Want To Make Music Like Imogen Heap? Jen, Her New AI Venture, Can Help

Jen, a tech startup backed by Grammy-winning musician/producer Imogen Heap and headed by music-tech veteran Shara Senderoff, today is launching StyleFilters, a tool that allows people to infuse the distinctive audio sensibility of an artist’s work into their own generative-AI musical creations, while ensuring the original artist gets paid.

“For so many years, people said, ‘If only we could bottle up that (musician’s) vibe,’” said Senderoff, Jen’s CEO. “No one ever said, ‘Well, what if we could?”

Each StyleFilter is somewhat akin to a character “skin” in a video game, Senderoff said, but in this case it brings the “vibe” and feel of a song into an AI music creation tool (don’t minimize the value of a vibe, by the way; it was the subject of a 2018 lawsuit and $5 million award to Marvin Gaye’s estate over similarities between Gaye’s Got to Give It Up and Blurred Lines, the similarly loose-limbed Robin Thicke and Pharrell Williams hit).

Artists who license one of their songs as a StyleFilter will receive 70 percent of the resulting revenues. The StyleFilters give a user 90 minutes of AI generation capability at one of two levels of “strength of influence,” priced at $4.99 and $7.99. Generating two minutes of audio now through Jen takes several seconds, Senderoff said, but that processing time likely will speed up to near-immediate responses in coming months, given the rapid progression of AI tools.

The modest pricing is “opening (AI music creation) to a whole new audience without wondering about whether it’s legal,” Senderoff said. “We have to change the (compensation) model. We have to create a model with a framework for digital scale.”

The pricing is roughly similar to that of many plug-ins, channel strips, loops, stabs, effects, and other audio bits and add-ons that producers and electronic musicians routinely buy and load into their digital audio workstation software, such as Acid, Pro Tools and Logic Pro.

In this case, though, the StyleFilter helps guide a generative AI model in creating a unique song with influences from a licensed artist. The user can then take the resulting audio, export it to a DAW, break it into “stem” components and further develop a song. Senderoff said.

Each StyleFilter is tied to a specific song, the launch offerings featuring five of Heap’s best-known songs, including 2005’s Headlock, which recently resurfaced on Billboard’s Hot 100 after a TikTok viral moment, Good Night and Go, and her newest single, What Have You Done to Me. The company said it plans to add StyleFilters from numerous other artists in coming months.

Jen’s base AI model is trained on about 40 licensed catalogs of songs, which ensures that the copyrighted material is legally cleared and paid for, a big issue for creatives trying to ethically make a living in the streaming era amid the disruptions of often-wholesale copying of their music by many AI companies.

“Not one note was unethically sourced,” Senderoff said. “If you go on Jen and ask it to create a song with dirty guitars and drums, you own that song. It’s fully licensed. We licensed everything.”

The London-based Heap has long been among the most tech-savvy of successful musicians. She broke through in 2002 as lead singer of the duo FrouFrou, which scored with Let Go and Breathe In. She then released a string of solo albums, including the hit Hide and Seek from 2005’s Speak for Yourself, notable for its clever use of vocoder.

Heap won a Grammy for engineering her 2009 album Ellipse, and was one of the producers on Taylor Swift’s Album of the Year-winning 1989. Four other Grammy nominations include one for music created for the Harry Potter theatrical spinoff, The Cursed Child.

In the years since her musical breakthrough, Heap has developed tech such as the Mi.Mu gloves, which allow musicians to manipulate sound with gestures; Mycelia, a blockchain-based music-distribution platform; and Mogen, an AI assistant modeled after Heap.

Heap also founded Auracles, a company providing verified artist identities and designed to streamline artist approvals of uses of their music. Senderoff said Auracles’ ID capabilities are a key underlying technology for the Jen StyleFilter architecture.

Heap memorably minted non-fungible tokens live on the audio app Clubhouse during the depths of the Covid lockdown, a process she’d been doing for several years before NFTs and even crypto and blockchain technologies were in wide use.

Like Heap, Senderoff has been in and around music, and the technologies transforming it, for a long time.

Her father worked for such major music hardware manufacturers as Gibson, Korg, and Marshall. Senderoff herself spent several years partnered with music super-manager Scooter Braun in a Los Angeles-based fund backing music-related tech companies, while functioning as something of an advisor without portfolio for labels and musicians trying to understand blockchain, cryptocurrencies, and now artificial intelligence.

The two have been collaborating over the past year to create connections between Heap’s company and Jen’s capabilities, Senderoff said. Heap has been intensely involved the past eight months architecting the business model that ensures originating artists get paid.