First impressions of the Fuji X-T2

I took the X-T2 out for its first photos yesterday to a favorite spot for this kind of testing: the sewage settling ponds on Radio Road in Redwood Shores. This is a great place for this kind of work because I can guarantee both close and distant birds, birds in flight, airplanes on landing pattern to SFO, and typically, clear skies with some humidity and haze. I went in the middle of the day, or course sticking the camera with harsh shadows and glare to deal with, because I assume if a camera can do decent work in poor light, it’ll kick tail with great light.

These are all screen captures from my Lightroom session as I imported and evaluated them. I’ve done nothing to the images; these are raw images straight out of the camera. The camera was overall in factory default state, using ISO 1600, Aperture mode, fixed aperture (primarily F8) and automatic white balance. I had exposure compensation set to 0 and didn’t touch it or any other in-camera adjustments during the shoot, so these images are as the camera created them with as few tweaks as I could manage, and no sharpening or lightroom adjustments. The camera was hand-held with image stabilization on.

Oh, and this area is full of very dark brown ducks leading to good tests for pulling detail out of shadows, and both white and white/black birds — the Black-Necked Stilt is nature’s own dynamic range test device.

A quick technical note: RAW images out of the camera are 50 Megabytes; in Lightroom, converting them to DNG reduces that to 39 Megabytes. I was able to fit 620 images onto a 32 Gigabyte memory card; that’s many fewer than you could fit on that card with an X-T1 by about a third. I forsee 64GB cards in my future — and remember, larger RAW images require more processing power on your computer, so if you’re system is older or slow, it might become untenable with these bigger sensor images without an upgrade. Just something to consider.

My General Impression

Here’s a sample image of a gull resting on one of the islands in the pond. These birds are almost a pure white with neutral grey feathering, and depending on the species, some deep black highlights. Almost a grey card with wings.


I’ve blown this up to 100% (aka 1:1 pixels). And remember, there’s no sharpening applied anywhere, and this is a raw image, so no sharpening in camera, either. That, to me, is a very sharp image before sharpening. There is a bit of luminance noise visible on the monitor, but it’s less than I see in similar setups on the Canon 7DmkII. To me, the colors look spot on.

For another look at color, here’s a Southwest Jet flying by:


You see some haze in the air, easily fixed in Lightroom. There’s again some luminance noise in the sky, but minimal. The colors look spot on and nicely saturated but not overdone. There’s none of the yellow/warm tone bias you see on Canon sensors.

Here’s our Black-Necked Stilt:


Look at the histogram: with both extreme blacks and extreme whites in the image the camera stuck the exposure square in the middle without highlights being blown or blacks being crushed; stop and think about just what the dynamic range of the bird is here, and it’s handled by this camera with room to spare. I don’t think I’ve ever seen that with a digital camera before.

This camera doesn’t blow highlights or crash blacks

In examining the histograms, I noticed something interesting; the camera never — not once — blew out a highlight or crushed the blacks. Every image with wide dynamic ranges got exposed for the middle of the histogram with room to spare on both sides. Images with less extreme dynamic ranges seemed to get a tweak pushing it as if I’d adjusted things to encourage ETTR (exposure to the right) — except I hadn’t.

It really looks to me like the camera is doing some in-camera dynamic range compression when necessary to give you a RAW image that has all of the information needed to process it. I’m sure I can set up a situation where it has to blow out highlights, but short of manual mode, it looks like the automated exposure modes have been programmed to avoid exposure problems that might affect later processing.

I’m impressed. This is a lighting and subject situation that can drive some aperture-mode cameras crazy, and I got 100% good photos in a 600 image shoot without a single blowout or crushed black. Stop and think about that for a second.

Fuji is doing something really interesting here. They’ve programmed the camera to handle nasty exposure situations for you, and it seems to do a good job. I want to do more testing, but I think this is huge; a huge improvement in automation that’ll help all photographers (except those that insist the one true way is manual exposure…)

Here is a photo where I came as close to blowing out highlights as I did; what happened was I started shooting a burst with the camera away from the white birds, which locked in exposure without factoring in the big blobs of white, and than panned over. And if you look at the histogram, we still didn’t blow out highlights: the Fuji did some kind of magic and pulled them back from being damaged, although there’s a bump right near the edge of the histogram.


Here’s that image after a couple of minutes of basic processing, a bit of chromatic abberation removal (overall, the 100-400 is quite good about that), and a basic sharpening and noise reduction pass.


That’s one baby step beyond just pushing the Auto button, and took me about three minutes.

