A Filmmaker’s Guide to Precision Autofocus in Video Cameras

Published:
Updated:

When shooting video, the mantra is usually “Always use manual focus.”

While it is true that manual focus gives you the most control over your project, autofocus technology has come a long way, especially in recent years.

In this article, I’ll look what type of autofocus is best for video, and also at the technologies behind different types of autofocus.

Autofocus vs continuous autofocus

There’s a big difference between autofocus for stills and continuous autofocus for video.

Autofocus for stills needs to quickly focus on a target to snap a photo. But auto focus for video needs to constinously track a moving object and quickly switch between objects.

So what we want in a video camera is good continuous autofocus.

There are certain situations in which you might need to utilize your camera’s continuos autofocus. For example:

  • When you’re a single-camera operator doing interviews or corporate videos
  • When you’re filming using a gimbal
  • When you’re filming dolly shots (because keeping focus on the subject can be a difficult and preoccupying task as the camera-to-subject distance changes).

You can see the cameras with the best continuous autofocus on the market today here.

The four types of Autofocus Technologies: Contrast Based, Phase Based, Dual Pixel, and Time of Flight

First, let’s talk about how autofocus works for your camera.

There are four leading technologies: contrast-based, phase-based, dual-pixel, and time-of-flight (or laser) autofocus.

However, outside of some types of phone cameras and some DIY experimental autofocus systems, laser autofocus is quite rare.

However, there are indications that we might soon start to see time of flight in prosumer cameras.

The two most widespread technologies are contrast-based autofocus and phase-based autofocus.

Hybrid Autofocus Systems

I just wanted to add a quick note on hybrid autofocus. Often, camera manufacturers today use a hybrid autofocus system that utilizes two or more of these autofocus systems.

Fx Panasonic has begun combining phase-detect and contrast-based autofocus, getting the best of both worlds.

Furthermore, these systems can be combined with features such as eye-tracking for even better accuracy.

The best autofocus systems available on the market today consist of these technologies:

  • Phase detection + contrast-based + eye detection (fx Panasonic)
  • Phase detection + eye detection (fx Sony)
  • Dual Pixel autofocus (Canon)

Below, you can read more about these in more detail.

Contrast Based Autofocus

Contrast-based autofocus is a technology that has been around for a while. It mimics what you would do when searching for focus manually.

I.e., when you look through your viewfinder, you bring the subject into or out of focus by moving the lens in or out – and you know you’re in focus when the edges of your subject are sharply and defined.

With contrast-based autofocus, your camera does essentially the same thing. The camera looks for micro-contrast along edges in a small image section.

If the edges surrounding the subject are out of focus, the dark and light pixels are blurred together, and there is low contrast. When the subject is brought into focus, the lights and darks are not blurred and there is higher contrast.

Contrast-based autofocus can do a great job of bringing your subject into very clear focus and even track the subject to a degree.

However, there are drawbacks.

When hunting for focus, you must gradually adjust the lens in and out, searching for the sweet spot. Your camera does the same thing.

The autofocus does not tell the camera whether the lens is front-focusing or back-focusing.

Consequently, the camera will move the lens back and forth, bringing the subject into focus, overshooting, and then readjusting to optimize.

This can result in subtle pulsing in the backgrounds as the camera continuously refocuses.

Another drawback to contrast-based autofocus is that it can take longer for the camera to focus on the subject than other technologies. This can be an issue when shooting video and there is lots of camera movement or in the frame.

The relative weakness of contrast-based autofocus when shooting video, along with the fact that it can remain very effective for focusing even in settings with a shallow depth-of-field or shooting through telephoto lenses, means this technology is better suited for keeping track of moving subjects in still photography.

Phase Detection Autofocus

The second major autofocusing technology we will look at is phase detection autofocus. This method works by using dedicated focusing sensors.

In some cameras, certain pixels are used for focusing rather than for capturing the image. The missing pixels are interpolated from the surrounding image.

In the newer dual-pixel autofocus technology, pixels can perform both tasks: focusing and image capture. But more on that later.

Phase detection autofocus works in stereo. A sensor receives signals from two separate diodes (the light-sensing bits in your camera). A microprocessor then analyzes the similarity in the signal.

