Adobe adds AI audio cleanup and 3D tools to Premiere and After Effects

IBC 2023 just kicked off, and Adobe isn't wasting any time showing video pros where their monthly subscription money is actually going.

Instead of chasing wild text-to-video gimmicks, Adobe rolled out heavy beta updates across Premiere Pro, After Effects, and Frame.io focused squarely on killing mundane busywork.

Cutting down the daily grind

If you edit talking heads, you know the drill: hunting down stray background noise and trimming hundreds of "ums" and "ahs."

Premiere Pro is taking a direct swing at both. The new Enhance Speech tool uses AI to strip out background rumble and make low-grade field mics sound like they were recorded in a professional booth. It even has a mix slider so you don't accidentally make the audio sound unnaturally sterile.

Meanwhile, the Text-Based Editing feature can now automatically spot filler words and pauses, letting you bulk-delete them across an entire interview in one shot. Audio Category Tagging also works in the background to label dialogue, music, and ambient noise, pulling up the right tools the second you click a clip.

Under the hood, Adobe says timeline drawing performance is now five times faster compared to the previous version, paired with an Effects Manager to isolate crashing plugins.

Real 3D in AE and direct-to-cloud shooting

Over in After Effects, motion designers get a true 3D workspace. You can now import 3D models natively without clunky third-party bridges, using image-based lighting to blend models naturally into live-action scenes.

AE's Roto Brush also received an upgraded AI model that drastically cuts down the pain of isolating tangled hair, waving limbs, and transparent fabrics frame by frame.

For remote collaboration, Frame.io launched Storage Connect, letting enterprise users link their existing AWS S3 buckets directly rather than moving all their assets to new servers. Its Camera to Cloud network also grew, picking up direct integrations with Fujifilm's new 102MP GFX-100 II, Atomos Ninja recorders, and Accsoon's SeeMo adapters that turn iPhones into cloud-uploading monitors.

All these updates are available in beta starting today, with general availability slated for later this fall.

Why it matters

While generative AI startups promise to replace creative directors overnight, Adobe is winning points by quietly automating the jobs nobody wanted to do anyway.

We think features like Enhance Speech and filler-word removal offer immediate, tangible value. If you edit podcasts, YouTube explainers, or documentary interviews, this easily saves an hour or two on rough cuts. You don't have to alter your creative instincts; the timeline just gets out of your way faster.

Strategically, Adobe is tightening its grip on the full production pipe. By connecting camera hardware directly to Frame.io proxies and feeding those straight into Premiere, they are turning video production into an unbroken loop that keeps rivals like DaVinci Resolve on defense.

The only real test will be reliability. Video editors have endured decades of crashes and bloat, and while a five-fold timeline speedup sounds fantastic on paper, Adobe still has to prove these beta toys won't crumble under real-world deadlines.