Battery Life

It’s too early for a real feeling for battery life, but on this first shoot I took about 600 images, many in burst mode (and fast burst is really fast), and at the end of it, the battery indicator had bumped down one notch. On the X-T2 the indicator is upgrade and now has five segments, and it seems to be more linear than on the X-T1, which had a tendency to wait a long time to show the first notch go away, and then have the battery run down quickly from there. If I’m right that this is more accurate about charge, then I’m guessing 1500 images for a battery is likely, but I need to test in the field. My sense is battery usage is improved in this camera.


One of the things that attracted me to this camera was the promise of improved autofocus. The X-T1 autofocus system is pretty good, especially once you practice with it enough to know how to use it. Overall, I’d put the X-T1 AF ahead of my old, venerable Canon 7D but not quite as good as the 7DmkII.

Fuji AF has three modes:

  • Manual, where it connects into the AF-Lock button on the back and lets you trigger it on demand; this is very similar to Canon back button focus. When in this mode, you get a programmable focus peaking on the screen that once you start using it, you wonder how you took photos before you had it.
  • Single, where it’ll lock on once when you press the shutter and stay at that point through the burst
  • Continuous, where it’ll lock in at the start, and then track focus through the burst.

As a long-time Canon shooter I swore by back button focus, but I found on the X-T1 I couldn’t reliably lock onto a flying bird with it, but I had a lot better luck with both Single and Continuous (single for birds flying parallel to you, continuous for approaching and leaving). On the X-T2, all three modes worked rather well for moving objects and almost perfect for stationary ones.

Autofocus can be configured for wide field sensing across all sensors, or for region sensing, or for single point autofocus. The size of the region and the location of it and the single point are all movable across the image, and the controls to do this are all on camera (no menu diving!) and easy to manipulate with a bit of practice.

What I found in early testing was very reliable AF even when I tried to confuse it across all three modes. In terms of focusing speed, Manual was fastest, Single was next, and Continuous was slower but not terribly, which is what I’d expect. Full field AF was slowest, region was next, and single point fastest.

How you use these is up to your subject and style, but I was finding I was getting best results in single/region or point AF, depending on situation, but you risk having your subject move out of your sharp depth of field. I’m not yet sure if I’ll prefer Continuous for that situation, or learning to release/press the shutter across a longer burst to use single to refocus each sub-burst. I’m currently leaning that way, but time will tell, but I’m rarely in a position where I need to refocus mid-burst and can’t interupt capture, so that may be what I’ll lean on.

In Summary

What I’m seeing here in general is that the automation tools for both Autofocus and Exposure are much better, and therefore reduce the number of situations where you need to go into manual mode to solve a problem are significantly reduced. That means you can focus on the subject and framing instead of the camera logistics, which I think will lead to better pictures all around. Manual mode isn’t obsolete and the X-T2 supports it well, but once you learn to trust the camera, you’ll likely find yourself in it less and less.

Overall, the camera feels very comfortable in my hands; if you’ve used an X-T1 you’ll feel right at home here. I think the autofocus systems are much improved over the X-T1, and well beyond what my 7DmkII can do; I think you would have to head into the 1D to match this autofocus system (it’ll be interesting to see how the 5DmkIV stacks up). The exposure systems just blew me away, and the more I looked at my results, the more magical it felt. It just refused to give me a bad exposure and I haven’t even started tweaking exposure compensation much less some of the deeper customizations. Truly blow-away.

Response is fast and reliable. AF is fast and reliable. The camera felt snappy and fast; fast burst (up to 10FPS) is really fast and I will likely rarely use it in my work because it’s almost too fast for me. You still do fill up the buffers after a while, but this camera is set up to get you the shot and make you happy with it, and to me, it exceeded the rather high expectations I had for it.

And as a bird/nature/landscape/critter photographer, it’s the first time I’ve felt — shooting with the Fuji platform — like I was shooting with no compromises. Not only no compromises, but I think the improvements in AF and exposure on this camera might make it the camera to beat in the market.

Ask me about that after another few thousand images and a rain storm or two…

But this camera is pretty awesome.

A few more images

Here are a few more images, all unprocessed, so you can get a sense of the detail and look at the histograms. I can’t think of a camera under automated exposure management that gave me this good an exposure on every image, every time. Seriously: in 600 images taken during this test, not one exposure ding. Now, to be honest, the light wasn’t changing at the time, and I want to see how this handles during low light and bad weather and with wildly changing light, but already, this camera is giving me a lot of confidence in it — without needing a lot of operator fiddling to make it happen.






By the way, any time you see me photographing the pelicans, I was shooting through some brush to see how the autofocus handled cluttered foregrounds, and it always locked on the birds and never one tried to focus on the grasses. Here’s a look at it with the zoom pulled back; this wasn’t heavy brush cover, but it’s enough that most autofocus systems would trip on it regularly, and this one handled it perfectly.


And yes, standing behind a bush to take pictures through it is the sort of thing you do when you’re testing, right?