Autofocus and performance

The a7R IV is more than responsive enough to capture this image of a man regretting his decision to feed a seagull from so near a distance.
Out of camera JPEG | Sony FE 24-105mm F4 @ 24mm | ISO 100 | 1/320 sec | F4
Photo by Carey Rose

The a7R IV comes with an updated autofocus system, comprising a total of 567 on-sensor phase detection autofocus points that cover 99.7% of the frame vertically and 74% horizontally. It also has Sony's latest implementation of Real-time Tracking AF, seamlessly moving between subject tracking and face-and-eye-detection.

The autofocus system generally performs well, especially compared to the older a7R III, and Real-time Tracking is a blessing for those that regularly photograph people. However, we did find some situations in which it would trip up, as well as some persistent issues with autofocus precision during the course of our testing.

Key takeaways:

  • Real-time Tracking continues to impress with ease-of-use and effectiveness
  • Incredible autofocus accuracy while shooting single drive, but accuracy drops during burst shooting
  • It's best to switch out of tracking in very low-light or backlit situations
  • Even slight mis-focus is made more obvious by the very high 60MP resolution
  • Burst shooting speeds drop noticeably when shooting uncompressed Raw
  • UHS-II slots do offer a speed benefit, but large uncompressed Raw file sizes mean the buffer still takes a while to clear

In-depth

The autofocus implementation in the a7R IV is the same one that made its debut in the a6400, and we continue to be big fans. Click here for an in-depth look at the system and how we'd set it up.

Basically, the autofocus system appears daunting at first glance, but because of how good the camera's tracking is, you can set it up one way for almost all types of photography, and just never change it. We generally leave the camera in AF-C or continuous autofocus, with 'Tracking: Flexible Spot M' as our autofocus area. You can place that area over any subject you like, initiate focus, and it should track that subject tenaciously.

Now, let's take a look at how the system performs.

Autofocus system performance

To test continuous AF performance, we first try to shoot a subject approaching at a steady speed using the central AF point. This lets us see how good the camera is at assessing subject distance and whether it can drive its lens to that point quickly.

Straight-on
Weave
All images captured using the Sony 70-200mm F2.8 GM at 200mm.

We then have the subject weave across the camera's AF region in a way the camera can't predict. This has the advantage that the approach rate varies as the subject changes direction. For this test we use the camera's Real-time Tracking feature by enabling one of its tracking modes, so it needs to identify and follow a subject around the scene, as well as trying to keep it in focus. In both situations, we placed the autofocus area over the reflective vest.

Both of these rollovers show the results using the maximum burst speed of 10 fps, but you can expect similar results if you slow down to 8 fps to get a live feed between shots. Basically, the camera performs acceptably, if not quite in line with the high standards we've come to expect from Sony's lower-resolution models.

These two runs are broadly representative of our findings, meaning that each and every run - both straight-on, and weave - contained an amount of slightly out-of-focus (and usually back-focused) images. It probably won't be a problem if you were to downsize, but it's certainly noticeable if you're looking at 100%, as we are here.

Real-world tracking

We did notice the a7 IV would struggle sometimes in very low lighting situations, as well as in brighter situations with strong backlighting. Basically, the camera will continue to draw a box around your subject's face or eye, but the focus is somewhere behind the subject. You can see this in the below rollover, with two images taken moments apart with tracking being initiated for both. Or in this backlit example.

In-focus Out-of-focus

On the other hand, in the vast majority of situations the camera seemed to perform exceedingly well, even with distant subjects, or very shallow depth-of-field images from very fast lenses. With a bit of work to get the camera set up, you can also make sure that you can access other autofocus modes with either a custom button, or have the shutter half-press and AF-ON buttons enable separate AF functions, for example (see this video tutorial). In any case, we don't expect most users to have serious issues with the a7R IV's autofocus performance.

