Audacity gets free AI plugins for music separation, generation and remixing

>It might be free and open-source, but Audacity certainly isn’t an audio editor that’s failing to move with the times. It recently got itself a new suite of plugins that are powered by Intel’s OpenVINO AI toolkit, and it sounds like some of them could be pretty useful.

First up, we have Music Generation and Music Style remix, both of which use the Stable Diffusion AI model – specifically in its Riffusion guise – to create either new music from a prompt, or existing music.

Music Separation, on the other hand, does what it says on the tin: breaks a song up into either vocal or instrumental parts, or into vocal, drums, bass and a combined ‘anything else’ part.

Stem Separation is increasingly common; not only does it form the basis of a number of desktop and browser-based applications, but it’s also being included in an increasing number of DAWs and other music production packages.

Among the other new features are Noise Suppression, which is designed to remove background noise from an audio file, and Whisper Transcription, a tool for transcribing spoken word audio or vocal recordings.

“We decided to add AI features to Audacity so that we could offer capabilities that were previously unavailable or just really difficult to achieve,” Audacity product manager Martin Keary told Intel Business. He also confirms that “the plugins run locally on your PC rather than from a server, addressing privacy concerns”.

According to Keary, there could be more to come, too. “This is just a first step,” he says. “We hope to continue partnering with Intel to develop all kinds of new AI tools in the future to help take Audacity to a level no one’s ever seen before.”

The only downside here is that the AI plugins are currently only available to Windows users, but we’re hoping that they’ll be coming to Mac and Linux, too.