Video Autofocus Tracking: How to Compare Continuous AF

How continuous autofocus works in video

In still photography, autofocus can lock, refocus, and lock again between shots. In video, the camera must continuously adjust focus while recording, because the subject or the camera is usually moving. This is called continuous autofocus, sometimes labeled AF-C or AF-F. The camera reads the sensor feed, estimates the distance to the subject, and drives the lens to keep that subject sharp as the scene changes.

The speed and smoothness of this process depend on three things: the autofocus sensor or phase-detect points on the imaging sensor, the processor that calculates focus direction and distance, and the lens motor that moves the glass. A slow lens motor or a processor that hesitates will cause the focus to lag behind the action, which is visible as brief soft periods in the footage.

For a deeper explanation of the different autofocus technologies, see our article on autofocus systems.

Video autofocus is also affected by the frame rate and resolution you record. Higher resolutions like 4K or 6K produce more data per second, which can increase the processing load. In practice, most modern cameras handle 4K autofocus well, but entry-level models may show slower response at higher resolutions. Check the camera's specs or the manual for any notes about autofocus performance in different recording modes.

• Continuous AF reads the sensor continuously during recording
• The processor estimates subject distance and drives the lens motor
• Focus smoothness depends on the lens motor and the camera's algorithm
• Higher resolutions can increase the processing load

What to look for in specs

Manufacturers often list autofocus specifications in terms of the number of phase-detection points, the autofocus sensitivity (measured in EV), and the subject recognition modes. Here is what each of those means in practice.

Phase-detection points on the sensor cover a certain area of the frame, called the autofocus coverage. A camera with wide coverage can track a subject that moves across most of the frame. Coverage is often stated as a percentage of the frame width or height. A camera with 90% coverage will keep a subject in focus even when it is near the edge of the frame.

Autofocus sensitivity, listed as a negative EV value, tells you how well the system can focus in low light. A rating of -4 EV means the camera can focus in very dark conditions. If you shoot indoors or at night, look for a camera with a low EV rating. For a full explanation of these ratings, read our autofocus sensitivity EV article.

Subject recognition is the feature that lets the camera lock onto a face, an eye, an animal, or a vehicle. The camera identifies the subject and keeps the focus point on it, even when you pan or when the subject moves. This is the single most useful feature for video, because it reduces the need to manually adjust the focus point while recording.

The number of autofocus points is less important than the coverage and the recognition algorithms. A camera with 400 points and 90% coverage will track better than a camera with 1000 points and 60% coverage. See our autofocus points and coverage article for more on this distinction.

Lens compatibility also matters. Some cameras can only use their fastest autofocus modes with native lenses. Third-party lenses may work, but at reduced tracking speed. Check the lens mount compatibility before you buy, especially if you plan to adapt lenses from another system. Our lens mount guide explains the differences.

Tracking performance and its limits

Even the best autofocus system has limits. A subject that moves directly toward the camera is harder to track than one moving across the frame, because the required focus distance changes quickly. A subject that is small in the frame, like a distant bird, is also difficult for the camera to detect reliably. Understanding these limits helps you choose a camera that meets your needs and set realistic expectations.

The lens plays a large role. A lens with a fast autofocus motor, such as a stepping motor or a linear motor, will respond more quickly than an older lens with a slower motor. If you often film moving subjects, invest in a lens that is known for fast autofocus.

The camera's processor and autofocus algorithm determine how the system behaves in challenging situations. Some cameras are conservative and hold focus on the subject even if it moves briefly behind an obstacle. Others are more aggressive and may jump to the background. Many cameras let you adjust this behavior through a custom setting, often called AF tracking sensitivity or AF response. You can set it to be more sticky (for subjects that are briefly occluded) or more responsive (for fast changes).

Focus breathing, the slight change in angle of view when the lens focuses, is also visible in video. It is not an autofocus issue, but it affects how natural the footage looks when the camera racks focus. See our lens focus breathing article for more.

How to test tracking before you buy

Specs are a starting point, but the only way to know if a camera can track the way you need is to test it in a realistic scenario. Here is a simple test you can do in a store or at home.

