Disclosure - I am a hobbyist photographer and a photo gear aficionado. I wrote this article to share my experience with dpreview.com forum members, who expressed their interest in my observations. This is my first gear review, so there may be some flaws in methods I employed or in their execution. This is not an attempt of a scientific review with a firm control of conditions and elements, but rather an observation of ad hoc use and results. I only address camera features relevant to me.
After 4 years with my Sony A7R III, I am itching for a new toy. Just to make it clear, there is no objective need driving the change. The A7R III is such an accomplished camera even by today's standards and such a good value for money, I am finding it hard to think of an alternative. I got so spoiled by the 42mp and nigh 15 stops of dynamic range that there have to be a significant advantages in other aspects of a camera to consider downgrading.
There were "just" four interchangeable lens cameras released in the past few years which I would actually want to own. Fujifilm GFX 100s ticks all my boxes, but it is too expensive as is the Sony A7R V. I like the OM Systems OM-1 for its superb IBIS and computational features which do away with the need for tripod in many situations, but it doesn't cut it in resolution and dynamic range department. Which brings me to Fujifilm X-T5. I owned an X-H1 and there were many things about it which I loved - very well thought out camera body and handling, the analogue controls, the colours..., but I found my unit unreliable, occasionally freezing up and after it reset itself to factory settings for the second time, I decided to move on. X-T5 offers the traditional Fujifilm experience with analogue controls and 2-axis tilt screen with the new 40mp sensor, and I got hooked.
I tried to find information on how the sensor performance compares to 40-ish mp full frame alternatives, but I could only see portrait photographers shooting the X-T5 or X-H2 (same sensor and processor) against Sony A7 IV or Canon R6 II. The results were not flattering for the 40mp Fujifilms being outperformed by lower mp full frame options in ability to resolve fine detail.
I was not convinced. These comparisons were shot with the top FE and RF primes, either 50mm f1.2 or 85mm f1.2 (f1.4 for Sony) versus XF 33mm f1.4 or the XF 56mm f1.2 II. The mentioned XF lenses are surely excellent in what they deliver, but it seemed to me that it would be a fairer comparison to shoot them against f1.8 or f2 full frame primes. This would bring closer equivalence considering the crop factor and optical formulas of the lenses. In addition, the comparisons were shot wide open with the Fujinons and relying on eye-AF - the Canon and the Sony lead the pack here and Fujifilm lags behind.
I wanted to see how the sensor performs with the camera on a tripod and a lens stopped down to landscape aperture. So, I took advantage of Fujifilm's generous 48 hour free loan offer and rented an X-H2 with XF 16-55mm f2.8 to shoot it side by side with my A7R III with FE 24-105mm f4. On a flop side, X-T5 was not available at the time, on a flip side, I consider X-H1 to be one of the best camera bodies I ever held, so I was curious how the second generation compares. Time was limited and light even more so, as is usually the case this time of year in the West Midlands.
UPDATE: Diffraction may be visible on 40mp APS-C sensor from f6.3, so having shot the X-H2 at f8 puts it at a slight disadvantage.
To compare the resolution, I shot 4 scenarios with different subjects at different distances. I used my Leofoto LS-324C tripod with LH-40 head, 2s self timer and Capture One 22 Pro to demosaic and edit the raw files. Framing is shifted by the difference in the distance from tripod mounting point to the lens mount centre between the two cameras. Each camera and C1 automatically apply presets and lens profiles, I adjusted only sharpening always to the higher value of the two files being compared. To be precise, Sony applies 20 points more sharpening by default on a scale of 1000 points with the same radius and threshold, so there is nothing radical happening here. I applied "Film Standard" curve to Sony files and "Provia Standard" to Fujifilm files. They look similar-ish this way, more so than using my favourite "Classic Chrome". A7R III colours are closer to real life. X-H2 uses different ISO rating and has base ISO of 125 vs ISO 100 on A7R III. The 1/3rd stop difference is carried over to ISO settings where ISO is a variable. We will be viewing the photos as a whole scene, then 100% crop, then 200% crop. Let's keep in mind - at 200%, we are talking serious, nigh pathological degree of pixel peeping, but it certainly helps to appreciate the fine margin in the performance of the two cameras. Zipped raw files are linked at the bottom, so feel free to play with them as you please.
