Advertisement

Adobe Premiere Pro can automatically reframe your videos

The new feature is powered by Adobe’s AI and ML tech, Adobe Sensei.

Reframing video footage to fit a different aspect ratio is a painful process. Simply changing your project's settings from a widescreen aspect ratio to a square one won't cut it; you'll also have to shift the footage around to keep all the action in the reduced frame. Auto Reframe in Adobe's Premiere Pro will do that shifting automatically from now on. Revealed today at the International Broadcasting Convention, this feature will help editors quickly create videos suitable for different platforms, something that can take hours when done manually.

If you were around during the VHS era, you may already be familiar with the process of changing aspect ratios. "Pan and scan" editing was used to convert 16:9 theatrical movie prints into 4:3 home videos. Editors had to move the footage frame by frame to make sure all relevant visuals stayed on-screen. Today, editors have to do the same when creating multiple versions of the same video. For example, an editor's main task may be to create a 16:9 video to be published on YouTube, but he may have to create a square version of the video for Instagram as well.

"With Auto Reframe, users can simply drag the effect onto the individual clip or clips they wish to reframe and it does the work for them, saving countless hours," says Adobe. The effect is powered by Adobe Sensei, the company's AI and machine learning engine that automates tasks throughout the Creative Cloud software suite. Sensei analyzes each frame of the footage and creates keyframes (basically time-based coordinates) that follow the action and adjust the framing accordingly. This will cut the time it takes to create a new version of a video from hours to minutes. And for editors, who are often freelancers or contractors, time is money.

Auto Reframe will launch on Premiere Pro later this year. Motion graphics templates will also be coming to After Effects, making animated graphics, text and lower third visuals a faster, easier process.