Set the camera to a video mode (4K if available) and turn on the face or eye recognition. Point the camera at a person walking toward you from about 5 meters away. Record the approach and check the footage for two things: whether the face stays sharp, and whether the focus pulses (oscillates in and out) while the subject is still. Pulsing is a sign that the algorithm is struggling.

Then pan the camera to follow the person as they walk across the frame. Watch for the focus point to stay locked on the eye or face. If the point jumps to the background, that is a sign of weak tracking.

Try a low-contrast or backlit scene, such as a person standing in front of a bright window. This tests the camera's ability to maintain focus when the subject has little contrast. See our low light autofocus guide for more specific tests.

Finally, test with the lens you plan to use. Autofocus performance can vary significantly between lenses, and a lens with a fast motor will feel completely different from a slow one. Bring a lens card or check the lens specifications to understand the motor type.

In a store, you cannot always control the environment, but you can at least verify that the camera locks onto a face, tracks a moving subject, and does not have obvious focus hunting. The manual will tell you how to change AF area modes and set tracking sensitivity, which is worth reading before you test.

• Record a subject walking toward the camera and check for focus pulsing
• Pan with a subject moving across the frame and watch the focus point
• Test in backlight to see if the camera loses the subject
• Use the lens you actually plan to buy
• Try the custom AF response settings to see if they change behavior

Choosing the right camera for your work

The camera you need for video autofocus depends on what you shoot. A vlogger who films themselves talking to the camera needs reliable face and eye detection at arm's length. A wildlife filmmaker needs a camera that can track a distant bird or animal, often with a long telephoto lens. A documentary shooter needs smooth focus transitions and a system that does not hunt.

For vlogging and self-recording, look for a camera with a flip screen and good face and eye AF, plus a lens with a fast and quiet autofocus motor. The flip screen lets you confirm focus while you record. You can compare models in our best compact cameras with 4K video guide, since many of those are well suited for vlogging.

For run-and-gun documentary work, a camera with a wide AF coverage area and a large sensor will give you flexibility. The best approach is to check the autofocus performance in the mode you will actually use, which for many people is 4K. Our best 4K cameras guide lists models that are strong all-rounders.

If you film fast-moving subjects, like sports or children, look for a camera with a high frame rate and reliable tracking. Many mirrorless cameras now have impressive speed, and the AF system is one of the key differentiators. The best mirrorless cameras with 4K video is a good place to start.

For professional video work, a camcorder or a cinema camera may have fewer autofocus features than a mirrorless camera, but they offer other benefits like built-in ND filters and XLR audio inputs. See the best video cameras guide for those options.

What to pick for your use

Questions

What is the difference between AF-S and AF-C for video?

AF-S (single autofocus) locks focus once and does not change it, while AF-C (continuous autofocus) keeps adjusting focus to follow a moving subject. For video, you should use AF-C (sometimes called AF-F) so the camera keeps the subject sharp as it moves.

Is phase-detection autofocus better than contrast-detection for video?

Phase-detection is generally faster and more responsive, because it calculates the focus direction directly from the sensor data. Contrast-detection works by searching for maximum contrast, which can cause the focus to hunt or pulse in video. Most modern cameras use hybrid systems that combine both.

Why does my camera's autofocus pulse or hunt in video?

Focus pulsing is often caused by the autofocus algorithm overshooting the focus point, or by a lens motor that cannot move precisely enough. It can also happen in low contrast or low light. Adjusting the custom AF response settings, like tracking sensitivity, can sometimes reduce pulsing.

Does the lens affect video autofocus performance?

Yes, the lens motor is critical. Lenses with stepping motors or linear motors are faster and quieter, which is important for video. Older lenses with slower motors may not keep up with fast-moving subjects, even on a camera with excellent autofocus.

What is subject recognition and do I need it?

Subject recognition lets the camera detect a face, eye, animal, or vehicle and keep the focus point on it. For video, it is very useful because it frees you from manually moving the focus point. If you film people, face and eye AF is almost essential.

How many autofocus points do I need for video?

The number of points matters less than the coverage area and the quality of the algorithm. A camera with fewer points but wide coverage and good subject recognition will track better than one with more points and narrow coverage. Look at the coverage percentage in the specs.