Sony A6000 Field Test Part II

Sony A6000 Field Test Part II

Focusing on a world in motion

by Eamon Hickey | Posted: 05/23/2014

In this shooter's report installment I'll backtrack a little and talk about the Sony A6000's feel and handling, and then move on to a deeper discussion and real world testing of the highly touted AF performance.

Handling. When I first unpacked the Sony Alpha 6000, it felt immediately familiar. To me, its shape, size, and weight seem similar to the NEX-6 and NEX-7. That's also true of the way it feels in my hand, and that's a good thing. I find all three models very easy to hold securely, even with just a couple of fingers. I've now carried the A6000 for more than 15 hours on 12 different outings, and I have no complaints at all about its portability and comfort. As we noted above, it feels reasonably solid, but I felt a little cautious about torquing it too much when the large Sony FE 70-200mm F4 G OSS lens was mounted. [Note: Image comparisons beside the NEX-6 can be seen in our A6000 overview.]

In my first hour with the Sony Alpha 6000, I spent a fair bit of time checking out the buttons, dials, and menus. The control placement works very well for me. I can quite easily reach the AEL button (which I use for autofocus) and both control dials with my thumb, and the custom and function buttons are also readily accessible, meaning I can make fast settings changes. We mentioned the new menu system in our overview, and I'll just add that it's as big an improvement in use as you would hope. I'd give a lot to get my NEX-7 updated to the new menu layout.

Autofocus. In part 1 of this report, I noted that the Sony A6000's single-shot AF (AF-S) is lightning fast, as Sony claims. I found it quite quick and decisive even in moderately low light when I shot some pictures of a friend in a restaurant. In extremely dim light, the AF did slow down, taking half a second or a second to focus sometimes, and it failed to acquire focus on some low contrast subjects in that light. But in every other circumstance where I've used AF-S, it's been terrific.

The headline news on the Sony A6000, however, is its claimed improvements in continuous AF (AF-C) on moving subjects. To test those claims, I've shot almost 2,500 photos in more than 150 burst sequences of flying birds, dogs at play, soccer, flag football, sprinters, joggers, and bicyclists in light ranging from bright daylight to dark overcast. I happen to also be reviewing the Nikon D4S, which I shot side-by-side with the A6000 as a kind of benchmark to measure it against. On the spec sheet, they both claim to be able to do continuous follow focus at 11 frames per second. [Please read the important caveats and qualifications that I note at the end of this section.]

After some initial warm up shoots (the birds and the dogs were mostly a bust, but I learned a lot about the camera's AF system), I took the Sony A6000 out one afternoon to a soccer field in my neighborhood. It was fairly dark overcast -- about EV 8 or 9 (ISO 100). I shot everything with the FE 70-200mm lens, using f/4.0 in order to give the shallowest possible depth-of-field and thus better reveal focusing errors. The burst speed was set to Continuous Hi (11 fps) and Optical Steady-Shot was on (mode 1).

As I shot, I suspected that both the Wide and Zone AF area modes covered too much of the frame for soccer; they were being fooled into focusing on players who were behind or beside the player with the ball. I had Focus Lock-On activated, which is supposed to track a subject's position within the frame, passing it from focus point to focus point as it moves across the frame. It works pretty well when there is an individual object against a distant background, but it didn't seem to help much in the chaotic action of soccer. The Center AF area mode was fooled less often, but I had more trouble keeping it on target.

When I picked up the Nikon D4S to shoot with it for awhile, I noticed something else that seemed very odd. I was seemingly better at predicting the action with the D4S -- I was catching more peak moments. My strong sense is that the electronic viewfinder of the A6000, with its slight delay, was causing me to react late when the ball was passed and the action shifted. The optical viewfinder of the D4S has no equivalent delay, but on the other hand, the camera does blackout the viewfinder briefly for every frame. Whatever the precise cause, there was no question that I was slightly, but noticeably, faster at following changing action with the D4S. It's of no real concern to me overall, but if shooting sports for money was in my game plan (it's not), this would drop the A6000 off my shopping list.

When I examined my pictures closely later, I saw that, yes, the Wide and Zone AF area modes were often fooled into focusing on the wrong subject. My first shot or three in any given sequence were also often not really sharp, although the camera would usually catch up in later frames. To tally up my approximate hit rate -- i.e. sharp shots vs. soft shots -- I didn't count aborted sequences or ones where I knew or strongly suspected that I had messed up. With those caveats stated, my hit rate in typical properly executed burst sequences was roughly 50%.

It's important to understand that number in context. A 50% hit rate at 11 frames-per-second works out to 5 or 6 sharp pictures per second. That's pretty good, and comparable or better than what most DSLR cameras in the A6000's price range can do.

Let me also note another important thing at this point: judging sharpness is partly subjective. I'm fairly sure I have a stricter standard for sharpness than many photographers. It's also important to remember that whether a photo is usably sharp depends a great deal on how large you plan to display it. A picture that is not pin sharp -- and would get a failing grade from me for this report -- will probably look fine in a smaller print or on Facebook. See the collage image of the runner that we've posted here for a sense of how I graded sharpness.

A few days after my soccer outing, I took a walk to East River Park and found a high school girls flag football game in full swing. The light was good -- sunny late afternoon, or about EV 14 (ISO 100). As before, I shot everything with the FE 70-200mm lens, wide open at f/4.0. Here again, the Sony Alpha 6000's available AF area modes were not ideal: both Wide and Zone area modes could be fooled by objects (i.e. players) I didn't want to focus on, and Center area was a little difficult to keep on target. When I looked at my images later, I tallied up 130 frames and a hit rate of around 40-50% in typical burst sequences. (I'm giving a rough number for hit rates in this report because even in a decently executed burst some misfocused images were probably my fault, not the camera's, and because the sharpness of many shots is arguable.)

