Fujifilm GFX 50R image with as it might be processed: noise reduction applied and warming gradient across foreground.
Fujinon 45mm F2.8 | ISO 100 | 1/25 sec | F10

All camera comparisons have potential for error or inconsistency between cameras. Here are the ones we identified in this test and tried to minimize:

Exposure

ISO 64 allows the Nikon to tolerate 2/3 more light per unit area of its sensor, compared with the base ISO of 100 on the Fujifilm. Conveniently, this 2/3EV difference is very close to the difference in F-number required to give both cameras the same depth-of-field.

The logical implication of this is that the amount of additional 56% light per unit area the Z7 sensor can tolerate comes close to countering the GFX's 68% larger sensor area (the crop factor, which is calculated from the diagonal of the sensors make it look like a closer match, but the different aspect ratios of the chips means the GFX's sensor is slightly larger than the crop factor implies).

Differing aspect ratios

A 4:3 image can't be perfectly overlapped with a 3:2 one, but the framing of these shots has an overlapping region that uses 95% of the GFX's image and 93% of the Nikon image, so the comparison is about as fair as it can be. The overall difference in sensor area means that, even with the exposure difference, the Fujifilm will have received around 8% more light over its whole image, than the Nikon.

ISO accuracy

We suspect Nikon's ISO 64 mode is achieved by using some of the non-linear response that is usually unused by most manufacturers (promising greater DR capability by at the risk of slightly unpredictable results near clipping). To ensure a fair comparison, we analyzed both Raw files using RawDigger and checked how far below clipping each tone in the image is. We found they matched to around 0.1EV difference.

Microlenses

Some of the sharpness difference between the two cameras is likely to stem from the microlens on the GFX sensor having gaps between them. This maximizes spatial separation between pixels (and hence increases apparent sharpness) but comes at the cost of a greater risk of moiré.

Processing logic

Both shots were processed using the same color mode, with brightness, highlight and shadow adjustments tweaked to match brightness across both images. Noise reduction and sharpening was minimized in each instance, with sharpening conducted manually in Photoshop before output.

These processing settings make use of more than 12EV of both camera's dynamic range and give an understanding of the performance you will get from both cameras if trying to utilize most of their dynamic range, and highlighting any potential areas of difference between the two cameras.

We don't expect anyone to process images using this little noise reduction but it reduces the risk that Adobe's profiling is hiding any difference between the two. It also means that any impact of PDAF banding in the Nikon image is not disguised.

Back to the Fujifilm GFX 50 R vs Nikon Z7 comparison