I have been using a FujiFilm GFX 100s medium format camera now for a couple of weeks; some observations were posted earlier:
Having become a little more comfortable with the operation of this beast, my attention has turned to whether I should try using my existing prime lenses, and so I ask myself:
- How well do they work on the GFX?
- Is it worth the effort or should I use only native GF lenses?
To answer these questions I decided to look at different configurations to gauge the performance, functionality and usability of different combinations of these non-GF lenses. My testing is not a rigorous scientific study and the results are subjectively assessed. I have neither the tools nor technical capabilities to perform a more exacting undertaking.
Image Capture
By image capture I implicitly mean how much information is being captured with each frame. Presumably, the more information the better. The table below summarizes the differences in the resulting images in terms of file and image size for three different scenarios. The first is that for the GFX in its native medium format mode. The GFX also offers a 35mm mode where it crops virtually the sensor to the 35mm format. This would be the setup one might be expected to use if attaching a 35mm lens to the GFX body. The last entry in the table provides a benchmark with the Leica M10-R for comparison with a full-frame camera body.
| Camera / Format | Image Size (Pixels) | Colour Depth | Sensor Area | File Size (mb) |
|---|---|---|---|---|
| GFX100s/Medium Format | 11,648 x 8736 (102mp) | 14 / 16 bit | 1,441mm2 | 108 (compressed) |
| GFX100s/35mm | 9,552 x 6368 (61mp) | 14 / 16 Bit | 66 (compressed) | |
| Leica M10-R/35mm | 7,864 x 5200 (41mp) | 14 Bit | 864mm2 | 47 (uncompressed) |
The GFX certainly collects more information, as shown by the image size and colour depth, and as suggested by the corresponding file size. Even in its 35mm mode, it collects more than does the Leica, probably due in part to the differences in pixel size: GFX = 3.76 microns; Leica M10-R = 4.59 microns. While a smaller pixel size is usually associated with increased noise, as noted previously in my observations on Technology Improvements, noise seems not to be a problem.
Lens Performance
Determining the performance of a lens is a complex matter. Sean Reid undertakes such studies, with far more rigour and understanding than I can ever do, in his newsletter Reid Reviews, so I will not attempt to recreate what he and others have already done. But for my purpose, my question is quite basic: is there a [visually] noticeable difference between my existing lenses and the GF 45-100mm lens I currently have? The answer is yes there is a difference primarily due to distortions around the edges of the image.
The image below shows the differences in the amount of the scene transmitted through the lens to the sensor. The test was conducted inside, with a 90mm focal length, and the subject was about 1.5 meters from the camera. The full image corresponds to that part of the scene transmitted by the medium-format lens. The area bounded by the red rectangle is the portion of the scene captured with a 35mm lens mounted in medium-format mode. The area bound by the blue rectangle is the portion of the scene captured with a 35mm lens mounted on the Leica M10-R.
However, I suspect the difference between the full image and the one bounded by the red box has to do with difference in the angle of view of each lens. As such the test is interesting but inconclusive.
The next series of images look at the distortions introduced by mounting 35mm lenses of various focal lengths on a medium format body. Lenses typically perform best through the centre of the glass, often softening towards the edges, which we see in the examples below.
Most of the lenses introduce vignetting. This can only be expected as they are designed to cover a 35mm sensor surface, not the larger medium format surface. But, it is interesting how little is actually lost. The Konica lens shows very little vignetting. While there is some distortion of the image around the edges of the image, again it’s not as bad as might be expected. I don’t see any colour distortions.
Overall, I’m pleasantly surprised.
Usability and Functionality
The Leica and Konica lenses fit well on the camera body and don’t feel unbalanced. There is certainly a usability argument for the Leica and Konica lenses in the sense of size, weight, etc. The Fujinon 45-100 is a relatively large and bulky lens. However, this usability argument becomes less clear-cut as one finds the need to carry several of the smaller prime lenses vs. the single Fujinon, not to mention the need to change lenses.
The prime lenses have distance markings on them allowing one to prefocus and/or use hyperfocal distance to quickly set to the maximum depth of field. I haven’t really figured out how to do this on the Fujinon lenses.
The major impact on the usability of the non-Fujinon lenses is their lack of integration with the camera body, meaning functionality (e.g.,automatic focusing, face detection, aperture setting) is not available.
Closing Thoughts
I am not arguing one camera vs. the other, rather I’m tending more towards fit-for-use.
While the GFX camera body is small relative to other medium format cameras, it is still much bulkier than the Leica. Coupled with its larger lenses the camera has a greater intimidation factor; the Leica is more compact and less in-your-face.
The prime lenses mounted on the GFX body do reduce the intimidation, but more importantly, they are more often much faster than anything currently available from FujiFilm. While lens speed is less a requirement because of the in-body stabilization and the high ISO performance, depth of field remains a consideration.
The functionality of the GFX camera body — auto focus, in-body stabilization, wide dynamic range, face / eye detection etc. — make this camera more usable for a wider set of applications, ranging from landscape, to portraiture. Neither is great for sports photography. Its large sensor and moveable LCD screen make the GFX more practical for product photography. In its 35mm format mode, it can capture a 61mp image with 16-bits colour depth, out performing the Leica M10-R. Yet, given vignetting and image distortion is less severe than I anticipated, other than reducing file size, I’m not sure there is any reason to use this mode, and if necessary, crop in post production.
Yet, the smaller size of the Leica makes it more appropriate for situations where bulk / intimidation is a problem, such as street photography. Being able to preset the focus makes this a fairly responsive camera; a point-and-shoot with no need to wait for the auto-focus to miss the subject.
This leads to the final point, that the Leica is largely a manual camera; certainly it is not the “computing platform” that one might argue is the FujiFilm. It is thus a matter of opinion as to whether this is better or not. Is there an argument that the Leica provides a closer connection to photography, while the FujiFilm hides it under a veil of technology? It might be like comparing a manual vs. automatic car.
Leave a Reply