People-Powered Code: How LANDR Sparks New Uses for a Tried-and-True Music Production Process by Tying Algorithm to Community

Instant online mastering is responding to more musicians’ needs, thanks to input from users like Of Montreal’s Bennett Lewis and the Grateful Dead’s Bob Weir

When community manager Rory Seydel joined LANDR, he discovered that on top of his job to blog about and engage with the musicians using LANDR to master their audio tracks, he would be working on customer support. At first, Seydel was surprised, “but then I saw that customer support was the key to the community,” he recalls. “Especially when the website was down, I began to hear about all sorts of different uses by all sorts of different musicians and producers.”

User feedback pointed to striking stories: Grateful Dead guitarist and founding member Bob Weir uses LANDR to master the prolific output of his studio, where a shifting cast of musicians gather regularly to jam in free-spirited collaboration. A London-based DJ’s high LANDR traffic—he mastered thousands of tracks in a handful of days—set off the service’s alarms, until Seydel discovered he was madly mastering a huge back catalog of dub plates, old dub remixes cut straight to vinyl.

“None of this would have been possible,” notes Seydel, “if these artists had had to book a studio and an engineer.”

LANDR’s reasonable price point and automated ease have done more than enable musicmakers with a lot of material or with a limited budget to get their work in sparkling final form. It has changed the way musicians and producers use the mastering stage of production, from a no-turning-back polish of final mixes, to an evolving feedback loop, a litmus test to judge a track or mix.

This is exactly how Of Montreal guitarist Bennett Lewis uses LANDR: To get a better idea of what’s working and what’s not in his mixes. “You create this song, mix it the best you can,” says Lewis, “and you get it to a point when you say, ‘This sounds good. I am going to send it to LANDR and let them do their thing’. When you get a master back you realize, ‘Oh wait, it was a long way from done.’”

“It’s like an objective set of ears,” explains Seydel, “that allows you to hear how the mix really sounds. You can go back and adjust your mix, or even rework your material.”

Meanwhile, LANDR’s A.I.-based approach learns, too, gaining more and more insight into what musicians from different genres want, by referencing a growing database of more than 800,000 past tracks. It could be a “hotter” file for an EDM artist or something a little lower in volume for a country track. Yet the bulk of this insight comes from human input and interaction, as users tell the LANDR team what’s working and what isn’t. “If we get a consensus in a genre, that’s what we try,” Seydel says. “A lot of what we do is based on user feedback. Then we make changes and get another round of feedback. There’s constant fine-tuning.”

“An algorithm is never just an algorithm, never just a bunch of code,” explains Seydel. “It’s the community around it that makes it alive, that makes it a real boon for musicians.” An automated mastering service has to have a serious human side: It has to learn to embrace genre difference, to grasp the subtleties, and deliver the clear beauty that sets mastered tracks apart from raw recordings. “This could never happen without people,” Seydel states.