The first two lenses I purchased with the Fujifilm GFX 100S are similar everyday focal lengths: 45mm and 50mm. The plan is to compare these and keep the one that works best. So I immediately went for a walk in my neighbourhood... who could resist!
Nonetheless, I return to my bookshelf for this article, since it provides a controlled environment with a variety of colours, textures, and printed text. Test shots allow me to get a quick idea of lens acuity / sharpness at different apertures. While I don't claim to be scientific, this is more controlled than mere guesswork.
.JPG)
The lenses
Let's start with the key lens characteristics. Both the GF 45mm and GF 50mm have a 62mm filter thread, aperture dial, and similar autofocus characteristics. While the focal length and maximum aperture are similar, they are distinguished by size. The 50mm is the most compact form factor in the entire system, while the 45mm is a more typical size in a system where lenses are huge! I should note that the 50mm has a rather strange minimal stubby hood which functions to recess the front element out of harm's way. The 45mm (pictured above) has a petal hood.
Fujifilm GF 50mm f/3.5 R LM WR
- 48 mm long, 54 mm with hood, 335g
- 55 cm close focus
- 40 mm f/2.8 equivalent
Fujifilm GF 45mm f/2.8 R WR
- 88 mm long, 142 mm with hood, 490g
- 45 cm close focus
- 36 mm f/2.2 equivalent
I indicate full-frame equivalence to aid those, like me, who are coming from such a system. If you wish to argue the relevance of this concept please first read my exhaustive article on the subject. For these calculations a crop factor of 0.8 was used. In any case both lenses are more similar than different when it comes to the rendered field of view. Personally I enjoy the more relaxed feeling of the 35 to 45 mm range compared with a typical normal lens. (Note that the GF 63 mm is normal for this system.)
Test method
For this test I fixed the camera to a tripod about 1.1m from my bookshelf. Image stabilisation was off and timed shutter release was used. I focused manually using the convenient magnification feature. The room had consistent natural sunlight, diffused by the central position in the room, so there were no hotspots or inconsistencies.
Though I had meant to hold at ISO 400 I accidentally changed this halfway through the test. Since I could see no difference in image quality I decided not to reshoot since the light had changed by the time I noticed my mistake. I judged this factor to be more important than a small ISO shift. I am not scientific but pragmatic! Any small differences in exposure are due to the camera metering itself, which I had set to centre.
I note that this test is not designed to display aberrations, bokeh characteristics, flaring, or other characteristics.
I shot lossless compressed 14-bit RAW files. All indications are that these are functionally equivalent to larger uncompressed 16-bit files except (perhaps) for the most extreme grading scenarios.
The files were developed in Affinity with no enhancements. But I did use the automatic lens correction profiles since this pipeline is typical. On inspection it is apparent that the 45mm needs more geometry correction than the 50mm but neither show significant vignetting.
Any real-world development would produce better images in terms of desired colours, contrast, sharpness, or other characteristics. This test provides a baseline but not optimal images.
I stacked the 11662 x 8746 pixel images and then reduced the overall size by 50% to enhance any apparent acuity. Very few people are going to be reproducing photos at the original resolution so this is a more realistic way to judge results. Then I cropped to a 1000 pixel square in the areas of interest.
I had originally also shot at f/16 but I omitted these images from the comparison. It was evident that all such photos had noticeable softening due to diffraction. I am not used to diffraction coming into play so early but... welcome to medium format! To be clear, this diffusion had not yet "ruined" the images but was enough to obscure small differences between the lenses. I would be OK shooting at f/16 in real world usage.
Test results
Here are the comparison grids. Click on each to expand to full size. As always, I recommend viewing on a computer, not a mobile screen.
The first grid shows crops from the centre of the image, where focus was also in the centre, on the word "SOUND" in white on the black background spine.
Second is a crop from the extreme left side of the frame, focus as above.
For the next two images the focus was at the extreme left side of the frame, or as near as I could comfortably manage, on the letters "THE LEGEND". First a crop from the focus area.
This is a centre crop of photos where the focus was as above, at the left edge.
I test with two focus points in order to assess field curvature. If a lens has blurry edges when focussed at the centre and a blurry centre when focused at the edge then field curvature is indicated. This optic could be OK for a portrait lens since it actually helps to emphasise a subject in the centre of the frame. An example is the Pentax FA 43mm Limited which has an incredible render once you learn how to use it. But for environmental shots, street photography, landscapes, etc. field curvature can be detrimental.
If a lens is not sharp at the edge of the frame even when focused at the edge, this is more indicative of off-axis aberrations.
Analysis
There is very little to say because both lenses are excellent. The 45mm is slightly crisper than the 50mm even wide open. Clarity improves even more when stopped down from f/2.8 to f/4. The performance across the frame is spectacular.
For fun I mounted my Carl Zeiss Planar T* 50mm f/1.7 but it couldn't compare. Besides distracting vignetting the lens was only sharp in the centre until stopped down significantly. Oh dear. (More adapted lenses will be tried in the future.)
Apparently everything I have read is true. The GFX system is amazing... to the point of rendering such tests moot.
Handling
A couple notes about handling. When mounted on the camera the 45mm does not quite sit flush on a flat surface. The hood is raised just a tiny amount so that any movement causes it to jiggle, if this makes sense! Never seen this with any other lens and it's surprisingly annoying. Though the lens is significantly larger it still fits in the hand comfortably for shooting.
The difference between close focus distances is decisive. One of the beneficial results of a 100 megapixel sensor is the ability to crop in a great deal and still have a detailed photo. In this way 45mm can act like 90mm or more. Often I like to get close to subjects and get a magnified view.
The 45 mm lens has 45 cm close focus, providing 33 cm clearance in front of the hood. This never got in the way of me getting the shot I wanted. The 50 mm lens has 55 cm close focus, providing 50 cm clearance in front of the hood. This required me to back off from a comfortable distance quite often. (An example is the shot accompanying this article.) Of course if you are not interested in getting close to your subject this won't matter.
Conclusion
The improved IQ, close focus, slightly wider field of view, and light capture are compelling reasons to favour the GF 45. But the extremely compact GF 50 argues against those factors. I will take more photos and see which is more practical.
No comments:
Post a Comment