings used.
I feel like a sucker, having purchased the Fujifilm GFX100S. Instead of versatility, I have a $6000 camera with apparently an inherent defect that offers strictly limited post processing options. Which is a shame, because without the striping problems the GFX100S would be the best overall camera on the market for monochrome image making, due to its incredible detail capture. The Fujifilm GFX100 shares this problem as per my findings in 2019.
Below, looks nice, eh? Not so much—fine for Instagram but I can do that with an iPhone. The image is unusable for any larger reproduction—marred by horizontal stripes in sky and land, which I show in the crops per the link above.
View southwest over Patriarch Grove
f8 @ 1/25 sec electronic shutter, ISO 100; 2021-05-31 19:15:34
Fujifilm GFX100S + GF80mmF1.7 R WR @ 65.5mm equiv (80mm) + polarizer Breakthrough Photography X4
ENV: White Mountains, altitude 11800 ft / 3597 m, 55°F / 12°C
RAW: vignetting corrected, +81 Shadows, -100 Highlights, +40 Whites, +20 Contrast, +50 Dehaze, +15 Clarity, diffraction mitigating sharpening
[low-res image for bot]
Prognosis
Can Fujifilm fix these horizontal white stripes with a firmware update? Doubtful.
The striping issue has existed since the release of the GFX100 as per my findings in 2019, yet it has not been fixed. Worse, IMO, Fujfilm has not been forthright about the issue, leaving me to suppose that it is an attempt to ignore the issue and hope it will not gain traction.
I do not have any Fujifilm contacts, nor has anyone from Fujifilm contacted me. I am stunned at how a camera company can ignore blogs like mine, which function essentially as a free outside quality control and R&D resource!
Joseph C writes:
Have you seen the PDAF pimples or horizontal stripes when using EFCS or mechanical shutter? I’m asking because I haven’t been able to duplicate your results with the GFS 100S (and LR/ACR).
DIGLLOYD: on my to do list are the 6 combinations of 14/16 bit X ES/EFC/MS X polarizer/none. The trick is finding a situation like the one above, verifying that the issue occurs (requires the computer), then taking all 12 combinations to see if any of those factors have an influence.
Thing is, EFC and MS are not viable for guaranteed full sharpness due to shutter shake, as Jim Kasson has shown—always a little less sharp than ES (electronic shutter). Might not matter at 30mm or 50mm, but surely matters at 110mm on up. So ES is the only way to go for fully sharp images, guaranteed, at least at some shutter speeds and more so with longer lenses, particularly the Fujifilm GF 250mm f/4.
I reproduced the horizontal lines at home at 500' elevation as the original proof. The recent findings on PDAF pimples and the example above were both at higher elevation (7600' and 11800' respectively), where there is considerably more UV/violet/blue light. In particular the shadow areas in the image above were getting a lot of skylight. However, use of a polarizer makes no difference, and since it cuts down blue light substantially, the blue light hypothesis seems dubious.
Proving and disproving
The existence of the problem is proven by my examples, and I could show dozens more. The only thing that remains is to determine what provokes it. Possibilities (items struck-through have been ruled out by testing):
- Camera specific eg "bad camera"? Ideally I’d have two GFX100S camera bodies I could shoot side-by-side. If both showed exactly the same results or conversely if one did and one did not, that would be very useful information.
- Lighting: UV/violet/blue, etc. Blue light seems involved given the sky, but maybe that is just because the sky makes it more obvious and it is such a pure hue.
- Subject matter and its colors, like blue sky. But even the land is affected at times, as per my example and that was true two years ago with the GFX100 also.
Polarizer.14/16 bit captures because the choice directly affects sensor transit time as well as different A/D conversion.Shutter mode: ES/EFCS/MS.Shutter speed.
UPDATES
On June 2, I systematically looked into some of the above potential variables.
- Shutter (ES, EFCS, MS) makes no difference.
- 14-bit and 16-bit lossless make no difference.
- Polarizer makes no difference.
- Tentatively, shutter speed does not appear to make any difference. I tested {1/110, 1/45, 1/25, 1/10, 1/6, 0.3} seconds with no difference seen.
For those who want to try to reproduce:
Find a patch of blue sky with a gradient, because visibility of the striping varies with density. Clear solid blue, no haze, because haze tends to mask the stripes.
Thanks to readers Tom W and James K for reporting a browser warning for this site, which was using TLSv1.1, which is now deprecated.
This cropped up because companies like Apple have updated their web browsers to issue a warning for TLSv1.1. That’s confusing to users and not important for this site, but that’s how it is.
As of this morning, I’ve hardened the site to use TLSv1.2 with TLSv1.3 optional. Wasn’t planning on it here at 11800' elevation in the White Mountains my van... such a lovely morning lost but it was worth doing.
Also, supported cipher suites have been hardened to only the following:
TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384
TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256
TLS_DHE_RSA_WITH_AES_128_GCM_SHA256
TLS_DHE_RSA_WITH_AES_256_GCM_SHA384
TLS_AES_256_GCM_SHA384
TLS_AES_128_GCM_SHA256
TLS_ECDHE_RSA_WITH_3DES_EDE_CBC_SHA
TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256
TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384
TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA256
TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA384
Very old web browsers might no longer be able to access this site as a result. There is no way to fix that (in general) without degrading site security and getting a 'B' rating. While a few cipher suites could be added and still get an 'A', those suites are ranked as 'weak', and I have chosen to exclude them for now.
SSLLabs.com report for diglloyd.com as of 2021-06-01
