Nine Inch Nails quietly used L-Acoustics' Source Intelligence on Trent Reznor's vocals on the Peel It Back Tour

Trent Reznor’s vocals on Nine Inch Nails’ Peel It Back Tour went through machine learning, and the band’s sound engineer kept that “hush-hush” for the whole run.

Billboard reports in its November 7, 2026 issue that front of house engineer Jamie Pollock used L-Acoustics’ Source Intelligence on the tour for the first time. Front of house is the person mixing what the crowd hears.

Source Intelligence makes no new audio. It’s nongenerative machine learning that picks out the singer’s voice in real time and removes the rest of what the microphone caught.

What Jamie Pollock did with Trent Reznor’s vocals

Pollock never announced it, but Billboard says fans noticed the difference anyway. Reznor’s voice came through clean on everything from a piano version of “Something I Can Never Have” on the B-stage to the heavy material.

That range is what makes a Nine Inch Nails vocal hard to mix.

“

Trent's whispering at times; he's trying to make it as intimate as possible. And then [the next] minute, he's screaming.

A whisper needs a lot of gain, and on a loud stage that gain brings everything around the mic up with it.

How L-Acoustics Source Intelligence cleans a live vocal

“The microphone that the singer is holding onstage is capturing more than the voice,” L-Acoustics CEO Laurent Vaissié told Billboard. “You have the drums, you have the guitar amps, you have the crowd.”

Add PA bleed, which is the sound from the speakers leaking back into the mic. That’s also where feedback comes from.

L-Acoustics says Source Intelligence continuously identifies the voice and strips out the rest, and that an engineer doesn’t need to set a threshold or change the console workflow.

If you’ve split a track in Pro Tools or Moises, you already know the idea. Mix calls it “stem-separation-like technology”. The difference is the deadline: a stem splitter on your laptop can take its time, and a live vocal can’t, so L-Acoustics spec’s end-to-end latency at less than 8.5 ms.

One caveat: the 40 dB and 8.5 ms figures are L-Acoustics’ own, and I haven’t tested it. You’d need a tour’s PA to do that.

Which artists and venues use L-Acoustics Source Intelligence

Nine Inch Nails were one of the first acts to test it, and L-Acoustics says it’s been in the field since Spring 2025. It launched at the L-Acoustics Keynote 2026, and here’s who else has it, per Billboard and L-Acoustics:

• Bruno Mars, Rüfüs Du Sol and Harry Styles have started using it on the road.
• Clair Global is a beta partner on the Benson Boone and Rosalía tours.
• The Broadway production Dog Day Afternoon is using it through PRG.
• The Hollywood Bowl runs it in a new system with 26 L-Acoustics speakers, which the Los Angeles Philharmonic first played through on June 20.

“It’s basically studio quality vocals now, which changes the game completely,” says Joey Diehl, front of house engineer for Benson Boone, in L-Acoustics’ press release.

There’s one limit. Source Intelligence runs only on the L-ISA Processor II for engineers mixing on an L-Acoustics PA, so you can’t load it into any rig.

This is one piece of a bigger shift in live music. The same Billboard feature covers AI in concert marketing and ticket bots, and Suno hired a Ticketmaster exec to run a live division in October. Vaissié’s verdict on how shows sound now: “The results are pretty mind-blowing.”