Anecdotal Evidence: Reviewing AF-C on the Fujifilm X-T2

On Sunday, I decided to give my X-T2 its first test of the ability to continuously autofocus during high speed burst shooting by attending the PAC-12 North Beach Volleyball Invitational Tournament held at Alki Beach in Seattle. As I plan on using this camera for a few airshows and motorsports events this season, specifically Seafair and motorcycle road racing, knowing what the Fujifilm X-T2 can, and can’t, do while shooting 11 FPS bursts is sort of important to me. My needs are quite simple and easily achievable with a professional-grade dSLR like the Nikon D5 or D850, but based on all of the blogs reviewing the X-T2, it may not be so simple. So, I went out to find out for myself.

Burst speed is mostly useless if you miss shots while waiting for images to be written to the memory card. The X-T2 has dual UHS-II SD slots, and with a proper pair of cards, is able to clear the buffer long before you press the shutter button again.

This is in no way scientific. This is just a consumer testing the capabilities of their consumer purchase in an environment that meets or exceeds the conditions of the intended purpose. Road racing, and acrobatic aerial demonstrations for that matter, have far more predictable movements due to physics at speed. Velocity dictates the equation and makes it easy to predict where the subject will begin and end to make it easier to frame a shot. However, variability increases as speed slows and forces like friction and deflection come into greater play, reducing predictability. Volleyball, like any other ball sport, has to be one of the least predictable sports to follow with a lens because of all these factors. Based on my assumption, if my X-T2 can work well shooting a volleyball match, it should work just fine for my intended purpose.

Because I didn’t want to bother with asking permission to shoot around the event, I shot as a spectator, limiting my framing and compositions. Because of the weather, I didn’t plan to stay long… just long enough to get truly familiar with my camera, it’s capabilities and a some shots I could review under Lightroom. I felt this would give a more accurate perspective as a consumer using this camera as a spectator, but more importantly, not be a nuisance since I wasn’t there to get paid and the last thing I wanted to do was possibly get in the way of the action.

To prevent warping of fast moving subjects, I chose to shoot with mechanical shutter only. To achieve a maximum 11 FPS burst rate with minimal time between bursts, I attached the Vertical Power Grip, selected “Boost Performance Mode” with the switch and chose “11 FPS” in the drive settings. Otherwise, you’re limited to 8 FPS high speed burst with mechanical shutter or 14 FPS with electronic shutter. For minimal time between bursts, I used Sony’s G Series SD cards, which utilize the UHS-II high speed bus and 299MB/s peak write speeds that reduce write times between long bursts significantly compared to UHS-I cards. I used a pair of these cards and set the camera to write uncompressed RAW to one and Fine JPEG 3:2 Large to the other simultaneously.

Never once did I find myself waiting on the camera to write to the card before I was able to fire off another burst. Simply put, these SD cards were fast enough to keep up with all the action.

Armed with my 50-140mm ƒ/2.8, Fujifilm’s 70-200mm professional equivalent, I found a good spot to sit on the concrete steps along Alki beach and began tracking the action with my EVF. The EVF’s 100 FPS framerate easily kept up with the scene with no ghosting, tearing or stuttering on the display. Movements were smooth and true to life, causing no dissonance in my brain as I shoot with both eyes open. Using the widetracking option, the camera seemed to be following the players easily as I watched the green boxes lock on, and stay locked on, to my intended subject and stick like white on rice.

My only complaint would be the slight “screen door” effect, something Fujifilm should fix in the next iteration with a higher resolution EVF. This is something the X-H1 already features.

Before firing off my first burst, I decided to double-check my settings. Front switch to AF-C, drive mode to High Speed Burst, Performance Mode to Boost, AF mode to Widetracking, OIS on. It was then I decided to adjust the AF-C Custom Settings. Normally I leave it on Custom Set 1, but I decided to set it to Set 6 and customize it after seeing the system get confused by the stationary net refs as I followed the ball. I put all settings to far right with center bias; this kept the AF on it’s toes as it would expect erratic movements from my subject while ensuring the lock didn’t stray toward a ref entering the frame or a player crossing between my lens and subject. Another half-press with the shutter release on a background player during a game showed the changes I made were an improvement.