If light is focused on the same place in each sensor, then the image is in focus. If not, the microprocessor calculates how far and in what direction the lens must be moved to bring the subject into focus.

The nice thing about this system is that the camera knows whether to move the lens forward or backward and just how much. This makes phase detection autofocus significantly faster than contrast-based autofocus.

For shooting video, this is a boon, as it means you can smoothly and continually maintain focus on your subject, even during complex camera maneuvers or a rapidly moving subject.

Unfortunately, due to the quantized nature of the sensors, there is a degree of rounding when computing the exact lens distance by your camera.

Usually, this causes no discernible effect, but for exact photography with a very shallow depth of field or when shooting with a telephoto lens, your camera might be slightly less in focus compared with contrast-based autofocus.

Phase Detection Autofocus is excellent for video.

Dual Pixel Autofocus

Dual pixel autofocus is a relatively new technology.

When Canon introduced it, they touted it as a revolutionary innovation. Whether or not the product lived up to the hype is debatable. However, it is becoming a widespread feature as numerous companies seek to introduce models equipped with similar technology.

At its core, dual-pixel autofocus is a form of phase detection autofocus. However, whereas traditional phase detection relies on pixels that capture the image OR focus on the subject, dual pixel autofocus does both simultaneously.

This is achieved by focusing the light input in each pixel onto two photodiodes (the material that senses light). There are then two input feeds per pixel that are sent to a microprocessor. The microprocessor then analyses the light data and sends a signal to reduce the phase shift between signals.

Dual pixel autofocus gives you the same advantages as phase detection technology. There is no need to hunt for focus, quicker focusing and keeping the subject in focus, and the ability to adjust focusing speed and sensitivity.

Because pixels can both capture images and determine focus distance, this technology eliminates the need to interpolate the missing image pixels.

Also, since each pixel can be used to determine focus distance, it is easier to focus on subjects in any area of the frame.

This technology is also beneficial for tracking and maintaining focus on a moving subject or shifting focus between different subjects.

Dual Pixel autofocus is great for video.

Time of Flight Autofocus

Besides contrast-based and phase-shift autofocus, there is one more major technology we will take a look at. Time-of-flight or laser autofocus is a less widespread technology that works similarly to sonar.

Few cameras, outside phone cameras, and a few DIY solutions are equipped with this autofocus technology. Still, it can’t hurt to understand how this technology works, so let’s take a quick look at it.

As mentioned, time of flight autofocus works a lot like sonar. The camera sends out a beam of visible or infrared light, which bounces off the subject and back to the sensor. The camera measures the time it takes for the emitted light to return to the sensor.

Because the speed of light is constant, the distance between the camera and the subject can be calculated. The camera can then adjust the lens to the appropriate distance to bring objects at that distance into focus.

Time of Flight autofocus is not widespread, but there are rumors that we might start to see it soon in prosumer and professional video cameras.

Conclusion

The technology and quality of built-in autofocus have come a long way in recent years and can be a useful tool in many situations.

Whether you are shooting stills and need to track a moving subject or filming video and need to refocus continuously, there are advantages and drawbacks to each technique.

Contrast-based autofocus systems can give you the highest focus on a subject, even in shallow depth-of-field situations. This technology can help shoot stills, even when using a tele lens.

Phase detection autofocus systems eliminate the need to hunt for focus. This makes it quicker and cleaner to keep your subjects sharp during a shoot. Evolving tech such as dual-pixel autofocus continues to refine this system, making it a powerful tool for a videographer.

At the time of writing, the two best technologies for continuous autofocus for video and filmmaking is phase detection autofocus and dual pixel autofocus.

Time of flight autofocus shows promise but hasn’t been integrated into professional and prosumer cameras. But that might soon change.

The next step is understanding the different autofocus settings available on video cameras.

But whether you are focusing by hand or utilizing your camera’s built-in features, what matters is your vision. So stay creative and keep shooting.


Author

    by
  • Cade Taylor

    Cade Taylor is a filmmaker and writer based out of Los Angeles. Originally from Seattle, he continues to work as the Outreach Coordinator for the Bigfoot Script Challenge, where he helps connect up-and-coming writers with industry professionals. When he’s not working on his own projects, helping out with Bigfoot, or covering desks, Cade loves to share what he knows with other filmmakers and promote great content.

    View all posts

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.