The first scenario is our bookcase shot from 2m with tripod adjusted so that the lens and the face of the shelves are perpendicular. Lenses are stopped down to f8 for optimal performance.
 |
| X-H2, 34mm f8, 1/1.3s, ISO 125 |
 |
A7R III, 50mm f8, 0.8s, ISO 100
|
 |
X-H2 100% crop
|
 |
| A7R III 100% crop |
 |
X-H2 200% crop
|
 |
| A7R III 200% crop |
Unsurprisingly, A7R III combo resolves more detail, but not by much. Viewing at 100%, the difference is minimal and only at 200% it becomes worth commenting on.
Let's have a look at some photos of a pixel-peeper's best friend shot from 2m with tripod adjusted so that the lens and the face of the wall are perpendicular. Lenses are stopped down to f8 for optimal performance.
 |
| X-H2, 34mm, f8, 1/15s, ISO 125 |
 |
| A7R III, 49mm, f8, 1/15s, ISO 100 |
Brick wall is ruthless and this time round the difference is a bit more obvious, but again, it is very close. With a very small adjustment adding 20 Clarity and 15 Structure to the Fujifilm, the photos are pretty much identical.
Let's have a look at some photos when focusing toward infinity. I overexposed A7R III shots by mistake and corrected for that in post. It does not have an effect on the result.
 |
X-H2, 35mm, f8, 1/250s, ISO 125
|
 |
A7R III, 50mm, f8, 1/200s, ISO 100
|
 |
X-H2 100% crop
|
 |
A7R III 100% crop
|
 |
X-H2 200% crop
|
 |
A7R III 200% crop
|
We see a repeating pattern here - A7R III in the lead, X-H2 right on its tail.
How will they deal with the UK winter foliage? I moved the tripod a bit when swapping the cameras, but I adjusted the crops so they are close. I again overexposed A7R III shots by mistake and corrected for that in post. It does not have an effect on the result.
No surprise happening here.
It is nice to see that the cameras perform consistently across the four scenarios. Overall, as far as resolution is concerned, Fujifilm's 40mp aps-c sensor is excellent, trailing full frame alternatives by a narrow margin. By increasing Clarity by 20 and Structure by 15 in the RAFs, the gap is pretty much closed.
I ran a series of shots to test noise level at ISO 3200, 6400 and 12800 with A7R III and ISO 4000, 8000 and 12800 with X-H2, but managed to underexpose by 1-2 stops, so I will not share the files. From what I could salvage from the photos, I gather that A7RIII has about a 2/3 to 1 stop advantage over the X-H2, which corroborates findings of other reviewers. The Sony maintained better colours while Fujifilm had less chromatic noise. I find chromatic noise ugly, so I found myself to prefer Fujifilm files to Sony's even if they were objectively noisier. There was one peculiar thing about the X-H2 noise though - at ISO 6400 and above, white, red, blue and green specks appear in the noise, probably something to do with x-trans layout:
 |
X-H2, ISO 8000, 200% crop
|
 |
X-H2, ISO 8000, 100% crop
|
I messed up exposure settings in the dynamic range test as well, so I won't be posting comparison images either. Fortunately for you, you will be able to find such images with other reviewers. From my findings, highlight recovery was about the same with both cameras retaining good level of detail up to 3 stops overexposure, losing blown areas at 4 stops and completely falling apart completely at 5 stops. In shadows recovery, A7R III had a clear advantage of about 1 stop. I refrain from pushing shadows more than 3 stops on A7R III. I was surprised how quickly the colour starts to fade with A7R III from 3 stop shadow recovery on, I haven't noticed this until running this test. Fujifilm seems to be a bit better here. As in high ISO test, X-H2 showed more noise, but less chromatic noise. Some shadows recovery examples - there is a 2/3 stop difference in exposure between the cameras, crops are at 100%.
 |
| The Scene |
 |
| A7R III, 1/8s, f8, ISO 100, 2 stops recovery |
 |
A7R III, 1/15s, f8, ISO 100, 3 stops recovery
|
 |
X-H2, 1/5s, f8, ISO 125, 2 stops recovery
|
 |
X-H2 1/10s f8 ISO 125, 3 stops recovery
|
To sum up, high ISO and dynamic range performance comparison findings are similar to comparisons of other aps-c vs full frame cameras with modern sensors, with full frame retaining ~1 stop advantage. A pleasant finding was X-H2's very good highlights recovery performance.
