L-Acoustics Introduces Source Intelligence for Vocal Separation in Live Environments

L-Acoustics has announced Source Intelligence, designed to isolate vocals and minimize smearing during live performances. Using machine learning, you can filter out sound sources such as instruments, crowd noise, and feedback from monitors while preserving the singer’s voice. This feature rejects up to 40 dB of extraneous signals, reducing the risk of feedback.

Machine learning algorithms accomplish this by continually identifying voices as different sound sources enter the vocal microphone, and filtering out those that don’t match the singer’s consistent tone and range.

“For decades, engineers have solved PA bleed and feedback problems with compromises like notch filters that color the sound, unpredictable expanders, and staging layouts that separate performers from the audience,” explains Ryan John, director of software product management at L-Acoustics. “Source Intelligence completely removes that trade-off. The machine listens continuously and operates continuously, allowing engineers to focus on the mix instead of fighting feedback, and performers to own the entire stage.”

This separation essentially allows for studio-level complexity during performance. Features like EQ and reverb are uninterrupted by other instruments or noise sources, allowing live engineers to adjust vocals based on their unique delivery and current environment. Engineers across the industry are already praising Source Intelligence, as vocals are a key element in almost any type of live performance.

“Source Intelligence delivers a vocal sound like no other. It’s basically studio-quality vocals and it completely changes the game,” says Joey Deal, FOH Engineer at Benson Boone.

“I was very impressed with the amount of headroom, and was able to get crisp, clean vocals in all scenarios, including very noisy stage environments and actors in front of a PA,” said Tony Award-winning sound designer Cody Spencer, who used the system on the Broadway production of Dog Day Afternoon.

Source Intelligence is currently only available within the L-Acoustics L-ISA Processor II. Once installed, it integrates seamlessly into your L-Acoustics workflow.

For more information, please visit: L-acoustics.

Harry Levin is a freelance journalist whose credits include SPIN, Billboard, MusicTech, Grammy.com, and Los Angeles Magazine. His musical journey began 20 years ago with a Led Zeppelin CD. He played jazz trombone throughout college, produced large-scale electronic music events, and now devotes his professional time to writing and editing.