“LALAL.AI unlocks previously unusable content": A Lead Video Editor on Embracing AI in Video Production

As Lead Video Editor at one of the world’s biggest agency networks, Daryl Yeoman from VML is no stranger to tight deadlines, challenging briefs, and audio files that show up broken, corrupted, or baked into oblivion. But instead of letting technical setbacks stall the creative process, he’s found ways to work smarter, without compromising on storytelling.

In this interview-slash-case-study, we spoke to the London-based editor to learn more about how the team works, how LALAL.AI fits into their daily grind, and what the future holds for editors in a world increasingly shaped by automation. From separating stems in multi-language interviews to breathing new life into archival ads with no working files, here’s how he’s using technology to make more room for creativity.

Inside VML UK: A Post Facility That Does It All

We asked Daryl to shed more light on VML UK and their video production team's work.

As the Lead Video Editor for our London team, my role actually extends beyond the local studio. I also serve as a crucial bridge, connecting our diverse editing hubs across the globe. A significant aspect of my responsibility involves fostering a collaborative environment where editors can share ideas, best practices, and cutting-edge tools.

I’m really interested in how we can use AI to make our work more efficient but not at the cost of creativity. Finding that sweet spot, where we can be more productive without losing that human touch, is what I’m always aiming for. It's about finding ways to use these new technologies to enhance our creative vision, not replace it.

Video Editor - Megan Leydon, Motion Graphics for DOOH - Michael Friddle

We’re really lucky to have WPP Open at our disposal; as a part of the wider network, all the agencies under WPP have access to the state-of-the-art technology platform. It houses a vast range of AI tools we use daily. The editors are often found generating videos for rough cuts, image generation for storyboards, and also utilising the platforms LLMs to generate transcriptions, or using the Canvas BETA to help us ideate different social outputs, or routes to a script.

The editing team in London is made up of five full-time editors, all with mixed disciplines. We also work with trusted freelancers and have up to five working with us at anytime. We undertake a vast range of content, from TVC, VOD, Social, Online, DOOH, Internal Comms, and the all-important mood films and case studies. We cover offline to online, making us a full-service in-house post facility.

We're always keeping an eye on trends, and what's behind those movements. We'll definitely look to follow and mimic some of those techniques when it makes sense for the project.

We like to mix and match, take what works, and then add our own twist. It's about staying current but also coming up with our own fresh ideas.

We want to be the ones setting the trends, not just following them.

We recently tackled a massive project with interviews in fifteen different languages. We discovered Riverside.fm. Honestly, it's become one of our favorite tools. Uploading hours of interviews and having it automatically detect the language, translate into English, and spit out a perfectly formatted SRT file. It’s magic!

Video Editor - Daryl Yeoman, Assistant Video Editor - Megan Leydon

We could instantly get those subtitles onto our timelines, which saved us an incredible amount of time. It allowed us to focus on the creative aspects of crafting the narrative. Tools like Riverside.fm make me excited about the future of video editing. They're not just about saving time; they're about unlocking that time for new creative possibilities.

Sound First: Why Audio Matters More Than Ever

It's no secret that audio is a huge part of how a video is perceived. Be it a movie trailer, an ad of a world-renowned brand or a video of a small TikTok creator, audio and music can either make or break it. We asked Daryl about their approach of working with sound.

Honestly, sound is everything. It's not just something you hear; it's something you feel. That old saying, 'if you can't hear it, it's working,' that's just wrong. Sound hits you, it creates a mood, it tells you what to feel. And with excellent sound design libraries like Soundly, they're like a secret weapon for editors, helping us inch a little closer to that magic that the world's best sound designers create.

Audio is the most critical part of any moving image work. And because of this, you have to ensure you start the project correctly. We have extensive ways of working at VML UK, and all editors adhere to strict guidelines on how to prepare audio throughout their project through to delivery.

Video Editors - Stephen McInerney & Daryl Yeoman

We're focusing on crafting stories that have room to breathe. We're slowing things down, giving the audience time to connect with what they're seeing. Sound plays a huge role in that, it's how we add that emotional layer. We're focusing on the simple things, like how we use text and animation. Making sure they enhance the story and aren’t just a bolt on.

"For our team, the core problem LALAL.AI solves is the ability to unlock and repurpose previously unusable content."

Surely, we couldn't help but ask Daryl how LALAL.AI helps the team in their routine.

We use LALAL.AI for a range of different projects. Mostly internal films and mood films, but occasionally, we would use on external facing projects. For our team, the core problem LALAL.AI solves is the ability to unlock and repurpose previously unusable content. When assembling mood films, we often find ourselves needing to utilise diverse material.

LALAL.AI's remarkable ability to separate audio stems allows us to extract the specific elements we require, whether it's dialogue or music, effectively salvaging valuable content.

Moreover, it's a lifesaver for transcreation projects! When dealing with files from external agencies, issues like corruption or missing assets can derail progress. LALAL.AI acts as a powerful backup, enabling us to isolate and remaster essential audio elements and ensure seamless project delivery. It's become a critical component of our toolkit, allowing us to overcome audio limitations and maintain our creative momentum.

When working with video materials that have “baked-in” audio, to put it bluntly, you are screwed.

When crafting mood films, we often need to repurpose existing footage, but the baked-in audio can hinder our vision. By extracting vocals, we gain the ability to build a truly immersive soundscape, layering our own music and sound design to create a more compelling and emotional experience.

That massive multilingual project, where we also relied on Riverside.fm, stands as a testament to the value of LALAL.AI. We received delivery assets that were not only in various languages but also had music baked into the interview. This presented a significant obstacle to achieving the clear, uncompromised audio we needed.

LALAL.AI's remarkable ability to isolate audio stems allowed us to effortlessly remove the baked-in music, restoring the interviews to their original, pristine quality. These tools played a pivotal role in ensuring the project's success, allowing us to deliver a polished and professional final product despite the initial audio challenges.

AI-powered stem separation already is an integral part of daily life for us.

The Future of Video Production & the Next Generation of Creators

Tools like OpenAI's Sora must have disrupted the video production community, with many wondering if there's a point in rolling into the field. We were curious about Daryl's stance on whether they really are a threat to video production teams.

These tools will be classed as a threat until the general public have too passed the Rubicon. But as professionals, we’re utilising these technologies to get ahead, push the boundaries of what’s possible, and free up more time to be even more creative.

We're definitely at a turning point, and it's going to be interesting to see how we use all these new tools. But what gets me going is thinking about the next generation of editors.

The kids who are already making amazing stuff in their bedrooms, they're just so creative in ways we’d never have even thought of. When they start coming into the industry in the years to come, it's going to be wild. They're going to change everything.


VML’s UK team is proof that AI tools like LALAL.AI aren’t replacing creativity, they’re enhancing it. From salvaging unusable audio to giving editors more space to focus on narrative and emotion, stem separation has become an everyday essential in their toolkit. And with an eye on both innovation and the next wave of creators, they’re setting the tone for where video editing is headed next.


Follow LALAL.AI on InstagramFacebookTwitterTikTokReddit, LinkedIn, and YouTube to keep up with all our updates and special offers.

💛
And if you want to share your own story about how LALAL.AI helped you and what you're using it for, feel free to fill in this form (it won't take long, we promise!). As a token of our gratitude, we'll add 300 minutes to your account!