A note on IBIS
Until arrival of the A7R V, Sony implementation of in-body image stabilisation has been known to be less impressive than that of its rivals. I remember being very happy with IBIS on X-H1 and found A7R III to be a step down in this respect. I expected the X-H2 to have at least a stop advantage over A7R III, but it was not the case... at least not in my hands. There is a caveat to this - the FE 24-105/4 is stabilised and the XF 16-55/2.8 is not. I confidently shoot my A7R III at ss 1/focal length and I don't hesitate to go a stop lower with great care and, if possible, multiple shots taken to be certain I have a sharp image. At 24mm, I had no problem producing sharp images at 1/10s with 80% hit rate. I managed to get a sharp shot (1 of 3) at 1/5s, but it is not something I would attempt unless I had no choice. At 16mm, I was confident with the X-H2 at 1/15s, but had to take extra care at 1/8 to retain that 80% hit rate. Any slower than that and the images were blurry, I wouldn't call any one I got "sharp-ish". Crops are at 100%:
 |
| The Scene |
 |
| A7R III, 24mm, 1/10s, f8, ISO 100 |
 |
| A7R III, 24mm, 1/5s, f11, ISO 100 |
 |
| X-H2, 16mm, 1/15s, f8, ISO 125 |
 |
X-H2, 16mm, 1/8s, f11, ISO 125
|
|
I would speculate it is the higher quantity of much smaller pixels, which puts higher demand on stability to achieve a sharp shot. A comparison with X-H2S would be interesting.
A note on autofocus
Autofocus performance is not high on my priority list, so this is a summary of a few things I observed when using the cameras. Ability to focus on different things under different lighting conditions is about the same. Either both cameras acquired focus easily or they both struggled. Eye-AF works very well on A7R III and when it can find an eye, which can be an issue in bad light, it almost always nails the shot. X-H2 tracks eyes very well too, switches nicely between face and eye when needed and the overall experience felt more fluid, no doubt because of advances in AF tracking over the past half decade. However, when the shutter was half pressed, the AF locked. This seems illogical, but may have been due to settings, which I did not care to dive into.
A note on handling
Side by side, the bodies with their lenses are pretty much the same size. With a battery, one card in slot 1, lens hood and a neck strap they weighed in 1408g for X-H2 and 1383g for A7R III. The top display is an excellent feature and one of the things I miss from my X-H1.
The grip on X-H2 is short and has no space for my pinkie, just like on A7R III. Interestingly enough, A7R III felt more comfortable and secure with the lens on, the X-H2 felt much more front-heavy. After a closer examination, this is due to the lens being mounted lower relative to the position of the hand on the grip on the X-H2 than on the A7R III and below the axis of the wrist. I find it amusing that reviews generally give the grip on X-H2 a high praise, where this had been one of the most criticised features on A7R III for some time after its release. If you use AF joystick, I much prefer X-H2's position by the viewfinder more as using it doesn't require a thumb to move too far down, which compromises the grip on A7R III.
Otherwise, as far as controls are concerned, it is very much a 'mayto-'matho situation. I used the EVF with performance boost on and it was very nice, just make sure you set the brightness to your liking. I set it to manual default value, but the shots turned out to be slightly underexposed. Comparing the X-H2 to X-H1 bodies, disregarding the traditional vs PASM controls layout and tech inside, I feel like the X-H2 is a downgrade. I prefer the larger grip, the soft shutter button, the clickable rear dial and the overall higher robustness of X-H1.
Conclusion
Resolution-wise, the sensor is excellent and considering an X-T5 at £1699 is £550 cheaper than A7R IIIA on high street at the time of writing, it is easy to find its appeal beyond retro controls and nice sooc colours. I suspect Sony will discontinue A7R IIIA shortly as there are already 2 newer generations in shops, so the nearest rivals for resolution will be Nikon Z7 II and Panasonic S1R at just under £3k, making the 40mp Fujifilms even more attractive. For the money you save, you are giving up the usual stop of dynamic range and high ISO performance compared to full frame sensors, however, it can buy you an extra lens, accessories or a nice photo trip. In landscape photography, GND filters and exposure bracketing where possible can close the gap even further.
Gear reviews are much more work than I expected and even if this one isn't perfect, I had fun doing it. Once I receive my deposit back, I will loan a GFX 50s II with GF 32-64/4 and give it another go!
X-H2 sample images (there are some photos where I was pushing IBIS or DR on purpose, so check file names):
Comments
Post a Comment