This image was taken using the a7R IV's tracking mode, and resulted in perfect focus, even at 60.2MP.
Out-of-camera JPEG | Sony 135mm F1.8 GM | ISO 100 | 1/5000 sec | F1.8
Photo by Rishi Sanyal

Aperture Drive in AF

Sony has a menu option called Aperture Drive in AF that changes the behavior of the lens iris as the camera focuses. In 'Standard' mode the camera acquires focus wide open (up to but not beyond F2) in AF-S, but stays stopped-down to the shooting aperture in AF-C. 'Silent Priority' forces the camera to always focus at the shooting aperture, to eliminate the sound of the iris opening and stopping back down during focus acquisition. Our concern with both these modes has - for years - been that focusing performance suffers in AF-C if you shoot stopped down, particularly in low light. It deprives the sensor of light and reduces the phase differences that the focus system relies on to calculate the amount of refocus required.

The a7R IV purportedly gains the option to force the camera to focus wide open in AF-C. This behavior is enabled by selecting the new 'Focus Priority' option under 'Aperture Drive in AF'. We've tested it, and its behavior is a bit different than Sony describes. The camera only opens up the aperture as much as it feels it needs to based on light level. The aperture stays open to this value, not only during focusing but always, so your depth-of-field preview may not be accurate to your shooting aperture.

In very low light... focus can be significantly faster

In broad daylight any gains are marginal, largely because the a7R IV tends to not open up the aperture much beyond your shooting aperture, and because the camera is now even better than its predecessors at focusing quickly despite a smaller aperture. In very low light, though, where the camera chooses to leave the aperture open (up to F2) focus can be significantly faster, a very welcome change.

Unfortunately, there's a downside: shutter lag. Sony mirrorless lenses stop down their irises far slower than conventional DSLR lenses (for reasons that elude us), so the smaller your shooting aperture, the longer the delay between shutter press and image capture.

We would need to conduct more varied real-world low light shooting to determine if the a7R IV generally opens up the aperture enough in this mode, or whether it would further benefit from always opening up the aperture to F2. But we also understand Sony's approach: it's trying to minimize slow iris actuations. What we'd like to understand, and see fixed, is the (painfully) slow iris actuation in the company's mirrorless lenses.

We'd recommend you enable 'Focus Priority' if you're finding AF to be slow in dim conditions where you wish to shoot stopped down to, for example, maintain focus on groups of people or give the focus system some 'wiggle room' for faster moving subjects. Just keep in mind the added shutter lag.

Performance and UHS-II cards

As seems to be the case with each new Sony camera, we find the a7R IV to be fractionally more responsive to your inputs than the third-generation Alpha cameras. This means that navigating menus or adjusting exposure settings happen just a bit quicker than before, but unfortunately, there's still more lag than there is on several competing cameras.

We are especially happy to see Sony's adoption of dual UHS-II SD card slots on the a7R IV, since our findings show that the faster cards do indeed make a measurable difference in how quickly the camera can get its immense files stored.

Here's a breakdown of buffer depth and write times compare between a Sony G-series UHS-II U3 card and a SanDisk Extreme Pro UHS-I U3 card.

Quality Burst rate (max) Buffer depth Buffer clear time
Uncompressed Raw + Xtra Fine JPEG ~6 fps ~29 shots

UHS-II: 27 sec

UHS-I: 54 sec

Compressed Raw + Xtra Fine JPEG 10 fps
(Forced 12-bit readout)
~61 shots

UHS-II: 55 sec

UHS-I: 68 sec

Xtra Fine JPEG 10 fps ~63 shots

UHS-II: 55 sec

UHS-I: 55 sec

Standard JPEG 10 fps ~58 shots

UHS-II: 27 sec

UHS-I: 27 sec

No surprise, it's with the largest files that the UHS-II card really makes sense. But we are a bit disappointed that decreasing the quality past a certain threshold doesn't give you increased buffer depth, nor faster write-times from the faster cards, indicating there's a bottleneck elsewhere. Still - these are more than respectable numbers for a camera firing off 60MP files which can be as large as 117MB on disk.

And don't forget, as we mentioned earlier in the review, shooting compressed Raw with any burst speed drops your files to 12-bit, further limiting the degree to which they can be manipulated.