Multi-player sports like soccer and football are where the smartest AF area modes that you see on higher-end (and some mid-level) DSLRs, and the sophisticated software algorithms that underlie them, are a big help. For both the soccer and the football, I used the smaller Dynamic AF area modes (d9 and d21) on the Nikon D4S, and it was much better at grabbing the player I wanted it to grab and then sticking with them. Because the action is so unpredictable, these sports also put a premium on initial focus acquisition -- i.e. when you swing the camera quickly to re-aim on a new point of peak action you want that first or second shot to be sharp. With the A6000, the first 2 to 3 frames of these sequences were frequently somewhat soft.

After the football, I found some sprinters, joggers, and bicyclists to shoot in other areas of East River Park. Here the Zone AF area mode on the Sony A6000 worked well (no other players to distract it). The position, pace, and direction of the subjects were also more predictable. In total, I photographed almost 20 burst sequences of these subjects, with several comprising 35-40 frames. On these subjects, my hit rate in typical sequences was a bit higher on average -- ranging from 50-65% or so, burst to burst.

Here again, the first few frames in a burst sequence were often soft, even when I had carefully pre-focused for a second or two on the approaching runner or cyclist. I also noticed that the sharp and soft images seem to come in clusters. The camera will get 2 or 3 sharp in a row, then 2 or 3 soft, then back to sharp again. The collage image of the runner gives a sense of this. In a fairly small minority of my burst sequences, I saw 7 or 8 consecutive soft frames before the camera recovered. In a perfect world, I'd rather see a more random distribution of sharp and soft, or quicker recovery to sharp, so as not to have peak moments pass by with, say, 5 consecutive soft images.

A quick side note: While shooting the bicyclists on the promenade at East River Park, I twice had a chance to put the custom settings memory function to work, both times when I looked up to notice ships passing on the river. As I jumped up from my bench and made the three-second dash to the promenade railing, it was very quick to switch from my "sports" AF-C, JPEG-only custom setup to my "standard" AF-S, RAW+JPEG setup and get the shot. I praised this feature in part 1 of this report, and this is the reason why.

On two later days, I went back to East River Park and shot another 70 sequences of soccer, sprinting, and bicycling to test some additional variables. Somewhat to my surprise, switching the drive mode to Continuous Mid, which shoots at about 6 frames per second, did not seem to affect my hit rate noticeably -- it remained about 40-60%, depending on the sport, yielding 2-4 sharp images per second. It also made soccer action, which requires frequent rapid re-aiming of the camera, harder to track (though it did not have that affect on sprinters and bicyclists).

So, in my tests, the faster Continuous Hi mode had a clear advantage, giving about twice as many sharp images in any given time period. Switching OSS off seemed to drop my hit rate modestly. I also tested Center AF area mode on sports that don't have multiple players (sprinters and bicyclists), but my hit rate was about the same as with the Zone AF area mode. I should also note that in my collection of more than 150 bursts, there are some outliers with no obvious explanation: two or three where the hit rate was 75% or better, and, conversely, maybe 8 or 10 where the hit rate was below 25%.

So what's the executive summary? My experience with the Sony Alpha 6000 is that it did a pretty decent job with multi-player fast action sports, but it would benefit from more sophisticated AF area modes and better initial focus acquisition. Still, I got many bursts of 8 to 12 images with 3 to 6 sharp shots. Not bad at all. For these sports, I would use Center AF area mode.

For action with fewer obstacles and steadier motion like running and bicycling, Zone AF area works fairly well, and the camera's hit rate went up a bit -- to about 50-60% in typical sequences -- although the first few frames of any given burst were still often soft, even with pre-focusing. But again, overall, this is very respectable for a camera at this price.

I won't belabor this, but I know many of you are curious, so I'll say that my success rate with the Nikon D4S was much, much higher. Its Dynamic AF area modes are much better for multi-player sports; it nails the first image nearly always; and on steady-state motion like bicycles, it almost never misses. But remember that the D4S would easily best a $650 Nikon DSLR, too. That the Sony A6000 can deliver 4-6 sharp images per second on a reasonably regular basis is very impressive for a mirrorless ILC camera costing 1/10th the price of the D4S.

Now some caveats you should keep in mind: Remember that I am one photographer. I have 25 years shooting experience, including action and sports with pro-level AF cameras and lenses, but I am not a pro sports photographer. The pros are a lot better than I am, trust me. I was working with one lens; other Sony lenses are undoubtedly different than the FE 70-200 (some will definitely be slower, others might be faster). I didn't test every sport or action subject and all have their unique characteristics. Even basketball and soccer differ from each other. Crucially, I didn't get to test indoor sports, where gym lighting can put extra stress on an AF system. [Editor's note: Our reviewer Dave Pardue shot the A6000 with this lens at an indoor hockey game and reports similar results with burst mode success -- citing that the first few frames were generally soft, and he got a success rate of ~35-45% of frames being sharp on average. A few of the sharper images are on display in our A6000 gallery.]

For installment 3 of this user report, I'll let go of my AF obsession (2500 frames? What was I thinking?), and look at the rest of the A6000's resume. In the meantime, be sure to add your comments and questions below.