Fire at will! After the first few bursts, I came to realize that action sports aren’t my thing and I was out of practice when trying to follow a subject with unpredictable movement. I had no ability to predict placement when zoomed in, something sports shooters typically do. Be it my ignorance of volleyball, the cold, damp weather or impatience, I was mostly following a split second behind the action with my lens, capturing shots just after they happened. Other times, I would try to predict the action and end up firing off a long burst of nothing and ending the burst a split second before the action I intended to capture happened. I tried not to focus on my incompetence and began to aim for achievable goals, like capturing serves and ace celebrations. To mix it up, I also threw out composition and aimed at the net for some action.

While slower burst speeds will let the more talented shooters still capture the intended moment in time, higher bursts ensure a higher hit rate for less experienced shooters. Many bursts managed to catch the split second a hand made contact with the ball and it’s deformation, a challenge for even the best photographers limited to Single shot, AF-S.

531 images later and with a rain shower on its way, I decided to pack it up. Normally I’d use the camera’s built-in wireless transfer to copy the images from my camera to my iMac but fortunately I had the presence of mind to forgo wireless transfer this time. Taking the cards out of the camera, I used the iMac’s SD card slot to transfer the images for Lightroom and back them up. 21.35GB later, I was glad I’d made that choice. It also exemplified why I should switch to using compressed RAW, especially when there’s little to no adverse effects on quality observed to this point, unlike Sony’s compressed RAW format.

What I discovered as I went through the shots in Lightroom was that the hit to miss ratio was extremely high in the X-T2’s favor. I did get a single burst where the whole series was out of focus but that was clearly my fault as I shot it unprepared for the action and had aimed at the sky. Tossing those, I found the camera’s hit rate to exceed 93%. However, keep in mind the conditions: midday, cloudy skies. Your camera settings also impact the results, so take the time to properly adjust autofocus settings to match the conditions you plan to shoot in. For me, the camera did exactly what I asked it to do and did it consistently. This allowed me to predict my shots and also predict how the camera would respond. Available settings allowed me to customize the camera’s response to a certain degree, further reducing the effort required from me while shooting in a fast-paced environment. The results spoke to me and made clear that even in an environment that’s more challenging than what I primarily intend to use it in, the Fujifilm X-T2 is far more capable than many bloggers had led me to originally believe. Continuous AF isn’t perfect in the X-T2 and there are cameras on the market that are better suited for this type of work, but the X-T2’s continuous autofocus ability is benefitting greatly from Fujifilm’s constant stream of kaizen product improvement. So, maybe AF-C wasn’t very good in the beginning of its product cycle, but it seems to be working excellently now.

There is of course a caveat to all of this, including obviously that your experiences may not completely reflect mine. The first requirement is to be realistic by tempering your expectations to keep them within the realm of reality. No camera was built for you; they are built for many and with the many options available, there is likely one out there that better meets your expectations… as long as you’re being realistic.

Please note that I’m “old school.” What that means is that my experience is shaped by my prior ownership of older digital and film cameras, so, my abilities today reflect decades of adapting to those limitations inherent in older, limited designs. That said, my ultimate impression of the X-T2 will also be shaped by those experiences. Predictive, continuous autofocus is ever evolving but has never been as good as it is today. Many years of adapting to shortcomings has made me develop techniques that probably influenced my hit rate in the positive, something that may not be true for you. Keep in mind that a younger, less experienced user, or a less patient one, may have a negative experience in this regard as no autofocus system is capable of mind reading. A rational user should hope for, and expect, a system that’s predictable and makes sense to them, preferably tunable, so that the limitations are less limiting and allows the user to still create the desired results consistently. There are some cameras that are better at continuous autofocus and many that are far worse. This post is simply my experience with my X-T2 after taking the time and effort to properly set it up for the task I had set out to perform with it. It’s definitely not perfect, but it’s far better than the loudest voices will have you believe. So, if you’ve had a poor experience with your X-T2 in this regard, maybe it’s just you.

Anyways, now for some of the examples taken from that day: